HORMONAL INFLUENCE ON RENAL FUNCTIONWITH PARTICULAR

REFERENCE TO DIABETES MELLITUS

Klaas Hoogenberg

CONTENTS

Chapter 1 Introduction. 1

Chapter 2 Abnormal plasma noradrenaline response and exercise-induced 37albuminuria in IDDM patients.Scand J Clin Lab Invest 1992; 52: 803-811

Chapter 3 Influence of ambient plasma norepinephrine on renal haemo- 49dynamics in IDDM patients and healthy subjects.Scand J Clin Lab Invest 1995; 55: 15-22

Chapter 4 Exogenous norepinephrine induces an enhanced micro- 61proteinuric response in microalbuminuric IDDM patients.J Am Soc Nephrol (in press)

Chapter 5 Norepinephrine-induced blood pressure rise and renal 81constriction is not attenuated by enalapril in microalbuminuricIDDM patients. Nephrol Dial Tansplant (in press )

Chapter 6 Effects of low dose dopamine on renal and systemic 91haemodynamics in during incremental norepinephrine infusion in healthy volunteers. Crit Care Med 1998 (in press)

Chapter 7 Effect of growth hormone (GH) and insulin-like growth factor I 101on urinary albumin excretion: studies in acromegaly and GH deficiency. Acta Endocrinologica Copenh 1993; 129: 151-157

Chapter 8 Insulin-like growth factor I and altered renal haemodynamics 111in growth hormone deficiency, acromegaly, and diabetes mellitus. Transplant Proc 1994; 26: 505-507

Chapter 9 Contributory roles of circulating glucagon and growth hormone 117to increased renal haemodynamics in IDDM patients. Scand J Clin Lab Invest 1993; 53: 821-828

Chapter 10 Increased urinary IgG/albumin index in normoalbuminuric 127IDDM patients: a laboratory artefact. Diabetic Medicine 1996; 13: 651-655

Chapter 11 Alterations in cortisol metabolism in IDDM patients: relationship 135with metabolic control and estimated blood volume and effect of angiotensin converting enzyme inhibition.J Clin Endocrinol Metab 1995; 80: 3002-3008

Chapter 12 Summary and conclusions. 149

Nederlandstalige samenvatting. 155

Information on the author, list of publications. 163

Dankwoord. 167

Chapter 12

Introduction

Background

Epidemiology of diabetic renal disease

Functional stages of diabetic renal involvement

Microalbuminuria: a predictor of DN and cardiovascular disease

Pathogenesis of diabetic renal disease Metabolic factors Renal haemodynamic factors Lipoproteins Genetic factors Renal effects of noradrenaline The growth hormone-insulin-like growth factor-I system and kidney function

Sodium and volume homeostasis in IDDM and the role of 11$-hydroxysteroid dehydrogenase

Treatment of diabetic nephropathy Arterial hypertension Metabolic control Dietary intervention

References

Purpose of the thesis

Abbreviations: IDDM: insulin-dependent diabetes mellitus; DN: diabetic nephropathy; ERPF:1

effective renal plasma flow; GBM: glomerular basement membrane; GFR: glomerular filtrationrate; GH: growth hormone; IGF-I: insulin-like growth factor-I; NE: norepinephrine; RAAS: renin-angiotensin-aldosterone system; SNS: sympathetic nervous system

CHAPTER 1

INTRODUCTION

Background

In patients with (insulin-dependent) diabetes mellitus (IDDM) , the development of1

nephropathy, hallmarked by the presence of proteinuria in excess of 0.5 g/day, is a seriouscomplication. Diabetic nephropathy (DN) leads to progressive deterioration in kidneyfunction. Moreover, DN is associated with a highly increased incidence of cardiovasculardisease. Renal failure and cardiovascular disease are in fact competing risks in thesepatients, making renal replacement therapy only necessary in those who survivecardiovascular complications. The natural course of DN has changed over the last decades.Early and aggressive treatment has been shown to retard renal function loss and, inparticular, to improve survival in these patients. Despite the large progress in the treatmentof DN, its etiology is yet incompletely understood and it is not possible to prevent thiscomplication. There is evidence that both hereditary, as well as metabolic andhaemodynamic factors contribute to its pathogenesis. Knowledge of these factors willidentify patients at risk for developing nephropathy. The concept of microalbuminuria, i.e.a urinary albumin excretion rate between 20 and 200 µg/min or 30 to 300 mg/day, as anearly clinical sign of diabetic renal involvement, has greatly improved our understandingof the natural course of diabetic renal disease and has enabled the development of earlyintervention and prevention strategies.

This thesis aims to evaluate the influence of norepinephrine (NE) and the growth-hormone insulin-like growth factor-I (GH-IGF-I) axis on renal function. Both substancesbelong to hormonal systems that control renal function in opposite directions: NE causesrenal vasoconstriction, whereas stimulation of the GH-IGF-I-system induces renalvasodilation. The early stages of diabetic renal involvement are characterised by imbal-ances in glomerular vasodilatation and vasoconstriction. Against this background, thepossible role of these hormonal factors in DN is investigated.

This chapter outlines the epidemiology, the functional stages, the pathogenesis andthe therapeutic aspects of renal disease in IDDM. Several aspects of the pathogenesis ofDN are more extensively overviewed in sections on the effects of NE and the GH-IGF-I-system on kidney function. Abnormalities in sodium and volume homeostasis in IDDM,and the role of 11$-hydroxysteroid dehydrogenase (11$-HSD) in protecting the mineral-ocorticoid receptor from activation by cortisol is briefly recapitulated.

Chapter 14

Epidemiology of diabetic renal disease

In the early cohorts of IDDM patients, diagnosed between 1933 to 1959, thecumulative incidence of nephropathy amounted to 41-43% after 25 years of diabetesduration [1,2]. In these cohorts, nephropathy was associated with a 10-year mortality rateof 50 to 77% [1,2]. Comparing the 25 years cumulative incidence of nephropathy inIDDM diagnosed between 1933-1942 and 1953-1962, a remarkable decrease was notedfrom 41% to 27% [3]. A very spectacular decline up to 28%, 8.9% and 5.8% in IDDMpatients diagnosed between 1961-1965, 1966-1970, and 1971-1975 has been reported inmetabolically wellcontrolled Swedish patients [4], but in Danish cohorts a 35% incidenceis still observed [5].

Survival in IDDM with proteinuria has improved dramatically. In Denmark, IDDMpatients with onset of proteinuria between 1957 and 1973 had a mortality rate that was 40times higher than in patients without proteinuria [6]. After onset of proteinuria, the 8 yearssurvival rate of such patients was only 48%. In comparison, 8 years survival in IDDMpatients with onset of proteinuria between 1974-1978 and 1979-1983 had increased to82% and 87%, respectively [7,8].

The decline of incidence in overt proteinuria and cardiovascular mortality has beenattributed to both improved blood pressure regulation and metabolic control [3-5]. In theolder epidemiological studies [1,2,6], arterial hypertension was not treated since it was notrecognised to have prognostic significance. Glycaemia could also not be strictly controlledsince home-based blood glucose and glycosylated haemoglobin measurements were notyet available [9]. Remarkably, the peak incidence in proteinuria (10 to 15 years after theonset of IDDM) has remained unchanged over the last decades [1-6].

It is noteworthy that there is also a decline in the incidence of progression frommicroalbuminuria to overt proteinuria in IDDM. Previously, 80 to 90% of microalbumi-nuric IDDM patients progressed to overt nephropathy [10,11]. Recent estimates revealedthat during the last 10 to 15 years only 30% of microalbuminuric patients progressed toclinical proteinuria [12-16]. This suggests that better metabolic control and blood pressureregulation are currently achieved in many microalbuminuric IDDM patients. The DiabetesControl and Complication Trial (DCCT) indeed showed that intensive insulin treatmentreduces the progression of microalbuminuria in IDDM [17,18]. Furthermore, several trialsunequivocally demonstrated that early treatment with antihypertensive drugs can arrest ordelay the progression of microalbuminuria in normotensive microalbuminuric IDDM [19-22]. Only 10% of microalbuminuric patients treated with ACE-inhibitors developed overtproteinuria over an 8 years period, which is remarkably different from the 40% incidenceof microalbuminuria in patients not treated with ACE-inhibition [22]. Longer follow-upwill clarify whether intensive insulin treatment [17,18] and early ACE-inhibition treatment[19-22] will really prevent or only postpone DN. From an optimistic point of view, theaforementioned estimates suggest a decline in incidence and prevalence of nephropathy inthe next decades.

Functional stages of diabetic renal involvement

Introduction 5

Table 1. Functional stages of diabetic renal involvement

GFR Ualb.V MAP

Stage 1 Hyperfunction 88 = (8) = Nephromegaly*

Stage 2 Silent phase 8/= =(8) = Early histological changes*

Stage 3 Microalbuminuria = 8(88) =(8) Albumin excretion 20-200 µg/min ** **

(30-300 mg/day)

Stage 4 Nephropathy 9 88 8 Dipstick positive (albuminuria > 200 µg/min, proteinuria>500 mg/day)

Stage 5 End stage renal failure 99 88 88 Dialysis support or transplantation

GFR: glomerular filtration rate, Ualb.V urinary albumin excretion rate, MAP: mean arterialpressure. * can be present during poor metabolic control, ** aggravates during exercise.

The renal changes in IDDM patients are classically divided into 5 functional stages[23] (Table 1). An increased kidney size due to glomerular enlargement and tubularhypertrophy and hyperplasia (renal hypertrophy/hyperplasia), with concomitant increasesin glomerular filtration rate (GFR) and renal blood flow (glomerular hyperfiltration/hyperperfusion) are typical for stage 1 diabetic renal involvement. Transient increases inurinary albumin excretion can be seen at diagnosis of IDDM and often reverse afterinstitution of insulin therapy. Physical exercise testing is associated with abnormalalbumin excretion rates at this stage of renal involvement. In a subset of patients, renalhyperfunction persist for years, especially during poor metabolic control.

Stage 2 is characterised by early histologic alterations such as glomerular basementmembrane (GBM) thickening and mesangial expansion, that are generally present after 2years of disease duration. Except for the aforementioned exercise provocation, albuminexcretion rate is normal, so that the glomerular filtration barrier against the loss ofmacromolecules is assumed to be intact. However, some studies have found an increasedurinary excretion of the much larger and neutrally charged IgG in normoalbuminuricIDDM patients [24]. In the context of an intact GBM this finding is not well understood.GFR is either normal or elevated, comparable to stage 1 involvement. There are noreliable methods to diagnose this stage of renal involvement and the early histologicabnormalities have been found to correlate poorly with future progression to DN [25].Also, the increase in urinary albumin excretion induced by exercise lacks any predictivevalue on the future development of nephropathy.

Stage 3 diabetic renal involvement, also designated as incipient diabetic nephropath-y, is characterised by a persistently raised albumin excretion rate (micro-albuminuria),which typically develops after 5 to 15 years of diabetes duration in a subset of IDDMpatients. The raised albumin excretion is ascribed to an early impairment of the glomerularfiltration barrier against the loss of macromolecules. A decrease in the negatively chargedheparan sulfate proteoglycan (HS-PG) of the GBM is one of the biochemical alterationsthat is likely to be responsible for the increased loss of albumin, but haemodynamic factorsmay be involved as well [26]. GFR is remarkably unaltered or may be elevated in

Chapter 16

subgroups of patients. Blood pressure may still be below the normal range, although slightincreases in night and day-time blood pressure have been reported with 24-hourambulatory blood measurements [27]. Exercise causes an exaggerated blood pressure rise,further indicating abnormalities in blood pressure regulation at this stage of renalinvolvement.

The clinical hallmark of stage 4 involvement is the presence overt proteinuria inexcess of 0.5 g/day. Arterial hypertension is almost always present and contributes impor-tantly to the loss of kidney function [28]. Unless arterial hypertension is treated, GFRdeclines at a rate of approximately 1 ml/min per month. Besides more outspoken thicken-ing of the GBM and mesangial expansion due to increased formation of extracellularmatrix, histologic examination now shows arteriolar hyalinosis and an increased numberof sclerosed glomeruli, appearing in a diffuse and nodular pattern, as first described byKimmelstiel and Wilson [29]. The progressive nature of nephropathy results in generalisedglomerulosclerosis ultimately leading to end-stage renal disease. Stage 5 diabetic renalinvolvement represents end stage renal failure requiring dialysis or renal transplantation.

Microalbuminuria: a predictor of DN and cardiovascular disease

Under normal circumstances small amounts of albumin pass through the glomerularfiltration membrane. Most of the filtered albumin undergoes tubular reabsorption, so thatthe final urinary albumin excretion rate is very low [30]. With the introduction of sensitiveassays in the early seventies it became possible to detect urinary albumin at these lowconcentrations [31,32]. It was soon established that elevations in urinary albuminexcretion (microalbuminuria) were typically present in the years preceding DN [10,11].Moreover, it has been established that the presence of microalbuminuria carried anincreased risk on cardiovascular complications [33,34]. These findings led to the conceptthat microalbuminuria represents the incipient stage of diabetic renal disease [23,35-39].The association with cardiovascular complications suggested that the presence ofmicroalbuminuria could also be an indicator of generalised vascular damage [37]. Atpresent, microalbuminuria is the first detectable clinical sign of an increased risk of DNand of cardiovascular disease in IDDM [36,38,39].

Despite the great advance to measure proteinuria at low levels with assays that havelow coefficients of variation [31], measurement of urinary albumin excretion rate iscomplicated by a large biological variability [38-40]. Day-to-day variation in albuminuriais as high as 30-50%, and there is considerable chance that a random urine sample willshow supra normal values. Thirty-eight % of patients with IDDM experience sporadicepisodes of microalbuminuria without developing persistent microalbuminuria [41].Recent diagnosis, worsened metabolic control, systemic illness, urinary tract infection andphysical exercise are potential confounding factors that temporarily raise albuminexcretion [10,38,39]. While single urine measurements suffice for screening purposes,there is general agreement that multiple urine collections are required to diagnosemicroalbuminuria reliably [36,38,39,42]. An albumin/creatinine ratio in a random or earlymorning urine sample of >3.5 mg/mmol is highly predictable for the presence ofmicroalbuminuria [38,40], although a lower cut-off level of 2.5 mg/mmol has beenproposed for firstly voided morning urine [39]. For definite evaluation of

Introduction 7

microalbuminuria, three timed overnight urine collections can be used to avoid the effectsof daytime physical activities, but 2-4 hour daytime or 24 hour urine collections also givereliable results [43]. Using overnight or timed day-time urine collections, albuminexcretion is expressed in µg/min and microalbuminuria is defined as levels between 20µg/min and 200 µg/min [36,38,39]. In 24 hour urine collections an albumin excretion ratebetween 30 to 300 mg/day indicates microalbuminuria. The lower level of 20 µg/min (or30 mg/day) is clearly above the upper normal limit of 10 to 12 µg/min found in healthysubjects [38,39,44]), but the cut-off level of 20 µg/min has been chosen because of itspredictive value to discriminate patients at risk to develop nephropathy [36,38,39]. Thusthere is a grey area between 10 and 20 µg/min (15-30 mg/day). The upper value of 200µg/min (300 mg/day) corresponds to a total protein excretion of 0.5 g per day.

If microalbuminuria is correctly diagnosed, 30% of these patients will progress toovert proteinuria in 10 years [12-16,19-22,36,38]. A normal albumin excretion rateexcludes progression to nephropathy with a 99.5% chance [41]. It has been reported thatpatients with an albumin excretion rate between 70 to 200 µg/day are particularly likely toprogress, whilst patients with a lower albumin excretion rate between 20 to 70 µg/day aremore likely to remain stable [12,42]. Thus, microalbuminuria is a very sensitive but notspecific measure to identify patients at risk of progression to nephropathy.

The increased risk of cardiovascular morbidity and mortality in microalbuminuricIDDM patients indicates that elevations in urinary albumin excretion have much broaderconsequences than representing a risk marker for the development of nephropathy[33,35,37-39]. This has been brought into a wider perspective by demonstrating thatmicroalbuminuria also predicts early mortality in non-insulin dependent diabetes mellitus(NIDDM) as well as in the general population [45-47]. Apart from other well establishedrisk factors, microalbuminuria appears to be a powerful indicator of cardiovascular disease[48]. The association between microalbuminuria and ischaemic heart disease is intriguing,and may be part of the metabolic syndrome of which alterations in blood pressureregulation are an important component [49]. The clear association between elevations inblood pressure and microalbuminuria in patients with IDDM as well as in patients withessential hypertension, and the fact that albumin excretion acutely falls after bloodpressure lowering, support a haemodynamic basis in the genesis of micro-albuminuria[9,42,50]. On the other hand, an atherogenic lipoprotein profile [51,52], increased plasmaconcentrations of clotting factors and decreased fibrinolysis [51,54], endothelial dysfunc-tion [53-55] and insulin resistance [56] are other manifestations of the metabolic syndromethat have been documented in microalbuminuric IDDM patients.

Pathogenesis of diabetic renal disease

The pathogenesis of structural and functional abnormalities in DN is likely to bemultifactorial (Table 2). In this section metabolic and haemodynamic abnormalities that

Chapter 18

Table 2. Mechanisms and factors implicated in the pathogenesis of diabetic nephropathy

1. Metabolic consequences of hyperglycaemia: Features: microcirculatory changes, glomerular basement membrane (GBM) thickening,

decreased heparan-sulfate-proteoglycan content of GBM, mesangial cell proliferation andextracellular matrix production

Possible pathways: - upregulation of diacylglycerol (DAG) and protein kinase-C (PKC)- nonenzymatic glycosylation: production advanced glycosylation products (AGE’s),- polyol pathway: sorbitol accumulation, altered cellular redox state

Implicated intrarenal growth factors and cytokines:- angiotensin II, endothelin, insulin-like growth factor I - platelet-derived growth factor-$, vascular endothelial growth factor,- transforming growth factor-$ and other cytokines as IL-1$, IL-6, IL-8, TNF, IFN-(

2. Altered renal haemodynamics: Features: increases in glomerular blood flow, intraglomerular pressure, filtration surface Mechanisms: diminished arteriolar resistance, imbalances in afferent/efferent tone,

mesangial dysfunction Contributing factors:

- hyperglycaemia and insulin - activated growth-hormone-insulin-like-growth-factor-I-axis- hyperglucagonaemia - inadequate suppressed renin-angiotensin II system, increased angiotensin II reactivity- altered sympathetic tone, increased norepinephrine reactivity - increased induction of nitric oxide versus disturbed endothelial function- abnormal prostaglandin metabolism, increased levels of atrial natriuretic factor, upregulation of kinins - augmented tubular sodium reabsorption, increases in total exchangeable sodium and extracellular volume

3. Elevated lipid levels: glomerular lipid accumulation in glomerulosclerosis resembling atherosclerosis

4. Genetic Predisposition: genes involved yet unknown, only polymorphism in the ACE-gene identified as a marker of progression

are considered to be implicated in the pathogenesis of DN are outlined. Furthermore, alter-ations in lipoprotein metabolism and genetic factors that may influence the developmentof DN are briefly described. Particular attention is paid to the renal effects of NE and theGH-IGF-I-system.

Metabolic factorsChronic hyperglycaemia is an inevitable consequence of IDDM and may induce

alterations in many cellular and molecular functions. The metabolic theories addressmechanisms by which elevated blood glucose levels may be causally involved in thedevelopment of microvascular complications.

First, the injurious effects of hyperglycaemia could be mediated via its effects on themicrocirculation. Elevated blood glucose levels induce arteriolar vasodilatation, increase

Introduction 9

blood flow and raise hydrostatic pressure, impair vasoregulation and thereby fail to protecttarget organs from increases in blood pressure. By such effects on the microcirculation,particularly on capillary pressure, hyperglycaemia may be responsible for leakage ofplasma proteins and deposition of proteins in the walls of arterioles and capillaries andthus induce damage to the kidneys [57,58].

A second pathway stresses the direct role for blood glucose to induce structuralglomerular abnormalities. Under experimental conditions, glucose has been demonstratedto cause GBM thickening and mesangial cell proliferation [59,60], to increase extracellularmatrix production and synthesis of type IV collagen [61,62], and to decrease GBM densityof the negatively charged HSPG [63]. These effects may in part be mediated by theexpression of a matrix-producing cytokin, transforming-growth-factor-$ (TGF-$) [64].More recently, it has been shown that upregulation of intracellular signal transduction viastimulation of diacylglycerol (DAG) and protein kinase C (PKC), as present in diabeticpatients, can raise TGF-$ and other growth factors, like vascular endothelial growthfactor, angiotensin II and endothelin [65]. Interestingly, inhibition of this system by anorally active PKC-$ isoform inhibitor has been shown to reverse the expression of TGF-$,to decrease the production of type IV and VI collagen and to restore haemodynamicabnormalities in diabetic rats [66]. These findings are in favour of an important role of thePKC-transduction system through which elevated levels of blood glucose may be involvedin the pathogenesis of DN.

Third, hyperglycaemia is also associated with an increased non-enzymaticglycosylation of long-lived proteins that may undergo Amadori rearrangement and therebylead to the irreversible formation of advanced glycosylation endproducts (AGE’s). AGE’shave been demonstrated to induce mesangial expansion and increase type IV collagensynthesis [67-69]. These effects may be mediated via specific AGE receptors [69], and areprevented by neutralising antibodies or by aminoguanidine in experimental diabetes[69,70].

Finally, hyperglycaemia causes an increased substrate delivery into the polyol path-way that results in the accumulation of sorbitol and changes the cytosolic redox state. Inthe polyol pathway, glucose is reduced to sorbitol by the enzyme aldose-reductase andsorbitol is oxidised to fructose by the enzyme sorbitol dehydrogenase (Figure 1). Underconditions of hyperglycaemia increased amounts of sorbitol are produced, as for instancedocumented in diabetic kidneys. The accumulation of this compound has been proposedto cause damage of renal tissue [71]. Another consequence of increased substrate deliveryinto the polyol pathway is the accumulation of NADP and NADH, leading to an increased+

cytosolic cell ratio of free NADH/NAD [72]. Such changes in redox state are also present+

in hypoxic tissues. Since vasodilation and increased blood flow are characteristic earlyvascular responses of tissue hypoxia and are also seen during hyperglycaemia, the so-called pseudohypoxia theory argues that an altered cytosolic redox state may be involvedin the haemodynamic alterations and subsequent microvascular complications of IDDM[72].

D-GLUCOSE

NADPH

6PG R5P

G6PDH 6PGDH

LDH

G6P LACTATE PYRUVATE

NAD NADHNADP + +

SORBITOLAR SDH

D-FRUCTOSE

Chapter 110

Figure 1. Reduction of glucose to sorbitol and oxidation of sorbitol to fructose in the sorbitol pathway.Reduction of glucose to sorbitol by aldose reductase (AR) is coupled to oxidation of NADPH toNADP . NADP is reduced to NADPH by the hexose monophosphate pathway. Oxidation of sorbitol+ +

to fructose by sorbitol dehydrogenase (SDH) is coupled to reduction of NAD to NADH. The cytosolic+

ratio of free NADH/NAD is in equilibrium with lactate and pyruvate. G6P: glucose-6-phosphate;+

6PG: 6-phosphogluconaat, and R5P: ribulose-5-phosphate.

Renal haemodynamic factors Elevations in GFR have long been recognised in patients with IDDM and can persist

for many years after the onset of diabetes [73-75]. The early stages of experimentaldiabetes are also characterised by a state of glomerular hyperfiltration. Its possiblepathogenetic role became apparent when glomerular hyperfiltration, consequently toexperimental renal ablation, was found to induce glomerulosclerosis [76]. In a similarway, a chronic increase in single nephron glomerular filtration (SNGFR) was associatedwith the development of glomerulosclerosis and renal function loss in diabetic rats [77].

According to the equation GFR= kf(ÎPc-ÎB), net hydraulic pressure (ªPc), netoncotic pressure (ªB), filtration surface area and hydraulic permeability (kf) determineglomerular ultrafiltration [30]. Since GFR changes linearly with glomerular blood flowunder conditions of pressure equilibrium in the rat [30] and is highly correlated witheffective renal plasma flow (ERPF) in man [78], renal blood flow is considered a determi-nant of GFR. Theoretically, a change in any of these factors could be involved in diabetichyperfiltration. Using the micropuncture technique, increases in glomerular blood flow,intraglomerular pressure and ultrafiltration coefficient have been documented inhyperfiltering diabetic rats [79-81]. Of these factors, the rise in intraglomerular capillarypressure was shown to play a key role as pharmacological amelioration of intraglomerularhypertension with ACE-inhibitors could be prevent glomerulosclerosis in these animals[82-83]. The rise intraglomerular capillary pressure in diabetes has been attributed toimbalances in afferent and efferent glomerular arteriolar tone, and to an increase insystemic arterial pressure [80,81]. A diminished glomerular afferent tone and an increasedglomerular efferent constriction are vascular abnormalities that have been implicated inthe glomerular hypertension associated with diabetes [80,81].

The original assumption that intraglomerular hypertension directly causes glomeru-lar damage appears to be an oversimplification since later studies have shown thathaemodynamic and non-haemodynamic factors are involved in the process of

Introduction 11

glomerulosclerosis [84-87]. Nevertheless, an important role is still attributed to glomerularhypertension, either as an initiator or as a conditional factor in a cascade of cellular eventsthat leads to glomerular damage. Several mechanisms have been proposed. First, chronicpressure overloading of the capillary endothelial cell layer may lead to cell detachment,GBM denudation, collagen exposure and consequently to platelet aggregation, fibrinaccumulation and intracapillary microthrombosis. Second, capillary dilation may disruptthe attachment of podocytes to the GBM with the subsequent formation of subendothelialdeposits. Third, continuous stretching may induce mesangial cell proliferation andextracellular matrix production by stimulating cytokin expression [85-87]. As a result,either of these mechanisms may impair the glomerular filtration barrier and enhanceglomerular protein passage, which, in turn, might be toxic and accelerate glomerulardamage [88]. In a similar way, glomerular accumulation of atherogenic lipoproteins couldcontribute to the process of glomerulosclerosis [89].

The neurohumoral stimulus for the hyperfiltration phenomenon in diabetes isunknown, although many factors could play a contributory role [81]. Moderate hypergly-caemia increases GFR [90,91] and chronical lowering of blood glucose reduces GFR inhyperfiltering IDDM patients [92]. The effect of glucose on GFR may result from adecrease in afferent glomerular arteriolar tone, mediated via the tubulo-glomerularfeedback (TGF) loop [90,91]. Insulin at doses that raise its plasma level 4 to 8 fold acutelyelevates ERPF [93,94], and its vasodilating properties are probably mediated via nitricoxide that directly affects glomerular arteriolar tone and mesangial function [95]. Anotherrelevant action of insulin is an increased tubular sodium reabsorption, leading to increasesin total exchangeable sodium and extracellular volume, which have been implicated indiabetic glomerular hyperfiltration [96,97]. Alterations in contra-regulatory hormones, GHand glucagon, which are well known renal vasodilators [98-100], could also play a rolesince these hormones are often elevated during suboptimal metabolic control [101-104].It is possible that intrarenal accumulation of IGF-I, as part of the GH-IGF-I-system, mayinduce the early functional and morphological renal changes in diabetes [105]. Renalhaemodynamics in diabetes may also depend on alterations in neurohormonal systems thatregulate glomerular tone, such as the renin- angiotensin-aldosterone-system (RAAS) andthe sympathetic nervous system (SNS) [81]. The presence of diabetes obviously affects theSNS, although various changes have been reported [106]. Furthermore, systemic bloodpressure responses to pressor agents like angiotensin II and NE are increased in diabetesmellitus [107-110]. Among other factors, vasoactive substances, such as prostaglandins,nitric oxide, kinins and atrial natriuretic peptide, have also been implicated in diabeticglomerular hyperfiltration [81]. Currently, none of these factors has been found to fullyaccount for the increase in GFR in IDDM patients. For instance, exogenous infusions ofglucose, insulin, GH, IGF-I and glucagon all increase GFR, but not to levels as generallyencountered in diabetic hyperfiltration.

Although animal studies provided an experimental basis for an increased GFR andelevated intraglomerular pressure as pathogenetic factors involved in the development ofDN, it should be stressed that there are caveats in extrapolating these data to the humansituation. First, glomerular pressure cannot be measured in man and it is thus unknownwhether glomerular hyperfiltration in human IDDM is accompanied by an increase inintraglomerular pressure. The only indirect proof stems from fingernail micropuncture

Chapter 112

studies that demonstrated capillary hypertension in human IDDM, but no differences werefound between normo- and microalbuminuric patients [111]. Second, it is clinicallydifficult to measure glomerular hyperfiltration in IDDM patients. Many clinical studiesdefined glomerular hyperfiltration as a GFR above the upper normal limit of a controlpopulation. Such an arbitrary definition does not discriminate subtle intrarenalhaemodynamic abnormalities in apparently normofiltering patients and has, therefore, thedisadvantage to underestimate the hyperfiltration phenomenon. This could explain whysome [112], but not all [113,114] clinical observations found an elevated GFR to be impli-cated in DN. Finally, in man renal insufficiency is an uncommon consequence of long-standing glomerular hyperfiltration per se, since subjects with one kidney [115-117] andpatients with acromegaly [118,119] are not at high risk of renal failure. This stronglysuggests that glomerular hyperfiltration alone does not result in important glomerularinjury in humans.

Lipoproteins Higher serum levels of low-density lipoprotein (LDL) cholesterol, apolipoprotein B,

triglycerides and lipoprotein (a), and lower levels of high-density lipoprotein (HDL)cholesterol have been observed in IDDM patients with nephropathy and even in patientswith microalbuminuria [51,52,120-122]. Such atherogenic lipoprotein changes are likelyto explain in part the increased cardiovascular risk in these patients. There exists muchcontroversy whether lipoprotein abnormalities are also involved in the pathogenesis of DN[123]. An independent association between elevated LDL cholesterol levels and progres-sion of microalbuminuria [124] and overt nephropathy [125] has been observed. Thissuggests a pathophysiological role of hyperlipidaemia analogous to atherosclerosis.Indeed, the lipid depositions and mesangial cell proliferation of glomerulosclerotic lesionsshow a remarkable resemblance with the lipid filled monocytes and vascular smoothmuscle cell proliferation in atherosclerotic plaques. Furthermore, experimental studiesshowed hypercholesterolaemia to aggravate glomerulosclerosis, which was prevented bycholesterol lowering therapy [126,127]. However, clinical support for the benefit ofcholesterol lowering therapy on progression of nephropathy is lacking. Short-termsimvastatin treatment did not decrease albuminuria in IDDM with nephropathy [128].Lovastatin, another HMG CoA reductase inhibitor, attenuated the rate of renal functionloss in NIDDM patients with nephropathy, but the lack of intergroup comparison in thisstudy has been criticised [129].

Genetic factorsThe fact that only a proportion of IDDM patients eventually develop overt proteinur-

ia and that this complication has a peak incidence 10 to 15 years after the onset ofdiabetes, supports the notion that specific susceptibility factors are involved in thepathogenesis of DN [1-3,5]. The observation that DN clusters in affected families furtherstrengthens the involvement of genetic factors [130-132]. IDDM siblings belonging tofamilies with a first-degree relative suffering from nephropathy have a life long risk of70% to develop nephropathy, whereas this risk is only 20% when there is no familyhistory of DN [132], and it has been suggested that one or two major genes determine

Introduction 13

susceptibility to DN. Since a familial predisposition to essential hypertension is associatedwith an increased risk of DN [133-135], candidate genes have been sought among lociinvolved in the regulation of blood pressure.

Recent attention has been given to polymorphisms in genes encoding for RAAScomponents. In this respect, the ACE gene polymorphism seems to be of relevance. Thispolymorphism consists of an 287 basepair insertion (I) or deletion (D) of intron 16 of theACE gene. The DD genotype has been shown to be associated with an increasedcardiovascular risk in non-diabetic populations [136,137], in non-insulin-dependentdiabetic (NIDDM) patients [138,139] and in IDDM patients with nephropathy [140]. DDhomozygotes have elevated serum [141] and tissue ACE levels [142], causing an increasedvascular conversion of angiotensin I to angiotensin II [143] and an increased pressorresponse to angiotensin I [144]. It is, therefore, hypothesised that increased angiotensin IIformation is involved in the increased cardiovascular risk in conjunction with the DDgenotype. The issue whether the DD genotype is also associated with DN is controversial.Some cross-sectional studies showed the DD genotype to be more prevalent among IDDMand NIDDM patients with nephropathy [145,146], while in other reports no association ofthe ACE gene polymorphism with DN could be demonstrated [147-149]. Recently it wasshown that in IDDM patients treated with an ACE inhibitor, the rate of decline in GFR isgreater in patients with the DD genotype compared to patients with the ID or II genotype[150,151] Similar findings have been reported in non-diabetic subjects with nephropathy[152], and that study also demonstrated a less effective antiproteinuric effect of ACEinhibition treatment in subjects with the DD genotype. Thus, the ACE gene polymorphismis more likely to be a marker of progression of DN than a susceptibility factor for DN.

The number of candidate genes for DN is growing and evaluation of their putativeroles will require large numbers of subjects, including sib-pairs discordant for DN [132].

Renal effects of norepinephrineAfter its release from the terminal nerve endings of the SNS, NE acts in an autocrine

fashion on local "-adrenoceptors, while at the same time small amounts leak into thecirculation [153,154]. This spilled-over NE is not an inert circulating neurotransmitter, buta hormonally active substance [155,156]. In the kidney, "-adrenoceptors are located alongthe interlobular, afferent and efferent glomerular arterioles, mesangial cells and tubularsegments [157-159]. This distribution pattern suggests that NE may control glomerularblood flow, glomerular capillary pressure and renal sodium handling [153,154]. Studieson renal sympathetic nerves have shown that low frequency stimulation results in sodiumretention and renin release and high frequency stimulation in a fall in ERPF and somedecline in GFR [160,161]. Exogenous NE infusions markedly reduce ERPF without muchchange in GFR in animals [162-165]. The NE-induced renal haemodynamic changes arelikely mediated via afferent and efferent glomerular arterioles [160-165]. These vessels arethe major sites of flow resistance in the kidney and importantly determine renal bloodflow, whereas they are also involved in the control of intraglomerular pressure [30].Micropuncture studies in the rat have indeed documented that NE causes a fall in renalblood flow by afferent and efferent glomerular vasoconstriction, and that NE evidentlyincreases intraglomerular pressure [162]. Interestingly, the prevailing blood pressure wasfound to determine the glomerular vessel response. There was a predominant increase in

Chapter 114

efferent tone when blood pressure was kept unchanged, whereas both afferent and efferentglomerular resistance increased when blood pressure was allowed to increase [162]. Thelack of change in GFR during NE infusion was explained by an increase inintraglomerular pressure offsetting the fall in glomerular blood flow [162]. Although, theprecise intrarenal effects of NE are unknown in man, there are obvious similarities withanimal data. Indeed, intravenous infusions of NE lower ERPF but have little effect onGFR [166, 167]. Consequently the filtration fraction (FF) rises which may reflect a changein pressure profile along the arterioles. It is therefore plausible that NE also increasesintraglomerular pressure in man, as supported by the finding that NE augmentedproteinuria in nephrotic patients [168].

The diabetic state has variably been associated with an increased, unchanged, oreven a decreased SNS activity and/or vascular reactivity to NE [81,106]. Theseinconsistencies may be attributed to differences in the species investigated, in bloodglucose and insulin levels, or in the vascular bed under study [81,106,169-171]. There isonly one study in kidney tissue taken from severely hyperglycaemic rats that has addressedthe putative role of NE in DN. In this study, afferent glomerular arteriolar responsivenesswas attenuated in experimental diabetes [165]. Most evidence, however, points towards anincreased vascular responsiveness in diabetes [81,106]. For instance, systemic bloodpressure responsiveness to exogenous NE has repeatedly been found to be exaggerateddiabetic patients [108-110]. Furthermore, the responsiveness to NE-inducedvasoconstriction of dorsal hand veins, which contain "-adrenoceptors like glomerulararterioles, is increased in moderately hyperglycaemic microalbuminuric IDDM patients[172]. These findings raise the possibility that glomerular vessels in IDDM are alsohyperresponsive to NE. Such an exaggerated renal responsiveness could, therefore,contribute to the elevations in intraglomerular pressure and albumin excretion rate, andthus play a pathogenesis role in DN [81,172,173]. However, no human study has evaluatedrenal NE responses in IDDM, and has established whether NE has the ability to increasemicroproteinuria.

The growth hormone-insulin-like growth factor-I system and kidney function IGF-I is a small peptide hormone (MW 7.6 kDa), which production is under

pituitary GH control [174]. The pituitary GH product of 21.5 kDa (191 amino acids) issecreted in pulsatile fashion with approximately 13 surges per day and has a short half-lifeof 20 minutes that is prolonged after binding to GH-binding protein. GH is a strongsecretagogue of IGF-I. GH simulates IGF-I gene transcription and increases IGF-Isynthesis in many tissues. Most IGF-I present in the circulation originates from the liver[175]. IGF-I, in turn, inhibits pituitary GH release by a negative feedback mechanism. Thebiological activity of IGF-I depends on the plasma levels of several binding proteins thatinterfere with IGF-I receptor interaction, as well as on IGF-I receptor expression [174].The IGF-I shares 70% homology with proinsulin and binds with high affinity to the IGF-I-receptor and with lower affinity to the insulin receptor. The plasma levels of IGF-I rangefrom 10-125 nmol per liter, which is much higher than insulin with fasting levels in thepicomolar range. Excessive stimulation of the insulin receptor is, however, prevented bythe fact that more than 99% of IGF-I is bound to specific IGF-binding proteins (IGFBP),and only a small amount of IGF-I is present its free form [174]. About 85% of IGF-I is

Introduction 15

bound to IGFBP-3 and forms a 150 kDa ternary complex after association with the acid-labile subunit (ALS), that does not pass through the capillary barrier [176]. The bindingto IGFBP-3 has such a high affinity that competition between IGFBP’s and the IGF-Ireceptor occurs. The 150 kDa complex can thus be viewed as a circulating IGF-I reservoir[174]. Both cleavage by proteases and phosphorylation impair the formation of the ternarycomplex, and enhance binding of IGF-I with its receptor. Another 20% of serum IGF-I isfound in smaller (±45 kDa) complexes, containing IGFBP-1, IGFBP-2, IGFBP-3 orIGFBP-4, which can pass through capillary endothelial membranes and deliver IGF-I tospecific tissue-binding sites.

In man, renal haemodynamic parameters covary with endogenous GH and IGF-Ilevels. GFR and ERPF are elevated in acromegaly, decline after GH lowering treatmentwith octreotide, and are decreased in GH deficiency [177-182]. GH stimulates renalhaemodynamics after a lag period allowing IGF-I levels to increase [183,184], whereasrhIGF-I infusion induces an immediate rise in GFR and ERPF [185-187]. Thus, GH seemsto increase renal haemodynamics indirectly via stimulating IGF-I synthesis.

IGF-I receptors as well as mRNA encoding for the different IGFBP’s are expressedin various structures of the kidney including glomerular arterioles [188]. In contrast, GHreceptors are not present on human glomerular vessels. Contradictory results have beenpresented with respect to IGF-I production in human nephrons [174,188], and it isunknown whether the GH receptor is expressed in other glomerular structures [174].

Based on micropuncture studies in the rat, rhIGF-I has been shown to decreaseefferent glomerular arteriolar resistance with a trend towards a reduction of afferentarteriolar resistance [185]. Exogenous rhIGF-I does not increase glomerular capillarypressure. The rises in SNGFR and in whole kidney GFR are fully accounted for byincrements in glomerular blood flow and in the filtration coefficient [185]. The IGF-I-induced renal changes are likely mediated via nitric oxide (NO), because IGF-I has beenshown to increase NO synthesis in cultured vascular endothelial cells [189], and the NOsynthase inhibitor, N -nitro-l-arginine methyl ester, abolishes renal vasodilation by IGF-IG

[190]. Moreover, IGF-I could also be involved in mesangial cell relaxation [191].Several experimental studies indicate that an enhanced GH-IGF-I-axis could be

involved in diabetes-associated hyperfiltration and plays a pathogenetic role in thedevelopment of glomerulosclerosis. Following unilateral nephrectomy in the rat, IGF-I hasbeen found to accumulate in hyperfiltering nephrons and the increase in SNGFR wasinhibited by anti-IGF-I-antibody administration [192]. It is of interest, that IGF-I has beenfound to accumulate in kidneys of diabetic rats during the initial phases of renal enlarge-ment and renal hyperperfusion [193]. Concomitant increases in renal IGF-I receptorexpression and receptor binding activity have also been documented in kidney tissue fromstreptozotocin-induced diabetic rats [194]. This renal IGF-I accumulation has been shownto be GH dependent since it is diminished in hypophysectomised diabetic rats and is inpart restored after GH replacement [195]. Thus, these findings suggest that renal IGF-I isinvolved in renal hypertrophy and raise the possibility that GH is necessary for this effect.

A possible role of the GH-IGF-I-system in the development of glomerulosclerosisis supported by observations in hGH-transgenic mice and in rats bearing GH-producingtumours which develop albuminuria, mesangial cell proliferation and prematureglomerulosclerosis [196,197]. However, IGF-I alone may not be completely responsible

Chapter 116

for GH-IGF-I-induced glomerulosclerosis, since mice transgenic for IGF-I do not developglomerulosclerosis [198]. These negative findings may be due to lower IGF-I levels in thetransgenic animals expressing IGF-I as compared to those expressing high levels of GH,but it is also possible that concomitant elevations in GH are necessary for IGF-I to induceglomerulosclerosis. Pituitary GH deficiency or GH lowering treatment with octreotideindeed modified the renal alterations after streptozotocin-induced diabetes in rats, assupported by inhibition of glomerular hypertrophy and lower albumin excretion rates inthese animals [199,200]. Of interest, high glucose levels increased IGF-I, IGF-I mRNA,and IGF-I receptor expression in cultured mesangial cells, which, in conjunction withraised TGF-$1 levels, enhanced extracellular matrix production [201]. Taken together, itis possible that abnormalities in the GH-IGF-I-system can contribute to the developmentof glomerulosclerosis in diabetes mellitus.

During poor metabolic control both glomerular hyperfiltration and elevated circulat-ing GH levels have been documented [75,92,101-103]. However, no difference in diurnalGH profile between normo- and hyperfiltering IDDM patients was observed [202].Nevertheless, an exaggerated GH-responses to GH-releasing hormone has been shown inhyperfiltering IDDM patients [203], indicating that glomerular hyperfiltration is indeedrelated to abnormalities in GH-release. Despite high circulating GH-levels, serum IGF-Ilevels are often low, again in relation to poor metabolic control [204]. This has beenascribed to GH-resistance at the hepatic level [205,206]. Lower circulating IGF-I levels,in turn, may contribute to GH-hypersecretion by insufficient inhibition at the pituitary.This may cause an adverse sequence of events: GH may worsen metabolic control andpoor control may elevate GH-levels. Intensive insulin treatment, and more importantly,restoration of hepatic insulinisation, have been shown to increase IGF-I levels [207,208]and reverse GH-hypersecretion [102]. This indicates that in IDDM patients relativeinsulinopenia may cause GH hypersecretion and lower the IGF-I level.

How can elevated GH and lower IGF-I levels be implicated in diabetic glomerularhyperfiltration? The association of increased renal haemodynamics with abnormalities inthe GH-IGF-I system, as shown in diabetic rats, results from intrarenal IGF-I accumul-ation, an increased IGF-I receptor expression and alterations in IGFBP’s [105,193-195,199,200]. Such alterations covary with insulin levels since they are outspokenly presentduring insulinopenia and partly prevented by insulin treatment [193,210]. Thus, despiteimpaired (hepatic) IGF-I synthesis, poor metabolic control could in fact enhance renalIGF-I accumulation. A maximally stimulated intrarenal IGF-I accumulation may,therefore, explain why GH administration does not increase GFR and ERPF in poorlycontrolled IDDM patients, whereas it augments GFR and ERPF in well-controlled IDDMpatients [99]. The mechanisms underlying renal IGF-I accumulation are not preciselyknown, but are conceivably due to an increased trapping of IGF-I from the circulation[195]. Local production of IGFBP’s, increased IGF-I receptor expression [193-195] andincreased IGFBP-3 protease activity that impairs the formation of the ternary complex[211] could all be involved in intrarenal IGF-I accumulation in IDDM.

Sodium and volume homeostasis in IDDM and the role of 11$-hydroxysteroiddehydrogenase

Introduction 17

An increase in exchangeable sodium accompanied by extracellular volume expan-sion is a well documented feature of diabetic patients and might contribute to elevationsin GFR, as well as to rises in blood pressure in association with microalbuminuria [96,110,212-217]. The mechanisms responsible for this abnormal sodium retention are incom-pletely understood. An enhanced renal tubular sodium reabsorption, possibly mediated byinsulin, may be involved [97]. In diabetic patients sodium excretion is attenuated follow-ing head out water immersion [218] and saline infusion [219]. Elevated plasma levels ofsodium retaining hormones, like angiotensin II, aldosterone en NE, are not encountered inIDDM and are unlikely to explain the tendency towards sodium retention[212,213,216,220,221].

Mineralocorticosteroids play a central role in extracellular sodium and fluidhomeostasis. Interestingly, the mineralocorticoid receptor has equal affinity for cortisoland aldosterone in vitro, but, in contrast, the renal tubules exclusively bind aldosterone invivo [222-225]. Recently, it has become clear that the mineralocorticoid receptor isprotected from being activated by cortisol by the intracellular enzyme 11$-hydroxysteroiddehydrogenase (11$-HSD) [224]. Two isoforms have been identified. 11$-HSD is1present in the liver. This enzyme is NADH/NAD dependent and catalyses the+

interconversion between cortisol and cortisone. 11$-HSD is expressed in the kidney and2unidirectionally catalyses the oxidation of cortisol to its inactive compound, cortisone.This isoenzyme is NADP dependent [225]. Thus, there is a link between cortisol metabo-+

lism and the regulation of volume and sodium homeostasis. For instance, geneticmutations in 11$-HSD that impair its activity cause hypokalaemic hypertension despite2undetectable aldosterone levels [226,227]. This so-called apparent mineralocorticoidexcess syndrome can also be acquired as a consequence of glycerrhethinic acid ingestion[228].

It is unknown whether 11$-HSD activity is altered in IDDM. Changes in thecortisol-cortisone shuttle towards cortisol could contribute to abnormal sodium retentionin IDDM. Alternatively, a shift towards cortisone could attenuate sodium retention.

Treatment of diabetic nephropathy

Arterial hypertension Arterial hypertension importantly contributes to the progression of proteinuria and

the loss of renal function in IDDM patients with nephropathy, and many studies havedemonstrated that blood pressure lowering reduces proteinuria and slows the rate ofdecline in GFR [28,229-233]. Even slight elevations in blood pressure have been shownto increase microalbuminuria in IDDM patients [9,39,42,112,234]. These findings haveresulted in the recommendation to start antihypertensive treatment in microalbuminuricIDDM patients when blood pressure values exceed 140/90 mmHg [235,236] or even130/85 mmHg [238]. It is, however, controversial whether one specific class ofantihypertensive agent is more effective than another. $-blockers [229,230,232], periph-eral vasodilators [229,230,302], diuretics [229,230,232] and ACE-inhibitors [231,233]have all been shown to effectively retard the decline in GFR in patients with overtnephropathy. Trials that compared the effectiveness of $-blockers and ACE inhibitorshave shown inconsistent results [238,239]. ACE-inhibition was found to be more effective

Chapter 118

than $-blockade, but in this study blood pressure reduction was better with the ACE-inhibitor [238]. Another study could not demonstrate a benefit of ACE-inhibition over $-blockade, but in that study baseline proteinuria was lower in the $-blocker group [239].Similar studies in non-diabetic renal disease either showed an increased [240] or nodifference in effectiveness of ACE-inhibitors as compared to $-blockers [152]. Interest-ingly, the latter study showed that the I/D polymorphism of the ACE gene was adeterminant of the renal protective outcome.

There are theoretical advantages of ACE inhibitors that are attributable to additionalrenoprotection as these agents have been documented to lower intraglomerular pressure inanimal studies [82,83]. Also, such a role for ACE inhibitors has been favoured in IDDMpatients with DN by demonstrating that captopril added to other antihypertensivemedications reduced the rate of renal function loss that could not be attributed to its bloodpressure lowering effects alone [241]. Furthermore, several meta-analyses comparingACE-inhibition treatment with other antihypertensive therapies pointed out that ACE-inhibitors induced larger reductions in proteinuria beyond their blood pressure loweringeffect [242,243].

So, should drugs interfering with the RAAS such as ACE-inhibitors and angiotensinII-antagonists be preferred, or will any rigorous blood pressure reduction, irrespective ofthe choice of drug, be sufficient to treat DN? It is important to realise that a clinical trialcomparing the effects on renal failure and mortality of these drugs has never beenperformed, but it is highly questionable whether such a study will ever be undertaken. Thelesson so far learned is that blood pressure should be rigidly lowered, since patients withthe largest blood pressure reduction will benefit the most [9]. ACE inhibitors haveunequivocally demonstrated to be effective and have gained an important place in thetreatment of DN [231,233,238,241]. As these drugs have also been shown to arrests ordelay progression of microalbuminuria [19-22], these patients should also be treated.Addition of diuretics may be useful to oppose the abnormal sodium retention in IDDM[213,244]. However, the potential advantage of $-blockers should not be overlookedbecause of proven secondary prevention of cardiovascular disease in non-diabetic subjects.Theoretically, calcium-antagonist may also have some advantages by modifying vascularreactivity to various endogenous pressor agents [245].

Although hypertension has been recognised as important prognostic indicator of renalfunction loss and blood pressure lowering therapy has proven renoprotective effectiveness,many patients still progress to end stage renal disease [229-233,241,246, 247]. A numberof post-hoc analyses on clinical trials in patients with renal disease of various origin, showedthat apart from blood pressure lowering, the severity of proteinuria was correlated with aworse prognosis with respect to long-term renal function [246,247]. Similar findings havebeen reported in patients with DN, in whom a relative large initial fall in albumin excretionafter $-blockers [248] and ACE-inhibitors [249] predicted a slower rate of decline in GFR.These studies suggest that clinical proteinuria is not only a marker of renal disease, butcould also be a pathogenetic factor in the process of glomerulosclerosis [250]. If it is truethat insufficient reduction in proteinuria and substantial residual proteinuria indicate apoorer prognosis, a more aggressive antiproteinuric therapy may possibly increase long-term renal function outcome in patients with renal disease. The advocated stringent bloodpressure targets [235-237] should then be individualised on the initial decline in

Introduction 19

proteinuria [251,252]. It is, however, unknown whether this is also true for proteinexcretion rates in the microalbuminuric range.

Metabolic controlAlthough not being the primary objective, the DCCT unequivocally showed that

intensive insulin treatment reduces the occurrence of microalbuminuria in IDDM patients[17,18]. In short, 1441 IDDM patients were allocated to either intensive or conventionalinsulin treatment. The goal of intensive therapy was to achieve near normal blood glucoselevels. The mean HbA1c level reached was 7.2%, which was approximately 2% lower thanthat in the conventionally treated IDDM patients. Mean follow-up was 6.5 years (range 3to 9 years). Most patients (n=1365) had a normal albumin excretion rate at baseline. Inthese patients, the estimated 9 year cumulative incidence in persistent microalbuminuria was10% and 26% in the intensive and conventional insulin treatment groups, respectively,indicating a reduction in risk of development of microalbuminuria of 60%. In 6% and 15%of these patients, respectively, microalbuminuria progressed to levels above 70 µg/min. Inall patients, including those with microalbuminuria at baseline, the calculated 9-yearcumulative incidence in clinical proteinuria was 3% and 7% with intensive and conventionaltreatment, respectively. This indicates a risk reduction in the development of overtnephropathy of 51% [17]. In a separate analysis of the 73 patients with microalbuminuriaat baseline, of which 38 were assigned to intensive therapy and 35 to conventional therapy,no difference in progression to clinical proteinuria was seen, which occurred in 8 patientsof each group [17].

The DCCT demonstrates that intensified insulin treatment reduces the risk ofdevelopment of microalbuminuria, and supports the role of hyperglycaemia in thepathogenesis of incipient nephropathy [58-72]. Accordingly, a meta-analysis of 16 studiesshowed that lowering blood glucose levels retarded progression urinary albumin excretionin normo- and microalbuminuric IDDM patients [253]. It should be noted that the lack ofeffect of intensive insulin therapy on the progression of microalbuminuria to overtnephropathy in the DCCT [17], was also found in a British collaborative study [254].Moreover, it is controversial whether improved metabolic control slows the rate of declinein GFR in patients with established DN [255-257]. Some reports have suggested thatglycaemic control loses its significance as a risk factor in established nephropathy[255,256], but these result have been challenged by the positive correlation betweenHbA1c levels and the rate of decline in GFR in patients with clinical nephropathy [257].It is conceivable that patients with overt DN will also benefit from strict metabolic control,although this will not reverse established renal disease [258,259]. Obviously, substantiallonger observation periods are required to determine whether sustained blood glucoselowering as achieved in the DCCT will prevent or delay renal failure over 10 to 20 years.

What degree of glycaemic control should be aimed for to minimise the risk ofcomplications? The major burden in managing IDDM patients to maintain blood glucosein the normal range is the risk of hypoglycaemia is [260,261]. This is illustrated by the 3fold increase in severe hypoglycaemic episodes as well as more nocturnal hypoglycaemiaand hypoglycaemia unawareness in the intensive insulin treated group of the DCCT[16,17]. The investigators of the DCCT recommended HbA1c levels of 7 to 8% since such

Chapter 120

a treatment goal would achieve a maximum benefitBto-risk ratio. Another reportpostulated, in contrast, that no clinical benefit results from a decrease of the HbA1c levelbelow 8.1%, since only higher HbA1c levels were associated an exponential increase inthe risk of developing microalbuminuria [262]. These authors proposed the presence of aglycaemic threshold for the risk of nephropathy. However, a post-hoc analysis of theDCCT did not support the existence of a glycaemic threshold, but in fact showed acurvilinear relationship between the HbA1c level and the logarithm of the risk ofmicroalbuminuria development [263]. Nonetheless, inherent to such a logarithmic relation,the reduction in risk is greatest when the highest HbA1c values are reduced. Weighing thisrisk of microalbuminuria against that of hypoglycaemia, and aware of the view thatelevated blood glucose levels alone do not cause nephropathy, but that haemodynamic andgenetic factors are also involved, it has been proposed that achievement of an HbA1c levelbelow 8.1% is a reasonable primary prevention strategy [264]. It was argued thatcontinuous monitoring will identify those patients in whom, despite improved glycaemiccontrol, progression of microalbuminuria takes place. These patients might then benefitfrom early aggressive blood pressure lowering therapy.

Dietary interventionProtein intake modulates renal function [265]. In humans, GFR and ERPF acutely

increase after an amino acid infusion and an oral protein load [265,266]. Experimentalstudies have shown that a high dietary protein intake contributes to a rise in the intra-glomerular pressure and GFR, and have documented that protein restriction ameliorates -glomerular hypertension and hyperfiltration, and retards the progression of renal functionimpairment [77,267]. These studies have led to the suggestion that protein restrictioncould be of clinical benefit in IDDM patients with nephropathy.

Only a few clinical studies addressed the effects of protein restriction in IDDMpatients with (incipient) nephropathy [234,268-271]. These studies demonstrated that alow protein diet decreased albuminuria, but variably affected the rate of decline in GFRwhich was either found to be unchanged [234,268,271] or indeed retarded by proteinrestriction [269, 270]. A meta-analysis of these studies showed that the relative risk ofprogression of albuminuria and decline in GFR was evidently lower in IDDM patientsusing a low protein diet compared with patients consuming a usual-protein diet [272]. Theanalysis also revealed that these favourable effects were not confounded by differences inblood pressure or metabolic control [272]. Accordingly, it is likely that IDDM patientsbenefit from a restriction in their daily protein intake. For this reason, the AmericanDiabetes Association has proposed to reduce dietary protein intake to 0.8 g/kg per day. Itshould be noted that such a protein restriction is difficult to maintain. In fact, a proteinintake of 0.8 g/kg per day was on average achieved in Dutch IDDM patients followingintensive dietary counselling [234]. Thus, the effectiveness of a low protein diet may belimited by suboptimal compliance. Clearly, there is currently no place to recommend -protein rich diets to IDDM patients.

References

1. Andersen AR, Sandahl Christiansen J, Andersen JK, Kreiner S, Deckert T. Diabetic nephropathy

Introduction 21

in Type 1 (insulin-dependent) diabetes: an epidemiological study. Diabetologia 1983; 25: 496-501

2. Krolewski AS, Warram JH, Christlieb AR, Busick EJ, Kahn CR. The changing natural history ofnephroapthy in Type I diabetes. Am J Med 1985; 78: 785-704

3. Kofoed-Enevoldsen A, Borch-Johnsen K, Kreiner S, Nerup J, Deckert T. Declining incidence ofpersistent proteinuria in Type I (insulin-dependent) diabetic patients in Denmark. Diabetes 1987;36: 205-209

4. Bojestig M, Arnqvist HJ, Hermansson G, Karlsberg BE, Ludvigsson J. Declining incidence ofnephropathy in insulin-dependent diabetes mellitus. N Engl J Med 1994; 330: 15-18

5. Rossing P, Rossing K, Jacobsen P, Parving HH. Unchanged incidence of diabetic nephropathyin IDDM patients. Diabetes 1995; 44: 739-743

6. Borch-Johnsen K, Andersen PK, Deckert T. The effect of proteinuria on relative mortality inType 1 (insulin-dependent) diabetes mellitus. Diabetologia 1985; 28: 590-596

7. Parving HH, Hommel E. Prognosis in diabetic nephropathy. Br Med J 1989; 299: 230-2338. Mathiesen ER, Borch-Johnsen K, Jensen DV, Deckert T. Improved survival in patients with

diabetic nephropathy. Diabetologia 1989; 32: 884-8869. Mogensen CE. Diabetic renal disease: The quest for normotension and beyond. Diab Med 1995;

12: 756-76910. Viberti GC, Jarrett RJ, Mahmud U, Hill RD, Argyropoulos A,Keen H. Microalbuminuria as a

predictor of clinical nephropathy in insulin-dependent diabetes mellitus. Lancet 1982; i: 1430-1432

11. Mogensen CE, Christensen CK. Predicting diabetic nephropathy in insulin-dependent diabetes.N Engl J Med 1984; 311: 89-93

12. Feldt-Rasmussen B, Mathiesen ER, Jensen T, Lauritzen T, Deckert T. Effects of improvedmetabolic control on loss of kidney function in type 1 (insulin-dependent) diabetic patients: anupdate of the Steno studies. Diabetologia 1991; 34:164-170

13. Forsblom CM, Groop P-H, Ekstrand A, groop LC. Predictive value of microalbuminuria inpatients with insulin-dependent diabetes of long duration. BMJ 1992; 305: 1051-1053

14. Almdal T, Nörgaard K, Feldt-Rasmussen B, Deckert T. The predictive value of microalbuminur-ia in IDDM. Diab Care 1994; 17: 120-125

15. Mathiesen ER, Rønn B, Storm B, Foght H, Deckert T. The natural course of microalbuminuriain insulin-dependent diabetes: a 10-year prospective study. Diab Med 1994; 12: 482-487

16. Rossing P, Hougaard P, Borch-Johnsen K, Parving HH. Progression from microalbuminuria todiabetic nephropathy. Diabetologia 1997; 40 [Suppl 1]: A528

17. The effect of intensive treatment of diabetes on the development and progression of long-termcomplications in insulin-dependent diabetes mellitus. The Diabetes Control and ComplicationTrial Research Group. N Engl J Med 1993; 29: 977-986

18. The diabetes control and complications (DCCT) research group. Effect of intensive insulintreatment on the development and progression of diabetic nephropathy in the Diabetes Controland Complications Trial. Kidney Int 1995; 47: 1703-1720

19. Marre M, Chattelier G, Leblanc H, Guyene TT, Menard J, Passa P. Prevention of diabeticnephropathy with enalapril in normotensive diabetics with microalbuminuria. Br Med J 1988;297: 1092-1095

20. Mathiesen ER, Hommel E, Giese J, Parving HH. Efficacy of captopril in postponing nephropathyin normotensive insulin dependent diabetic patients with microalbuminuria. Br Med J 1991; 303:81-87

21. Viberti CG, Mogensen CE, Groop LC, Pauls JF; for the European Microalbuminuria CaptoprilStudy Group. Effect of captopril on progression to clinical proteinuria in patients with insulin-dependent diabetes mellitus and microalbuminuria. JAMA 1994; 271: 275-279

Chapter 122

22. Mathiesen ER, Hommel E, Smith U, Parving HH. Efficacy of captopril in normotensive diabeticpatients with microalbuminuria:8-years follow-up. Diabetologia 1995;38[Suppl 1]: A46

23. Mogensen CE, Christensen CK, Vittinghus E. The stages in diabetic renal disease with emphasison the stage of incipient diabetic nephropathy. Diabetes 1983; 32 [Suppl 2]: 64-78

24. Deckert T, Feldt-Rasmussen B, Djurup R, Deckert M. Glomerular size and charge selectivity ininsulin-dependent diabetes mellitus. Kidney Int 1988; 33:100-106

25. Østerby R. Glomerular structural changes in Type 1 (insulin-dependent) diabetes mellitus:causes, consequences, and prevention. Diabetologia 1992; 35: 803-812

26. Deckert T, Kofoed-Enevoldsen A, Nørgaard K, Borch-Johnsen K, Feldt-Rasmussen B, Jensen T.Microalbuminuria: Implication for micro- and macrovascular disease. Diab Care 1992; 9: 1181-1191

27. Hansen KW, Pedersen M, Marshall SM, Christiansen JS, Mogensen CE. Circadian variation inblood pressure in patients with diabetic nephropathy. Diabetologia 1992; 35: 1074-1079

28. Mogensen CE. Progression of nephropathy in long-term diabetics with proteinuria and effect ofinitial anti-hypertensive treatment. Scab J Clin Lab Invest 1976; 36: 383-388

29. Kimmelstiel P, Wilson C. Intercapillary lesions in the glomeruli of the kidney. Am J Pathol 1936;12: 83-105

30. Maddox DA, Brenner BM. Glomerular ultrafiltration. In: Brenner BM, Rector FC (eds) Thekidney WB. Saunders Co., Philadephia, 1996, pp 286-333

31. Keen H, Chlouverakis C. An immunoassay method for urinary albumin at low concentrations.Lancet 1963; ii: 913-916

32. Mogensen CE. Urinary albumin excretion in early and long-term juvenile diabetes. Scand J ClinLab Invest 1971; 28: 101-109

33. Messent JWC, Elliott TG, Hill RD, Jarrett J, Keen H, Viberti G. Prognostic significance ofmicroalbuminuria in insulin-dependent diabetes mellitus: a twenty-three year follow-up study.Kidney Int 1992; 41: 836-839

34. Deckert T, Yokoyama, Mathiesen E, Rønn B, Jensen T, Feldt-Rasmussen B, Borch-Johnsen K,Skov-Jensen J. Cohort study of predictive value of urinary albumin excretion for atheroscleroticvascular disease in patients with insulin-dependent diabetes mellitus. BMJ 1996; 312: 871-874

35. Viberti GC, Keen H. The patterns of proteinuria in diabetes mellitus. Relevance to pathogenesisand prevention of diabetic nephropathy. Diabetes 1984; 33:686-692

36. Mogensen CE. Prediction of clinical diabetic nephropathy in IDDM patients. Alternatives tomicroalbuminuria? Diabetes 1990; 39: 761-767

37. Deckert T, Feldt-Rasmussen B, Borch-Johnsen K, Jensen T, Kofoed-Enevoldsen A. Albuminu-r-ia reflects widespread vascular damage. The Steno hypothesis. Diabetologia 1989; 32: 219-226

38. Feldt-Rasmussen B, Borch-Johnsen K, Deckert T, Jensen T, Skov Jensen J. Microalbuminuria:An important diagnostic tool. J Diab Comp 1994; 8: 137-145

39. Mogensen CE, Keane WF, Bennett PH, Jerums G, Parving HH, Passa P, Steffes MW, Striker GE,Viberti CG. Prevention of diabetic renal disease with special reference to microalbuminuria.Lancet 1995; 346; 1080-1084

40. Dullaart, RPF, Roelse H, Sluiter WJ, Doorenbos H. Variability of albumin excretion in diabetes.Neth J Med 1989; 34: 287-296

41. Mathiesen ER. Microalbuminuria in insulin-dependent diabetes: strategies to prevent diabeticnephropathy. Nephron 1993; 65: 511-521

42. Mathiesen ER, Røn B, Jensen T, Storm B, Deckert T. Relationship between blood pressure andurinary albumin excretion in development of microalbuminuria. Diabetes 1990;39: 245-249

43. Marshall SM. Screening for microalbuminuria: which measurement? Diab Med 1991;8:706-71144. Dullaart RPF, Beusekamp BJ, Meijer S, Hoogenberg K, Van Doormaal JJ, Sluiter WJ. Long-term

effects of linoleic-acid-enriched diet on albuminuria and lipid levels in Type 1 (insulin-

Introduction 23

dependent) diabetic patients with elevated urinary albumin excretion. Diabetologia 1992; 35:165-172

45. Mogensen CE. Microalbuminuria predicts clinical proteinuria in and early mortality in maturity-onset diabetes. N Engl J Med 1984; 310: 356-360

46. Yudkin JS, Forrest RD, Jackson CA. Microalbuminuria as predictor of vascular disease in non-diabetic subjects. Lancet 1988; ii: 530-533

47. Winocour PH, Harland JOE, Millar JP, Laker MF, Alberti KGMM. Microalbuminuria andassociated cardiovascular risk factors in the community. Potential marker for increasedcardiovascular risk factors in non-diabetic subjects? Arteriosclerosis 1990; 10: 727-731

48. Yudkin JS. Coronary heart disease in diabetes mellitus: three new risk factors and a unifyinghypothesis. J Int Med 1995; 238: 21-30

49. Reaven G. Role of insulin resistance in human disease. Diabetes 1988; 37: 1595-160750. Bianchi S, Bigazzi R, Baldari G, Camapese VM. Microalbuminuria in patients with essential

hypertension: effect of several antihypertensive drugs. Am J Med 1992; 93: 525-52851. Jensen T, Stender S, Deckert T. Abnormalities in plasma concentrations of lipoproteins and

fibrinogen in type 1 (insulin-dependent) diabetic patients with increased urinary albuminexcretion. Diabetologia 1988; 31: 142-145

52. Dullaart RPF. Plasma lipoprotein abnormalities in type 1 (insulin-dependent) diabetes mellitus.Neth J Med1995; 46: 44-54

53. Jensen T, Bjerre-Knudsen J, Feldt-Rasmussen B, Deckert T. Features of endothelial dysfunctionin early diabetic nephropathy. Lancet 1989; i: 461-463

54. Gruden G, Cavallo-Perin P, Bazzab M, Stella S, Vuolo A, Pagano G. PAI-I and factor VIIactivity are higher in insulin-dependent diabetes mellitus patients with microalbuminuria.Diabetes 1994; 43: 426-429

55. Stehouwer CDA, Fischer HRA, Van Kuijk AWR, Polak BCP, Donker AJM. Endothelialdysfunction precedes development of microalbuminuria in IDDM. Diabetes 1995; 44: 561-564

56. Yip J, Mattock MB, Morocutti A, Sethi M, Trevisan R, Viberti CG. Insulin resistance in insulin-dependent diabetic patients with microalbuminuria. Lancet 1993; 342: 883-887

57. Tooke JE. Microvascular haemodynamics in diabetes mellitus. Clin Sci 1986; 70: 119-12558. Parving HH, Viberti GC, Keen H, Christiansen JS, Lassen NA. Haemodynamic factors in the

genesis of diabetic microangiopathy. Metabolism 1993; 32: 943-95059. Kitamura M, Mitarai T, Maruyama N, Nagasawa R, Yoshida H, Sakai O. Mesangial cell

behaviour in the three-dimensional extracellular matrix. Kidney Int 1991; 40: 653-66160. Ellis EN, Steffes MW, Goetz FC, Sutherland DER, Mauer SM. Glomerular filtration surface in

type 1 diabetes mellitus. Kidney Int 1986; 29: 889-89461. Haneda M, Kikkawa R, Horide N, Togawa M, Koya D, Kajiwara N, Ooshima A, Shigeta Y.

Glucose enhances type IV collagen production in cultured glomerular cells. Diabetologia 1991;34: 198-200

62. Silbiger S, Schlondorff D, Crowley ST, Rosenberg L, Choi H, Hatcher V, Gordon P. The effectof glucose on proteoglycans produced by cultured mesangial cells. Diabetes 1993; 42: 1815-1822

63. Van Det NF, Van den Born J, Tamsma JT, Verhagen NAM, Berden JHM, Bruijn JA, Daha MR,Van der Woude FJ. Effect of high glucose on the production of heparan sulfate proteoglycan bymesangial and epithelial cells. Kidney Int 1996; 49: 1079-1089

64. Yamamoto T, Nakamura T, Noble NA, Ruoslahti E, Border W. Expression of transforminggrowth factor $ is elevated in human and experimental diabetic nephropathy. Proc Natl Acad SciUSA 1993; 90: 1814-1818

65. King GL, Brownlee M. The cellular and molecular mechanisms of diabetic complications.Endocrinol Metab Clin North Am 1996; 25: 255-270

Chapter 124

66. Ishii H, Jirousek MR, Koya D, Takagi C, Xia P, Clermont A, Bursell SE, Kern TS, Ballas LM,Heath WF, Stramm LE, Feener EP, King GL. Amelioration of vascular dysfunctions in diabeticrats by an oral PKC beta inhibitor. Science 1996; 272: 728-731.

67. Crowley ST, Brownlee M, Edelstein D, Satriano JA, Mori T, Singhal PC, Schlondorff DO.Effects of nonenzymatic glycosylation of mesangial matrix on proliferation of mesangial cells.Diabetes 1991; 40: 540-547

68. Cohen MP, Ziyadeh FN. Amadori glucose adducts modulate mesangial cell growth and collagengene expression. Kidney Int 1995; 45: 475-484

69. Cohen MP, Ziyadeh FN. Role of Amadori-modified nonenzymatically glycated serum proteinsin the pathogenesis of diabetic nephropathy. J Am Soc Nephrol 1996; 7: 183-190

70. Soulis T, Cooper ME, Vranes D, Bucala R, Jerums G. Effects of aminoguanidine in preventingexperimental diabetic nephropathy are related to the duration of treatment. Kidney Int 1996; 50:627-634

71. Beyer-Mears A, Ku L, Cohen MP. Glomerular polyol accumulation in diabetes and its preventionby oral sorbinil. Diabetes 1984; 33: 604-607

72. Williamson JR, Chang K, Frangos M, Hasan KS, Ido Y, Kawamura T, Nyengaard JR, Van denEnden M, Kilo C, Tilton RG. Hyperglycemic pseudohypoxia and diabetic complications.Diabetes 1993; 42: 810-813

73. Ditzel J, Junker K. Abnormal glomerular filtration rate, renal plasma flow, and renal proteinexcretion in recent an short-term diabetics. Br Med J 1972; 2: 13-19

74. Christiansen JS, Gammelgaard J, Frandsen M, Parving HH. Increased kidney size, glomerularfiltration rate and renal plasma flow in short-term insulin-dependent diabetics. Diabetologia1981; 20: 451-456

75. Mogensen CE, Christensen CK, Pedersen MM, Alberti KGMM, Boye N, Christensen T,Christiansen JS, Flyvbjerg A, Ingerslev J, Schmitz A, Ørskov H. Renal and glycemic determi-nants of glomerular hyperfiltration in normoalbuminuric diabetics.J Diab Comp 1990;4:159-165

76. Hostetter TH, Olsen JL, Rennke HG, Venkatachalam MA, Brenner BM. Hyperfiltration ofremnant nephrons: a potentially adverse response to renal ablation. Am J Physiol 1981; 241: F85-F93

77. Zats R, Meyer TW, Rennke HG, Brenner BM. Predominance of hemodynamic rather thanmetabolic factors in the pathogenesis of diabetic glomerulopathy. Proc Natl Acad Sci USA 1981;82: 5963-5967

78. Brenner BM, Humes HD. Mechanics of glomerular ultrafiltration. N Engl J Med 1977; 297: 148-154

79. Hostetter TH, Rennke HG, Brenner BM. The case for intrarenal hypertension in the initiation andprogression of diabetic and other glomerulopathies. Am J Med 1982; 72: 375-380

80. Viberti GC, Wiseman MJ. The kidney in diabetes mellitus: significance of the early abnormali-ties. Clin Endocrinol metab 1986; 15: 753-782

81. Bank N. Mechanisms of diabetic hyperfiltration. Kidney Int 1991; 40: 792-80782. Zats R, Dunn B, Meyer TW, Andersen CM, Rennke HG, Brenner BM. Prevention of diabetic

glomerulopathy by pharmacological amelioration of glomerular capillary hypertension. J ClinInvest 1986; 77: 1925-1930

83. Andersen S, Rennke HG, Garcia DL, Brenner BM. Short and long term effects of antihypertensi-ve therapy in the diabetic rat. Kidney Int 1989; 36: 526-536

84. Steffes MW, Østerby R, Chavers B, Bauer SM. Mesangial expansion as a central mechanism forloss of kidney function in diabetic patients. Diabetes 1989; 38: 1077-1081

85. Riser BL, Cortes P, Zhao X, bernstein J, Dumler F, Narins RG. Intraglomerular pressure andmesangial stretching stimulate extracellular matrix formation in the rat. J Clin Invest 1992; 90:1932-1943

Introduction 25

86. Shankland SJ, Ly H, Thai K, Scholey JW. Increased glomerular capillary pressure altersglomerular cytokin expression, Circ Res 1994; 96: 844-853

87. Isaka Y, Fujiwara Y, Ueda N, Kaneda Y, Kamada T, Imai E. Glomerulosclerosis induced by invivo transfection of transforming growth factor-beta or platelet-derived growth factor gene intothe rat kidney. J Clin Invest 1993; 92: 2597-2601

88. Benigni A, Zoja C, Remuzzi G. The renal toxicity of sustained glomerular protein traffic. LabInvest 1995; 73: 461-468

89. Keane WF. Lipids and the kidney. Kidney Int 1994; 46: 910-92090. Wiseman MJ, Mangili R, Alberetto M, Keen H, Viberti GC. Glomerular response mechanisms

to glycemic changes in insulin-dependent diabetics. Kidney Int 1987; 31: 1012-1018.91. Dullaart RPF, Meijer S, Sluiter WJ, Doorenbos H. Renal haemodynamic changes in response to

moderate hyperglycaemia in Type 1(insulin-dependent) diabetes mellitus. Eur J Clin Invest 1990;20: 208-213.

92. Wiseman MJ, Saunders AJ, Keen H, Viberti GC. Effect of blood glucose control on increasedglomerular filtration rate and kidney size in insulin-dependent diabetes. N Engl J Med 1985; 312:617-21.

93. Stenvinkel P, Bolinder J, Alvestrand A. Effects of insulin on renal haemodynamics and theproximal and distal sodium handling in healthy volunteers. Diabetologia 1992; 35: 1042-1048

94. Pelikánová T, Smrcková I, Krízová J, Stríbrná J, Lánská V. Effect of insulin and lipid emulsionon renal haemodynamics and renal sodium handling in IDDM patients. Diabetologia 1996; 39:1074-1082

95. Scherrer U, Randin D, Vollenweider P, Vollenwieder L, Nicod P. Nitric oxide release accountsfor insulin’s vascular effects in humans. J Clin Invest 1994; 94: 2511-2515

96. Brøchner-Mortensen J, Ditzel J. Glomerular filtration rate and extracellular volume in insulin-dependent patients with diabetes mellitus. Kidney Int 1982; 21: 696-698

97. Gans ROB, Van der Toorn L, Bilo JJP, Heine RJ, Donker AJM. Renal and cardiovascular effectsof exogenous insulin in healthy volunteers. Clin Sci 1991; 80: 219-225

98. Sandahl Christiansen J,Gammelgaard J, Orskow H, Andersen AR, Telmer S,Parving HH. Kidneyfunction and size in normal subjects before and during growth hormone administration for oneweek. Eur J Clin Invest 1981; 11: 487-491

99. Christiansen JS, Gammelgaard J, Frandsen M, Ørskov H, Parving HH. Kidney function and sizein Type 1 (insulin-dependent) diabetic patients before and during growth hormone administrationfor one week. Diabetologia 1982; 22: 333-337.

100. Parving HH, Christiansen JS, Noer I, Tronier B, Mogensen CE. The effect of glucagon infusionkidney function in short-term insulin-dependent juvenile diabetics. Diabetologia 1980; 19: 350-354.

101. Hansen AP. Abnormal serum growth hormone response to exercise in juvenile diabetics. J ClinInvest 1970; 49: 1467-1472.

102. Tamborlane WV, Sherwin RS, Koivisto V, Hendler R, Genel M, Felig P. Normalization of thegrowth hormone and catecholamine response to exercise in juvenile-onset diabetic subjectstreated with a portable insulin infusion pump. Diabetes 1979; 28: 785-788

103. Asplin CM, Faria ACS, Carlsen EC, Vaccaro VA, Barn RE, Iranmanesh A, Lee MM, VeldhuisJD, Evans WS. Alterations in the pulsatile mode of growth hormome release in men and womenwith insulin-dependent diabetes mellitus. J Clin Endocrinol Metab 1989; 69: 239-245

104. Gerich JE, Tsalikian E, Lorenzi M, Schneider V, Bohanon NV, Gustavson G, Karam JH.Normalization of fasting hyperglucagonemia and excessive glucagon responses to intravenousarginine in human diabetes mellitus by prolonged infusion of insulin. J Clin Endocrinol Metab1975; 41: 1178-1180

105. Flyvbjerg A. Growth factors and diabetic complications. Diab Med 1990; 7: 387-399

Chapter 126

106. Tomlinson KCSM., Gardiner SM, Hebden RA, Bennett T. Functional consequences ofstreptozotocin-induced diabetes mellitus, with particular reference to the cardiovascular system.Pharmacol Rev 1992; 44: 103-150

107. Drury PL, Smith GW. Increased pressor responsiveness to angiotensin II in uncomplicated type1 (insulin-dependent) diabetes. J Hypertension 1983; 1: 313-316

108. Christlieb AR, Janka H-U, Kraus B, Gleason RE, Icasas-Cabral EA, Aiello LM, Cabral BV,Solano A. Vascular reactivity to angiotensin II and to norepinephrine in diabetic subjects.Diabetes 1976; 25: 268-274

109. Beretta-Piccoli C, Weidmann P. Exaggerated pressor responsiveness to norepinephrine innonazotemic diabetes mellitus. Am J Med 1981; 71: 829-835

110. Weidmann P, Beretta-Piccoli C, Trost B. Pressor factors and responsiveness in hypertensionaccompanying diabetes mellitus. Hypertension 1985; 7 [Suppl II]: 33-42

111. Sandeman DD, Shore AC, Tooke JE. Relation of skin capillary pressure in patients with insulin-dependent diabetes mellitus and metabolic control. N Engl J Med 1992; 327: 760-764

112. Rudberg S, Persson B, Dahlquist G. Increased glomerular filtration rate as a predictor of diabeticnephropathy. An 8-year prospective study. Kidney Int 1992; 41: 822-828

113. Lervang HH, Jensen S, Brøchner-Mortensen J, Ditzel J. Early glomerular hyperfiltration and thedevelopment of late nephropathy in Type 1 (insulin-dependent) diabetes mellitus. Diabetologia1988; 31: 723-729

114. Jones SJ, Wiseman MJ, Viberti GC. Glomerular hyperfiltration as a risk marker for diabeticnephropathy: Five year report of a prospective study. Diabetologia 1991; 34: 59-60

115. Hakim RM, Goldszer RC, Brenner BM. Hypertension and proteinuria: long term sequelae ofuninephrectomy in in humans. Kidney Int 1984; 25: 930-936

116. Robitaille P, Mongeau J, Lorti LP. Long-term follow-up of patients who underwent unilateralnephrectomy in childhood. Lancet 1985; ii: 1297-1299

117. Schmitz A, Christensen CK, Christensen T, Sølling K. No microalbuminuria or other adverseeffects of long standing hyperfiltration in humans with one kidney Am J Kidney Dis 1989; 13:131-136

118. Molitch ME. Clinical manifestations of acromegaly. Endocrinol Metab Clin North Am 1992; 21:597-614.

119. Newbold KM, Howie AJ, Girling AJ, Kizaki T, Bryan RL, Carey MP. A simple method forassessment of glomerular size and its use in the study of kidneys in acromegaly and compensa-tory renal enlargement. J Pathol 1989;158:139-146

120. Lahdenperä S, Groop PH, Tilly-Kiesi M, Kuusi T, Elliott TG, Viberti GC, Taskinen MR. LDLsubclasses in IDDM patients: relation to diabetic nephropathy. Diabetologia 1994; 37: 681-688

121. Dullaart RPF, Groener JEM, Dikkeschei LD, Erkelens DW, Doorenbos H. Increasedcholesterylester transfer activity in complicated Type 1 (insulin-dependent) diabetes mellitus BIts relationship with serum lipids. Diabetologia 1989; 32: 14-19

122. Jenkins AJ, Steele JS, Janus ED, best JD. Increased plasma lipoprotein (a) in IDDM patients withmicroalbuminuria. Diabetes 1991; 40: 787-790

123. Gröne EF, Walli AK, Gröne HJ, Miller B, Seidel D. The role of lipids in nephrosclerosis andglomerulosclerosis. Atherosclerosis 1994; 107: 1-13

124. Coonrod BA, Ellis D, Becker DJ, Bunker CH, Kelsey SF, Lloyd CE, Drash AL, Kuller LH,Orchard TJ. Predictors of microalbuminuria in individuals with IDDM. Pittsburg epidemiologyof diabetes complication study. Diab Care 1993; 16: 1376-1383

125. Krolewski AS, Warram JH, Chrislieb AR. Hypercholesterolemia B A determinant of renalfunction loss and deaths in IDDM patients with nephropathy. Kidney Int 1994; 45: S125-S131

126. Keane WF, Kasiske BL, O’Donnell MP. Lipids and progressive glomerulosclerosis. A modelanalogous to atherosclerosis. Am J Nephrol 1988; 8: 261-271

Introduction 27

127. Kasiske BL, O’Donnell MP, Garvis WJ, Keane WF. Parmacologic treatment of hyperlipidaemiareduces glomerular injury in rat 5/6 nephrectomy model of chronic renal failure. Circulation Res1988; 62: 267-274

128. Hommel E, Andersen P, Gall MA, Nielsen F, Jensen B, Rossing P, Dyerberg J, Parving HH.Plasma lipoproteins and renal function during simvastatin treatment in diabetic nephropathy.Diabetologia 1992; 35: 447-451

129. Lam KSL, Cheng IKP, Janus ED, Pang RWC. Cholesterol-lowering therapy may retard theprogression of diabetic nephropathy. Diabetologia 1995; 38: 604-609

130. Seaquist ER, Goetz FC, Rich S, Barbosa J. Familial clustering of diabetic kidney disease.Evidence for genetic susceptibility to diabetic nephropathy. N Engl J Med 1989; 320:1161-1165

131. Borch-Johnsen K, Nørgaard K, Hommel E, Mathiesen ER, Jensen JS, Parving HH. Is diabeticnephropathy an inherited complication? Kidney Int 1992; 41: 719-722

132. Quin M, Angelico MC, warram JH, Krolewski AS. Familial factors determine the developmentof diabetic nephropathy in patients with IDDM. Diabetologia 1996; 39: 940-945

133. [C4] Viberti GC, Keen H, Wiseman MJ. Raised arterial pressure in parents of proteinuric insulindependent diabetics. Br Med J 1987; 295: 515-517

134. Krolewski AS, Canessa M, Warram JH, Laffel LMB, Christlieb AR, Knowler WC, Rand LI.Predisposition to hypertension and susceptibility to renal disease in insulin-dependent diabetesmellitus. N Engl J Med 1988; 318: 140-145

135. Earle K, Walker J, Hill C, Viberti GC. Familial clustering of cardiovascular disease in patientswith insulin-dependent diabetes and diabetic nephropathy. N Engl J Med 1992; 326: 673-677

136. Cambien F, Poirier O, Lecerf L, Evans A, Cambou JD, Arveiler D, Luc D, Bard JLM, Ricard S,Tiret L, Amouel P, Alhenc-Gelas F, Soubrier F. Deletion polymorphism in the gene forangiotensin-converting enzyme is a potent risk factor for myocardial infarction. Nature 1992;159: 641-644

137. Tiret L, Kee F, Poirier O, Nicaud V, Lecerf L, Evans A, Cambou JP, Arveiler D, Luc G,Amouyel P, Cambien F. Deletion polymorphism in angiotensin-converting enzyme geneassociated with parenteral history of myocardial infarction. Lancet 1993; 341: 991-992

138. Ruiz J, Blanche H, Cohen N, Velho G, Cambien F, Cohen D, Passa P, Froguel P. Inser-tion/deletion polymorphism of the angiotensin-converting enzyme gene is strongly associatedwith coronary heart disease in non-insulin-dependent diabetes mellitus. Proc Natl Acad Sci USA1994; 91: 3662-3665

139. Keavney BD, Dudley CRK, Stratton IM, Holman RR, Matthews DR, Ratcliffe PJ, Turner RC.UK prospective diabetes study (UKPDS) 14: association of angiotensin-converting enzyme inser-tion/deletion ploymorphism with myocardial infarction in NIDDM. Diabetologia 1995; 38: 948-952

140. Tarnow L, Cambien F, Rossing P, Nielsen FS, Hansen BV, Lecerf L, Poirier O, Danilov S,Boelskifte S, Borch-Johnsen K, Parving HH. Insertion/deletion polymorphism in the angiotensin-I-converting enzyme gene is associated with coronary heart disease in IDDM patients withdiabetic nephropathy. Diabetologia 1995; 38: 798-803

141. Rigat B, Hubert C, Alhenc-Gelas F, Cambien F, Corvol P, Soubrier F. An insertion/deletionpolymorphism in the angiotensin I-converting enzyme gene accounting for half of the varianceof serum enzyme levels. J Clin Invest 1990;86:1343-1346

142. Danser JAH, Schalenkamp MADH, Bax WA. Angiotensin converting enzyme in the humanheart. Effect of the deletion/insertion polymorphism.Circulation 1995; 92: 1387-1388

143. Buikema H, Pinto YM, Rooks G, Granjean JG, Schunkert H, Van Gilst WH. The deletionpolymorphism of the angiotensin converting enzyme gene is related to phenotypic differences inhuman arteries. Eur Heart J 1996; 17: 787-794

Chapter 128

144. Ueda S, Elliott HL, Morton JJ, Connell JMC. Enhanced pressor response to angiotensin I innormotensive men with the deletion genotype (DD) for angiotensin-converting enzyme.Hypertension 1995;25:1266-1269

145. Marre M, Bernadet P, Gallois Y, Savagner F, Guyene TT, Hallab M, Cambien F, Passa P,Alhenc-Gelas F. Relationships between angiotensin I, converting enzyme gene polymorphism,plasma levels, and diabetic retinal and renal complications. Diabetes 1994; 43: 384-388

146. Mitzuiri S, Hemmi H, Inoue A, Yoshikawa H, Tanegashima M, Fushimi T, Ishigami M,Amagasaki Y, Ohara T, Shimatake H, Hasegawa A. Angiotensin-converting enzymepolymorphism and development of diabetic nephropathy in non-insulin-dependent diabetesmellitus. Nephron 1995; 70: 455-459

147. Doria A, Warram JH, Krolewski AS. Genetic predisposition to diabetic nephropathy. Evidencefor a role of the angiotensin I-converting enzyme gene. Diabetes 1994; 43: 690-695

148. Tarnow L, Cambien F, Rossing P, Nielsen FS, Hansen BV, Lecerf L, Poirier O, Danilov S,Parving HH. Lack of relationship between an insertion/deletion polymorphism in the angiotensinI-converting enzyme gene and diabetic nephropathy and proliferative retinopathy in IDDMpatients. Diabetes 1995; 44: 489-49

149. Schmidt S, Schone N, Ritz E, and the Diabetic Nephropathy Study Group. Association of ACEgene polymorphism and diabetic nephropathy? Kidney Int 1995; 47: 1176-1181

150. Parving HH, Jacobsen P, Tarnow L, Rossing P, Lecerf L, Poirier O, Cambien F. Effect ofdeletion polymorphism of angiotensin converting enzyme gene on progression of diabeticnephropathy during inhibition of angiotensin converting enzyme: observational follow up study.BMJ 1996; 313: 591-594

151. Tarnow L, Rossing P, Nielsen FS, Cambien F, Parving HH. Progression of diabetic nephropathyduring long standing angiotensin converting enzyme inhibition B the impact of polymorphismsin the renin angiotensin system. Diabetologia 1997; 40 [Suppl 1]: A19

152. Van Essen GG, Rensma PL, De Zeeuw D, Sluiter WJ, Scheffer H, Apperloo AJ, De Jong PE.Association between angiotensin-converting-enzyme gene polymorphism and failure ofrenoprotective therapy. Lancet 1996; 347: 94-95

153. Di Bona GF. The function of renal nerves. Rev Physiol Biochem Pharmacol 1982; 94: 74-181154. Di Bona GF. Neural control of renal function: role of renal alpha adrenoceptors. J Cardiovasc

Pharmacol 1985; 7 [Suppl 8): S18-S23155. Silverberg AB, Shah SD, Haymond MW, Cryer PE. Norepinephrine: hormone an neuro-

transmitter in man. Am J Physiol 1978; 3: E252-E256156. Chang PC, Kriek E, Van der Krogt JA, Blauw GJ, Van Brummelen P. Haemodynamic effects of

physiological concentrations of circulating noradrenaline in man. Clin Sci 1988; 75: 469-475157. Barajas L, Wang P. Localization of tritiated norepinephrine in the renal arteriolar nerves. Anat

Rec 1979; 195: 525-534158. Stephenson JA, SummersRJ. Autoradiographic evidence for a heterogeneous distribution of "1-

adrenoceptors labelled by [ H] prazocin in rat, dog, and human kidney. J Auton Pharmac 1986;3

6: 109-116159. Edwards RM, Trizna W. Characterization of "-adrenoceptors on isolated rabbit renal arterioles.

Am J Physiol 1988; 254: F178-F183160. Hermansson K,Larson M, Kallskog O, Wolgast M. Influence of renal nerve activity on arteriolar

resistance, ultrafiltration dynamics and fluid reabsorption. Pflugers Arch 1981; 389: 85-90161. Kon V, Ichikawa I. Effector loci for renal nerve control of cortical microcirculation. Am J

Physiol 1983; 245: F545-F553. 162. Meyers BD, Deen WM, Brenner BM. Effects of norepinephrine and angiotensin II on the

determinants of glomerular ultrafiltration and proximal tubule fluid reabsorption in the rat. CircRes 1975; 37: 101-110

Introduction 29

163. Edwards RM. Segmental effects of norepinephrine and angiotensin II on isolated renalmicrovessels. Am J Physiol 1983; 244: F526-F534

164. Ito S, Arima S, Lin Ren Y, Juncos LA, Carretero OA . Endothelium-derived relaxing factor/nitricoxide modulates angiotensin II action in the isolated microperfused rabbit afferent but notefferent arteriole. J Clin Invest 1993; 91: 2012-2019

165. Ohishi K, Okwueze MI, Vari RC, Carmines PK. Juxtamedullary microvascular dysfunctionduring the hyperfiltration stage of diabetes mellitus. Am J Physiol 1994; 267: F99-F105

166. Smythe McC, Nicke JF, Bradley SE . The effect of epinephrine (usp), l-epinephrine, and l-norepinephrine on glomerular filtration rate, renal plasma flow and the urinary excretion ofsodium, potassium and water in normal man. J Clin Invest 1952; 31: 499-506

167. Lang CC, Rahman AR, Balfour DJK, Struthers AD. Effect of norepinephrine on renal sodiumand water handling in euhydrated and overhydrated man. Clin Sci 1993; 85: 487-494

168. Lathem W. Renal circulatory dynamics and urinary protein excretion during infusions of l-norepinephrine and l-epinephrine in patients with renal disease. J Clin Invest 1965; 35: 1277-1285

169. Gans ROB, Bilo HJG, Van Maarschalkerweerd WWA, Heine RJ, Nauta JJP, Donker AJM.Exogenous insulin augments in healthy volunteers the cardiovascular reactivity to norepinephrinebut not angiotensin II. J Clin Invest 1991 88: 512-518

170. McNally PG, Watt PAC, Rimmer T, Burden AC, Hearnshaw JR, Thurston H. Impairedcontraction and endothelium-dependent relaxation in isolated resistance vessels from patientswith insulin-dependent diabetes mellitus Clin. Sci 1994; 87: 31-36

171. McNally PG, Lawrence IG, Watt PAC, Hillier C, Burden AC, Thurston H. The effect of insulinon vascular reactivity of isolated resistance arteries taken from healthy volunteers. Diabetologia1995; 38: 467-473

172. Bodmer CW, Patrick AW, How TV, Williams G. Exaggerated sensitivity to NE-inducedvasoconstriction in IDDM patients with microalbuminuria. Possible etiology and diagnosticimplications. Diabetes 1992; 41: 209-214

173. Poortmans JR. Postexercise proteinuria in humans. JAMA 1982; 253: 236-240174. Feld S, Hirschberg R. Growth hormone, the insulin-like growth factor system and the kidney.

Endocrine Reviews 1996; 17: 423-480175. Schwander JC, Hauri C, Zapf J, Froesh ER. Synthesis and secretion of insulin-like growth factor

and its binding protein by the perfused liver rat liver: dependence on growth hormone status.Endocrinology 1983; 113: 297-305

176. Baxter RC, Maretin JL, Beniac VA. High molecular weight insulin-like growth factor bindingprotein complex: purification and protein properties of the acid-labile subunit from human serum.J Biol Chem 1989; 264: 11843-11848

177. Ikkos D, Ljunggren H, Luft R. Glomerular filtration rate and renal plasma flow in acromegaly.Acta Endocrinol (Copenh) 1956; 21 :226-236

178. Corvilain J, Abramow M, Bergans A. Some effects of human growth hormone on renalhemodynamics and on tubular phosphate transport in man. J Clin Invest 1962; 41: 1230-1235

179. Eskildsen PC, Parving HH, Mogensen CE, Christiansen JS. Kidney function in acromegaly. ActaMed Scand 1979; Suppl 624: 79-82

180. Dullaart RPF, Meijer S, Marbach P, Sluiter WJ. Renal reserve capacity in growth hormonedeficient subjects. Eur J Clin Invest 1992; 22: 562-568

181. Dullaart RPF, Meijer S, Marbach P, Sluiter WJ. Effect of a somatostatin analogue, octreotide, onrenal haemodynamics and albuminuria in acromegalic patients. Eur J Clin Invest 1992; 23: 494-502

182. Pedersen MM, Christensen SE, Christiansen JS, Pedersen EB, Mogensen CE, Ørskov H. Acuteeffects of a somatostatin analogue on kidney function in Type 1 diabetic patients. Diab Med1990; 7: 304-309

Chapter 130

183. Hirschberg R, Rabb H, Bergamo R, Kopple JD. The delayed effect of growth hormone on renalfunction in humans. Kidney Int 1989; 35: 865-870

184. Hirschberg R, Kopple JD. Increase in renal plasma flow and glomerular filtration rate duringgrowth hormone treatment may be mediated by insulin-like growth factor I. Am J Nephrol 1988;8: 249-254

185. Hirschberg R, Kopple JD, Blantz RC, Tucker BJ. Effects of human recombinant insulin-likegrowth factor I on glomerular dynamics in the rat. J Clin Invest 1991; 87: 1200-1206

186. Hirschberg R, Brunori G, Kopple JD, Guler H-P. The effects of insulin-like growth factor I onrenal function in normal man. Kidney Int 1993; 43: 387-397

187. Giordano M, DeFronzo R. Acute effects of human recombinant insulin-like growth factor I onrenal function in humans. Nephron 1995; 71: 10-15

188. Chin E, Michels K, Bondy CA. Partition of insulin-like growth factor (IGF)-binding sitesbetween IGF-I and IGF-II receptors and IGF-binding proteins in the human kidney. J ClinEndocrinol Metab 1992; 78: 156-164

189. Tsukahara H, Gordienko DV, Tönshoff B, Gelato MC, Goligorski MS. Direct demonstration ofinsulin-like growth factor-I-induced nitric oxide production by endothelial cells. Kidney Int 1994;45: 598-604

190. Haylor J, Singh I, El Nahas AM. Nitric oxide synthesis inhibitor prevents vasodilation by insulingrowth factor I. Kidney Int 1991; 39: 333-335

191. Inishi Y, Okuda T, Arakawa T, Kurokawa K. Insulin attenuates intracellular calcium responsesand cell contraction caused by vasoactive agents. Kidney Int 1994; 45: 1318-1325

192. Haylor J, McKillop I, El Nahas M. IGF-I antibody inhibits hyperfiltration following unilateralnephrectomy. J Am Soc Nephrol 1995; 6: 679 (Abstract)

193. Flyvbjerg A, Thorlacius-Ussing O, Naeraa R, Ingerslev J, Ørskov H. Kidney tissue somatomedinC and initial renal growth in diabetic and uninephrectomized rats. Diabetologia 1988; 31: 310-314

194. Werner H, Shen-Orr Z, Stannard B, Burguera B, Roberts CT, LeRoith D. Experimental diabetesincreases insulin growth factor I and II receptor concentration and gene expression in kidney.Diabetes 1990; 39: 1490-1497

195. Ørskov H, Flyvbjerg A, Frystyk J, Grøbaek H, Foegh M, Harris AG. IGF-I and acute renalhypertrophy: inhibition by somatostatin analogues. Transplant Proc 1993; 2: 2061-2062

196. Quaife CJ, Mathews LS, Pinkert CA, Hammer RE, Brinster RL, Palmiter RD. Histopathologyassociated with elevated levels of growth hormone and insulin-like growth factor I in transgenen-ic mice. Endocrinology 1989; 124: 40-48

197. Kawaguchi H, Itoh K, Mori H, Hayashi Y, Makino S. Renal pathology in rats bearing tumour-secreting growth hormone. Pediatr Nephrol 1991; 5: 533-538

198. Doi T, Striker LJ, Gibson CC, Agodoa LYC, Brinster RL, Striker GE. Glomerular lesions in micetransgenic for growth hormone and insulin like growth factor-I. Am J Pathol 1990;137: 541-552

199. Flyvbjerg A, Frystyk J, Østerby R, Ørskov H. Kidney IGF-I and renal hypertrophy in GH-deficient diabetic dwarf rats. Am J Physiol 1992; 262: E956-962

200. Flyvbjerg A, Frystyk J, Thorlacius-Ussing O, Ørskov H. Somatostatin analogue administrationprevents increase in kidney somatomedin C and inital renal growth in diabetic anduninephrectomised rats. Diabetologia 1989; 32: 261-265

201. Pugliese G, Pricci F, Locuratolo N, Romeo G, Romano G, Giannini S, Gresci B, Galli G, RotellaCM, Di Mario U. Increased activity of the insulin-like growth factor system in mesangial cellscultured in high glucose conditions. Relation to glucose-enhanced extracellular matrixproduction. Diabetologia 1996; 39: 775-784

202. Wiseman MJ, Redmond S, House F, Keen H, Viberti GC. The glomerular hyperfiltration ofdiabetes is not associated with elevated plasma levels of glucagon and growth hormone.Diabetologia 1985; 28: 718-721

Introduction 31

203. Blankestijn PJ, Derkx FHM, Birkenhäger JC, Lamberts SWJ, Mulder P, Verschoor L,Schalenkamp MADH, Weber RFA. Glomerular hyperfiltration in insulin-dependent diabetesmellitus is correlated with enhanced growth hormone secretion. J Clin Endocrinol Metab1993;77: 498-502

204. Press M, Tamborlane WV, Sherwin RS. Importance of raised growth-hormone levels inmediating the metabolic derangements of diabetes. N Engl J Med 1984; 310: 810-815

205. Clayton KL, Holly JM, Carlsson LM. Loss of the normal relationships between growth hormone,growth hormone-binding protein and insulin-like growth factor I in adolescents with insulin-dependent diabetes mellitus. Clin Endocrinol 1994; 41: 517-524

206. Ishii DN, Guertin DM, Whalen LR. Reduced insulin-like growth factor-I mRNA content in liver,adrenal glands and spinal cord of diabetic rats. Diabetologia 1994; 37: 1073-1081

207. Tamborlane WV, Hintz RL, bergman M, Genel M, Felig P, Sherwin RS. Insulin-infusion-pumptreatment of diabetes. Influence of improved metabolic control on plasma somatomedin levels.N Engl J Med 1981; 305: 303-307

208. Hanaire-Broutin H, Sallarin-Caute B, Poncet ME, Tauber M, bastide R, Chalé JJ, Rosenfeld R,Tauber JP. Effect of intraperitoneal insulin delivery on growth hormone binding protein, insulin-like growth factor (IGF)-I, and IGF-binding protein-3 in IDDM. Diabetologia 1996; 39: 1498-1504

209. Scott CD, baxter RC. Production of insulin like growth factor I and its binding protein in rathepatocytes cultured from diabetic and insulin-treated diabetic rats. Endocrinology 1986; 119:2346-2353

210. Raz I, Haramati Z, Hirschberg B, Moshe R, Nephesh I, Weiss O Flyvbjerg A. The IGF systemin early diabetic nephropathy in normoinsulinemic hyperglycaemic rats. Diabetologia 1997; 40[Suppl 1]: A512

211. Bereket A, Lang CH, Fan J, Frost RA, Wilson TA. Insulin-like growth factor binding protein(IGFBP)-3 proteolysis in children with insulin-dependent diabetes mellitus: a possible role forinsulin in the regulation of IGFBP-3 protease activity. J Clin Endocrin Metab 1995; 80: 2282-2288

212. Brøchner-Mortensen J. Glomerular filtration rate and extracellular fluid volumes duringnormoglycaemia and moderate hyperglycaemia in diabetics insulin-dependent patients withdiabetes mellitus. Scand J Clin Lab Invest 1973; 32: 311-316

213. Weidman D, Beretta-Piccoli C, Keusch G, Glück Z, Mujagic M, grimm M, Meier A, ZieglerWH. Sodium-volume factor, cardiovacular reactivity and hypotensive mechanism od diurestictherapy in mild hypertension associated with diabetes mellitus. Am J Med 1977; 67: 779-784

214. O'Hare JA, Ferris JB, Brady D, Twomey B, O'Sullivan DJ. Essential hypertension andhypertension in diabetic patients without nephropathy. Hypertension 1983; 1 [Suppl 2]: 200-203

215. O'Hare JA, Ferris JB, Brady D, Twomey B, O'Sullivan DJ. Exchangeable sodium and renin inhypertensive diabetic patients with and without nephropathy. Hypertension 1985; 7 [Suppl 2]: 43-48

216. Feldt -Rasmussen B, Mathiesen ER, Deckert T, Giese J, Christensen NJ, Bent-Hansen, NielsenMD. Central role for sodium in the pathogenesis of blood pressure changes independent ofangiotensin, aldosterone and catecholamines in Type 1 (insulin-dependent) diabetes mellitus.Diabetologia 1987; 30: 610-617

217. Weidman D, Ferrari D. Central role for sodium in hypertension in diabetic subjects. DiabetesCare 1991; 14: 220-232

218. O'Hare JP, Roland JM, Walters G, Corrall RJM. Impaired sodium excretion in response tovolume expansion induced by water immersion in insulin-dependent diabetes mellitus. Clin Sci1986; 71: 403-409.

219. Roland JM, O'Hare JP, Walters G, Corrall RJM. Sodium retention in response to saline infusionin uncomplicated diabetes mellitus. Diab Res 1986; 3: 213-215.

Chapter 132

220. O'Hare JA, Ferris JB, Twomey BM, Gonggrijp H, O'Sullivan DJ. Changes in blood pressure,body fluids, circulating angiotensin II and aldosterone, with improved diabetic control. Clin Sci1982; 63: 415s-418s.

221. Ferris JB, Sullivan PA, Gonggrijp H, Cole M, O’Sullivan D. Plasma angiotensin II andaldosterone in unselected diabetic patients. Clin Endocrinol (Oxf) 1982; 17: 261-269

222. Arriza JL, Weinberger C, Cerelli G. Cloning of human mineralocorticoid receptor com-plementary DNA: structural and functional kinship with the glucocorticoid receptor. Science1987; 237: 268-275

223. Sheppard K, Funder JW. Mineralocorticoid specificity of renal type 1 receptors: in vivo bindingsstudies. Am J Physiol 1987; 252: E224-E229

224. Walker BR, Edwards CRW. 11$-Hydroxysteroid dehydrogenase and enzyme-mediated receptorprotection: life after liquorice? Clin Endocrinol (Oxf) 1991; 35: 281-289

225. Seckl JR. 11$-Hydroxysteroid dehydrogenase isoforms and their implications for blood pressureregulation. Eur J Clin Invest 1993; 22: 589-601

226. Wilson RC, Harbison MD, Krozowski ZS, Funder JW, Shackleton CHL, Hanauske-Abel HM,Wei JQ, Hertecant J, Moran A, Neiberger RE, Balfe JW, Fattah A, Daneman D, Licholai T, NewMI. Several homozygous mutations in the gene for 11$-hydroxysteroid dehydrogenase type 2 inpatients with apparent mineralocorticoid excess. J Clin Endocrinol Metab 1995; 80: 3145-3150

227. Stewart PM, Krozowski ZS, Gupta A, Milford DV, Sheppard MC Whorwood CB. Hypertensionin the syndrome of apparent mineralocorticoid excess due to mutation of the 11$-hydroxysteroiddehydrogenase type 2 gene. Lancet 1996; 347: 88-91

228. Stewart PM, Valentino R, Wallace AM, Burt D, Shackleton CHL, Edwards CRW.Mineralocorticoid activity of liquorice 11ß-hydroxysteroid dehydrogenase activity comes of age.Lancet 1987; ii: 821-824

229. Mogensen CE. Long-term anti-hypertensive treatment inhibiting the progression of diabeticnephropathy. Br Med J 1982; 285: 685-687

230. Parving HH, Andersen AR, Smidt UM, Svendsen PA. Early aggressive of antihypertensivetreatment reduced rate of decline in kidney function in diabetic nephropathy. Lancet 1983; 1:1175-1179

231. Bjørk S, Nyberg G, Mulec H, Granerus G, Herlitz H, Aurell M. Beneficial effects of angiotensinconverting enzyme inhibition on renal function in patients with diabetic nephropathy. Br Med J1986; 293: 471-474

232. Parving HH, Andersen AR, Smidt UM, Hommel E, Mathiesen ER, Svendsen PA. Effect ofantihypertensive treatment on kidney function in diabetic nephropathy. Br Med J 1987; 294:1443-1447

233. Parving HH, Hommel E, Smidt UM. Protection of kidney function and decrease in albuminuriaby captopril in insulin dependent diabetic patients with nephropathy. Br Med J 1988; 297: 1086-1091

234. Dullaart RPF, Beusekamp BJ, Meijer S, Van Doormaal JJ, Sluiter WJ. Long-term effects ofprotein-restricted diet on albuminuria and renal function in IDDM patient without clinicalnephropathy and hypertension. Diabetes Care 1993; 16: 483-492

235. Viberti GC, Mogensen CE, Passa P, Bilous R, Mangilli R. St Vincent declaration 1994:guidelines for the prevention of diabetic renal failure. In: Mogensen CE, ed. The kidney andhypertension in diabetes mellitus, 2nd ed. Boston: Dordrecht/London: Kluwer, 1994: 515-527

236. [H8] Anon. Prevention of diabetes mellitus: report of a WHO study group. WHO tech Rep Ser844. Geneva WHO, 1994, 55-59

237. Consensus development conference on the diagnosis and management of nephropathy in patientswith diabetes mellitus. Diab Care 1994; 17: 1357-1361

238. Bjørk S, Mulec H, Johnsen SA, Nordén G, Aurell M. Renal protective effects of enalapril indiabetic nephropathy. BMJ 1992; 304: 339-343

Introduction 33

239. Elving LD, Wetzels JFM, Van Lier HJJ, De Nobel E, Berden JHM. Captopril and atenolol areequally effective in retarding progression of diabetic nephropathy. Diabetologia 1994; 37: 604-9

240. Hannedouche T, Landais P, Goldfarb B, El Esper N, Fournier A, Godin M, Durand D, ChanardJ, Mignon F, Suc JM, Grünfeld JP. Randomised controlled trial of enalapril and B-blockers innon-diabetic chronic renal failure. BMJ 1994; 309: 833-837

241. Lewis EJ, Hunsicker LG, Bain RP, Rohde RD. The effect of angiotensin-converting-enzymeinhibition on diabetic nephropathy. N Engl J Med 1993; 329: 1456-1462

242. Kasiske BI, Kalil RSN, Ma JZ, Liao M, Keane WF. Effect of antihypertensive therapy on thekidney in patients with diabetes: A meta-regression analysis. Ann Int Med 1993; 1118: 129-138

243. Gansevoort RT, Sluiter WJ, Hemmelder MH, De Zeeuw D, De Jong PE. Antiproteinuric effectof blood-pressure-lowering agents: A meta-analysis of compartive trials. Nephrol Dial Transplant1995; 10: 1963-1974

244. Weidmann P, Beretta-Piccoli C, Keusch G, Z. Glück, Mujagic M, Grimm M, Meier A, ZieglerWH. Sodium-volume factor, cardiovascular reactivity and hypotensive mechanism of diuretictherapy in mild hypertension associated with diabetes mellitus. Am J Med 1979; 67: 779-784

245. Bakris GL, Barnhill BW, Sadler R. Treatment of arterial hypertension in diabetic humans:Importance of therapeutic selection. Kidney Int 1992; 41: 912-919

246. Peterson JC, Adler S, Burkart JM, Greene T, Hebert LA, Hunsicker LG, King AJ, Klahr S,Massry SG, Seifter JL for the MDRD Study Group: Blood pressure control, proteinuria and theprogression of renal disease. Ann Intern Med 1995; 123: 754-762

247. Locatelli F, Marcelli D, Comelli M, Alberti D, Graziani G, Buccianti G, Redaelli B, GiangrandeA, and the Nothern Italian Cooperative Study Group. Prospective, randomized, multicenter trialof effect of protein restriction on progression of chronic renal insufficiency. Nephrol DialTransplant 1996; 11: 461-467

248. Rossing P, Hommel E, Smidt UM, Parving HH. Impact of arterial pressure and albuminuria onthe progression of diabetic nephropathy in insulin-dependent diabetic patients. Diabetes 1993; 42:715-719

249. Rossing P, Hommel E, Smidt UM, Parving HH. Reduction in albuminuria predicts a beneficialeffect on diminishing the progression of human diabetic nephropathy during antihypertensivetreament. Diabetolgia 1994; 37: 511-516

250. Remuzzi G, Bertain T: Is glomerulosclerosis a consequence of altered glomerular permeabilityto macromolecules? [Editorial] Kidney Int 1990; 38: 384-394

251. Navis GJ, De Jong PE, De Zeeuw D: A comparison of progression in diabetic and non-diabeticrenal disease: similarity of progression promoters. In: The kidney in hypertension and diabetesmellitus, 3rd ed., edited by Mogensen CE, Boston/Dordrecht/London, Kluwer 1996,p 551-560

252. Mogensen CE, Hansen KW, Østerby R, Damsgaard EM. Blood pressure elevation versusabnormal albuminuria in the genesis and prediction of renal disease in diabetes. Diabetes Care1992; 15: 1192-1204

253. Wang PH, Lau J, Chalmers TC. Meta-analysis of effects of intensive blood-glucose control onlate complications of type I diabetes. Lancet 1993; 341: 1306-1309

254. Microalbuminuria Collaborative Study Group, United Kingdom. Intensive therapy andprogression to clinical albuminuria in patients with insulin dependent diabetes meliitus andmicroalbuminuria. BMJ 1995; 311: 973-977

255. Viberti CG, Bilous RW, Mackintosh D, Bending JJ, Keen H. Long term correction ofhyperglycaemia and progression of renal failure in insulin dependent diabetes. Br Med J 1983;286: 598-602

256. Parving HH, Smidt UM, Friisberg B, Bonnevie-Nielsen V, Andersen AR. A prospective study ofglomerular filtration rate and arterial blood pressure in insulin-dependent diabetics with diabeticnephropathy. Diabetologia 1981; 20: 457-461

Chapter 134

257. Nyberg G, Blohmé G, Nordén G. Impact of metabolic control in progression of clinical diabeticnephropathy. Diabetologia 1987; 30: 82-86

258. Reichard P, Nilsson BY, Rosenqvist U. The effect of long-term intensified insulin treatment onthe development of microvascular complications of diabetes mellitus. N Engl J Med 1993; 329:304-309

259. Manto A, Cotroneo P, Marra G, Magnani P, Tilli P, Greco AV, Ghirlanda G. Effect of intensivetreatment on diabetic nephropathy in patients with type I diabetes. Kidney Int 1995; 47:231-235

260. Santiago JV. Lessons from the Diabetes Control and Complication trial. Diabetes 1993; 42: 1549-1554

261. Dahl-Jørgensen K, Brinchmann-Hansen O, Bangstad HJ, Hanssen KF. Blood glucose control andmicrovascular complications B What do we do now? Diabetologia 1994; 37: 1172-1177

262. Krolewski AS, Laffel LMB, Krolewski M, Quinn M, Warram JH. Glycosylated hemoglobin andthe risk of microalbuminuria in patients with insulin-dependent diabetes mellitus. N Eng J Med1995; 332:1251-1255

263. The Diabetes Control and Complications Trial Research Group. The absence of a glycemicthreshold for the development of long-term complications: The perspective of the DiabetesControl and Complications Trial. Diabetes 1996; 45: 1289-1298

264. Viberti CG. A glycaemic threshold for diabetic complications? N Eng J Med 1995; 332:1293-1294

265. Hostetter TH. Human renal response to a meat meal. Am J Physiol 1986; 250: F613-F618266. Ter Wee, Geerlings W, Rosman JB, Sluiter WJ, Van der Geest S, Donker AJM. Testing renal

reserve filtration capacity with an amino acid solution. Nephron 1985; 41: 193-199267. Brenner BM, Meyer TW, Hostetter TH. Dietary protein intake and the progressive nature of

kidney disease: the role of hemodynamically mediated glomerular injury in the pathogenesis ofprogressive glomerulosclerosis in aging, renal ablation and intrinsic renal disease. N Engl J Med1982; 307: 652-659

268. Chiavarella A, Di Mizzio G,stefoni S, Borgnino LC, Vannini P. Reduced albuminuria afterdietary protein restriction in insulin-dependent diabetic patients with clinical nephropathy.Diabetes Care 1987; 22: 67-71

269. Barsotti G, Clardella F, Morelli E, Cupisti A, Mantovanelli A, Glovannetti S. Nutrional treatmentof renal failure in type 1 diabetic nephropathy. Clin Nephrol 1988; 29; 280-287

270. Walker JD, Dodds RA, Murrells TJ, Bending JJ, Mattock MB, Keen H. Restriction of dietaryprotein and progression of renal failure in diabetic nephropathy. Lancet 1989; ii: 1411-1415

271. Zeller K, Whittaker E, Sullivan L, Raskin P, Jacobson HR. Effect of restricting dietary protein onthe progression of renal failure in patients with insulin-dependent diabetes mellitus. N Engl JMed 1991; 324: 78-84

272. Pedrini MT, Levey AS, Lau J, Chalmers TC, Wang PH. The effect of dietary protein restrictionon the progression of diabetic and nondiabetic renal diasease: Ameta-analysis. Ann Int Med1996; 124: 627-632

Purpose of the thesis

This thesis aims to answer to following questions:

1. Does an increase in plasma NE levels contribute to the exercise-induced rise in albuminuria and are there differences between normo- and microalbuminuricIDDM patients and healthy subjects?

2. Are ambient plasma NE levels related to renal haemodynamic parameters in normo-and microalbuminuric IDDM patients and in healthy subjects?

Introduction 35

3. Do exogenous NE infusions, given at doses that induce predetermined rises in bloodpressure, cause a microproteinuric response in normo- and microalbuminuric IDDMpatients and healthy subjects? What are the determinants of such a putativemicroproteinuric response? Are differences in renal haemodynamic NE responsive-ness responsible for possible differences in NE-stimulated microproteinuria amongthese subjects?

4. Does ACE-inhibition treatment attenuate the NE-induced blood pressure rise andrenal vasoconstriction in microalbuminuric IDDM patients.

5. Does low dose dopamine oppose NE-induced renal vasoconstriction in healthysubjects?

6. Are abnormal GH-IGF-I levels, as encountered in (un)treated acromegalic patientsand GH deficient patients, associated with differences in urinary protein excretion ascompared to healthy subjects?

7. Is renal functional reserve as assessed by amino acid infusion inversely related tobaseline renal haemodynamic parameters among GH deficient patients, healthysubjects, patients with (un)treated acromegaly, and normo- and hyperfiltering IDDMpatients? Are basal renal haemodynamic parameters positively related to the plasmaIGF-I levels in these subjects?

7. Do IDDM patients with an elevated GFR have an exaggerated GH-response afterexercise and a higher plasma glucagon level and as compared with IDDM patientswith a normal GFR? Is exercise-stimulated GH secretion and the glucagon levelrelated to GFR, ERPF and kidney size in IDDM patients?

8. Is an increased urinary IgG excretion a genuine feature of normoalbuminuric IDDMpatients or are laboratory artefacts responsible for this thus far unexplained observation?

9. Are there differences in urinary cortisol and cortisone metabolites in normo- andmicroalbuminuric IDDM patients compared to healthy subjects? Does this indicate ashift in the so-called cortisol-cortisone shuttle towards cortisol that could be involvedin sodium retention and volume expansion in IDDM? Does ACE inhibition treatmentinfluence the cortisone-cortisol shuttle in microalbuminuric IDDM patients?

Department of Endocrinology, University Hospital Groningen, The Netherlands1

Scand J Clin Invest 1992;52: 803-811

CHAPTER 2

ABNORMAL PLASMA NOREPINEPHRINE RESPONSE ANDEXERCISE-INDUCED ALBUMINURIA IN IDDM PATIENTS

K. Hoogenberg and R.P.F. Dullaart1 1

Submaximal exercise provokes an abnormal elevation in albuminuria in patients with IDDM.Plasma catecholamines might be involved in this phenomenon by a renal vasoconstrictive effect.Twelve healthy subjects (Controls: albuminuria<10µg/min), 13 normoalbuminuric IDDM patients(DNormo:albuminuria<10µg/min) and 13 microalbuminuric IDDM patients (DMicro: albuminuria10-200µg/min) performed a fixed bicycle workload (600 kpm for 20 min + urine collection 40 minpostexercise). None of the patients suffered from autonomic neuropathy or hypertension. Fractionalalbumin clearance (FalbCl) rose in DNormo (p=0.02) and DMicro (p=0.01) but not in the Controls(p=0.40). Basal plasma epinephrine and norepinephrine (NE) were not different in the three groups.The increments in NE were more pronounced in DNormo and DMicro than in Controls(Controls<Dnormo, p<0.05; Controls< DMicro, p<0.01). The changes in FalbCl weresignificantly correlated with the changes in NE (all subjects r=0.65, p<0.001). The increments inepinephrine were not different in the IDDM groups compared to the controls, and were not relatedto the changes in FalbCl. Multiple regression analysis showed that changes in plasma NE (p<0.002)and in mean arterial pressure(MAP, p<0.005) independently contributed to the changes in FalbCl(multiple R=0.73).It is concluded that the exercise induced plasma NE response is increased innormo- and microalbuminuric IDDM patients. NE appears to contribute in the exercise-inducedchanges in renal protein handling, possibly by its effect on renal haemodynamics.

IntroductionSubmaximal exercise has been reported to result in a marked elevation of

albuminuria in patients with insulin-dependent diabetes mellitus (IDDM) with eithernormal or slightly elevated urinary albumin excretion at rest [1-7]. This albuminuricresponse is thought to be related to an increased glomerular passage of macromolecules,and possibly indicates an early impairment in the glomerular filtration barrier in IDDM[8]. Changes in tubular protein reabsorption are not considered to be of great importanceduring moderately strenuous exercise since the excretion of $2-microglobulin, a markerof tubular protein handling, does not increase under these circumstances [2,3].

The mechanisms by which exercise leads to a pronounced increase in albuminuriain IDDM patients are still poorly understood. During exercise the effective renal plasmaflow (ERPF) decreases more than the glomerular filtration rate (GFR). Consequently,the filtration fraction i.e. the GFR divided by the ERPF increases [4,6,9]. A rise in the

Chapter 238

intraglomerular pressure, which is assumed to be reflected by this increase in filtrationfraction [9], might be an important factor that contributes to the enhanced excretion ofurinary albumin. An abnormal rise in systolic blood pressure has also been implicatedin the exercise induced rise in albuminuria in IDDM particularly in conjunction withmicroalbuminuria [5], although this has not always been reported [2,6]. Studies in therat have shown that norepinephrine (NE) increases renal vascular resistance predomi-nantly via efferent glomerular vasoconstriction [10]. Both epinephrine and NE increaserenal vascular resistance in humans [11-13]. It has therefore been proposed that the renalhaemodynamic changes during exercise are related to increased circulatory levels ofplasma catecholamines [9,14]. By a renal haemodynamic mechanism, plasma catechol-amines thus could possibly affect renal protein handling during exercise in IDDMpatients.

The purpose of the present study was to determine whether an altered plasmacatecholamine response could play a role in the exaggerated rise in albuminuria inIDDM with normo- and microalbuminuria during moderately strenuous exercise.

Subjects and methods

SubjectsAll subjects consented to the procedure after explanation of the purpose of the

study, which was approved by the local medical ethics committee. Three groups ofsubjects were investigated (Table 1): 12 healthy control subjects (urinary albuminexcretion rate (Ualb.V<10 µg/min, Controls); 13 IDDM patients with normo-albuminuria (Ualb.V<10 µg/min, DNormo), 13 IDDM patients with microalbuminuria(Ualb.V ranging from 10 to 200 µg/min, DMicro).

Ualb.V was determined in three consecutive overnight urine collections obtainedwithin three months prior to the study. Three urine samples were used to classify theIDDM patients because of the high variability of Ualb.V [15]. A level of Ualb.V above10 µg/min was considered to be abnormal (above the 97.5 percentile for 50 healthycontrols subjects). Urinary tract infection was excluded by bacterial culture. All IDDMpatients suffered from ketosis prone diabetes mellitus and were considered to be insu-lin-dependent on clinical grounds. All subjects had a serum creatinine level below 120µmol/l and were normotensive (systolic blood pressure #140 mmHg and diastolic bloodpressure #90 mmHg). No medication other than insulin or oral contraceptives wereallowed. The IDDM patients did not have autonomic neuropathy as determined byValsalva manoeuvre, beat to beat variation during deep breathing and blood pressureresponse to standing [16]. A questionnaire showed no differences in physical activityamong the three groups. As is shown in Table 1 the three groups were closely com-parable in age. There were no significant differences in sex distribution. The twoIDDM groups were comparable with respect to duration of disease, metabolic controland retinopathy. All subjects were studied in the fasting state while the morning insulindose was withheld. Blood pressure was measured using a sphygmomanometer(Baumano- meter, W.A. Braun Inc, N.Y., USA). Korotkoff phase 5 was taken as thediastolic blood pressure. Blood pressure and pulse rate were measured at 5 min intervals

Plasma norepinephrine and albuminuria during exercise 39

Table 1. Clinical characteristics of the study groups.

Controls DNormo DMicro(n=12) (n=13) (n=13)

Age (years) 25±3 26±6 27±5

Duration of disease (years) -- 10±4 12±5

Sex (males/females) 6/6 5/8 10/3

Ualb.V (µg/min) 3.1 (1.6-7.8) 5.1(2.6-9.8) 21.3(10.2-166.7)a c

Serum creatinine (µmol/l) 79±8 76±12 89±17d

Overnight creatinine clearance 109±27 110±29 102±23d

(ml/min per 1.73m )2

Body mass index (kg/m ) 21.6±1.7 22.9±2.1 23.4±2.32

HbA (%) 5.3±10.4 7.7±1.6 7.6±1.21e

Fasting blood glucose (mmol/l) 4.6±0.4 11.9±4.7 12.1±5.2e

Retinopathy (O/B/P) -- 11/1/1 6/4/3b

Systolic blood pressure change to 1 (-11 to 12) 3 (-12 to 15) 1 (-13 to 12)

standing (mmHg)

Beat to beat variation (beats/min) 35±7 39±13 32±12

Valsalva ratio 1.76±0.28 1.72±0.28 1.78±0.32

Controls: control subjects; DNormo: IDDM patients with Ualb.V<10 µg/min; DMicro: IDDM

patients with Ualb.V>10 µg/min and <200 µg/min. Ualb.V: urinary albumin excretion rate;a

Retinopathy: O: absent; B:background; P:proliferative. Values are given in mean±SD, exceptb

Ualb.V and systolic blood pressure change to standing which are given in median(range); denotesc

p<0.001 from Controls, DNormo; denotes p<0.05 from DNormo; denotes p<0.01 from DNormo,d e

DMicro.

during

the exercise and the recordings were averaged for analysis. Mean arterial pressure(MAP) was calculated as b diastolic + a systolic blood pressure. Experimental design

The exercise protocol of Mogensen [1,2] was used, except that the subjectsexercised only at 600 kpm for 20 min. The subjects drank 250 ml water per 20 minthroughout the test from 0800 hours until 1200 hours to promote diuresis. Before theexercise the subjects were sitting for three hours. Urine was collected at 20 min intervalsfrom 1000 hours until 1200 hours, including two postexercise collections. The exercisewas performed on a bicycle ergometer from 1100 hours until 1120 hours. Blood sampleswere taken at regular intervals from a cannula inserted into an antecubital vein whichwas kept patent with a saline drip. During the study, the fractional clearance of albumin(FalbCl), calculated as the albumin clearance divided by the creatinine clearance, wasused instead of the Ualb.V. This allowed us to take account of differences in GFRbetween the individuals and possible changes in GFR during the test, to correct for

Chapter 240

changes in serum albumin concentration due to exercise induced haemoconcentration[17], and for possible errors in urine collection due to incomplete bladder emptying[15]. It was assumed that urinary albumin excretion had returned to baseline levels aftertwo hours water loading [18] and so the 20 min urine collection directly before theexercise test was used as the reference period. FalbCl during the 20 min of exercise until40 min thereafter was averaged for each subject to evaluate the effect of exercise.

Laboratory methods Urine samples were stored at -20 C for a maximum of 2 weeks until analysis.o

Samples from each subject were determined in one run. Urinary albumin was measuredusing a commercially available double-antibody radioimmunoassay (Diagnostic ProductsCorporation, Apeldoorn, The Nether-lands, cat no KHAD2). The lower detection limitwas 0.07 mg/l. Serum albumin was measured on a SMAC autoanalyzer (TechniconInstruments Inc. Tarrytown, NY, USA). Urinary and serum creatinine were measuredon SMA(C) auto-analyzers. Blood glucose was measured on a Yellow Springs glucoseanalyser (Model 23A, Yellow Springs Inc., Yellow Springs, Ohio, USA). HbA1c wasdetermined by colorimetry [19]. Plasma epinephrine and NE concentrations were mea-sured using high-performance liquid chromotograhpy as previously described [20].

Statistical analysisResults are expressed as mean±SD for parametrically distributed data and as

median (interquartile ranges) for non-parametrically distributed data. Comparisons ofvariables between groups were carried out using analysis of variance for parametricallyand non- parametrically distributed data as appropriate. Changes of variables withingroups were assessed by paired Student's t-tests or Wilcoxon tests. Adjustment formultiple comparisons was carried out using Duncan's method [21]. Differences inprevalence of clinical variables were analyzed by chi-square statistic. Correlations weresought using Spearman's rank correlation. Multiple regression analysis was used todisclose the independent contribution of parameters. P-values less than 0.05 were con-sidered to be significant.

Results

Blood glucoseBlood glucose concentrations, measured directly before and after the exercise,

were 4.5±0.4 and 4.4±0.6; 11.2±3.1 and 11.7±2.8; 11.1±3.6 and 10.7±4.2 mmol/lin Controls, DNormo and DMicro, respectively.

Blood pressure and pulse rateSystolic and diastolic blood pressure were similar in the three groups at rest

(Table 2). During exercise systolic blood pressure was higher in DMicro than inDNormo whereas diastolic blood pressure was lower in DNormo than in the other twogroups. The increments in systolic blood pressure and MAP were larger in DMicro thanin the other groups. The changes in Controls and in DNormo were similar. The pulse

Plasma norepinephrine and albuminuria during exercise 41

rate was significantly higher in the two IDDM groups compared with the controls butthe increment in response to exercise were comparable in all groups (Table 2).

Chapter 242

Table 3. Effect of exercise on fractional albumin clearance.

Fractional albumin clearance (×10 ) exercise+postexercise-6

preexercise exercise+postexercise % of prexercise

Controls (n=12) 0.69 (0.54-1.44) 0.82 (0.51-2.19) 111 (70-296)

DNormo (n=13) 0.75 (0.33-1.53) 2.04 (0.79-6.69) 272 (87-1599)a,e e

DMicro (n=13) 2.99 (1.77-11.97) 10.93 (3.38-26.25) 215 (104-524)c,d f,gb,d,f

Controls, DNormo, DMicro: see text and Table I. Data are given in median (interquartile ranges). denotes DNormo>Controls, p<0.05; denotes DMicro>DNormo, p<0.05; denotes

a b c

DMicro>DNormo, p<0.001; denotes DMicro>Controls, p<0.001; denotes p=0.02 from pre-d e

exercise; p=0.01 from preexercise; % change in Dmicro>Controls, p<0.05.f g

Table 2. Blood pressure and pulse rate response to exercise.

preexercise during exercise increase

Bloodpressure (mmHg) systolic/diastolic systolic/diastolic systolic MAP

Controls (n=12) 122±11 / 79±8 154±16 / 73±7 32±8 6±2a a

DNormo (n=13) 118±10 / 73±6 150±17 / 66±9 32±15 6±5c a a

DMicro (n=13) 120±10 / 76±9 166±14 / 74±8 46±15 15±9b a,d a,d

Pulse rate (beats/min)

Controls (n=12) 67±11 128±24 61±19e e a

DNormo (n=13) 79±9 150±21 71±22a

DMicro (n-13) 80±10 145±13 69±12a

Controls, DNormo, DMicro: see text and Table I. MAP: mean arterial pressure. Data are given inmean±SD. denotes p<0.001 from preexercise; denotes DMicro>DNormo, p<0.05;

a b

denotes DNormo<Controls, DMicro, p<0.05; denotes Dmicro>Controls, DNormo, p<0.01; c d e

denotes Controls<DNormo, DMicro, p<0.01

Fractional albumin excretionIn the preexercise period no difference in FalbCl could be demonstrated between

DNormo and Controls, but FalbCl was significantly higher in DNormo than in Controlsduring the exercise+postexercise period (Table 3). In DMicro, FalbCl was significantlyhigher compared with DNormo and Controls during both periods (Table 3). During theexercise + postexercise period, FalbCl increased in DNormo (p=0.02) and in DMicro(p=0.01) but not in Controls (p=0.40). The relative change in FalbCl, expressed as apercentage of the preexercise values, was greater in DMicro than in Controls (p<0.05).

Plasma norepinephrine and epinephrine concentrationsAt baseline there were no differences in plasma epinephrine and NE concentrations

among the three groups (Table 4). In all groups, plasma epinephrine and NE concentra-tions increased significantly in response to exercise. The exercise-induced increments in

Plasma norepinephrine and albuminuria during exercise 43

Table 4. Plasma norepinephrine and epinephrine concentrations.

preexercise exercise postexercise1000 hours 1100 hours 1120 hours 1200 hours

Norepinephrine (nmol/l)

Controls (n=12) 2.3 (2.0-3.7) 2.5 (2.0-4.3) 3.5 (2.7-5.1) 1.9 (1.2-2.5)a

DNormo (n=13) 2.4 (1.4-3.1) 2.9 (1.9-3.3) 5.2 (3.7-6.7) 2.3 (1.7-2.9)b,c

DMicro (n=13) 2.1 (1.6-2.3) 2.2 (1.5-3.1) 5.8 (4.4-6.6) 2.1 (1.6-2.4)b,d

Epinephrine (nmol/l)

Controls (n=12) 0.14 (0.08-0.21) 0.11 (0.09-0.18) 0.40 (0.20-0.50) 0.16 (0.13-0.23)b

DNormo (n=13) 0.13 (0.10-0.21) 0.12 (0.10-0.16) 0.30 (0.18-0.43) 0.15 (0.06-0.25)b

DMicro(n=13) 0.10 (0.08-0.26) 0.14 (0.07-0.29) 0.88 (0.43-0.97) 0.18 (0.07-0.29)b,e

Controls, DNormo, DMicro: see text and Table I. Data are given in median (interquartile ranges). denotes p<0.05 and denotes p<0.01 from preexercise; denotes DNormo>Controls, p<0.05;

a b c d

denotes DMicro>Controls, p<0.02; denotes DMicro>Controls and DNormo, p<0.05.e

plasma NE concentration were greater in both IDDM groups (DNormo: 2.5(1.0- 4.6)nmol/l; DMicro: 3.2(2.1-3.8) nmol/l) than in Controls (1.0(0.0-2.0) nmol/l) (Dnormo>Controls, p<0.05; DMicro>Controls, p<0.01) (Figure 1A). The exercise inducedincrements in plasma epinephrine were not different in Controls (0.24(0.11-0.35)nmol/l) compared with DNormo (0.18(0.08-0.29) nmol/l) and DMicro (0.52(0.22-0.70)nmol/l), but the increments in DMicro were greater than in DNormo (p<0.05) (Figure1B).

Relationships between plasma catecholamines, haemodynamics and glycaemia The relative changes in FalbCl were positively related to the increase in plasma

NE in all groups (All subjects r=0.65, p<0.001; Controls r=0.62, p<0.05; DNormor=0.63, p<0.05; DMicro r=0.56, p<0.05) (Figure 2). The exercise induced changesin FalbCl were also correlated with the increase in MAP in DNormo (r=0.64, p<0.01)and in DMicro (r=0.60, p<0.02) but not in Controls (r=0.25, p=0.22). Multipleregression analysis demonstrated that both changes in plasma NE (p<0.002) and inMAP (p<0.005) independently contributed to the exercise induced alterations in FalbCl(multiple r=0.73, n=38). Multiple regression analysis in two IDDM groups showedsimilar independent effects of plasma NE (p<0.02) and MAP (p<0.005) on thechanges in FalbCl (multiple r=0.75, n=26). When multiple regression analysis wascarried out using the increments in systolic blood pressure instead of MAP the contribu-tion of blood pressure rise was not significant (p=0.20).

1.2

1.0

0.8

0.6

0.4

0.2

0

-0.2Controls DNormo DMicro

Cha

nge

in P

lasm

a E

pine

phrin

e (n

mol

/l)

***

**

7.0

6.0

5.0

4.0

3.0

2.0

0

-1.0Controls DNormo DMicro

Cha

nge

in P

lasm

a N

orep

inep

hrin

e (n

mol

/l)

***

*1.0

***

Chapter 244

Figure 1. Changes in plasma norepinephrine and epinephrine concentrations in response to exercise.Controls F, DNormo F, DMicro M: see text and Table 1. Bars are median values. Upper panel:Increments in plasma epinephrine concentration. * denotes p<0.01 from preexercise; ** denotes in-crement in DMicro>DNormo, p<0.05. Lower panel: Increments in plasma norepinephrine con-centration. * denotes p<<0.05 from preexercise; ** denotes increment in DNormo andDmicro>Controls, p<<0.05.

Plasma norepinephrine and albuminuria during exercise 45

Baseline plasma NE concentrations and pulse rate were not correlated. Duringexercise a positive relationship was observed between plasma NE and pulse rate in Con-trols (r=0.77, p<0.01) as well in the combined IDDM groups (r=0.53, p<0.01). Nosignificant correlations were found between plasma epinephrine concentrations andeither FalbCl, pulse rate or blood pressure during exercise in any of the groups. In theIDDM groups the changes in plasma epinephrine and NE concentrations were not relatedto actual glycaemia and HbA1.

DiscussionExercise has been proposed to provoke abnormalities in renal protein handling in

IDDM [1,2]. It is unclear whether plasma catecholamines are involved in this abnormalrise in albuminuria. The subjects participating in the present study did not have systemichypertension or signs of autonomic neuropathy. Patients with an abnormal urinaryalbumin excretion rate above 10 µg/min were designated microalbuminuric but it shouldbe noted that levels higher than 20 µg/min are considered to represent a risk marker fordiabetic nephropathy [22].

Exercise induced an abnormal rise in albuminuria in both IDDM groups in accor-dance with many previous studies employing a comparable fixed workload [1-5]. Themechanisms which are responsible for this abnormal rise in albuminuria are still notprecisely understood. Systemic blood pressure rise, alterations in renal haemodynamicsand in glomerular perm selectivity or a combination of these factors are likely to beinvolved [1-9]. As earlier studies have shown [2,6] no difference in blood pressureresponse was observed between the normoalbuminuric IDDM and non-diabetic subjects,making it unlikely that an abnormal increase in blood pressure per se was responsible forthe marked albuminuric response in this group of patients. In the microalbuminuricIDDM subjects, the increase in blood pressure was larger than in the other groups, inaccordance with other data [5], suggesting that alterations in systemic haemodynamicscould have contributed to the abnormal rise in albuminuria in these patients. This studyshowed a significant relationship between the increase in plasma NE and the albuminuricresponse during exercise. In addition, the rise in plasma NE was significantly greater innormo- and microalbuminuric IDDM patients than in the normal subjects, although therewas a considerable overlap in individual responses. Multiple regression analysis substan-tiated that both the exercise induced rise in plasma NE and in MAP had an independenteffect on the albuminuric response. It is noteworthy that, although used by some investi-gators [7], diastolic blood pressure recordings are probably underestimated whenmeasured by the auscultatory method during exercise [23]. Several studies have showna relationship between (change in) systolic blood pressure and albuminuria duringexercise [3,5], but this has not consistently been reported [4,6]. In this investigation,excluding patients with hypertension, the contributory effect of systemic blood pressurewas not significant when using systolic blood pressure instead of MAP in the multipleregression analysis.

It has been documented that the filtration fraction rises more and is higher duringexercise in IDDM subjects, especially in patients with microalbuminuria [4,6]. It istherefore probable that there is a link between the increased glomerular passage of

10

-2

Change in Plasma Norepinephrine (nmol/l)

0 2 4 6 8

4

103

102

101

Cha

nge

in F

albC

l (%

of p

reex

erci

se)

33.240

Chapter 246

Figure 2. Changes in plasma norepinephrine concentration and relative changes in fractionalalbumin clearance in response to exercise. Controls ", DNormo ", Dmicro •: see text and Table1. Relative changes in fractional albumin clearance are expressed as % of pre-exercise values. Allsubjects (n=38):r=0.65, p<0.001; Controls (n=12) F:r=0.62, p<0.05 DNormo (n=13)F:r=0.63, p<0.05; DMicro (n=13) !:r=0.56, p<0.05

albumin and the rise in filtration fraction during exercise in IDDM patients. Experimen-tal studies in the rat have shown that infusion of NE results in an increase in filtrationfraction in parallel with directly measured increments in intraglomerular pressure [10],whereas the degree of albuminuria in response to NE is closely related to changes infiltration fraction [24]. Taken together, the previously reported renal haemodynamicchanges [4,6] and the presently shown significant relationship between changes inalbuminuria and changes in plasma NE concentrations during exercise would support thehypothesis that NE is involved in the increase in albuminuria during exercise [9,14].This present observations also suggest that an altered catecholamine response couldpossibly contribute to an abnormal rise in albuminuria in IDDM.

Several factors might be involved in the abnormal rise in plasma NE duringexercise in IDDM. A fixed workload was employed in the expectation that this would

Plasma norepinephrine and albuminuria during exercise 47

discriminate IDDM and non-diabetic subjects with respect to their albuminuric response[1-5]. It has been suggested that the maximal working capacity is reduced inmicroalbuminuric but not significantly so in normoalbuminuric IDDM patients [6].Aerobic working capacity was found to be decreased in patients with autonomic neurop-athy [17,25] and microalbuminuria whereas normoalbuminuric patients showed a normalmaximal oxygen uptake [26]. It seems therefore probable that the fixed workload wasmore stressful for the microalbuminuric patients but it is unlikely that the relativeworkload was larger for the normoalbuminuric IDDM patients. In addition, it has beenshown that an exaggerated catecholamine response can be related to suboptimal meta-bolic control [27,28], but no relationship between plasma catecholamines and HbA1c oractual glycaemia could be demonstrated in this study. The expected increase in pulse rate[2,7] raises the possibility of an abnormal regulation of the autonomic nervous systemin IDDM [29]. Moreover, an enhanced vasopressor responsiveness to infusion of NEhas been demonstrated in IDDM patients either without overt microvascular compli-cations [30] or with microalbuminuria [31]. Whether glomerular vascular structuresexhibit such an increased sensitivity to circulatory catecholamines remains to be eluci-dated.

In conclusion, the present observations suggest that renal protein handling duringexercise is related to both circulating plasma NE and blood pressure in IDDM. Theexaggerated albuminuric response, even in the absence of an increased urinary albuminexcretion at rest, was associated with abnormalities in systemic catecholamineregulation. Direct experiments are needed to demonstrate a causal relationship betweenthese phenomena. In addition, the rise in plasma NE was significantly greater in normo-and microalbuminuric diabetic patients than in the normal subjects, although there was aconsiderable overlap in individual responses.

References

1 Mogensen CE, Vittinghus E. Urinary albumin excretion during exercise in juvenile diabetes.Scand J Clin Lab Invest 1975; 35: 295-300

2 Viberti GC, Jarret RJ, McCartney M, Keen H. Increased glomerular permeability to albumininduced by exercise. Diabetologia 1978; 14: 293-300

3 Mogensen CE, Vittinghus E, Sølling K. Abnormal albumin excretion after two provocativerenal tests in diabetes: physical exercise and lysine injection. Kidney Int 1979; 16: 385-393

4 Vittinghus E, Mogensen CE. Albumin excretion and renal haemodynamic response to physicalexercise in normal and diabetic man. Scand J Clin Lab Invest 1981; 41: 627-632

5 Christensen CK. Abnormal albuminuria and blood pressure rise in incipient diabetic nephropat-hy induced by exercise. Kidney Int 1984; 25: 819-823

6 Feldt-Rasmussen B, Baker L, Deckert T. Exercise as a provocative test in early renal diseasein Type 1 (insulin-dependent) diabetes: albuminuric, systemic and renal haemodynamicresponses. Diabetologia 1985; 28: 389-396

7 Watts GF, Williams I, Morris RW, Mandalia S, Shaw KM, Polak A. An acceptable exercisetest to study microalbuminuria in Type 1 diabetes. Diabetic Med 1989; 6: 787-792

8 Mogensen CE, Christensen CK, Vittinghus E. The stages in diabetic renal disease with empha-sis on the stage of incipient diabetic nephropathy. Diabetes 1983; 32 (Suppl 2): 64-78.

9 Poortmans JR. Postexercise proteinuria in humans. JAMA 1985; 253: 236-240 10 Meyers BD, Deen WM, Brenner BM. Effects of norepinephrine and angiotensin II on the

Chapter 248

determinants of glomerular ultrafiltration and proximal tubule fluid reabsorption in the rat. CircRes 1975; 37: 101-110

11 Smythe McC, Nickel JF, Bradley SE. The effect of epinephrine (usp), l-epinephrine, and l-norepinephrine on glomerular filtration rate, renal plasma flow and the urinary excretion ofsodium, potassium and water in normal man. J Clin Invest 1952; 31: 499-506

12 Poortmans JR. Exercise and renal function. Exerc Sport Sci Rev 1977; 5: 255-29413 Mimran A. Regulation of renal blood flow. J Cardiovasc Pharmacol 1987; 10[Suppl 5]: S1-S914 Drury PL, Smith GM, Ferriss JB. Increased vasopressor responsiveness to angiotensin II in

Type 1 (insulin-dependent) diabetic patients without complications. Diabetologia 1984; 27: 174-179

15 Dullaart RPF, Roelse H, Sluiter WJ, Doorenbos H. Variability of albumin excretion in diabe-tes. Neth J Med 1989; 34: 287-296

16 Ewing DJ, Clarke BF. Diagnosis and management of diabetic autonomic neuropathy. Br MedJ 1982; 285: 916-918

17 Hilsted J, Galbo H, Christensen NJ, Parving HH, Benn J. Haemodynamic changes duringgraded exercise in patients with autonomic neuropathy. Diabetologia 1982; 22: 318-23.

18 Viberti GC, Mogensen CE, Keen H, Jacobsen FK, Jarret RJ, Christensen CK. Urinary excre-tion of albumin in normal man: the effect of waterloading. Scand J Clin Lab Invest 1982; 42:147-151

19 Fluckiger R, Winterhalter KH. In vitro synthesis of hemoglobin A1C. FEBS Lett 1976; 71:356-360

20 Smedes F, Kraak JC, Poppe H. Simple and fast solvent extraction system for selective andquantitative isolation of adrenaline, noradrenaline and dopamine from plasma and urine. JChromatogr 1982; 231: 25-39

21 Duncan DBL. Multiple range and multiple F-tests. Biometrics 1955; 11: 1-42.22 Mogensen CE. Prediction of clinical diabetic nephropathy in IDDM patients. Alternatives to

microalbuminuria? Diabetes 1990; 39: 761-76723 Karlefors T. Haemodynamic studies in male diabetics. Acta Med Scand 1966[Suppl]449:45-8024 Brenner BM, Bohrer MP, Baylis C, Deen WM. Determinants of glomerular permselectivity:

Insights derived from observations in vivo. Kidney Int 1977; 12: 229-23725 Hilsted J, Galbo H, Christensen NJ. Impaired cardiovascular responses to graded exercise in

diabetic autonomic neuropathy. Diabetes 1979; 28: 313-31926 Jensen T, Richter EA, Feldt-Rasmussen B, Kelbaek H, Deckert T. Impaired aerobic work

capacity in insulin dependent diabetics with increased urinary albumin excretion. Br Med J1988; 296: 1352-1354

27 Christensen NJ. Plasma norepinephrine and epinephrine in untreated diabetics, during fastingand after insulin administration. Diabetes 1974; 23: 1-8

28 Tamborlane WV, Sherwin RS, Koivisto V, Hendler R, Genel M, Felig P. Normalization of thegrowth hormone and catecholamine response to exercise in juvenile-onset diabetic subjectstreated with a portable insulin infusion pump. Diabetes 1979; 28: 785-788

29 Hilsted J. Pathophysiology in diabetic autonomic neuropathy: cardiovascular, hormonal andmetabolic studies. Diabetes 1982; 31: 730-737

30 Beretta-Piccoli C, Weidmann P. Exaggerated pressor responsiveness to norepinephrine innonazothemic diabetes mellitus. Am J Med 1981; 71: 829-835

31 Bodmer CW, Patrick AW, How TV, Williams G. Exaggerated sensitivity to NE- inducedvasoconstriction in IDDM patients with microalbuminuria. Possible etiology and diagnosticimplications. Diabetes 1992; 41: 209-214

Plasma norepinephrine and albuminuria during exercise 49

Acknowledgments

This study was in part supported by a grant from the Dutch Diabetes ResearchFund. We are indebted to AL Aalders, MD for her technical assistance. The Departmentof Pulmonary Physiology is acknowledged for the provided hospitality.

Departments of Endocrinology, Nephrology, General Medicine, and Surgical Intensive1 2 3 4

Care Unit, Groningen State University Hospital, The Netherlands

Scand J Lab Invest 1995; 55: 15-22

CHAPTER 3

INFLUENCE OF AMBIENT PLASMA NOREPINEPHRINE ONRENAL HAEMODYNAMICS IN IDDM PATIENTS AND

HEALTHY SUBJECTS

K. Hoogenberg , A.R.J. Girbes , C.A. Stegeman , W.J. Sluiter , 1 4 2 1

W.D. Reitsma and R.P.F. Dullaart 3 1

Imbalances in renal vasodilatory and vasoconstrictive mechanisms are responsible for the renalhaemodynamic changes observed in IDDM patients. Animal experiments have shown thatnorepinephrine (NE) infusion increases the intraglomerular pressure by predominantly efferentarteriolar vasoconstriction. The relationships between ambient plasma NE levels and renalhaemodynamics were studied in 18 healthy control subjects (group C), in 17 normoalbuminuric(albumin excretion rate (Ualb.V) <20 µg/min; group D1) and in 17 microalbuminuric (Ualb.V 20-200µg/min; group D2) IDDM patients without overt autonomic neuropathy. Supine glomerular filtrationrate (GFR (ml/min per 1.73m )) and effective renal plasma flow (ERPF (ml/min per 1.73m )) were2 2

determined over a 2 h period using constant infusions of I-iothalamate and I-hippuran, respec-125 131

tively. The subjects were studied in the fasting state. The IDDM patients were investigated during nearnormoglycaemia. Data are given as means±SD. In group D1, GFR and ERPF (126±15 and 538±89,respectively) were elevated as compared to group C (108±15 and 478±73; p<0.01 and p<0.05, respec-tively). In group D2, GFR (124±25, p<0.05) but not ERPF (515±104) was higher than in group C. GFRand ERPF were negatively correlated with venous plasma NE in C (r=-0.61, p<0.005 and r=-0.64,p<0.001, respectively), in group D1 (r=-0.54, p<0.03 and r=-0.63, p<0.005, respectively) and in groupD2 (r=-0.53, p<0.03 and r=-0.60, p<0.01, respectively). Multiple regression analysis disclosed thatdiabetes per se, independent from plasma NE, had a positive contribution to GFR. In contrast, ERPFwas only related to plasma NE levels. In conclusion, GFR and ERPF are inversely related to venousplasma NE levels both in healthy and in IDDM subjects, supporting the hypothesis that plasma NE isa vasoconstrictive substance. The independent positive effect of diabetes as a categorial variable onGFR, suggests that concomitant vasodilating mechanisms play a role in the renal haemodynamicalterations in IDDM patients.

Introduction

In patients with insulin-dependent diabetes mellitus (IDDM) the presence of micro-albuminuria does not only predict the future development of diabetic nephropathy [1], butis also associated with generalized vascular damage [2]. Microalbuminuria is thought toresult from an increased glomerular leakage of macromolecules. Both changes in theglomerular permselectivity properties as well as in systemic and intrarenal haemo-dynami-

Chapter 350

cs factors have been demonstrated to underlie elevations in urinary albumin excretion [3].Before and during the microalbuminuric phase renal haemodynamic changes, commonlyreflected by an increase in glomerular filtration rate (GFR), are frequently observed inIDDM patients [4-7]. In animal models of diabetes mellitus an increase in GFR isassociated with the development of glomerulosclerosis and the progressive loss of kidneyfunction [8]. Some clinical observations also support the notion that an elevated GFR perse might contribute to the subsequent development of diabetic nephropathy [5,9].

The elevated GFR associated with IDDM has been attributed to an increase ineffective renal plasma flow (ERPF) [5-7], an increase in the glomerular filtration surfacearea [10] and possibly an increase in the intraglomerular pressure [5]. Many vasodilatoryand vasoconstrictive substances are able to modulate renal haemodynamics by influencingthe glomerular afferent and efferent arteriolar tone [6,7,11]. Glomerular arterioles contain"-adrenoreceptors which mediate vasoconstriction induced by norepinephrine (NE)released at the terminal nerve ending of the sympathetic nervous system (SNS).Intravenous infusion to reach high physiological NE levels lowers ERPF, whereas the fallin GFR is less marked [11-14]. Consequently, the filtration fraction (FF) (i.e. the quotientof GFR and ERPF) rises, reflecting a change in pressure profile along the arterioles.Recent data suggest that IDDM with microalbuminuria are hypersensitive to NE-inducedvasoconstriction of dorsal hand veins which also carry "-adrenoreceptors [15]. Thusalterations in SNS activity might be implicated in the pathogenesis of diabetes-associatedrenal haemodynamic abnormalities. However, the relation between plasma NE and renalhaemodynamic parameters is uncertain. In a small group of adolescent IDDM patients astrong negative relationship between the plasma NE level and GFR was observed [16],whereas others did not find such a relationship [17,18].

Therefore, the present cross sectional study was conducted to establish the relation-ships between circulatory NE levels and renal haemodynamic changes in normo- andmicroalbuminuric IDDM patients as compared with matched healthy subjects.

Subjects and methods

Subjects

All participants consented to the procedure after explanation of the purpose of thestudy, which was approved by the local medical ethics committee. Eligible subjects hadan age between 21 and 50 years, a serum creatinine #120 µmol/l, a urinary proteinexcretion # 500 mg/day, and a body mass index <27 kg/m . Subjects with hypertension2

(systolic blood pressure >160 mmHg and/or diastolic blood pressure >95 mmHg), usingantihypertensive or anti-inflammatory drugs were excluded from participation. The IDDMwere considered to be insulin-dependent since their glucagon stimulated C-peptide levelswere <0.2 nmol/l. They had a disease duration of at least 5 years. Autonomic function wasassessed with beat-to-beat variation (abnormal if variation less than 10 beats per minute),valsalva manoeuvre (abnormal if lower than 1.11) and systolic blood pressure response tostanding (abnormal if exceeding 30 mmHg) [19]. None of the

Ambient plasma noradrenaline and renal haemodynamics 51

Table 1. Clinical characteristics of the study groups.

Group C Group D1 Group D2 (n=18) (n=17) (n=17)

Age (years) 30.8 ± 7.7 32.0 ± 8.3 35.8 ± 8.0

Duration of disease (years) - 14.8 ± 6.0 18.8 ± 5.3

Sex (males/females) 13 / 5 15 / 2 14 / 3

Ualb.V (µg/min) n.d. 9.1(6.4-11.5) 34.4(22.9-105.3)a c

Mean arterial blood pressure (mmHg) 92 ± 9 95 ± 7 100 ± 8d

Body mass index (kg/m ) 22.6 ± 2.3 22.5 ± 2.0 23.7 ± 2.02

HbA1c (%) n.d. 8.6 ± 1.4 8.3 ± 1.0

Urinary sodium excretion (mmol/day) 182 ± 76 156 ± 63 203 ± 82

Urinary urea excretion (mmol/day) 425 ± 108 435 ± 127 384 ± 113

Insulin dose (U/kg per day) - 0.74 ± 0.23 0.79 ± 0.22

Retinopathy (O/B/P) - 11/4/2 8/5/4b

Valsalva ratio n.d. 1.57 ± 0.26 1.62 ± 0.38

Beat-to-beat variation during

deep breathing n.d. 29 ± 13 28 ± 10

Change in systolic blood pressure

in response to standing (mmHg) n.d. -5 (-10 to 5) -4 (-8 tot 2)a

Group C: control subjects; group D1: IDDM patients with UalbV <20 µg/min; group D2: IDDMpatients with UalbV > 20 µg/min and <200 µg/min. n.d. denotes not determined; Ualb.V: urinary a

albumin excretion rate; Retinopathy: O: absent; B: background; P: proliferative. Values are givenb

in mean ± SD, except for Ualb.V and blood pressure response to standing which are given in median(interquartile range). denotes p<0.001 from group D1; denotes p<0.01 from groups C and D1

c d

subjects suffered from overt autonomic neuropathy defined as at least 1 abnormal test.Because of the high variability of urinary albumin excretion [20], three timed overnighturine collections were obtained to measure urinary albumin excretion rate (Ualb.V).Microalbuminuria was defined as Ualb.V ranging 20 to 200 µg/min in at least two urinesamples [1].

Three groups of subjects were included in the study: 18 healthy controls (group C);17 IDDM patients with normoalbuminuria (group D1), 17 IDDM patients with micro-albuminuria (group D2). Table 1 shows the clinical characteristics of the study groups anddata of the autonomic function tests. The groups were closely comparable with respect toage and there was no difference in sex distribution. Mean urinary sodium and ureaexcretion was similar in the groups and none of the subjects used a protein (<0.8 g/day perkg bodyweight) or a sodium restricted (<50 mmol/day) diet. The two IDDM groups didnot differ in diabetes duration and metabolic control. The proportion of patients withretinopathy was similar in each IDDM group.

Procedure of the renal haemodynamic measurements

Chapter 352

All participants were studied after an overnight fast and remained fasting during theprocedure. Starting at 0800 h, the subjects were studied in the supine position and were only allowed to stand on voiding. Diuresis was promoted by a waterload of 300 ml/h,and no other liquids than water were permitted to drink. The IDDM patients received theirlast regular insulin dose 8 to 12 h before the start of the study.

The IDDM patients were studied during near normoglycaemia to minimize thepossible effects of actual glycaemia on GFR and ERPF [21]. Thus, a 5% glucose solutionwas initially infused at a rate 1 ml(0.28 mmol)/h per kg bodyweight together with regularacting insulin (Velosulin H.M., Novo-Nordisk, Bagsvaerd, Denmark). The amount ofinsulin infused was 1% of the total daily requirements per hour. Blood glucose wasmeasured at 30 min intervals. If blood glucose exceeded 8.3 mmol/l extra insulin wasgiven intravenously until one hour before the start of the renal haemodynam/icmeasurements. Further insulin boluses were not given since insulin can acutely stimulatecatecholamine release [22]. After a stabilisation period of at least 2 h the renal clearancestudies were carried out.

GFR and ERPF were measured simultaneously during a 2 h clearance period, usingprimed infusions of I-iothalamate and I-hippuran, respectively [23]. I-iothalamate125 131 125

and I-hippuran were infused at a constant rate of 4.8µCi/h (178 kBq/h) and 12 µCi/h131

(444 kBq/h), respectively. The clearances were calculated using the formulas U.V/P andI.V/P, respectively. U.V represents the urinary excretion rate of the tracer, I.V representsthe infusion rate of the tracer, P represents the mean tracer value of 2 plasma samplestaken at the beginning and at the end of the clearance period. Errors in the estimation ofthe GFR due to incomplete bladder emptying and dead space were corrected bymultiplying the clearance of I-iothalamate with the formula: clearance of I-hippuran125 131

(I.V/P) / clearance of I-hippuran (U.V/P). The coefficients of variation for GFR and131

ERPF are 2.2% and 5.0%, respectively [23]. The GFR and ERPF were corrected to 1.73m2

of body-surface area. The FF was calculated as the ratio of the GFR and the ERPF. Blood pressure was recorded every hour, using a sphygmomanometer. Korotkoff

phase I and V were taken as the systolic and the diastolic blood pressure, respectively. Atthe beginning and the end of the clearance period venous blood was drawn from a catheterinserted into an antecubital vein (kept patent with 0.9% NaCl 20 ml/h).

Laboratory methods

Blood glucose was measured using a Yellow Springs glucose Analyser (Model 23A,Yellow Springs Inc,. Yellow Springs, Ohio, USA). HbA1c was determined by colorimetry(24) (reference values: 4.5 to 5.8%). Serum creatinine, albumin, and urinary sodium andurea were measured on SMA(C) autoanalysers (Technicon Instru-ments Inc. Tarrytown,N.Y., USA). Urinary albumin was measured by radioimmuno-assay (Diagnostic ProductsCorporation, Apeldoorn, The Netherlands; coefficient of variation 8%). Plasma insulinand plasma C-peptide concentrations were measured by radioimmuno-assay. Plasmaepinephrine and NE concentrations were determined by high-performance liquidchromatography (coefficient of variation 4%) [25].

Statistical analysis

Results are expressed as mean±SD for parametrically distributed data and as median

Ambient plasma noradrenaline and renal haemodynamics 53

(interquartile ranges) for non-parametrically distributed data. Comparisons of variablesbetween groups were carried out using analysis of variance for para-metrically andnon-parametrically distributed data as appropriate. Changes of variables within groupswere assessed by paired Student's t-tests or Wilcoxon tests. Adjustment for multiplecomparisons was carried out using Duncan's method [26]. Differences in prevalence ofclinical variables were analysed by chi-square statistic. Correlations were studied usingSpearman's rank analysis. Multiple stepwise regression analysis was used to disclose theindependent contribution of the hormonal and metabolic parameters to GFR and ERPF.Two-sided p-values less than 0.05 were considered to be significant.

Results

Blood glucose, plasma insulin and renal haemodynamic parameters

During the 2-h clearance period, insulin and blood glucose concentrations were notdifferent between the two IDDM groups and did not change significantly (Table 2).

GFR was significantly higher in both IDDM groups as compared to group C (p<0.01and p<0.05 for groups D1 and D2, respectively, Table 2), whereas ERPF was only higherin group D1 than in group C (p<0.05, Table 2). FF was significantly elevated in group D2as compared to group C (p<0.05) (Table 2). In the combined IDDM groups, ERPF as wellas FF were higher than in group C (p<0.05, for both, Table 2).

Plasma norepinephrine and epinephrine.

Plasma NE levels were higher in group C than in groups D1 and D2 at the beginning(p<0.05 for both, Table 3) but not significantly so at the end of the 2 h clearance period(Table 3). The differences in averaged plasma NE levels did not reach significance(p=0.08 for group C compared with groups D1 and D2, Table 3). Plasma epinephrineconcentrations were not significantly different between the groups. No significant changeswere observed in plasma NE and epinephrine during the 2 h study period.

Correlation analysisThe averaged values of plasma NE and epinephrine were used for the correlation

analysis. GFR and ERPF were negatively correlated with plasma NE levels in group C(r=-0.61, p<0.005 and r=-0.64, p<0.001, respectively), in group D1 (r=-0.54, p<0.03 andr=-0.63, p<0.005, respectively) and in group D2 (r=-0.53, p<0.03 and r=-0.60, p<0.01,respectively, Figure 1A,B). In the combined IDDM groups, a positive correlation wasfound between FF and plasma NE (r=0.34, p<0.05, Figure 1C) which was not present ingroup C (r=0.11, NS, not shown). GFR was significantly related to HbA1c(r=0.34,p<0.05) in the combined IDDM groups, whereas no significant relation existedbetween HbA1c, ERPF (r=0.24, p=0.17) and FF (r=0.02, NS). Plasma epinephrine wasnot correlated with renal haemodynamics.

Chapter 354

Table 2. Renal haemodynamics and corresponding plasma glucose and insulin levels.

Group C Group D1 Group D2 IDDM all (n=18) (n=17) (n=17) (n=34)

Glomerular filtration rate 108 ± 15 126 ± 15 124 ± 25 125 ± 20a b b

(ml/min per 1.73m )2

Effective renal plasma flow 478 ± 73 538 ± 89 515 ± 104 527 ± 96b b

(ml/min per 1.73m )2

Filtration fraction 0.22 ± 0.02 0.23 ± 0.03 0.24 ± 0.03 0.24 ± 0.03b b

Fractional sodium excretion 0.95 ± 0.53 0.78 ± 0.31 0.93 ± 0.42 0.88 ± 0.33

Plasma glucose (mmol/l)08.00 h 4.6 ± 0.4 12.2 ± 5.5 11.6 ± 4.4 11.9 ± 4.9

a a a

begin n.d. 6.6 ± 1.1 6.9 ± 1.4 6.7 ± 1.1end n.d. 6.5 ± 1.1 6.5 ± 1.2 6.5 ± 1.1

Free plasma insulin (mU/l) begin n.d. 28 (23 - 49) 26 (18 - 35) 28 (19 - 35)

end n.d. 32 (19 - 43) 23 (13 - 34) 26 (17 - 40)

Group C: control subjects; group D1: IDDM with Ualb.V <20µg/min; group D2: IDDM patients withUalbV >20 µg/min and <200 µg/min, IDDM all: combined IDDM groups; n.d. not determined;Values are given in mean±SD, except for free insulin which is given in median (interquartile range). denotes p<0.01 and denotes p<0.05 from group C

a b

Table 3. Plasma norepinephrine and epinephrine concentrations during the renal haemodynamicmeasurements.

Group C Group D1 Group D2 IDDM combined (n=18) (n=17) (n=17) (n=34)

Plasma norepinephrine (nmol/l)

Begin 1.56 ± 0.45 1.32 ± 0.56 1.28 ± 0.40 1.29 ± 0.48a

End 1.52 ± 0.53 1.30 ± 0.50 1.33 ± 0.49 1.31 ± 0.49Mean 1.54 ± 0.50 1.31 ± 0.45 1.30 ± 0.44 1.31 ± 0.44

Plasma epinephrine (nmol/l)

Begin 0.15 ± 0.10 0.20 ± 0.14 0.19 ± 0.10 0.19 ± 0.12End 0.13 ± 0.10 0.18 ± 0.13 0.20 ± 0.10 0.19 ± 0.11Mean 0.14 ± 0.11 0.18 ± 0.12 0.22 ± 0.10 0.19 ± 0.11

Group C: control subjects; group D1: IDDM patients with Ualb.V <20 µg/min; group D2: IDDMpatients with Ualb.V >20 µg/min and <200 µg/min. Values are given in mean ± SD. denotes p<0.05a

from groups D1 and D2

Ambient plasma noradrenaline and renal haemodynamics 55

Multiple regression analysis was carried out to evaluate whether renalhaemodynamics were independently related to the plasma NE level and to establish ifplasma NE contributed to the differences in GFR and ERPF between the IDDM and thenon-diabetic subjects. In the analyses plasma NE as well age, diabetes duration, HbA1c,sodium and urea excretion, and blood pressure were the independent variables. In thecombined groups (n=52), ERPF was only related to plasma NE (r=-0.61, P<0.001). Themodel was not improved by including the presence of IDDM as a categorial variable(p=0.30, multiple r=0.59). Thus IDDM per se did not influence ERPF, independently fromthe level of NE. GFR was related to ERPF both in the combined groups (r=0.79, p<0.001)as well as in IDDM groups (r=0.86, p<0.001). However in the combined groups, IDDM,included in the model as a categorial variable, had an independent effect on GFR (r=0.22,p<0.01; multiple r=0.88), indicating that diabetes associated vasodilation influenced GFR.The other covariates did not contribute to either ERPF or GFR in a model includingplasma NE and IDDM.

Discussion

The elevations in GFR and ERPF associated with IDDM are considered to reflect animbalance between glomerular vasodilatory and vasoconstrictive mechanisms [5-7,11,17].This study demonstrates that GFR and ERPF are inversely related to plasma NE levels innormo- and microalbuminuric IDDM patients as well as in healthy subjects. Our resultsextend recent observations in a small group of adolescent IDDM patients showing anegative correlation between plasma NE and GFR [16] and the possibility arises thatplasma NE is a determinant of renal haemodynamics.

No differences were observed in the inverse correlations between plasma NE andERPF in the IDDM patients as compared to the control subjects. This raises the possibilitythat the presently observed slightly lower plasma NE levels in the IDDM subjects couldcontribute to the elevations in ERPF. Clearly, our study design does not permit aconclusion with respect to potential differences in renal NE sensitivity in association withIDDM or microalbuminuria [15].

Both animal and human studies indicate that renal blood flow is an importantdeterminant of GFR [5-7,27]. Our data support this view since GFR was strongly relatedto ERPF. However, the elevated GFR cannot be fully accounted for by an increase inERPF, since the filtration fraction was higher in the IDDM patients, particularly in thosewith microalbuminuria. GFR is also determined by the intraglomerular pressure, theoncotic pressure and the ultrafiltration coefficient and any of these factors or a combina-tion of them might explain the IDDM related increase in filtration fraction [5]. Indeed, theintraglomerular pressure is elevated in experimental diabetes [8] and capillary hyperten-sion has been documented in IDDM patients [28]. In addition, the glomerular filtrationsurface area is increased in IDDM [10] which will influence the ultrafiltration coefficient.The present study did not differentiate between these mechanisms but multiple regressionanalysis disclosed an independent effect of the diabetic state per se on GFR, thussuggesting that concomitant vasodilating mechanisms are present in IDDM patients.

200

180

160

140

80

Norepinephrine (nmol/l)

Norepinephrine (nmol/l)

Norepinephrine (nmol/l)

0.5 1.0 1.5 2.0 2.5 3.0

A

120

100

0

800

700

600

500

00.5 1.0 1.5 2.0 2.5 3.0

ER

PF

(mm

m)

2in

per

173.

l/G

FR

(mm

m)

2in

per

173.

l/

B

400

300

0.32

0.28

0.24

0.20

0.000.5 1.0 1.5 2.0 2.5 3.0

Filt

ratio

n fr

actio

n

C

Chapter 356

Figure 1. Relationships between renal haemodynamics and plasma norepinephrine levels (NE). F Controlsubjects, F IDDM patients with Ualb.V<20 µg/min,! IDDM patients with Ualb.V>20 µg/min and<200µg/min. A: plasma NE and GFR in F r=-0.61, p<0.005; in F r=-0.54, P<0.03; in ! r=-0.53, p<0.03;B: plasma NE and ERPF in F r=-0.64, p<0.001; in F r= -0.63, p<0.005; and in! r=-0.60, p<0.01; C: plasma NE and FF in the combined IDDM groups r=0.34, p<0.05.

Ambient plasma noradrenaline and renal haemodynamics 57

NE constricts renal vasculature via stimulation of "1-adrenoreceptors, which arelocated along afferent and efferent renal arterioles [11,14,29]. As compared with otherisolated renal vessels, efferent glomerular arterioles from the rabbit have the highestsensitivity to the vasoconstrictive action of NE [30] and micropuncture studies in the rathave shown that NE lowers glomerular blood flow and increases the transglomerularpressure [13]. In humans NE infusion results in a decrease in ERPF and a rise in filtrationfraction [12]. Moreover, an increase in plasma NE appears to contribute to exercise-induced changes in microalbuminuria possibly by a renal haemodynamic mechanism [31].

The pathophysiological mechanisms responsible for the relation between the basalplasma NE level and renal haemodynamics are unknown. NE, present in venous forearmblood, is derived both from local skeletal muscular nerve activity and from the arterialcirculation [32]. The forearm venous plasma NE concentration is the net result of thesesources minus local neuronal uptake. Thus, the peripheral venous forearm NE concentra-tion is influenced by other factors than SNS activity alone. NE acts both locally as aneurotransmitter and distantly as a hormonal factor [33]. It cannot be discriminatedwhether the relationship between plasma NE and renal haemodynamics indicates a distanteffect or reflects a generalized or local neurotransmitter spillover of the SNS. To study therole of SNS activity in diabetes related renal haemodynamic abnormalities it would benecessary to obtain arterial and venous renal blood samples simultaneously with GFR andERPF determinations.

In conclusion, GFR and ERPF are inversely correlated with the venous plasma NElevel in healthy subjects as well as in IDDM patients with normo- and microalbuminuria,supporting the hypothesis that circulatory NE can be a renal vasoconstrictive factor. Thepositive contribution of IDDM per se to GFR, independently from the negative effects ofplasma NE, suggests that concomitant vasodilating mechanisms play a role in the renalhaemodynamic alterations in IDDM patients.

References

1 Mogensen CE. Prediction of clinical diabetic nephropathy in IDDM patients. Alternatives tomicroalbuminuria? Diabetes 1990; 39: 761-767.

2 Deckert T, Feldt-Rasmussen B, Borch-Johnsen K, Jensen T, Kofoed-Enevoldsen A. Albuminuriareflects widespread vascular damage. The Steno hypothesis. Diabetologia 1989; 32: 219-226.

3 Mogensen CE, Christensen CK, Vittinghus E. The stages in diabetic renal disease with emphasison the stage of incipient diabetic nephropathy. Diabetes 1983; 32 (Suppl 2): 64-78.

4 Mogensen CE. Early glomerular hyperfiltration in insulin-dependent diabetics and latenephropathy. Scand J Clin Lab Invest 1986; 46: 201-206.

5 Viberti GC, Wiseman MJ. The kidney in diabetes mellitus: significance of the early abnormalities.Clin Endocrinol metab 1986; 15: 753-782.

6 Hoogenberg K, ter Wee PM, Lieverse AG, Sluiter WJ, Dullaart RPF. Insulin-like growth factorand altered renal haemodynamics in growth hormone deficiency, acromegaly, and Type 1diabetes mellitus. Transpl Proc 1994; 26: in press.

7 Hoogenberg K, Dullaart RPF, Freling NJM, Meijer S, Sluiter WJ. Contributory roles of glycatedhaemoglobin, circulatory glucagon and growth hormone to increased renal haemodynamics inType 1 (insulin-dependent) diabetes mellitus. Scand J Clin Lab Invest 1993; 53: 821-828.

8 Hostetter TH, Rennke HG, Brenner BM. The case for intrarenal hypertension in the initiation andprogression of diabetic and other glomerulopathies. Am J Med 1982; 72: 375-380.

Chapter 358

9 Rudberg S, Persson B, Dahlquist G. Increased glomerular filtration rate as a predictor of diabeticnephropathy. An 8-year prospective study. Kidney Int 1992; 41: 822-828.

10 Østerby R, Parving H-H, Nyberg G, Jørgensen HE, Løkkegaard H, Svalander C. A strong correla-tion between glomerular filtration rate and flitration surface in diabetic nephropathy. Diabetologia1988;31: 265-270.

11 Bank N. Mechanisms of diabetic hyperfiltration. Kidney Int 1991; 40: 792-807.12 Smythe McC, Nickel JF, Bradley SE. The effect of epinephrine (usp), l-epinephrine, and l-

norepinephrine on glomerular filtration rate, renal plasma flow and the urinary excretion ofsodium, potassium and water in normal man. J Clin Invest 1952; 31: 499-506.

13 Meyers BD, Deen WM, Brenner BM. Effects of norepinephrine and angiotensin II on thedeterminants of glomerular ultrafiltration and proximal tubule fluid reabsorption in the rat. CircRes 1975; 37: 101-110.

14 Mimran A. Regulation of renal blood flow. J Cardiovasc Pharmacol 1987; 10 (Suppl 5): S1-S9.15 Bodmer CW, Patrick AW, How TV, Williams G. Exaggerated sensitivity to NE-induced

vasoconstriction in IDDM patients with microalbuminuria. Possible etiology and diagnosticimplications. Diabetes 1992; 41: 209-214.

16 Johansson B-L, Berg U, Freyschuss U, Hall K, Troell S. Factors related to renal haemodynamicsin young type-1 diabetes mellitus patients. Pediatr Nephrol 1990; 4: 589-592.

17 Estmatjes E, Fernandez MR, Halperin I, Camps J, Gaya J, Arroyo V, Rivera F, Figuerola D. Renalhaemodynamic abnormalities in patients with short-term insulin-dependent diabetes mellitus: roleof renal prostaglandins. J Clin Endocrinol Metab 1985; 60: 1231-1236.

18 Feldt-Rasmussen B, Mathiesen ER, Deckert T, Giese J, Christensen NJ, Bent-Hansen L, NielsenMD. Central role for sodium in the pathogenesis of blood pressure changes independent ofangiotensin, aldosterone and cathecholamines in Type 1 (insulin-dependent) diabetes mellitus.Diabetologia 1987; 30: 610-617.

19 Ewing DJ, Clarke BF. Diagnosis and management of diabetic autonomic neuropathy. Br Med J1982; 285: 916-918.

20 Dullaart RPF, Roelse H, Sluiter WJ, Doorenbos H. Variability of albumin excretion in diabetes.Neth J Med 1989; 34: 287-296.

21 Dullaart RPF, Meijer S, Sluiter WJ, Doorenbos H. Renal haemodynamic changes in response tomoderate hyperglycaemia in type 1(insulin-dependent) diabetes mellitus. Eur J Clin Invest 1990;20: 208-213.

22 Rowe JW, Young JB, Minaker KL, Stevens AL, Pallotta J, Landsberg L. Effect of insulin andglucose infusions on sympathetic nervous system activity in normal man. Diabetes 1981; 30: 219-225.

23 Donker AJM, Van der Hem GK, Sluiter WJ, Beekhuis H. A radioisotope method for simultaneousdetermination of the glomerular filtration rate and effective renal plasma flow. Neth J Med 1977;20: 97-103.

24 Flückiger R, Winterhalter KH. In vitro synthesis of haemoglobin A1C. FEBS Lett 1976; 71:356-360.

25 Smedes F, Kraak JC, Poppe H. Simple and fast solvent extraction system for selective andquantitative isolation of adrenaline, NA and dopamine from plasma and urine. J Chromatogr1982; 231: 25-39.

26 Duncan DB. Multiple range and multiple F-test. Biometrics 1955; 11: 1-42.27 Brenner BM, Humes HD. Mechanics of glomerular ultrafiltration. N Engl J Med 1977; 297: 148-

154.28 Sandeman DD, Shore AC, Tooke JE. Relation of skin capillary pressure in patients with insulin-

dependent diabetes mellitus and metabolic control. N Engl J Med 1992; 327: 760-764.29 Stephenson JA, Summers RJ. Autoradiographic evidence for a heterogeneous distribution of

alpha1-adrenoreceptors labelled by [ H]prazosin in rat, dog and human kidney. J Auton Pharmac3

Ambient plasma noradrenaline and renal haemodynamics 59

1986; 6: 109-116.30 Edwards RM. Segmental effects of norepinephrine and angiotensine II on isolated renal

microavessels. Am J Physiol 1983; 244: F526-534.31 Hoogenberg K, Dullaart RPF. Abnormal plasma noradrenaline response and exercise induced

albuminuria in type 1 (insulin-dependent) diabetes mellitus. Scan J Clin Lab Invest 1992; 52: 803-811.

32 Hjemdahl P. Plasma catecholamines - analytical challanges and physiological limitatations. ClinEndocrinol Metab 1993; 2: 307-353.

33 Silverberg AB, Shah SD, Haymond MW, Cryer PE. Norepinephrine: hormone and neurotransmit-ter in man. Am J Physiol 1978; 234(3): E252-E256.

Acknowledgments

We are indebted to M. van Kammen, F. Kranenburg-Nienhuis and A. Drent-Bremerfor their skilful assistance.

Department of Endocrinology, Nephrology, Angiology and General Medicine, Groningen1 2 3 4

State University Hospital, Groningen, The Netherlands J Am Soc Nephrol (in press)

CHAPTER 4

EXOGENOUS NOREPINEPHRINE INDUCES AN ENHANCEDMICROPROTEINURIC RESPONSE IN

MICROALBUMINURIC IDDM PATIENTS

K. Hoogenberg , W.J. Sluiter , G. Navis , T.W. van Haeften , A.J. Smit , W.D.1 1 2 1 3

Reitsma and R.P.F. Dullaart 4 1

Exogenous norepinephrine (NE) increases intraglomerular pressure in animal experiments, butit is unknown whether NE induces a microproteinuric response in man. Moreover, it has not beenstudied whether possible microproteinuric and renal haemodynamic changes induced by NE arealtered in IDDM complicated by microalbuminuria. Therefore, the microproteinuric and renalhaemodynamics responses to exogenous NE infusions were measured in 8 matched normoalbuminuricIDDM patients (group D1), microalbuminuric IDDM patients (group D2) and control subjects (groupC). As anticipated, mean arterial pressure (MAP)-NE dose response curves were significantly shiftedleftwards in groups D1 and D2 compared to group C (p<0.05), indicating a higher systemic NEresponsiveness in IDDM. On separate days, NE or placebo was infused at individually determined NEthreshold (ÎMAP = 0mmHg, T), 20% pressor (ÎMAP = 4mmHg, 20%P) and pressor doses (ÎMAP =20mmHg, P) with measurement of urinary albumin and IgG excretion (Ualb.V and UIgG.V), and ofglomerular filtration rate (GFR, by I-iothalamate) and effective renal plasma flow (ERPF, by I-125 131

hippurate). At NE pressor dose Ualb.V and UIgG.V rose in all groups (p<0.05 to 0.01), whereasurinary $2-microglobulin was unchanged. The increases in Ualb.V and UIgG.V were morepronounced in the microalbuminuric group than in the other groups (p<0.05). An NE dose-dependentfall in ERPF and rise in filtration fraction (FF) was found in all groups (p<0.05 to 0.001 for all),whereas GFR did not change significantly. The renal haemodynamic dose-response relationship wassimilar in the groups. In conclusion, exogenous NE acutely promotes glomerular protein leakage, andit is plausible that intraglomerular NE effects contribute to this phenomenon. The microproteinuricresponse is enhanced in microalbuminuric IDDM despite unaltered renal haemodynamicresponsiveness, which may reflect a specific NE response or a general effect of vasopressor stimuli topromote glomerular protein leakage in patients with a preexistent defect in glomerular permselectivity.

Introduction

Microalbuminuria predicts the future development of nephropathy in insulin-dependent diabetes mellitus (IDDM) [1-3]. Both structural changes in the glomerularbasement membrane and changes in systemic and renal haemodynamics are thought to beinvolved in the development and progression of microalbuminuria in IDDM [2-5].Diabetes-associated glomerular hyperfiltration and elevations in intraglomerular pressure

Chapter 462

may contribute to the progression to overt diabetic nephropathy [6-10]. Severalneurohormonal systems, such as the renin-angiotensin system (RAS) and the sympatheticnervous system (SNS) are involved in the regulation of glomerular pressure [11].Norepinephrine (NE) increases intraglomerular pressure in the rat [12]. In accordance witha renal vasoconstrictive effect of NE, the effective renal plasma flow (ERPF) is correlatednegatively with the ambient plasma NE level in man [13], and in non-diabetic subjectsexogenous NE lowers ERPF without much change in glomerular filtration rate (GFR) [14-16].

Exogenous NE may increase proteinuria in non-diabetic subjects with overtproteinuria [15], but it is unknown whether exogenous NE affects microproteinuria inIDDM. In an uncontrolled study in healthy subjects, NE did not evoke an albuminuricresponse [17]. Exercise, a manoeuver that increases NE levels, is well known to increasemicroproteinuria in normo-and microalbuminuric IDDM patients and healthy subjects.Increases in urinary albumin excretion during exercise are related to changes in plasma NEand systemic blood pressure [18]. Systemic and intraglomerular effects of NE may,therefore, be involved in the increased microproteinuria following strenuous activities[18,19]. It has, moreover, been proposed that an exaggerated renal haemodynamic NEresponsiveness could play a role in the pathogenesis of diabetic renal disease [20].

This randomized, placebo controlled study was performed to test whether NE evokesan increase in microproteinuria and if so whether such an increase is more pronounced innormo- and microalbuminuric IDDM patients. Therefore, we compared the micro-proteinuric and renal and systemic haemodynamic responses during stepwise exogenousNE infusions in IDDM patients with and without microalbuminuria and healthy subjects.Since an enhanced systemic blood pressure response to NE was anticipated in IDDM [21-23], the subjects were studied at individually determined NE threshold, 20% pressor andpressor doses.

Methods

Subjects

All subjects consented to participate in the study, which was approved by the localmedical ethics committee. Eligible subjects had an age between 30 and 55 years, a serumcreatinine <120µmol/l, a urinary protein excretion <500 mg/day, a body mass index <28kg/m , no overt hypertension (systolic blood pressure <160 mmHg and diastolic blood2

pressure <95 mmHg), no signs of clinical autonomic neuropathy assessed by beat-to-beatvariation during deep breathing (abnormal if difference <10 beats per min), Valsalvamanoeuver (abnormal if ratio <1.1) and systolic blood pressure response to standing(abnormal if decline >20 mmHg) [24]. Diabetic patients were insulin-dependent,glucagon-stimulated plasma C-peptide level was <0.2 nmol/l in all of them. They had adisease duration of at least 15 year. Patients with untreated proliferative retinopathy,symptomatic coronary heart disease, or peripheral vascular disease were excluded.Microalbuminuria was defined as a persistently elevated urinary albumin excretion rate(Ualb.V) between 20 and 200µg/min in at least 2 of the 3 overnight urine collectionsobtained over 1 year [1,25]. Urinary tract infection was excluded in all subjects bybacterial culture.

Norepinephrine, microproteinuria and renal haemodynamics 63

Table 1. Prestudy clinical characteristics of control and IDDM subjects.

Non-diabetic Normoalbuminuric Microalbuminuric Subjects IDDM patients IDDM patients

Group C Group D1 Group D2

Number 8 8 8Age (years) 44.4 ± 3.2 45.5 ± 3.6 46.7 ± 3.5Duration of disease (years) - 21.6 ± 2.1 27.6 ± 2.2Sex (males/females) 7 / 1 7 / 1 7 / 1Cigarette smokers (n) 3 3 3Urinary albumin excretion(µg/min) 4.0 (2.6-6.2) 4.5 (2.5-8.4) 65.9 (40.1-108.1) a

Serum creatinine (µmol/l) 84 ± 4 83 ± 4 88 ± 7Creatinine clearance (ml/min) 108 ± 5 114 ± 7 116 ± 12Mean arterial pressure (mmHg) 94.8 ± 2.5 92.4 ± 3.6 101.9 ± 1.5b

Body mass index (kg/m ) 23.2 ± 1.0 24.5 ± 1.1 25.6 ± 0.9 2

Glycated haemoglobin (%) 5.7 ± 0.2 8.3 ± 0.3 8.2 ± 0.3c

Sodium excretion (mmol/day) 109 ± 13 103 ± 9 109 ± 7Protein intake (g/kg per day) 1.11 ± 0.07 0.94 ± 0.07 1.03 ± 0.05Insulin dose (U/kg per day) - 0.74 ± 0.10 0.79 ± 0.09Retinopathy (absent/background/proliferative) - 5 / 2 / 1 0 / 5 / 3d

Valsalva ratio 1.63 ± 0.08 1.59 ± 0.07 1.52 ± 0.07Beat-to-beat variation 25 ± 3 24 ± 4 16 ± 3 during deep breathing Change in systolic blood pressure -9.4 ± 2.6 -6.9 ± 3.5 -7.3 ± 2.8in response to standing (mmHg)

Values are given in mean±SEM, except UalbV which is given in geometric mean and 95%confidence interval. p<0.001 and p<0.05 from group C and D1; p<0.001 from group D1 anda b c

D2; and p<0.05 from group D1 by ð analysis.d 2

Eight healthy control subjects (group C), 8 normoalbuminuric (group D1) and 8microalbuminuric IDDM patients (group D2) were included (Table 1). They wereindividually matched for gender, age (within 5 yr), body mass index (within 2.5 kg/m )2

and smoking habits. Metabolic control and daily insulin requirement was similar in thenormo- and microalbuminuric diabetic groups. Retinopathy was more severe in themicroalbuminuric group. Creatinine clearance was similar in the groups. Mean arterialpressure (MAP) was higher in patients with microalbuminuria (Table 1). Sevenmicroalbuminuric diabetic patients used an angiotensin converting enzyme (ACE)inhibitor that was stopped 6 weeks before the study. No other medication was used.

To avoid bias due to differences in sodium and protein intake on NE sensitivity andrenal haemodynamics [26,27], all participants adhered to diet containing 100 mmolsodium and 1 g protein/kg body wt per day 1 week before the studies. Protein intake wascalculated from urinary urea excretion assuming nitrogen balance as described [28].

Chapter 464

Dietary compliance was acceptable in the participants and similar in the groups (Table 1).The 3 women were studied in the luteal phase of the menstrual cycle.

Study design

The experiments were carried out during 3 consecutive days (see below) in a roomwith a controlled temperature of 22 C. The subjects were studied after an overnight fasto

and remained so during the experiments. Smoking was prohibited and no other liquid thanwater was allowed to drink. After completion of each study day meals were provided at1700 h and 2000 h. The diabetic patients received an intravenous insulin infusion(Velosulin H.M., Novo-Nordisk, Bagsvaerd, Denmark) at a rate of 30 mU/kg per h,starting at 0600 h. Blood glucose was measured every 5 to 10 min and clamped at 5mmol/l by varying the infusion rate of a 20% glucose solution using the euglycaemicclamp technique [29]. Potassium chloride (10 mmol per 100 g of glucose) was added tothe glucose bottles. The euglycaemic insulin clamp was used to minimize effects ofdifferences in actual glycaemia on renal haemodynamics [30] and to avoid that changes ininsulin levels during the experiments would potentially influence NE sensitivity [31]. Thediabetic subjects had been admitted to the hospital the evening prior to the study. Bloodglucose was kept between 7 and 11 mmol/l in the evenings and nights before and betweenstudy days. The controls received a 5% glucose infusion (without insulin) to keep bloodglucose unchanged from fasting levels, to prevent a decline in plasma insulin duringprolonged fasting and to prevent increases in endogenous catecholamine release as aconsequence of glucose counter regulation. The minimal glucose infusion rate was 30ml/h, because a 5% glucose solution was used as vehicle for renal tracers as well as for NEand placebo.

NE dose-response curves (day 1)

On day 1, individual dose-response curves to exogenous NE started at 1230 h asdescribed [32]. After a 30 min baseline period, an NE infusion (norepinephrine tartrate,Centrafarm BV, Leiden, The Netherlands, dissolved in 5% glucose) was given in the rightantecubital vein. The NE infusion rate was increased at predetermined 25 min steps of 10,20, 40, 80, 100, 150, 200, 250, 300, 400 and 600 ng/kg per min. The maximal infusionrate was reached if MAP was increased by 20 to 25 mmHg. Blood pressure was measuredevery 2 min at the left upper arm with a semi-automated sphygmomanometer (Dinamap®1846, Critikon, Tampa, FL, USA). Pulse rate was recorded continuously on an ECGoscilloscope (Hewlett Packard, Boblingen, Germany). MAP was calculated as a systolicblood pressure+b diastolic blood pressure (mmHg).

The pressor dose was defined as the NE infusion rate required to induce a rise inMAP of 20 mmHg above baseline and was calculated from a semi-logarithmic plot of theNE dose vs the individual changes in MAP. The individual regression lines were con-structed from the linear part of the dose-response curve [32]. For each participant the NEthreshold dose (0 mmHg change in MAP), 20% pressor dose (increase in MAP of 4mmHg) and pressor dose (increase in MAP of 20 mmHg) were calculated from theindividually obtained regression equations.

Norepinephrine, microproteinuria and renal haemodynamics 65

Renal haemodynamic studies and proteinuria assessment (day 2 and 3)

On day 2 and 3, NE or placebo were infused in randomized order in a single-blindedfashion at threshold, 20% pressor and pressor doses. The subjects remained supine andwere only allowed to stand on voiding. At 0700 h, each subject drank 600 ml water topromote diuresis. Thereafter, urinary volume loss was supplemented by oral water aftersubtraction of the amount of intravenous solutions. Note that urinary albumin excretiontemporarily rises after a water load and that its level has returned to baseline after a 2 hperiod [33]. Between 0900 h and 1630 h, renal haemodynamics were measured during 10consecutive clearance periods each lasting 45 min. Two baseline clearance periods werefollowed by 6 periods of NE (2 periods each at threshold, at 20% pressor and at pressordoses) or placebo. Finally, 2 determinations were obtained after cessation of NE orplacebo infusion (recovery period).

GFR and ERPF were measured simultaneously as the urinary clearances ofI-iothalamate and I-hippurate, respectively, and corrected for incomplete urine125 131

collections as described previously [34]. The coefficients of variation for GFR and ERPFare 2.2% and 5.0%, respectively [34]. The GFR and ERPF were corrected to 1.73 m of2

body-surface area. Filtration fraction (FF) was calculated as the quotient of GFR andERPF.

Laboratory measurements

Venous blood samples were taken at the end of each clearance period with thesubjects in supine position. Blood was drawn from a needle inserted into the left antecubit-al vein that was kept patent with a saline drip (0.9% NaCl, 30 ml per h). Blood sampleswere immediately centrifuged at 4EC and the samples were stored at -20EC before assay.Breakdown and in vitro generation of angiotensin II (AII) was prevented by the additionof o-phenantroline, enalapril and neomycin. Aliquots of urine for measurement of urinaryproteins were stored at -20EC and analysed within 2 weeks. For proper determination ofIgG, urine was 1:1 diluted with buffered glucose (0.1 M NaCl, 40 mM phosphate, sodiumazide (0.2%), 5% bovine serum albumin and 100 mM glucose adjusted to pH 7.4) [35].

Plasma insulin was determined by RIA. Plasma NE was analysed by HPLC [13,18].Plasma total renin and plasma active renin were determined with commercially availabledouble-antibody RIA’s (Nichols Institute Diagnostics, San Juan Capistrano, CA, USA).The antibody to renin has a cross-reactivity with prorenin of 0.2%. Prorenin wascalculated as the difference between total renin and active renin levels. Serum ACEactivity was measured with an HPLC technique. AII was determined by RIA [36] bycourtesy of P.M.H. Schiffers (Dept. of Pharmacology, State University Maastricht, TheNetherlands). Cross-reactivity of the AII antibody with angiotensin I is <0.1%. Serumaldosterone was determined by RIA. Serum albumin and IgG were measured usingELISA. Blood glucose was measured using a Yellow Springs glucose Analyser (Model23A, Yellow Springs Inc,. Yellow Springs, Ohio, USA).

Urinary albumin was measured with ELISA using rabbit anti-human albuminantibody (DAKO). The detection limit is 1 µg/l. Urinary IgG was measured using anin-house developed ELISA using polyclonal goat-antihuman-IgG (gammachain) (Tago,Burlingame, CA, USA, cat no 4100) and goat-antihuman-IgG peroxidase (Nordic,

Chapter 466

Tilburg, The Netherlands) as conjugate [35]. The detection limit is 1.5 µg/l. Urinary $2-microglobulin was measured by nephelometry.

Sodium, potassium, urea and creatinine in serum and urine were measured onSMA(C) autoanalysers (Technicon Instruments Inc. Tarrytown, N.Y., USA). Glycatedhaemoglobin was measured by HPLC (Bio-Rad, Veenendaal, The Netherlands).

Statistical analysis

Results are expressed as mean±SEM for normally distributed data and as geometricmean (95% confidence interval, CI) for not-normally distributed data unless statedotherwise. The 2 clearance periods obtained at baseline, threshold, 20% pressor, pressorand recovery were averaged for analysis. The NE effects were analysed after subtractionof the data obtained during the placebo day. Within-group and between-group differenceswere analysed with one-way ANOVA and repeated measurements ANOVA, forparametrically or non-parametrically distributed data where appropriate. Adjustment formultiple comparisons was carried out using Duncan's method. Bivariate correlations weresought using Spearman's rank correlation. If correlations were significant, logisticregression lines were constructed using standard curve fitting models where appropriate.Multiple regression analysis evaluated independent contributions of parameters on thetime-controlled observations during the 3 NE infusion periods. In case of not-normallydistributed data, logarithmically transformed values were used. P-values less than 0.05were considered to be significant.

Results

Blood glucose and insulin concentrations during NE and placebo administration

In the diabetic groups, blood glucose remained unchanged and was similar duringNE and placebo infusions (Table 2). In group C, blood glucose rose modestly withincreasing NE dose to levels that were slightly higher than in group D1 and D2 (Table 2).A reduction of glucose infusion rate was, however, not possible (see study design). In allgroups, plasma insulin was stable during NE and placebo administration. As expected,plasma insulin levels were consistently higher in group D1 and D2 compared to group C(p<0.001, Table 2).

NE dose response curves, blood pressure and venous plasma NE during renal haemodynamic studies

The dose-response curves of the calculated threshold, 20% pressor and pressor doseswere significantly shifted leftwards in group D1 and D2 compared to group C (p<0.05 forboth, Figure 1A). As a consequence of the lower NE doses required, lower plasma NElevels were attained in group D1 and D2 compared to group C during the NE infusion day(p<0.05, Table 3). There was a close relationship between the NE infusion rate andincrements in plasma NE levels in all groups (C: r =0.93, p<0.001 (n=24); D1: r =0.91,s s

p<0.001 (n=24); and D2: r =0.88, p<0.001 (n=24), Figure 1B). As this relationship wass

similar in the 3 groups, the averaged venous plasma NE-pressor curves were alsosignificantly shifted leftwards in both diabetic groups compared to group C (Figure 1C).

Norepinephrine, microproteinuria and renal haemodynamics 67

Table 2. Blood glucose and plasma insulin during norepinephrine (NE) and placebo infusions.

Non-diabetic Normoalbuminuric MicroalbuminuricSubjects (n=8) IDDM patients (n=8) IDDM patients (n=8)

Group C Group D1 Group D2

Period NE Placebo NE Placebo NE Placebo

Blood Glucose (mmol/l)

Baseline 5.1±0.1 5.1±0.1 4.9±0.1 5.3±0.1 5.2±0.2 5.1±0.1Threshold 5.8±0.2 5.0±0.1 5.0±0.1 4.9±0.1 5.2±0.2 5.0±0.1

b,d

20%Pressor 6.1±0.2 5.0±0.1 5.1±0.1 5.1±0.1 5.3±0.1 5.0±0.1c,e

Pressor 6.8±0.4 4.9±0.1 5.1±0.1 5.1±0.1 5.3±0.1 5.0±0.1c,e

Recovery 6.1±0.3 4.9±0.1 5.1±0.1 5.0±0.1 5.2±0.1 5.1±0.1 b,e

Plasma Insulin (mU/l) a

Baseline 10 ± 2 11 ± 2 30 ± 3 31 ± 3 31 ± 4 31 ± 3Threshold 11 ±2 12 ± 2 30 ± 3 28 ± 4 30 ± 4 30 ± 320% Pressor 12 ± 2 13 ± 2 31 ± 3 31 ± 3 30 ± 4 30 ± 3Pressor 13 ± 1 12 ± 2 32 ± 4 29 ± 3 32 ± 4 31 ± 4Recovery 19 ± 4 12 ± 2 30 ± 3 31 ± 3 30 ± 3 29 ± 3b

Values are given in mean±SEM. p<0.001 for all values in group C from group D1 and D2; a b

p<0.05, p<0.001 from baseline after correction for placebo day; p<0.05, p<0.01 from group D1c d e

and D2

NE infused at the calculated threshold, 20% pressor and pressor doses resulted in similarchanges in MAP in the 3 groups with adequate achievement of target blood pressures(Figure 1C, Table 3), while MAP was unaltered with placebo. Baseline pulse rate wassimilar among the groups (group C: 63±4, group D1: 62±4, group D2: 67±4 (beats/min)).In group C, pulse rate declined with each infusion step (threshold dose: -5±2, p<0.05; 20%pressor dose: -7±2, p<0.01; pressor dose: -11 ±2 (beats/min), p<0.001). In group D1 andD2 this fall was only present at NE pressor dose ( -7±2 and -6±2 (beats/min), p<0.01,respectively). The overall pulse rate response was less pronounced in group D1 and D2compared to group C (p<0.05).

Plasma renin and angiotensin II

During the placebo day, plasma renin and prorenin levels were higher in group D2than in group C and D1 (renin, see Table 3; averaged prorenin 173 (125 to 241), 281 (196to 403) and 512 (297 to 884) mU/L (geometric mean and 95%CI) in group D1, D2 and C,respectively). During NE pressor dose, renin increased in all groups (p<0.05) andremained so after cessation of NE infusion (p<0.05 for all, Table 3).

Chapter 468

Table 3. Blood pressure, venous plasma norepinephrine (NE),plasma active renin andangiotensin II concentrations during NE and placebo infusions.

Non-diabetic Normoalbuminuric MicroalbuminuricSubjects (n=8) IDDM patients (n=8) IDDM patients (n=8)

Group C Group D1 Group D2

Period NE Placebo NE Placebo NE Placebo

Mean Arterial Pressure (mmHg)

Baseline 92.7±2.0 90.2±2.7 91.4±3.0 90.3±2.6 97.4±1.3 96.5±2.3

ª at Threshold 2.0±0.5 -0.2±0.7 2.3±0.6 -1.0±0.4 2.5±0.6 0.1±0.9c

ª at 20%Pressor 4.7±1.1 -1.7±0.8 5.7±0.7 -1.8±0.9 6.0±1.0 -1.3±0.8d e d

ª at Pressor 20.6±1.6 -0.6±1.0 18.5±1.3 -2.0±0.7 20.8±1.4 -2.1±0.8e e e

ª at Recovery -7.0±3.0 0.5±1.6 -3.3±1.8 -3.6±1.3 -2.8±2.1 -0.8±0.7

Venous Plasma NE (nmol/l) a

Baseline 2.6(2.0-3.4) 2.6(1.7-3.8) 2.3(1.7-3.2) 2.3(1.6-3.2) 3.0(2.2-4.0) 2.7(2.2-3.4)

Threshold 7.0(4.0-12.4) 2.6(1.9-3.8) 4.8(3.4-6.9) 2.4(1.7-3.3) 5.3(4.2-6.5) 2.9(2.3-3.6)e e e

20%Pressor 10.4(6.1-17.5) 2.6(2.0-3.5) 5.9(3.7-9.4) 2.5(1.8-3.6) 6.7(5.3-8.4) 2.7(2.0-3.5)e ee

Pressor 29.2(20.8-41) 2.5(1.9-3.3) 16.0(8-32.2) 2.6(2.0-3.5) 17.3(11-27.1) 2.6(2.2-3.1)e e e

Recovery 3.6(2.7-4.9) 2.9(2.1-4.1) 3.1(2.1-4.6) 2.5(2.0-3.1) 3.6(2.6-5.1) 3.2(2.4-4.2)

Plasma Active Renin (mU/l)b

Baseline 40 (27-60) 41 (30-55) 37 (20-69) 37 (16-61) 57 (36-98) 74 (48-113)

Threshold 42 (30-59) 40 (32-49) 36 (19-67) 29 (15-59) 53 (33-84) 59 (34-103)

20%Pressor 46 (33-65) 39 (29-53) 40 (22-75) 28 (14-55) 61 (34-109) 59 (34-101)

Pressor 57 (38-85) 40 (30-53) 52 (30-90) 31 (16-59) 87 (48-159) 66 (41-107)c c d

Recovery 55 (33-99) 37 (27-51) 55 (28-108) 31 (14-65) 84 (50-139) 71 (47-107)d d c

Plasma Angiotensin II (pmol/l)

Baseline 16.2±2.3 15.2±2.1 13.9±2.4 13.6±2.1 16.3±2.0 18.1±2.2

Threshold 16.3±1.6 14.8±2.4 13.6±2.1 14.1±2.3 16.1±2.0 16.9±2.4

20% Pressor 18.1±1.6 14.8±2.7 15.3±2.4 13.3±1.9 18.1±2.6 17.6±2.7

Pressor 24.6±3.5 15.8±2.5 18.3±3.1 14.6±2.6 25.1±4.6 17.9±2.8c c d

Recovery 22.6±5.2 14.3±1.8 20.1±4.6 14.4±2.3 20.9±3.3 16.5±1.8c c

Values are given in mean±SEM, except plasma NE and plasma active renin which are given ingeometric mean and 95% confidence interval. Overall values in group C>groups D1 and D2

a

(p<0.05); overall values in group D2>groups C and D1 (p<0.05); p<0.05, p<0.01 and p<0.001b c d e

from baseline, after correction for placebo day.

10 50 100 500 1000NE dose (ng/kg/min)

P A

20%P

T

chan

ge in

pla

sma

NE

(nm

ol/l)

chan

ge in

MA

P (m

mH

g)

1 2 3 4 5 10 20 304050 100

change in plasma NE (nmol/l)

25 C

10

15

20

5

0

10 50 100 500 1000NE dose

100 B

1

10

0.1

Norepinephrine, microproteinuria and renal haemodynamics 69

Figure 1. Relationships between norepinephrine (NE) infusion, plasma NE increment and changes inmean arterial pressure (MAP). A. Systemic pressor responses to exogenous NE infusions at threshold(T), 20% pressor (20% P) and pressor (P) dose. B. NE infusion dose and changes in plasma NE at NEthreshold, 20% pressor and pressor dose. Spearman's rank correlation coefficients: control subjectsr =0.93, p<0.001; normoalbuminuric IDDM patients r =0.91, p<0.001 and microalbuminuric IDDMs s

patients r =0.88, p<0.001. The logistic regression lines are constructed by curve fitting. There were nos

differences in the relationships. C. Actual changes in MAP vschanges in plasma NE at T, 20% P and P dose during NE infusions. Data in geometric means ±antilog SEM. ! control subjects (group C), Q normoalbuminuric (group D1) and Îmicroalbuminuric(group D2) IDDM patients; * denotes p<0.05 from normo- and microalbuminuric IDDM.

Chapter 470

During placebo infusion, serum ACE activity and plasma angiotensin II were not differentamong the groups (Table 3). During NE pressor infusion, plasma AII levels increasedsimilarly in all groups (p<0.05 for all).

Urinary excretion of albumin, IgG and $2-microglobulin.

Urinary albumin and IgG excretion rates (Ualb.V and UIgG.V) were higher in groupD2 than in groups C and D1 at all observation periods (p<0.01, Table 4). During theplacebo day UIgG.V fell in all groups (p<0.05 to 0.01). Ualb.V also fell significantlyduring placebo in group D2 (p<0.05 to 0.01) with a tendency to fall in the other groups aswell (Table 4). Interestingly, during NE pressor infusion, Ualb.V and UIgG.V rosesignificantly in all groups (p<0.05 for all, Table 4) and some of these parameters were stillelevated at recovery in group C and D1. The increments in Ualb.V and UigG.V (correctedfor placebo) were more pronounced in group D2 (31.3 (11.8-83.1) and 4.0 (2.0-8.1)µg/min, respectively) than in group C (4.2 (1.3-13.7) and 0.6 (0.2-2.8) µg/min,respectively) and D1 (3.8 (1.4-10.1) and 0.6 (0.2-2.0) µg/min, respectively) (p<0.05, Table4 and Figure 2). Fractional albumin and IgG excretion also increased during NE pressorinfusion in all groups (p<0.05 to 0.01, Table 4). $2-Microglobulin excretion was notdifferent in the groups during placebo and remained stable during NE infusion (Table 4,Figure 2).

Multiple stepwise regression analysis was carried out to evaluate which parameterscontributed independently to the changes in microproteinuria and renal haemodynamicsin response to exogenous NE. Baseline Ualb.V, change in MAP and change in plasma NEindependently contributed to 37% (p<0.001), 8% (p<0.01) and to 6% (p<0.05) of thevariance in change in Ualb.V (multiple r=0.56, 72 data sets). The variance in change inUIgG.V resulted for 27% from baseline UIgG.V (p<0.001), for 7% from the change inMAP (p<0.02) and for 5% from the change in plasma NE (p<0.05, multiple r=0.48, 72data sets). No independent effects of plasma insulin, blood glucose and angiotensin II onmicroproteinuria were demonstrated (p>0.20, for all).

When the relative increments in Ualb.V and UIgG.V at NE pressor infusion werecompared, no differences in the responses were observed between the 3 groups (% incre-ment in Ualb.V and UIgG.V at NE pressor dose in group C: 233(122-344)% and 233(104-361)% (geometric mean and 95% confidence interval), in group D1 202(127-278)% and154(115-193)%, in group D2 166(133-202)% and 166(117-214)%, respectively). Further-more, multiple regression analysis showed that the relative changes in Ualb.V and UIgG.Vwere not independently related to baseline microproteinuria (p>0.20 for both). Thus, theextent of baseline glomerular protein leakage was the main determinant of the absolutemicroproteinuric response to NE.

Renal haemodynamic studies

During the placebo day, GFR and ERPF were not significantly different between the3 groups, whereas FF was slightly but significantly higher in group D2 compared to theother groups (p<0.05). Glomerular hyperfiltration, defined as GFR>130 ml/min per 1.73m [9], was present in 2 patients of group D1 and 3 patients of group D2. GFR remained2

essentially unaltered in the groups during NE infusion (Table 5).

Norepinephrine, microproteinuria and renal haemodynamics 71

Table 4. Urinary excretions of albumin, IgG and $2-microglobulin during norepinephrine (NE) and placebo infusions.

Non-diabetic Normoalbuminuric MicroalbuminuricSubjects (n=8) IDDM patients (n=8) IDDM patients (n=8)

Group C Group D1 Group D2

Period NE Placebo NE Placebo NE Placebo

Urinary albumin excretion rate (µg/min)a

B 4.6(3.5-6.2) 5.6(3.5-7.6) 4.2(2.9- 6.0) 5.6(3.4-9.3) 47.9(23.2-98.8) 52.2(22.3-122.2)T 4.7(2.9-7.8) 4.6(3.2-6.6) 4.3(2.9- 6.6) 5.1(3.1-8.3) 47.6(23.9-94.7) 38.9(16.2-93.4)S 5.5(3.5-8.8) 4.3(2.9-6.3) 5.6(3.7- 8.6) 5.2(3.2-8.7) 50.3(24.9-101.5) 36.8(14.8-91.8)

b

P 9.7(4.6-20.2) 4.5(3.1-6.7) 7.3(4.8-11.2) 4.6(2.8-7.4) 58.5(33.6-101.7) 33.8(14.9-72.2)e e d,f c

R 8.2(4.0-16.7) 4.6(3.1-6.9) 7.4(5.1-11.0) 4.1(2.6-6.4) 43.7(24.4-78.1) 31.3(15.2-64.6)d e b c

Fractional albumin clearance(10 )-6 a

B 1.1(0.8-1.4) 1.1(0.8-1.5) 1.0(0.8-1.4) 1.2(0.8-1.8) 11.7(4.9-28.1) 12.8(4.8-34.2) T 1.0(0.7-1.5) 1.1(0.8-1.4) 1.0(0.8-1.4) 1.2(0.9-1.8) 11.5(4.8-27.9) 9.6(3.2-28.8)S 1.3(0.9-1.8) 1.0(0.7-1.3) 1.3(0.9-1.8) 1.2(0.8-1.9) 12.4(5.3-29.1) 8.6(2.9-25.4)P 2.2(1.1-4.3) 1.0(0.8-1.4) 1.7(1.2-2.6) 1.1(0.7-1.7) 15.3(7.4-31.7) 7.9(3.0-21.0)

e d d,f b

R 1.9(1.0-3.4) 1.0(0.7-1.5) 1.9(1.3-2.6) 1.0(0.6-1.4) 11.7(5.6-24.3) 7.7(3.0-19.9)e c

Urinary IgG excretion rate(µg/min)a

B 1.1(0.9-1.3) 1.1(0.9-1.4) 1.2(0.8-1.6) 1.4(1.1-1.9) 4.4(3.0-6.6) 4.9(2.4-9.9) T 1.1(0.8-1.5) 1.0(0.7-1.3) 1.2(0.8-1.7) 1.1(0.7-1.8) 4.7(3.2-6.7) 4.1(2.4-6.9)S 1.1(0.8-1.6) 1.0(0.7-1.2) 1.3(0.9-1.1) 1.2(0.9-1.7) 5.1(3.3-7.7) 3.6(2.3-5.8)

b b

P 1.9(0.9-4.0) 0.9(0.7-1.2) 1.4(0.8-2.5) 1.1(0.9-1.4) 5.7(4.4-7.3) 3.2(2.0-5.1)e b d c d,f c

R 1.4(0.8-2.8) 0.9(0.7-1.2) 1.3(0.8-2.1) 1.0(0.7-1.5) 3.8(2.2-6.5) 3.2(2.3-4.6)b d c c

Fractional IgG clearance(10 )-6 a

B 1.3(1.1-1.7) 1.3(1.1-1.5) 1.2(1.0-1.4) 1.3(1.1-1.4) 4.9(2.7-9.2) 5.2(2.2-12.5)T 1.3(0.9-1.7) 1.2(0.9-1.5) 1.2(1.0-1.4) 1.1(0.9-1.3) 5.1(2.8-9.3) 4.3(1.9-9.5)S 1.4(1.0-2.0) 1.1(0.9-1.4) 1.2(0.9-1.6) 1.1(0.9-1.3) 5.5(3.1-9.5) 3.6(1.7-7.5)P 2.4(1.2-5.6) 1.1(0.9-1.3) 1.5(1.0-2.2) 1.0(0.9-1.2) 6.7(4.2-10.8) 3.2(1.5-6.9)

d d b d,f c

R 1.8(0.9-3.6) 1.0(0.8-1.3) 1.3(0.8-1.9) 0.9(0.7-1.3) 4.4(2.2-9.0) 3.3(1.6-6.7)b b b

$2-Microglobulin excretion(µg/min)

B 68(53-87) 72(43-122) 62(43-89) 59(43-82) 94(53-167) 88(48-163)T 53(31-91) 65(45-93) 52(36-75) 52(38-72) 89(54-145) 84(53-133)S 34(20-56) 63(36-112) 43(30-62) 50(37-68) 76(45-128) 73(48-110)P 37(24-58) 61(37-101) 39(26-57) 46(33-66) 63(33-119) 66(39-113)R 38(22-65) 57(37-88) 30(22-42) 44(29-66) 38(18-80) 69(47-100)

Values are given as geometric mean and 95% confidence interval. B: baseline; T: threshold; S: 20%pressor; P: pressor; R: recovery. All values in group D2>group C and D1 (p<0.01); p<0.05, p<0.01

a b c

from baseline(decline at placebo day); p<0.05, p<0.01 from baseline(increase during NE correctedd e

for placebo day); p<0.05 from groups C and D1 (absolute increase).f

2-m

icro

glob

ulin

(g

/min

)

1 2 3 4 5 10 20 30 40 50 100plasma NE (nmol/l)

150

75

100

125

50

25

UIg

G.V

(g

/min

)

10.0

5.0

7.5

2.5

0.0

Ual

b.V

(g

/min

)

10

15

30

60

90

120

5

0

Chapter 472

Figure 2. Relationship between plasma norepinephrine (NE) and A. urinary albumin (Ualb.V), B. IGG (UIgI.V) and C. $2-microglobulin excretion. Data in geometric mean ±antilog SEM, valuescorrected for placebo day. ! control subjects (group C), Q normoalbuminuric (group D1) and Îmicroalbuminuric (group D2) IDDM patients; * denotes p<0.05 and ** p<0.01 from baseline; *** p<0.05 increase in group D2 > group C and D1.

Norepinephrine, microproteinuria and renal haemodynamics 73

Table 5. Renal haemodynamics during norepinephrine (NE) and placebo infusions.

Non-diabetic Normoalbuminuric MicroalbuminuricSubjects (n=8) IDDM patients (n=8) IDDM patients (n=8)

Group C Group D1 Group D2

Period NE Placebo NE Placebo NE Placebo

Glomerular Filtration Rate (ml/min per 1.73m2)

Baseline 103±5 103±4 105±5 107±6 104±8 108±8 Threshold 106±6 105±5 109±5 109±8 108±9 109±9 20%Pressor 104±5 103±4 109±6 111±8 108±9 111±8 Pressor 103±4 103±4 108±5 108±7 106±19 109±8 Recovery 92±5 102±3 94±5 107±8 95±8 108±8

c d d

Effective Renal Plasma Flow (ml/min per 1.73m2)a

Baseline 458±24 461±17 451±22 464±23 442±40 435±38Threshold 399±17 454±16 411±21 467±24 408±39 435±38 c d e

20%Pressor 368±14 452±16 385±26 460±26 385±35 438±35 e e e

Pressor 303±15 446±13 323±20 453±23 308±22 432±36 e e e

Recovery 441±39 434±8 396±17 457±24 386±32 428±38e d

Filtration Fraction (%)a,b

Baseline 21.8±0.7 22.0±0.9 23.4±0.7 23.0±0.8 23.5±0.8 24.6±0.8Threshold 26.8±1.1 23.1±1.0 26.7±1.1 23.0±0.9 26.9±1.0 25.4±1.0 c d c

20%Pressor 28.2±1.0 22.9±0.8 28.7±1.2 23.7±0.7 28.4±1.0 25.6±0.8d e d

Pressor 34.4±1.7 23.2±0.9 33.8±1.3 23.6±0.7 34.2±1.0 25.3±0.8e e e

Recovery 21.3±1.0 23.5±0.7 23.8±0.5 23.8±0.5 24.6±1.0 25.4±0.8

Values are given in mean±SEM. Overall responses at threshold, 20% pressor and pressor perioda

in group C >groups D1 and D2 (p<0.05); p<0.05 from group C and D1 at placebo day;b

p<0.05, p<0.01 and p<0.001 from baseline, after correction for placebo day.c d e

ERPF progressively decreased in all groups with increasing NE infusion rates(p<0.05 to 0.001). As a result, FF rose significantly with all NE infusion steps (Table 5).The overall changes in ERPF and FF during NE infusions were more pronounced in groupC compared to the diabetic groups (p<0.05 and p<0.01, respectively). In each group, thechanges in ERPF and FF were strongly correlated with the changes in plasma NE levels(group C: r =-0.82 and r =0.85, n=24, p<0.001; group D1: r =-0.87 and r =0.85, n=24,s s s s

p<0.001; group D2: r =-0.76 and r =0.83, n=24, p<0.001, respectively, but no differencess s

in the relationships between the 3 groups were observed (Figure 3A-B). A similarrelationship was observed with NE infusion dose (data not shown). Thus, for a givenincrease in plasma NE level the changes in ERPF and FF were similar in group D1 and D2compared to group C and the differences in NE dose fully accounted for the betweengroup differences in ERPF and FF responses. These relationships were not confounded byvariation in plasma insulin and blood glucose (data not shown).

0.1 1 10 100

50

-50

-100

-150

-200

-250

-300

0

change in plasma NE (nmol/l)

group C

group D2group D1

A

0.1 1 10 100

25

20

15

10

5

0

-5

change in plasma NE (nmol/l)

chan

ge

in F

F (%

)

group D2

group D1

group C

B

Ech

ange

inR

PF

(mm

m)

2in

per

173.

l/

Chapter 474

Figure 3.A. Relationships between changes in plasma norepinephrine (NE) and change in effectiverenal plasma flow (ERPF), and B. and filtration fraction (FF) following exogenous NE infusions. Thelogistic regression lines are constructed by curve fitting (3 data sets per subject). No differencein ERPF and FF responses between the groups were demonstrated. ! control subjects (group C), R normoalbuminuric (group D1) and Î microalbuminuric (group D2) IDDM patients.

Norepinephrine, microproteinuria and renal haemodynamics 75

Discussion

The present study is the first to document that exogenous NE increasesmicroproteinuria in conjunction with a systemic blood pressure rise and renalvasoconstriction in normo-and microalbuminuric IDDM patients and healthy subjects. Thelack of effect on urinary $2-microglobulin indicates that NE did not affect tubular proteinhandling. The rises in absolute and fractional urinary albumin and IgG excretion,therefore, support the notion that NE increases glomerular protein leakage. Furthermore,in absolute terms the microalbuminuric response was more pronounced in micro-albuminuric IDDM patients than in normoalbuminuric IDDM patients and healthycontrols. By contrast, the renal haemodynamic responses to NE were similar in the 3groups, as indicated by similar relationships between plasma NE increments and renalhaemodynamic responses.

As systemic blood pressure is an important determinant of microproteinuria inIDDM [4,5,18,37], this study was designed to evaluate the microproteinuric and renalhaemodynamic effects of NE given in doses that induced predetermined blood pressureincrements. We were thus able to avoid confounding effects of differences in bloodpressure responses to NE. The enhanced blood pressure responsiveness in the diabeticgroups is in keeping with earlier reports [21-23] and underscores the relevance of thisapproach. The less pronounced decline in heart rate in the IDDM patients, which mayreflect impaired baroreflex function, could have contributed to this enhanced systemic NEresponsiveness [38].

The prevailing plasma insulin and blood glucose levels can affect systemic vascularNE responsiveness, as well as ERPF and GFR [30,31,39-41]. Exogenous insulin acutelyelevates renal plasma flow at levels of 90 mU/L, but not significantly so at levels of 30 to40 mU/l [40,41], and rises in blood glucose from normal to moderately elevated levelsmodestly increases GFR [30]. In our study, plasma insulin was kept constant during NEinfusion in all groups. Its level was higher in the IDDM patients (approximately 30 mU/l)than in the control subjects (approximately 10 mU/l). This difference was intentionallyallowed since plasma insulin levels of 30 mU/L are normally encountered in IDDM, butnon-physiologically high in fasting healthy subjects [13]. The normoglycaemiccircumstances allowed evaluation of the NE effects unaffected by the prevailing bloodglucose. The slight increase in blood glucose during NE infusion in the control subjects isvery unlikely to be of relevance in the interpretation of the NE effects, as supported by thelack of effect of this parameter in statistical analyses.

The similar regression lines between NE doses and venous plasma NE incrementsin all groups, is in keeping with an unaltered plasma NE clearance in diabetic patients[42]. Although venous antecubital NE measurement has the disadvantage of beingconfounded by forearm NE spillover and extraction [43], it is thus improbable that theinterpretation of our results was biassed by using venous plasma NE levels. Yet, our studydoes not allow conclusions to potential differences in intrarenal NE handling between(microalbuminuric) IDDM patients and control subjects.

Based on observations in hand vein studies, which like renal arterioles also contain"-adrenoceptors, it has been suggested that an exaggerated renal vascular responsivenessto NE might be involved in elevations in albumin excretion rate in IDDM [20]. In this

Chapter 476

study, we found no evidence for an altered renal vascular NE responsiveness in normo-and microalbuminuric IDDM as indicated by similar the relationships between incrementsin plasma NE and renal haemodynamic changes in the 3 groups. In comparison, thediabetic state has been variably associated with either an increase, no change or a decreasein vascular reactivity to vasoconstrictive agents [44]. These discrepancies may beattributed to differences in the species investigated, in the blood glucose and insulin levelsduring the experiments, or the vascular bed under study.

Posture and diurnal rhythm affect glomerular protein leakage. Our study clearlydemonstrates that urinary protein excretion falls under conditions of supine rest. Interest-ingly, for albumin this fall was particularly apparent in the microalbuminuric patients.This indicates that it is mandatory to include time-control data to properly evaluate theeffects of acute interventions on microalbuminuria. The previous reported lack of effectof NE to increase microalbuminuria in healthy subjects may thus have been due to theuncontrolled data use in that study [17]. Two recent studies report that exogenous angio-tensin II does not influence microalbuminuria in non-diabetic [45] and diabetic subjects[46], in spite of a clear-cut effect on renal haemodynamics. This could indicate that lowdoses of exogenous angiotensin II, unlike our findings with NE, are indeed unable toinduce a rise in albuminuria in man. On the other hand, the absence of time-control datain those studies also hampers a direct comparison with our data.

The microproteinuria promoting effect of NE may have arisen from its systemicblood pressure elevating effects as well as from an intrarenal effect, like a rise in glomer-ular pressure. Animal experiments have shown that NE, irrespective of systemic bloodpressure, induces a rise in glomerular pressure [12], thereby promoting glomerular proteinleakage. Although glomerular pressure can not be measured directly in man, our findingsof a well-preserved GFR despite a pronounced decrease in ERPF during NE are stronglysuggestive for a rise in filtration pressure. To identify the determinants of thismicroproteinuric response, we performed multiple regression analysis. This analysisrevealed blood pressure to be an independent determinant of the microproteinuricresponse, suggesting that the rise in blood pressure is causally involved in the rise inmicroproteinuria. Microalbuminuria was also independently related to the plasma NElevel, indeed suggesting a blood pressure independent effect of NE as well.

The exaggerated rise in microproteinuria in microalbuminuric IDDM could not beexplained by an altered blood pressure response or renal haemodynamic responsiveness toNE nor by a difference in NE-induced angiotensin II release. Instead, the baseline level ofmicroproteinuria appeared to be the main determinant of the microproteinuric response toNE across the groups. In accordance, no between group differences in relative incrementsin urinary albumin and IgG excretion were seen. This underscores that the preexistentdefect in glomerular perm selectivity is crucial in the absolute microproteinuric effect ofNE. Our data do not allow a conclusion whether such an effect is specific for NE, oralternatively, whether any proteinuria-promoting manoeuver is inclined to induce a largerresponse when a defect in glomerular capillary permselectivity is already present. Theenhanced microproteinuric response was found both for the negatively charged albuminand for the much larger and predominantly neutrally charged IgG, raising the possibilitythat a loss of anionic charge of the glomerular basement membrane is not the onlyabnormality at this early stage of diabetic nephropathy [47].

Norepinephrine, microproteinuria and renal haemodynamics 77

What is the clinical relevance of our findings? The present results are consistent withthe concept that NE contributes to a rise in microproteinuria as for instance observedduring strenuous exercise [18,19]. The larger absolute response in micro-albuminuricIDDM may implicate that this phenomenon contributes to persistent microproteinuricglomerular protein leakage. As glomerular protein leakage is presumed to be involved inthe progression of diabetic nephropathy [48,49], such small but repetitive increases inglomerular protein leakage could be involved in progressive renal damage in the long run.

In conclusion, exogenous NE causes a microproteinuric response in association witha rise in blood pressure, renal vasoconstriction and possibly intraglomerular hypertension.In microalbuminuric IDDM, this microproteinuric response is enhanced, which mayreflect either a specific response to NE or a general effect of vasopressor stimuli topromote glomerular protein leakage in patients with a preexistent defect in glomerularpermselectivity.

References

1 Mogensen CE. Prediction of clinical diabetic nephropathy in IDDM patients. Alternatives tomicroalbuminuria? Diabetes 1990; 39: 761-767

2 Mogensen CE, Christensen CK, Vittinghus E. The stages in diabetic renal disease with emphasison the stage of incipient diabetic nephropathy. Diabetes 1983; 32 [Suppl 2]: 64-78

3 Viberti CG, Keen H. The patterns of proteinuria in diabetes mellitus. Relevance to pathogenesisand prevention of diabetic nephropathy. Diabetes 1984; 33: 686-692

4 Parving HH, Andersen AR, Smidt UM, Svendsen PA.Early aggressive antihypertensive treat-ment reduces rate of decline in kidney function in diabetic nephropathy. Lancet 1983; i: 1175-1179

5 Mogensen CE, Hansen KW, Østerby R, Damsgaard EM. Blood pressure elevation versusabnormal albuminuria in the genesis and prediction of renal disease in diabetes. Diabetes Care1992; 15: 1192-1204

6 Hostetter TH, Rennke HG, Brenner BM. The case for intrarenal hypertension in the initiation andprogression of diabetic and other glomerulopathies. Am J Med 1982; 72: 375-380

7 Viberti GC, Wiseman MJ. The kidney in diabetes mellitus: significance of the early abnormalities.Clin Endocrinol Metab 1986; 15: 753-782

8 Bank N: Mechanisms of diabetic hyperfiltration. Kidney Int 1991; 40: 792-8079 Hoogenberg K, Dullaart RPF, Freling NJM, Meijer S, Sluiter WJ. Contributory roles of glycated

haemoglobin, circulatory glucagon and growth hormone to increased renal haemodynamics inType(insulin-dependent) diabetes mellitus. Scand J Clin Lab Invest 1993; 53: 821-828

10 Hoogenberg K, Ter Wee PM, Lieverse AG, Sluiter WJ, Dullaart RPF. Insulin-like growth factorand altered renal haemodynamics in growth hormone deficiency, acromegaly, and Type 1diabetes mellitus. Transpl Proc 1994; 26: 505-507

11 Maddox DA, Brenner BM. Glomerular ultrafiltration. In: The kidney, 5th ed., edited by BrennerBM, Philadelphia, WB. Saunders, 1996, pp 286-333

12 Meyers BD, Deen WM, Brenner BM. Effects of norepinephrine and angiotensin II on thedeterminants of glomerular ultrafiltration and proximal tubule fluid reabsorption in the rat. CircRes 1975; 37: 101-110

13 Hoogenberg K, Girbes ARJ, Stegeman CA, Sluiter WJ, Reitsma WD, Dullaart RFP. Influence ofambient plasma norepinephrine on renal haemodynamics in Type 1 (insulin-dependent) diabeticpatients and healthy subjects. Scand J Lab Invest 1995; 55: 15-22

14 Smythe McC, Nicke JF, Bradley SE. The effect of epinephrine (usp), l-epinephrine, and l-norepinephrine on glomerular filtration rate, renal plasma flow and the urinary excretion of

Chapter 478

sodium, potassium and water in normal man. J Clin Invest 1952; 31: 499-50615 Lathem W: Renal circulatory dynamics and urinary protein excretion during infusions of l-

norepinephrine and l-epinephrine in patients with renal disease. J Clin Invest 1956;35:1277-128516 Lang CC, Rahman AR, Balfour DJK, Struthers AD. Effect of norepinephrine on renal sodium and

water handling in euhydrated and overhydrated man. Clin Sci 1993; 85: 487-49417 Mogensen CE, Hansen KW, Sommer S, Klebe J, Christensen CK, Marshall S, Schmidz A,

Pedersen EB, Christiansen JS, Bjerregaard Pedersen E. Microalbuminuria: studies in diabetes,essential hypertension, and renal disease as compared with the background population. AdvNephrol Necker Hosp 1991; 20: 191-228

18 Hoogenberg K, Dullaart RPF. Abnormal plasma noradrenaline response and exercise inducedalbuminuria in type 1 (insulin-dependent diabetes mellitus. Scand J Lab Invest 1992;52: 803-811

19 Poortmans JR: Postexercise proteinuria in humans. JAMA 1982; 253: 236-24020 Bodmer CW, Patrick AW, How TV, Williams G. Exaggerated sensitivity to NE-induced

vasoconstriction in IDDM patients with microalbuminuria. Possible etiology and diagnosticimplications. Diabetes 1992; 41: 209-214

21 Beretta-Piccoli C, Weidmann P. Exaggerated pressor responsiveness to norepinephrine innonazotemic diabetes mellitus. Am J Med 1981; 71: 829-835

22 Weidmann P, Beretta-Piccoli C, Trost B. Pressor factors and responsiveness in hypertensionacompanying diabetes mellitus. Hypertension 1985; 7 [Suppl 2]: 33-42

23 Weidmann P, Beretta-Piccoli C, Keusch G, Glück Z, Mujagic M, grimm M, Meier A, ZieglerWH. Sodium-volume factor, cardiovascular reactivity and hypotensive mechanism of diuretictherapy in mild hypertension associated with diabetes mellitus. Am J Med 1979; 67: 779-784

24 Ewing DJ, Clarke BF. Diagnosis and management of diabetic autonomic neuropathy. Br Med J1982; 285: 916-918

25 Dullaart, RPF, Roelse H, Sluiter WJ, Doorenbos H. Variability of albumin excretion in diabetes.Neth J Med 1989; 34: 287-296

26 Dimsdale JE, Graham RM, Ziegler MG, Zusman RM, Berry CC. Age, race, diagnosis, andsodium effects on the pressor response to infused norepinephrine. Hypertension 1987; 10: 564-569

27 Bosch JP, Lauer A, Glabman S: Short-term protein loading in assessment of patients with renaldisease. Am J Med 1984; 77: 873-879

28 Maroni BJ, Steinman TI, Mitch WE. A method for estimating nitrogen intake of patients withchronic renal failure. Kidney Int 1985; 27: 58-65

29 De Fronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulinsecretion and resistance. Am J Physiol 1979; 237: E214-E223

30 Dullaart RPF, Meijer S, Sluiter WJ, Doorenbos H. Renal haemodynamic changes in response tomoderate hyperglycaemia in type 1(insulin-dependent) diabetes mellitus. Eur J Clin Invest 1990;20: 208-213

31 Gans ROB, Bilo HJG, Van Maarschalkerweerd WWA, Heine RJ, Nauta JJP, Donker AJM.Exogenous insulin augments in healthy volunteers the cardiovascular reactivity to norepinephrinebut not angiotensin II. J Clin Invest 1991; 88: 512-518

32 Grimm M, Weidmann P, Keusch G, Glück Z. Norepinephrine clearance and pressor effect innormal and hypertensive man. Klin Wochenschr 1980; 58: 1175-1181

33 Viberti GC, Mogensen CE, Keen H, Jacobsen FK, Jarrett RJ, Christesen CK.Urinary excretion ofalbumin in normal man: the effect of water loading. Scand J Clin Invest 1982; 42:147-151

34 Apperloo, AJ, De Zeeuw, Donker AJM, De Jong PE. Precision of glomerular filtration ratedeterminations for long-tem slope calculations is improved by simultaneous infusions of I-125

iothalamate and I-hippuran. J Am Soc Nephrol 1996; 7: 567- 672131

35 Hoogenberg K, Visser P, Marrink J, Sluiter WJ, Dullaart RPF. Increased urinary IgG/albuminindex in normoalbuminuric IDDM: a laboratory artefact. Diabetic Med 1996; 13: 651-655

Norepinephrine, microproteinuria and renal haemodynamics 79

36 Suzuki H, Ferrario CM, Speth RC, Broshnikan KB, Smeby RR, De Silva P. Alterations in plasmaand cerebrospinal fluid norepinephrine and angiotensin II during the development of renalhypertension in concious dogs. Hypertension 1983; 5: S139-S148

37 Eckberg DL, Harkins SW, Fritsch JM, Musgrave GE, Gradner DF. Baroreflex control of plasmanorepinephrine and heart period in healthy subjects and diabetic patients. J Clin Invest 1986;78:366-374

38 Dullaart RPF, Beusekamp BJ, Meijer S, Hoogenberg K, Van Doormaal JJ, Sluiter WJ. Long-termeffects of protein-restricted diet on albuminuria and renal function in IDDM patient withoutclinical nephropathy and hypertension. Diabetes Care 1993; 16: 483-492

39 McNally PG, Lawrence IG, Watt PAC, Hillier C, Burden AC, Thurston H. The effect of insulinon vascular reactivity of isolated resistance arteries taken from healthy volunteers. Diabetologia1995; 38: 467-473

40 Stenvinkel P, Bolinder J, Alvestrand A. Effects of insulin on renal haemodynamics and theproximal and distal sodium handling in healthy volunteers. Diabetologia 1992; 35: 1042-1048

41 Pelikánová T, Smrcková I, Krízová J, Stríbrná J, Lánská V. Effect of insulin and lipid emulsionon renal haemodynamics and renal sodium handling in IDDM patients. Diabetologia 1996; 39:1074-1082

42 Dejgaard A, Hilsted J, Christensen NJ. Norepinephrine and isoproterenol kinetics in diabeticpatients with and without autonomic neuropathy. Diabetologia 1986; 29: 773-777

43 Hjemdahl P: Plasma catecholamines - analytical challenges and physiological limitations. ClinEndocrinol Metab 1993; 7: 307-353

44 Tomlinson KCSM., Gardiner SM, Hebden RA, Bennett T. Functional consequences ofstreptozotocin-induced diabetes mellitus, with particular reference to the cardiovascular system.Pharmacol Rev 1992; 44: 103-150

45 Erley CM, Grau C, Furian Tch. Subpressor doses of angiotensin II do not increase albuminexcretion in humans. Clin Nephrol 1996; 46: 312-318

46 Spooren PFMJ, Gans ROB, Adèr HJ, Vermes I, Donker AJM. Urinary albumin excretion rateduring angiotensin II infusions in microalbuminuric patients with insulin and non-insulindependent diabetes mellitus. Nephrol Dial Transplant 1997; 12: 281-285

47 Deckert T, Feldt-Rasmussen B, Djurup R, DeckertM. Glomerular size and charge selectivity ininsulin-dependent diabetes mellitus. Kidney Int 1988; 33: 100-106

48 Remuzzi G, Bertain T: Is glomerulosclerosis a consequence of altered glomerular permeability tomacromolecules? [Editorial] Kidney Int 1990; 38: 384-394

49 Navis GJ, De Jong PE, De Zeeuw D: A comparison of progression in diabetic and non-diabeticrenal disease: similarity of progression promoters. In: The kidney in hypertension and diabetesmellitus, 3rd ed., edited by Mogensen CE, Boston/Dordrecht/London, Kluwer 1996, pp 551-560

Acknowledgments

This study was supported by a grant from the Dutch Diabetes Fund (Grant No.94130). We are indebted to A. Drent-Bremer and M. van Kammen for their skilfulassistance in the renal haemodynamic studies. The analytical help of W. de Jonge, P.Visser, H.J. de Jong, H. Breukelman and E. Venekamp-Hoolsema is appreciated. Dr. R.Henning is thanked for his help with the curve fitting models.

Department of Endocrinology and Nephrology, Groningen University Hospital,1 2

The Netherlands. Nephrol Dial Tranplant (in press)

CHAPTER 5

NOREPINEPHRINE-INDUCED BLOOD PRESSURE RISEAND RENAL VASOCONSTRICTION IS NOT ATTENUATED

BY ENALAPRIL IN MICROALBUMINURIC IDDM

K. Hoogenberg , G. Navis and R.P.F. Dullaart1 2 1

In non-diabetic subjects an attenuated systemic norepinephrine (NE) responsiveness maycontribute to the mechanisms of action of angiotensin-converting enzyme (ACE) inhibitor treatment.We determined whether ACE-inhibitor treatment influences systemic and renal haemodynamicresponsiveness to exogenous NE, as well as urinary albumin excretion during NE, in micro-albuminur-ic IDDM patients, representing a patient category that benefits by strict blood pressure control. In 7microalbuminuric IDDM patients, systemic and renal responsiveness to NE, infused at individuallydetermined threshold (Îmean arterial pressure (MAP)=0mmHg), 20% pressor (ÎMAP =4mmHg) andpressor (ÎMAP=20mmHg) doses, were compared before and after 8 weeks treatment with enalapril,10 mg daily. Blood glucose was clamped at 5 mmol/l and insulin was infused at 30 mU/kg per h.Enalapril decreased MAP (p<0.05) and microalbuminuria (p<0.05), whereas effective renal plasmaflow (ERPF) increased (p<0.01) and glomerular filtration rate (GFR) remained unaltered. Filtrationfraction tended to decline (p=0.09). The ACE inhibitor-induced fall in MAP disappeared at NE pressordose, and the overall mean increase in MAP in response to NE was even higher with than withoutenalapril (p<0.05). After enalapril, ERPF remained higher at all NE doses (p<0.05), but the magnitudeof the NE-induced fall in ERPF was not altered by ACE inhibition treatment. Overnight urinaryalbumin excretion fell by ACE-inhibition (p<0.05), but this effect was not seen during NE infusion.The angiotensin II/active renin ratio and serum aldosterone levels remained lower with enalapril at allNE doses (p<0.05). In conclusion, enalapril does not attenuate systemic and renal vascularresponsiveness to exogenous NE in microalbuminuric IDDM despite adequate inhibition of the renin-angiotensin-aldosterone system. These findings suggest that the effect of NE on vasoconstriction is noteffectively counteracted by ACE inhibition treatment alone.

Introduction

Blockade of the renin-angiotensin-aldosterone system (RAAS) by angiotensin-con-verting enzyme (ACE) inhibitors lowers blood pressure, (micro)albuminuria and slowsprogression of diabetic and non-diabetic renal insufficiency, although such treatment isunable to completely prevent this serious complication [1-3]. The mechanism of action ofthese agents is still incompletely understood. Both angiotensin II-dependent andindependent mechanisms are thought to be involved. The sympathetic nervous system(SNS) and RAAS are closely related [4] and ACE inhibition may influence systemicnorepinephrine (NE) reactivity. Experimental studies indeed have shown that systemic

Chapter 582

vascular NE responsiveness is attenuated by ACE inhibitors [5,6]. A decrease in systemicNE responsiveness after ACE inhibition treatment has been demonstrated in hypertensivesubjects [7-9]. Moreover, in healthy subjects the renal vasoconstrictive response has beenfound to be reduced after ACE inhibition [10]. This would suggests that attenuation ofrenal NE responsiveness may be involved in the renoprotective effect of ACE inhibitorsin non-diabetic subjects.

In insulin-dependent and non-insulin-dependent diabetes mellitus (IDDM andNIDDM), the systemic vasopressor response to exogenous NE has been repeatedly foundto be enhanced [11-13]. The effect of ACE inhibition on systemic and renal NE respon-siveness is less well characterised in diabetes mellitus. In patients with NIDDM and in aheterogeneous group of NIDDM and IDDM patients, treatment with enalapril or captoprildid not normalise the exaggerated systemic NE responsiveness [14,15], but this has notbeen studied in IDDM complicated by microalbuminuria. It is of particular relevance toevaluate the NE responsiveness after ACE-inhibition treatment in microalbuminuricIDDM patients, because control of factors that contribute to a blood pressure rise in suchpatients appears to be indicated to prevent progression to overt nephropathy. The purposeof our study was, therefore, to evaluate the effect of enalapril treatment on systemic andrenal haemodynamic responsiveness to exogenous NE in IDDM patients withmicroalbuminuria.

Subjects and methods

Subjects

The study was approved by the local medical ethics committee and all participantsgave written informed consent. The patients were considered insulin-dependent, andglucagon stimulated C-peptide levels were less than 0.2 nmol/l. Seven patients (6 men, 1women) participated. They had a mean age of 48±9 (mean±SD) years and a diabetesduration of 29±5 years. None had severe obesity (body mass index ranging from 23.5 to27.9 kg/m ). All of the patients had persistent microalbuminuria (defined as an urinary2

albumin excretion rate (Ualb.V) between 20 to 200 µg/min in at least two out of threeovernight urine collections for a 1-year period). All patients were treated with enalapril forelevated blood pressure related to the presence of incipient nephropathy (systolic/ diastolicblood pressure >140/90 mmHg), but none had essential hypertension before developmentof microalbuminuria. ACE inhibition treatment was stopped for 6 weeks before the startof the study. Without enalapril, 6 patients had a systolic blood pressure > 140 mmHg and1 patient had a diastolic blood pressure >90 mmHg. None of the patients had untreatedproliferative retinopathy, symptomatic coronary heart disease, peripheral vascular disease,or clinical autonomic neuropathy as evaluated by standard tests (beat-to-beat variationduring deep breathing, Valsalva manoeuvre and systolic blood pressure response tostanding). All participants were instructed by dietician to adhere to a diet containing 100mmol sodium and 1 gram protein/kilogram body weight per day throughout the studies.Three timed overnight urine collections were obtained directly prior to the studies.

Enalapril and norepinephrine reactivity in IDDM 83

Table 1.Baseline characteristics before and after 8 weeks treatment with enalapril.

before ACEi with ACEi

HbA1c (%) 8.4 ±0.2 8.5 ± 0.2

Body weight (kg) 83.2 ± 2.7 83.3 ±2.7

Urinary urea excretion (mmol/day) 279 ±26 282 ± 16

Sodium excretion (mmol/day) 110 ±8 123 ± 10

Mean of 3-timed overnight urinary 71.4 (33.5 - 152.2) 50.9 (36.8 - 97.2)a

albumin excretion rates (µg/min)

Data in mean±SEM, except for urinary albumin excretion rate which is geometric mean (95%confidence interval). ACEi: angiotensin converting enzyme inhibition. denotes p<0.05

a

Study design

NE-infusions were performed on three separate study days. On the first day,systemic MAP-NE dose response curves were made. The individual NE threshold (Îmeanarterial pressure (MAP)=0mmHg), 20% pressor (ÎMAP=4mmHg) and pressor doses(ÎMAP=20mmHg) were calculated from MAP-NE dose response curves. On a separateday, the stepwise NE infusions were given with renal function measurements during 10consecutive 45 min clearance periods. Two baseline periods were followed by 6 periodsof NE (2 periods for threshold, 20% pressor and pressor doses, respectively), after which2 determinations were obtained after cessation of the infusion (recovery period). Bloodpressure was recorded every 5 min (Dinamap). These systemic and renal haemodynamicstudies were repeated after 8 weeks of treatment with the ACE inhibitor enalapril (10mg/day), given at 0800 h. The same NE doses were administered during this secondevaluation. The patients also participated in another study that aimed to compare the renalhaemodynamic and microproteinuric NE responses in micro- and normoalbuminuricIDDM patients and control subjects, and the pre-enalapril data were also used for thatstudy [13]. One initially studied patient declined to participate in the second evaluation.

The experiments were carried out in a temperature-controlled room kept at 22 C.o

The subjects were studied after an overnight fast and remained so during the experiments.Smoking was prohibited and no liquid other than water was allowed to drink. The subjectsremained in supine position and were only allowed to stand on voiding. Two hours beforethe start of the renal haemodynamic measurements, each subject drank 600 ml water topromote diuresis. Thereafter, urinary volume losses were suppleted by water drinking. Aneuglycaemic insulin clamp was used to minimise effects of differences in actual glycaemiaon renal haemodynamics [16] and to avoid that changes in insulin levels during the experi-ments would potentially influence NE sensitivity. Insulin was infused (Velosulin H.M.,Novo-Nordisk, Bagsvaerd, Denmark) at a rate of 30 mU/kg per h. Blood glucose was keptat 5 mmol/l by varying the infusion rate of a 20% glucose solution. Potassium chloride (10mmol per 100 g of glucose) was added to the glucose vials.

GFR and ERPF were measured simultaneously using primed infusions of I-125

iothalamate and I-hippurate, respectively [16]. The clearances were calculated using the131

formula U.V/P and I.V/P, respectively. U.V represents the urinary excretion rate of the

Chapter 584

tracer, I.V represents the infusion rate of the tracer, and P represents the tracer value ofplasma samples taken at the end of each clearance period. Errors in the estimation of GFRdue to incomplete bladder emptying and dead space were corrected by multiplying theclearance of I-iothalamate with the formula: clearance of I-hippurate (I.V/P) /125 131

clearance of I-hippurate (U.V/P). The coefficients of variation for GFR and ERPF are131

2.2% and 5.0%, respectively [16]. The GFR and ERPF were corrected to 1.73m of body-2

surface area. Filtration fraction (FF) was calculated as the quotient of I-iothalamate and125

I-hippurate clearances. 131

Analytical methods

Blood samples were taken from an intravenous catheter at the end of each clearanceperiod with the subjects in supine position. Blood was immediately centrifuged at 4 C ando

samples were stored at -20 C before assay. Aliquots of urine for measurement of urinaryo

proteins were stored at -20 C and analysed within 2 weeks. Plasma insulin was determinedo

by RIA. Plasma NE was analysed by HPLC [17]. Active plasma renin was determinedwith a commercially available double-antibody RIA (Nichols Institute Diagnostics, SanJuan Capistrano, CA, USA) with a cross-reactivity with prorenin of 0.2%. Serum ACEactivity was measured with an HPLC technique. Angiotensin II was determined by RIA(Dept. of Pharmacology, State University Maastricht, The Netherlands). Cross-reactivityof the angiotensin II antibody with angiotensin I is <0.1%. Serum aldosterone wasdetermined by RIA. Blood glucose was measured using a Yellow Springs glucoseAnalyser (Model 23A, Yellow Springs Inc, Yellow Springs, Ohio, USA). Urinary albuminwas measured with ELISA using rabbit anti-human albumin antibody (DAKO). Thedetection limit is 1 µg/l. Sodium, potassium, creatinine and serum albumin in serum andurine were measured on SMA(C) autoanalysers (Technicon Instruments Inc. Tarrytown,N.Y., USA).

Statistical analysis

Results are expressed as mean±SEM for parametrically distributed data and asgeometric mean (95% confidence intervals) for non-parametrically distributed data unlessstated otherwise. The 2 clearance periods obtained at baseline, threshold, 20% pressor,pressor and recovery were averaged for analysis. Where appropriate, repeated measure-ments ANOVA for parametrically or non-parametrically distributed data was used toanalyse the effects of the NE-infusions before and after ACE inhibition. Paired t-tests orWilcoxon tests were then used to establish differences between corresponding observationperiods. Differences in NE effects on MAP and ERPF before and after ACE inhibitionwere evaluated by comparing the overall mean changes of each individual. p-Values lessthan 0.05 were considered to be significant.

Enalapril and norepinephrine reactivity in IDDM 85

Results

Metabolic control, body weight, urinary urea and sodium excretion were similarbefore and after treatment with enalapril (Table 1). Enalapril lowered baseline MAP by5%, but did not change pulse rate (Table 2). GFR was unaltered, whereas ERPF increasedby 9% (Table 2). FF tended to fall (p=0.09). Overnight microalbuminuria fell by 28%(Table 1). Enalapril increased plasma active renin and decreased serum ACE activity andserum aldosterone levels (Table 3). Baseline plasma AII did not significantly change, butthe AII/active renin ratio profoundly decreased, indicating adequate RAAS inhibition.

Plasma NE concentrations during NE infusions were similar before and after ACEinhibition (Table 2). Before enalapril, NE increased MAP at 20%pressor and pressor dosesto target levels (Table 2). With enalapril, NE infusion also increased MAP. At NE pressordose there was no difference in MAP compared to before enalapril. The overall meanincrement in MAP was greater during than before enalapril (p<0.05, Figure 1A). Atrecovery, MAP was again lower with than without enalapril. ACE inhibition treatment didnot influence the fall in pulse rate at NE-pressor dose and its rise at recovery (Table 2).

GFR remained essentially unaltered during NE infusions before and after enalapril,but declined at recovery (Table 2). ERPF progressively decreased at 20% pressor and atpressor dose both before and during enalapril, and corresponding rises in FF were seen.During all observation periods, ERPF remained significantly higher with than beforeenalapril, and the magnitude of the NE-induced decrease in ERPF was not influenced byenalapril treatment (Figure1B). The overall urinary albumin excretion rate during NEinfusion was 51.5 (24.3-109.1) µg/min before and 57.6 (26.3-126.3) µg/min after enalapril(p>0.20).

Before enalapril, plasma active renin, angiotensin II and serum aldosterone levels in-creased at NE pressor dose, whereas the angiotensin II/active renin ratio remained una-ltered (Table 3). With enalapril, plasma active renin and serum aldosterone levels rose atNE-pressor dose, but plasma angiotensin II did not significantly increase. At NE- pressordose a slight fall was seen in the angiotensin II/active renin ratio. The overall mean levelsof plasma active renin, serum ACE activity, angiotensin II, the angiotensin II/active reninratio and serum aldosterone were lower after enalapril.

Mean blood glucose and plasma insulin concentrations were 5.2±0.1 and 5.1±0.1mmol/l and 230±22 and 251±19 pmol/l, before and after enalapril, respectively (NS).

Discussion

As expected, enalapril lowered blood pressure and overnight urinary albuminexcretion, and induced renal vasodilatation in microalbuminuric IDDM patients. Thesystemic haemodynamic and the renal responsiveness to exogenous NE, as well as urinaryalbumin excretion during NE, were however not altered by ACE inhibition treatment.

Chapter 586

Table 3. Parameters of the renin-angiotensin II-aldosterone system during stepwisenorepinephrine infusions without and with enalapril

Plasma active Angiotensin- Angiotensin II Angiotensin II/ SerumRenin (mU/l) convering- (pmol/l) plasma active aldosterone

a a

enzyme (U/l) renin ratio (nmol/l)a a

without enalaprilBaseline 61 (33-111) 29 ± 3 15 ± 2 0.26 ± 0.05 360 ± 40Threshold 55 (32-94) 28 ± 3 14 ± 1 0.28 ± 0.05 330 ± 4020% Pressor 64 (49-187) 28 ± 4 16 ± 1 0.28 ± 0.06 420 ± 40Pressor 95 (49-187) 28 ± 4 25 ± 5 0.25 ± 0.05 710 ± 130

e e e

Recovery 91 (51-160) 25 ± 3 21 ± 4 0.24 ± 0.03 460 ± 50d

with enalaprilBaseline 121 (41-359) 6 ± 2 14 ± 4 0.10 ± 0.02 90 ± 20

b b b

Threshold 127 (49-335) 6 ± 2 11 ± 2 0.09 ± 0.02 100 ± 30b c b

20% Pressor 137 (54-485) 7 ± 2 12 ± 2 0.09 ± 0.02 110 ± 20b b b

Pressor 201 (85-485) 8 ± 2 18 ± 4 0.08 ± 0.01 220 ± 30e b b,d b,e

Recovery 198 (98-402) 9 ± 2 15 ± 2 0.07 ± 0.01 170 ± 40c,d b c,e b

Data in mean±SEM and geometric mean (95% confidence intervals). p<0.01 overall mean valuesa

without vs with enalapril, p<0.05 and p<0.01 from corresponding infusion period beforeb c

enalapril, p<0.05 and p<0.01 from baseline valuesd e

Table 2. Systemic and renal hemodynamic parameters during stepwise norepinephrineinfusions without and with enalapril.

Plasma Mean arterial Pulse Glomerular Effective renal Filtration norepinephrine pressure rate filtration rate plasma flow fraction

(nmol/l) (mmHg) (beats/min) (ml/min/1.73m ) (ml/min/1.73m ) (%) 2 2

without enalaprilBaseline 2.9 (2.1-4.2) 98± 2 70 ± 5 107 ± 9 460 ± 40 23 ± 1Threshold 5.1 (4.0-6.6) 100 ±1 66 ± 4 111 ±10 427 ± 40 26 ± 120% Pressor 6.6 (5.1-8.6) 102 ± 2 66 ± 4 110 ± 9 400 ± 34 28 ± 1

c d d

Pressor 17.7(10.3-30.2) 118 ±1 63 ± 4 109 ± 9 318 ± 22 34 ± 2d d d d

Recovery 3.6 (2.4-5.4) 95 ± 3 76 ± 5 97 ± 10 398 ± 32 24 ± 1 d d d

with enalaprilBaseline 2.9 (2.0-4.0) 92 ± 3 67 ± 4 111 ± 11 500 ± 44 22 ± 1

a b

Threshold 5.7 (4.8-6.9) 96 ± 2 65 ± 4 117 ± 12 463 ± 38 25 ± 1a a d

20% Pressor 6.3 (4.8-8.2) 98 ± 2 65 ± 4 116 ± 11 431 ± 35 27 ± 1a,c a,c d

Pressor 18.7 (11.1-31.5) 116 ± 2 62 ± 4 115 ± 10 352 ± 27 33 ± 1d d a,d d

Recovery 4.2 (2.9-6.0) 87 ± 3 74 ± 4 102 ± 9 460 ± 34 22 ± 1 b d c b

Data in mean±SEM and geometric mean (95% confidence intervals). p<0.05 and p<0.01 froma b

corresponding infusion period without enalapril, denotes p<0.05 and denotes p<0.01 fromc d

baseline

50 1A

0 10 20 30 40 50

40

30

20

10

0

chan

ge in

MA

P a

fter

enal

april

(m

mH

g)

change in MAP before enalapril (mmHg)

0 1B

-300 -250 -200 -150 -100 -50 0

-50

-100

-150

-200

-250

-300

chan

ge in

ER

PF

afte

r en

alap

ril(m

mm

)2

inpe

r1

73.l/

change in ERPF before enalapril(m m m )2in per 1 73.l/

Enalapril and norepinephrine reactivity in IDDM 87

Figure 1(A,B). Plot comparing individual changes in mean arterial pressure (MAP, A) and effectiverenal plasma flow (ERPF, B) in response to stepwise norepinephrine infusions before and after ACEinhibition. The lines indicate identical changes. Overall mean change in MAP was greater with thanwithout enalapril (n=7, p<0.05 by paired Wilcoxon test). No difference in overall ERPF response wasobserved.

Chapter 588

Apart from general factors like basal vasomotor tone and vascular reactivity, sodiumand volume homeostasis, as well as baroreceptor feedback control are determinants of the systemic reactivity to exogenous NE, and may thus contribute to an enhancedsystemic NE responsiveness in diabetes mellitus [8,11,13,18]. Such an exaggeratedresponsiveness has been found to be associated with an increased exchangeable sodiumand a higher extracellular volume, and this hyperreactivity decreases after diuretictreatment [11]. In addition, microalbuminuric IDDM patients have an attenuated pulserate decline during NE infusion [13], which suggests that an altered baroreceptorfeedback control may also contribute to the systemic NE hyperresponsiveness in IDDM[18].

The observation that enalapril treatment failed to diminish systemic vascular NEreactivity in microalbuminuric IDDM is in keeping with previous findings in diabeticpatients [11,12]. By contrast, in essential hypertension, the systemic responsiveness to NEappears to be attenuated after ACE inhibition treatment [7-9]. Several factors could beinvolved in the lack of attenuation of the systemic NE responsiveness after ACE inhibitionin diabetes mellitus. First, ACE inhibition therapy is unlikely to have influenced sodiumand volume status in the present study since body weight remained unchanged. Indeed,captopril does not significantly decrease exchangeable sodium in diabetes mellitus [14].In this respect, it is noteworthy that ACE inhibition treatment induces a negative sodiumbalance in essential hypertension [19]. Second, a similar heart rate decline during NEbefore and after enalapril was observed, suggesting no important alterations inbaroreceptor sensitivity, in agreement with findings in hypertensive diabetic patients [14].Third, although enalapril induced adequate RAAS suppression, it cannot be excluded thata more complete inhibition of angiotensin II formation is required to blunt systemic NEreactivity. Finally, we consider a type II error very unlikely to explain the lack of attenua-tion of enalapril on systemic NE reactivity, since the mean increment in MAP in responseto NE was even higher after ACE inhibition.

This study is the first to document the effect of ACE-inhibition treatment on NE-induced renal vasoconstriction in microalbuminuric IDDM patients. In IDDM, ACEinhibitors increase renal plasma flow without having much effect on glomerular filtrationrate. As a consequence, filtration fraction tends to decline [20]. These changes are ascribedto a predominantly postglomerular dilatation which is in part due to decreased angiotensinII formation. In our study, the renal vasodilatory effects of enalapril remained presentduring NE infusion, a finding thus far only documented in healthy volunteers [10].Enalapril prevented the rise in angiotensin II levels during NE pressor infusion. It maythus be inferred that this partial RAAS blockade accounted for the ongoing renalvasodilatation of the ACE inhibitor during NE infusion as compared to pretreatment.Although difficult to demonstrate in humans, animal studies have unequivocally docu-mented bidirectional interactions between the RAAS and the SNS in a way such that eachsystem is able to potentiate the other [4]. ACE inhibition treatment could, therefore,induce a decrease in SNS activation. In the isolated kidney, ACE inhibitors indeed havebeen demonstrated to diminish renal vascular NE responsiveness [5,6], and neuronally-induced renal vasoconstriction [21], suggesting interference in the RAAS-SNS interaction.The present in vivo study does, however, not support the assumption of an intrarenalsympathicolytic effect of enalapril in microalbuminuric IDDM patients, since the

Enalapril and norepinephrine reactivity in IDDM 89

exogenous NE-induced fall in ERPF was not altered by ACE inhibition treatment. Our findings may have clinical implications for the future design of renoprotective

strategies in IDDM patients. Currently, ACE inhibition treatment is considered to be themain tool to retard the decline in renal function loss in IDDM, both by lowering systemicblood pressure and by decreasing intraglomerular pressure [1-3]. Despite this provenrenoprotective potential, many patients still progress to end stage renal failure. Albeit ina small number of patients, the present study is consistent with the possibility thatmicroalbuminuric IDDM patients are not sufficiently protected by ACE inhibitiontreatment against rises in systemic blood pressure due to bouts of SNS activation duringdaily life activities. Although the effects of exogenous NE cannot be directly extrapolatedto physiological SNS stimulation, it is noteworthy that increases in urinary albumin excre-tion are well related to rises in blood pressure and plasma NE levels during exercise,which represents a strong endogenous SNS stimulus [17].

In conclusion, we found that enalapril does not attenuate systemic and renal vascularresponsiveness to exogenous NE in microalbuminuric IDDM patients. These findingssuggest that a potentially adverse effect of NE on systemic and renal vascular tone is noteffectively counteracted by ACE inhibition treatment alone.

References

1 Viberti GC, Mogensen CE, Groop LC, Pauls JF. Effect of captopril on progression to clinicalproteinuria in patients with insulin-dependent diabetes mellitus and microalbuminuria. JAMA1994; 271: 275-279

2 Lewis EJ, Hunsicker LG, Bain RP, Rohde RD. The effect of angiotensin-converting-enzymeinhibition on diabetic nephropathy. N Engl J Med 1993; 329: 1456-1462

3 Maschio G, Alberti D, Janin G, Locatelli F, Mann JFE, Motolese M, Ponticelli C, Ritz E, Zucchel-li P. Effect of the angiotensin-converting-enzyme inhibitor benezepril on the progression ofchronic renal insufficiency. N Engl J Med 1996; 334: 939-945

4 Reid IA. Interactions between ANG II, sympathetic vervous system, and baroreceptor reflexes inregulation of blood pressure. Am J Physiol 1992; 262: E763-E778

5 Casellas D, Mimran A, Dupont M, Chevil-Lard C. Attenuation by SQ 14,255 (captopril) of thevascular response to noradrenaline in the rat isolated kidney. Br J Clin Pharmacol 1980; 10: 621-623

6 Chiba S, Quilley CP, Quilley J, McGiff JC. Captopril decreases vascular reactivity independentlyof changes in converting enzyme activity and prostaglandin release in the rat isolated kidney. EurJ Pharmacol 1982; 83: 243-252

7 Fruncillo RJ, Rotmensch HH, Koplin JR, Swanson BN, Ferguson RK. Effect of captopril andhydrochlorothiazide on the response to pressor agents in hypertensives. Eur J Clin Pharmacol1985; 28: 5-9

8 Uehlinger DE, Ferrier CP, Matthieu R, Reuter K, Gnaedinger MP, Saxenhofer H, Shaw S,Weidmann P. Antihypertensive contribution of sodium depletion and sympathetic axis duringchronic angiotensin II converting enzyme inhibition. J Hypertens 1989; 7: 901-907

9 Malini PL, Strocchi E, Valtancoli G, Ambrosioni. ACE inhibition and pressor responsiveness tonorepinephrine in hypertensive patients. J Clin Pharmacol 1990; 30: 422-424

10 Lang CC, Rahman AR, Balfour DJK, Struthers AD. Enalapril blunts the antinatriuretic effect ofcirculating noradrenaline in man. J Hypertens 1993; 11: 565-571

11 Weidmann P, Beretta-Piccoli C, Keusch G, Z. Glück, Mujagic M, Grimm M, Meier A, ZieglerWH. Sodium-volume factor, cardiovascular reactivity and hypotensive mechanism of diuretic

Chapter 590

therapy in mild hypertension associated with diabetes mellitus. Am J Med 1979; 67: 779-784 12 Beretta-Piccoli C, Weidmann P. Exaggerated pressor responsiveness to norepinephrine in

nonazotemic diabetes mellitus. Am J Med 1981; 71: 829-835 13 Hoogenberg K, Sluiter WJ, Navis G, Van Haeften TW, Smit AJ, Reitsma WD, Dullaart RPF.

Exogenous norepinephrine induces an enhanced microproteinuric response in microalbuminuricinsulin-dependent diabetes mellitus. J Am Soc Nephrol (in press)

14 Beretta-Piccoli C, Ziegler WH, Antonini P, Bernasconi-Zapf M, Kressebuch H, Pianezzi F.Cardiovascular regulation during angiotensin coverting enzyme inhibition with captopril indiabetes-associated hypertension. J Hypertens 1990; 8: 671-678

15 Ciavarella A, Mustacchio A, Ricci C, Capelli M, Vannini P. Enhanced pressor responsiveness tonorepinephrine in type II diabetes. Effect of ACE inhibition. Diabetes Care 1994; 17:1 484-1487

16 Dullaart RPF, Meijer S, Sluiter WJ, Doorenbos H. Renal haemodynamic changes in response tomoderate hyperglycaemia in type 1 (insulin-dependent) diabetes mellitus. Eur J Clin Invest 1990;20: 208-213

17 Hoogenberg K, Dullaart RPF. Abnormal plasma noradrenaline response and exercise inducedalbuminuria in type 1 (insulin-dependent diabetes mellitus). Scand J Lab Invest 1992; 52: 803-811

18 Eckberg DL, Harkins SW, Fritsch JM, Musgrave GE, Gradner DF. Baroreflex control of plasmanorepinephrine and heart period in healthy subjects and diabetic patients. J Clin Invest 1986; 78:366-374

19 Navis GJ, De Jong PE, Donker AJM, an der Hem GK, De Zeeuw D. Diuretic effects of ACE-inhibition; comparison of low and liberal sodium diet in hypertensive patients. J CardiovascPharmacol 1987; 9: 743-748

20 Marre M, Chatellier G, Leblanc H, Guyene TT, Menard J, Passa P. Prevention of diabeticnephropathy with enalapril in normotensive diabetics with microalbuminuria. Br Med J 1988;1092-1095

21 Mimran A, Casellas D, Chevillard C, Dupont M, Jover B. Evidence for a postsynaptic effect ofcaptopril in isolated perfused rabbit kidney. Am J Cardiol 1982; 49: 1540-1541

Acknowledgments

Plasma angiotensin II was determined by courtesy of P.M.H. Schiffers, Departmentof Pharmacology, State University Maastricht, The Netherlands. This study was in partsupported by grants from the Dutch Diabetes Foundation.

Department of Endocrinology, Angiology and Surgical Intensive Care, University Hospital1 2 3

Groningen, The Netherlands.

Crit Care Med 1998 (in press)

CHAPTER 6

EFFECTS OF LOW DOSE DOPAMINE ON RENAL ANDSYSTEMIC HAEMODYNAMICS DURING INCREMENTAL

NOREPINEPHRINE INFUSION IN HEALTHY VOLUNTEERS

K. Hoogenberg , A.J. Smit and A.R.J. Girbes1 2 3

Dopamine is widely used to prevent a fall in renal function and blood flow in septic patients,especially when other pressor agents like norepinephrine (NE) are used. No human renalhaemodynamic studies exits to support this. Therefore, the effects of low dose dopamine on NE-induced renal and systemic vasoconstriction in were assessed in 7 normotensive healthy volunteers. Onseparate days, either a low dose dopamine (4 µg/kg per min) or a placebo (5% glucose) infusion, in asingle blinded randomized order, was added to incremental NE infusions of 40, 80, and 150 ng/kg permin given for 60 min each. Blood pressure and heart rate were measured with a semi-automateddevice, glomerular filtration rate (GFR) and effective renal plasma flow (ERPF) were determined withconstant infusions of I-iothalamate and I-hippurate, respectively. NE-alone progressively125 131

increased mean arterial pressure (MAP) to pressor levels, whereas this effect was attenuated byaddition of dopamine (p<0.05 vs NE-alone). GFR increased during lower nor-epinephrine doses anddid not decline at the highest NE dose. Addition of dopamine further increased GFR. ERPF fell witheach NE-alone infusion step, but this decline was completely prevented by concomitant dopamineinfusion (p<0.01 vs NE-alone). Sodium excretion tended to decrease with NE, but increased 2 to 3 foldafter addition of dopamine (p<0.01 from NE-alone). In conclusion, NE causes a large decrease inERPF but not in GFR in healthy man. Concomitant dopamine administration prevents this fall in ERPF,increases sodium excretion, and also attenuates the NE-induced rise in MAP. These findings warrantfurther clinical evaluation of the effect of concomitant low-dose dopamine and NE administration incritically ill patients.

Introduction

Norepinephrine (NE) is increasingly recognized as a valuable agent in the treatmentof septic shock [1-6]. By eliciting a strong vasoconstrictive response, NE has beenreported to restore blood pressure and tissue perfusion more efficaciously than high dosedopamine [1,4,6]. However, the NE-induced reduction in renal blood flow [7-9] makemany clinicians reluctant to use this agent in hypotensive patients since an alreadycomprimised renal function may further deteriorate under these circumstances. Intrave-nous dopamine at dosages from 1 to 4 µg/kg per min augments renal blood flow in healthyman [10,11] as well as in mechanically ventilated patients [12]. These low infusion ratesof dopamine have been proposed to oppose the renal vasoconstrictive actions of NE -[13,14]. For this reason low dose dopamine is widely used as a renoprotective agent

Chapter 692

during NE therapy. However, the renal supportive action of dopamine to overcome NE-induced renal vasoconstriction has not been proven in these situations, and thehaemodynamic effects of combined infusions has received little attention. In the dog,dopamine has been documented to increase in renal blood flow during NE infusions, butthese observations cannot be easily extrapolated to humans since dogs have a differentrenal haemodynamic profile during NE infusion from man [15]. In healthy subjects, veryrecent results suggested that dopamine may blunt the renal vasoconstrictive effect of NEthat was given as a single pressor dose [16]. In the present study, we evaluated whetherdopamine could counteract NE's actions on renal haemodynamics by adding low dosedopamine to stepwise incremental NE doses in normotensive healthy subjects.

Materials and methods

Written informed consent was obtained after explanation of the purpose of the studywhich was approved by the local medical ethics committee. Eligible subjects were maleswith an age between 20 to 35 years, a body mass index <27 kg/m , a systolic/diastolic2

blood pressure <140/85 mmHg, a serum creatinine concentration< 90 µmol/l, no historyof cardiovascular disease and no use of any medications. Seven healthy males with a meanof age 28 (range 22-35) years, mean body mass index of 23.8 (range 20.9-26.7) kg/m , and2

mean blood pressure of 123/76 (range 112-132/66-85) mmHg, participated in the studies.A 24 h urine collection revealed a normal creatinine clearance (mean 127 (range 117-161)ml/min/1.73 m ) in all of them. Mean sodium excretion amounted to 255 (range 189-321)2

mmol/day and was considered appropriate in the studies so that all subjects were advisedto adhere to their usual dietary habits. They were only requested to avoid large proteinmeals 24 h prior to the studies.

The experiments were carried out on 2 separate days, with a 1 wk interval, in atemperature controlled room of 22 C. The subjects were studied after an overnight fast ando

remained fasting during the experiments. Smoking was prohibited and water was the onlyliquid available to drink. Measurements were made with the subjects in the supine posi-tion, and they were only allowed to stand on voiding. To promote diuresis, each subjectdrank 600 ml water at 0730 h after which urinary volume losses were supplemented byoral water. A cannula was inserted in an antecubital vein for infusions of renal tracers, NE,dopamine or placebo. Venous blood was collected from the contra lateral arm. After anequilibration period of 120 min, baseline measurements were made from 1000 h to 1130hrs. From 1130 hrs to 1430 hrs, NE was infused at 3 incremental doses of 40, 80, 150ng/kg per min, respectively, with each infusion lasting 60 min. In a single blinded,randomized order, either dopamine (4 µg/kg/min) or placebo (5% glucose) was added toNE infusions. After cessation of the infusions at 1430 hrs, measurements were continuedduring a 60 min recovery period.

Glomerular filtration rate (GFR) and effective renal plasma flow (ERPF) were mea-sured with constant infusions of I-iothalamate and I-hippurate, respectively, as125 131

described previously [17]. After a priming dose, the radiopharmaceuticals were allowedto equilibrate for 120 min. Thereafter, hourly clearances were calculated using the for-mulas U.V/P and I.V/P, respectively. U.V represents the urinary excretion rate of thetracer, I.V represents the infusion rate of the tracer, and P represents the plasma tracer

Renal effects of combined norepinephrine and dopamine 93

level. Values were corrected to 1.73m of body-surface area. Filtration fraction (FF) was2

calculated as the ratio of the iothalamate and hippurate clearances. Blood pressure wasrecorded every 5 minutes with a semiautomated device (Dinamap 1846, Critikon, Tampa,FL, USA). Mean arterial pressure (MAP) was calculated as a systolic blood pressure + bdiastolic blood pressure (mmHg). Pulse rate was recorded continuously on anelectrocardiography oscilloscope (Hewlett Packard, Boblingen, Germany). Renal vascularresistance (RVR) in dyn.sec/cm per 1.73 m was calculated as MAP×80/ (1/haematocrit)5 2

× ERPF).Blood samples were taken in the middle and the end of each clearance period while

the subjects were in supine position. Plasma samples for determination of NE and dopa-mine were collected in pre-chilled tubes containing 10% EDTA, immediately centrifugedat 4 C, and stored at -20 C until HPLC analysis [18]. Electrolytes and creatinine in serumo 0

and urine were measured on SMA(C) autoanalysers (Technicon Instruments Inc.Tarrytown, N.Y., USA).

Based on previous studies in healthy volunteers, ERPF was expected to decline 20-30% with the maximal NE dose [19]. The hypothesis was tested that dopamine counter-acted this NE-induced decline in ERPF. A sample size of 7 subjects was large enough todemonstrate a difference between no change vs. 20-30% change in ERPF with a power of85% and a two-sided p-value<0.05. The coefficients of variation of ERPF and GFR are<5% [17]. Results are expressed as mean±SEM for parametrically distributed data andgeometric mean (95% confidence interval) for non-parametrically distributed data. Withinday changes were evaluated using ANOVA for repeated measurements for parametricallyand non-parametrically distributed data as appropriate. Between infusion differences weretested separately for each period with paired Student t-tests for parametrically andWilcoxon test for non-parametrically distributed data. p-Values less than 0.05 wereconsidered significant.

Results

On both study days, plasma NE concentrations increased similarly with theinfusions. At recovery from the combined NE/dopamine infusions plasma NE was stillelevated from baseline values. The dopamine infusions led to high circulating levels ofplasma dopamine (Table 1).

Baseline blood pressure and heart rate were not different on both study days.Systolic blood pressure increased similarly with both infusion regimens (p<0.05 to 0.01,Table 2). Diastolic blood pressure was evidently increased during 80 and 150 ng/kg permin NE-alone infusion (p<0.01), while only a modest rise was present during thecombined NE/dopamine at the 150 ng/kg per min infusion dose (p<0.05). Diastolic bloodpressure was lower with NE/dopamine compared with NE-alone during all infusionperiods (p<0.01, Table 2). Consequently, NE/dopamine augmented pulse pressure (differ-ence in systolic and diastolic blood pressure) more than NE-alone (p<0.05, Table 2), andMAP was lower with NE/dopamine than with NE-alone (p<0.05, Figure 1A). Heart ratedeclined with all NE-alone infusions (p<0.05 to 0.01), whereas no change in heart rate waspresent with the NE/dopamine infusions (Table 2).

Chapter 694

Table 2. Systolic and diastolic blood pressure, pulse pressure and heart rate.

Period: Baseline Norepinephrine infusion rate Recovery

40 ng/kg 80 ng/kg 150 ng/kg per min

Systolic blood pressure (mmHg)

NE-alone 123 ± 3 133 ± 3 145 ± 4 163 ± 6 131 ± 4a b b

NE+dopamine 121 ± 3 137 ± 2 148 ± 4 163 ± 5 126 ± 4a b b

Diastolic blood pressure (mmHg)NE-alone 76 ± 4 80 ±3 84 ± 4 91 ± 5 74 ± 4

d b,d b,d

NE+dopamine 75 ± 3 73 ± 3 74 ± 3 81 ± 3 73 ± 3b

Pulse pressure (mmHg)NE-alone 47 ± 3 53 ± 3 60 ± 3 72 ± 2 56 ± 3

a b

NE+dopamine 46 ± 4 64 ± 3 74 ± 5 82 ± 5 54 ± 3b,e b,e b,e

Heart rate (beats/min)NE-alone 59 ± 3 54 ± 3 51 ± 3 51 ± 4 71 ± 4

c a,c b,c b

NE+dopamine 58 ± 4 61 ± 3 61 ± 3 62 ± 3 82 ± 2b

Data in mean±SEM. NE: norepinephrine denotes p<0.05 and denotes p<0.01 from baselinea b

values; denotes p<0.05 and denotes p<0.01 vs. NE+dopamine; denotes p<.05 vs. NE-alonec d e

Table 1. Plasma norepinephrine and plasma dopamine concentrations.

Period Baseline Norepinephrine infusion rate Recovery

40 ng/kg 80 ng/kg 150 ng/kg per min

Plasma norepinephrine (nmol/l)

NE-alone 1.7(1.1-2.8) 5.1(3.3-7.9) 9.6(6.2-14.9) 13.4(8.6-21.1) 2.3(1.2-4.3)b b b

NE+dopamine 1.4(0.9-2.1) 5.4(4.5-6.5) 9.7(7.2-13.2) 13.4(11.3-23.9) 4.1(1.9-8.4)b b b a

Plasma dopamine (nmol/l)

NE-alone n.d. n.d. n.d. n.d. n.d.

NE+dopamine n.d. 414(195-1039) 612(350-1073) 620(380-1013) n.d.b b b

Data in geometric mean and 95% confidence interval. NE: norepinephrine, n.d. not detectable, a

denotes p<0.05 and denotes p<0.01 from baseline values.b

Renal effects of combined norepinephrine and dopamine 95

Table 3. Urine flow and fractional sodium excretion.

Period Baseline Norepinephrine infusion rate Recovery

40 ng/kg 80 ng/kg 150 ng/kg per min

Urine flow (ml/min)

NE-alone 6.4 ± 0.8 9.4 ± 1.1 8.1 ± 1.2 6.5 ± 1.1 3.6 ± 0.9NE+dopamine 6.0 ± 0.7 10.0 ±1.1 8.1 ± 1.1 9.8 ±1.0 3.0 ± 0.8

a a,d

Fractional sodium excretion (%)

NE-alone 0.98±0.35 0.91±0.23 0.69±0.09 0.72±0.21 1.02±0.21

NE+dopamine 0.84±0.17 2.29±0.17 2.44±0.43 1.99±0.34 0.81±0.20b,c b,c

Data in mean±SEM. NE: norepinephrine, denotes p<0.05 and denotes p<0.05 from baselinea b

values; denotes p<0.05 and denotes p<0.01 vs NE-alonec d

Baseline renal haemodynamic parameters were similar on both study days. WithNE-alone, GFR rose slightly but significantly at the 40 and 80 ng/kg per min infusion dose(p<0.05), whereas no change from baseline values was present at the 150 ng/kg per mindose (Figure 1B). The combined NE/dopamine infusions increased GFR progressively(p<0.01) to values that exceeded the NE-alone infusions (p<0.05, Figure 1B). ERPF de-clined in a dose dependent fashion with each NE-alone infusion step (p<0.05 to 0.01, Fig-ure 1C), and was accompanied by increases in FF (p<0.01, Figure 1D). These alterationswere opposed by simultaneous dopamine infusion, which in fact induced a modest rise inERPF at the 40 and 80 ng/kg per min NE-dose (p<0.05, Figure 1C). FF remained un-changed with NE/dopamine infusion, and was significantly lower compared with the NE-alone infusions (p<0.01, Figure 1D). RVR progressively increased with NE-alone(p<0.01), but changed only marginally with the combined NE/dopamine infusion. Conse-quently, RVR reached higher levels with NE-alone compared with NE/dopamine (p<0.05to .01, Figure 1E).

Urine flow did not change with NE-alone, while a slight increase was seen duringNE/dopamine at 40 and 150 ng/kg/min dose (p<0.05, Table 3). Fractional sodiumexcretion was similar at the start of both infusions days. A tendency towards a decease infractional sodium excretion was noted with the highest NE dose (p<.10). In contrast,addition of dopamine stimulated fractional sodium excretion (p<0.01), reaching 2 to 3 foldhigher levels than during NE-alone (p<0.05 to 0.01, Table 3).

Discussion

In the present study, we demonstrated that a 4 µg/kg/min dopamine infusion bluntedthe renal vasoconstrictive and overrided the sodium preserving effects of exogenous NEinfusions in normotensive healthy subjects. The dopamine infusion also diminished therise in blood pressure, enlarged pulse pressure and blunted the fall in heart rate which suggests that dopamine also influenced systemic haemodynamics during the NEinfusions.

12,000 E

10,000

8,000

6,000

4,000

Baseline NA:40 NA:80 NA:150 Recovery

Ren

al V

ascu

lar

Re

sist

ance

40

35

30

25

20

15

D

Filt

ratio

n F

ract

ion

(%

)

700

600

500

400

300

C

Effe

ctiv

e R

ena

l Pla

sma

Flo

wG

lom

eru

lar

Filt

ratio

n R

ate

160

140

120

100

B

130

120

110

100

90

80

A

Mea

n A

rter

ial P

ress

ure

(m

m H

g)(m

mm

)2

inp

er

173.

L/

(mm

m)

2in

per

173.

L/

(m

dyn

ee

cc

s.m

)2

5pe

r1

73

./

Chapter 696

Figure 1.A: Mean arterial pres-sure; B: glomerular filtrationrate; C: effective renal plasmaflow; D: filtration fraction; E:renal vascular resistance atbaseline, at the incremental nor-epinephrine (NE) infusions of40 (NE:40), 80 (NE:80) and150 (NE:150) ng/kg per min, -and at recovery from the infu-sions. !NE-alone infusion, "NE in combination with lowdose dopamine (4 µg/kg permin). * denotes p<0.05 and †denotes p<0.01 from baseline;‡ denotes p<0.05 and § denotesp<0.01 vs NE-alone infusion.

Renal effects of combined norepinephrine and dopamine 97

Exogenous NE is well known to decrease ERPF and to increase FF without substan-tially altering GFR, and its administration is generally associated with increases in tubularsodium reabsorption [7-9]. In contrast, dopamine enhances ERPF and increases GFR to alesser extent leading to a fall in FF. Moreover, dopamine strongly augments sodium excre-tion [10-12]. These renal effects are observed with dopamine doses from 1 to 4 µg/kg/min.In this study, a 4 µg/kg/min dopamine dose was given with the objective to producemaximal antagonistic action on the renal NE responses. Dopamine prevented the decreasein ERPF, increased GFR and fractional sodium excretion during the NE infusions. Whenour study was in progress a report in healthy subjects appeared showing that dopamine ata somewhat lower infusion rate of 3 µg/kg/min blunted the decline in ERPF and increasedsodium excretion when added to a single NE pressor infusion [16]. Except for this slightdifference in study design, the main outcome is remarkably similar. Since hippurate clear-ances have been found to reliably measure both NE- and dopamine-induced changes inrenal blood flow [20], our findings as well as those in the aforementioned study [16] areunlikely to be biased by the methods used. Data from animal studies have shown that bothsubstances influence renal vascular resistance on the level of the glomerular arterioles viastimulation of "- and DA-adrenoreceptors, respectively [21,22]. While NE constricts,dopamine dilates pre- and postglomerular vessels concomitantly. Since these vessels arethe major sites of resistance to blood flow in the kidney, NE is supposed to reduce ERPFvia glomerular arteriolar vasoconstriction, whereas dopamine changes vessel tone and flowin the opposite direction.

Apart from direct renal vasodilation, indirect effects by changes in cardiac outputcould also be involved. Cardiac output and heart rate increase by $1-adrenergic receptorstimulation at dopamine doses as low as 3 µg/kg/min [23]. In this study, dopamine attenu-ated the rise in MAP, enlarged pulse pressure and blunted the decline in heart rate duringthe NE infusions, and these effects may have resulted from $1-adrenoreceptor stimulation.If true, an increase in cardiac output could have stimulated ERPF. We [12] previouslysuggested such a mechanism in postoperative mechanically ventilated patients in whomthe ratio increase in ERPF/increase in cardiac output was not altered by 4 to 8 µg/kg/mindopamine infusions. Apart from direct renal haemodynamic or systemic haemodynamiceffects of dopamine and/or NE on intraglomerular pressure, other changes in e.g. oncoticpressure, filtration area and the diffusion coefficient [24] and a combination of them,could have caused alterations in GFR during the infusions. Elevations in intraglomerularpressure have been demonstrated in rat studies with NE infusions when blood pressure wasallowed to increase [25]. Renal vasodilation by dopamine could have resulted in anincreased transmission of systemic blood pressure to the renal vascular bed and conse-quently raised intraglomerular pressure to higher levels than NE alone. Redistribution ofrenal blood flow and increases in filtration area are other mechanisms by which dopaminemay increase GFR [26].

The finding that dopamine increased natriuresis during NE illustrates anotherimportant renal action of this substance. The well documented natriuretic effects [7-9]have been ascribed to stimulation of specific dopamine-receptors at the proximal tubule[27] and may occur independent of renal vasodilation [28]. Conversely, NE generallyproduces antinatriuresis by promoting tubular sodium reabsorption [7-9], but in this studythe number of subjects may have been too small to demonstrate such an effect on renal

Chapter 698

sodium handling. Dopamine's potential to counteract any possible antinatriuretic effect ofNE, is probably due to inhibition of the high energy-dependent sodium reabsorptionprocess in the proximal tubule [27].

NE has many characteristics to fulfill treatment goals in septic shock. It quicklyrestores perfusion pressure and thereby oxygen supply to hypoperfused tissues. It is,therefore, not suprising that NE is a valued treatment option in this condition [1-6].However, fear of renal function loss with NE monotherapy is widely distributed, but hasin fact only been demonstrated with suprapharmacological doses in experimental studies[29]. Nonetheless, low-dose dopamine is often proposed to conserve renal function duringNE therapy [13,14]. The present results, although obtained in healthy volunteers and thusdifficult to extrapolate to septic patients, support these assumptions to variable extent: NEreduced ERPF, but not GFR, and addition of dopamine was indeed able to restore bothGFR and ERPF above baseline. Although little direct evidence is known about the effectsof NE and dopamine on renal haemodynamics in septic patients, some of the tubulareffects have been studied in these patients.

A recent clinical study reported a loss of effectiveness of dopamine to stimulatediuresis and natriuresis in patients with sepsis syndrome whereas no effects could beelicited in patients with septic shock [30]. These findings contrast with our recent observa-tions that dopamine (2, 4 or 6 µg/kg/min) on top of a constant background NE infusion,induced a dose-dependent increase in urine output and fractional sodium excretion inseptic patients [31]. Moreover, dopamine increased diuresis in oliguric resuscitatedsurgical intensive care patients [32] and critically ill patients considered at risk for renalinsufficiency [33], but was not effective in patients after major abdominal surgery [34].That human studies addressing renal haemodynamics and not only diuresis and natriuresisin such patients are warranted, is supported by a recent study which reported disappointingeffects of the benefit of dopamine on prevention of renal failure [35].

In conclusion, in healthy volunteers low-dose dopamine prevented the NE-inducedfall in renal blood flow, and it increased sodium excretion. The so often invoked unfavour-able renal effects of NE were not manifested by a fall in GFR. Future studies have toextend these findings in patients with sepsis, and should in particular assess whether or notlow-dose dopamine improves clinical outcome in these patients.

References

1 Meadows D, Edwards JD, Wilkins RG, Nightingale P. Reversal of intractable septic shock withnorepinephrine therapy. Crit Care Med 1988; 16: 663-666

2 Desjars P, Pinaud M, Potel G, Tasseau F, Touze MD. A reappraisal of norepinephrine therapy inhuman septic shock. Crit Care Med 1987; 15: 134-137

3 Martin C, Saux P, Albanèse J, Gouin F. A new look at norepinephrine to treat humanhyperdynamic septic shock. Anaesthesiology 1987; 67: A648

4 Desjars P, Pinaud M, Bugnon D, Tasseau F. Norepinephrine therapy has no deleterious renaleffects in human septic shock. Crit Care Med 1989; 17: 426-429

5 Martin C, Eon B, Saux P, Gouin F. Renal effects of norepinephrine used to treat septic shockpatients. Crit Care Med 1990; 18: 282-285

6 Martin C, Papazian L, Perrin G, Saux P, Gouin F. Norepinephrine or dopamine for the treatmentof hyperdynamic septic shock? Chest 1993; 103: 1826-1831

Renal effects of combined norepinephrine and dopamine 99

7 Smythe McC, Nickel JF, Bradley SE. The effect of epinephrine (usp), l-epinephrine, and l-norepinephrine on glomerular filtration rate, renal plasma flow and the urinary excretion ofsodium, potassium and water in normal man. J Clin Invest 1952; 31: 499-506

8 Hollenberg NK, Solomon HS, Adams DF, Abrams HL, Merrill JP. Renal vascular responses toangiotensin and norepinephrine in normal man. Circ Res 1972; 31: 750-757

9 Lang CC, Rahman AR, Balfour DJK, Struthers AD. Effect of norepinephrine on renal sodium andwater handling in euhydrated and overhydrated man. Clin Sci 1993; 85: 487-494

10 McDonald RH, Goldberg LI, McNay JL, Tuttle EP. Effect of dopamine in man: augmentation ofsodium excretion, glomerular filtration rate, and renal plasma flow. J Clin Invest 1964; 43: 1116-1124

11 Smit AJ: Dopamine and the kidney. Neth J Med 1989; 34: 47-5812 Girbes ARJ, Lieverse AG, Smit AJ, Van Veldhuisen DJ, Zwaveling JH, Meijer S, Reitsma WD.

Lack of specific renal haemodynamic effects of different doses of dopamine in mechanicallyventilated patients after infrarenal aortic surgery. Br J Anaesth 1996; 77: 753-757

13 Barnard MJ, Linter SPK. Acute circulatory support. BMJ 1993; 307: 35-4014 Parker MM. The heart in sepsis. In: Shoemaker WC, Ayres SM, Grenvik A, Holbrook PR, eds.

Textbook of Critical Care. Philadephia, WB Saunders Co., 1995, pp 596-60415 Schaer G, Fink MP, Parillo JE. Norepinephrine alone versus norepinephrine plus low-dose

dopamine: Enhanced renal blood flow with combination pressor therapy. Crit Care Med 1985; 13:492-496

16 Richer M, Robert S, Lebel M. Renal hemodynamics during norepinephrine and low-dose dopa-mine infusions in man. Crit Care Med 1996; 24: 1150-1154

17 Donker AJM, Van der Hem GK, Sluiter WJ, Beekhuis H. A radioisotope method for simultaneousdetermination of the glomerular filtration rate and effective renal plasma flow. Neth J Med 1977;20: 97-103

18 Hoogenberg K, Girbes ARJ, Stegeman CA, Sluiter WJ, Reitsma WD, Dullaart RPF. Influence ofambient plasma norepinephrine on renal haemodynamics in Type 1 (insulin-dependent) diabeticpatients and healthy subjects. Scand J Lab Invest 1995; 55: 15-22

19 Hoogenberg K, Sluiter WJ, Van Haeften TW, Smit AJ, Reitsma WD, Dullaart RPF. No differencein renal haemodynamic and albuminuric response to noradrenaline in IDDM and healthy subjects.Diabetologia 1996; 39 [Suppl 1]: A294

20 Visscher CA, De Zeeuw D, De Jong PE, Sluiter WJ, Huisman RM. Drug induced changes in renalhippurate clearance as a measure of renal blood flow. Kidney Int 1995; 48: 1617-1623

21 Edwards RM. Segmental effects of norepinephrine and angiotensin II on isolated renalmicrovessels. Am J Physiol 1983; 244: F526-F534

22 Edwards RM. Response of isolated renal arterioles to acetylcholine, dopamine, and bradykinin.Am J Physiol 1985; 248: F183-189

23 Goldberg LI. Dopamine and new dopamine analogs: Receptors and clinical applications. J ClinAnesth 1988; 1: 66-74

24 Maddox DA and Brenner BM. Glomerular ultrafiltration. In: Brenner BM, Rector FC (eds) Thekidney. WB. Saunders Co., Philadephia, 1996, pp 286-333

25 Meyers BD, Deen WM, Brenner BM. Effects of norepinephrine and angiotensin II on thedeterminants of glomerular ultrafiltration and proximal tubule fluid reabsorption in the rat. CircRes 1975; 37: 101-110

26 Hollenberg NK, Adams DF, Mendell P. Renal vascular responses to dopamine: haemodynamicand angiographic observations in normal man. Clin Sci Mol Med 1973; 45: 733-742

27 Aperia A, Bertorello A, Seri I. Dopamine causes inhibition of NaKATPase from rat proximaltubule segments. Am J Physiol 1987; 252: F39-F45

28 Smit AJ, Meijer S, Wesseling H, Donker AJM, Reitsma WD. The effect of metoclopramide ondopamine-induced changes in renal function in healthy controls and in patients with renal disease.

Chapter 6100

Clin Sci 1988; 75: 421-428 29 Conger JD, Robinette JB, Guggenheim SJ. Effect of acetylcholine on the early phase of reversible

norepinephrine-induced acute renal failure. Kidney Int 1981; 19: 399-40930 Lherm T, Troché G, Rossignol M. Renal effects of low-dose dopamine in patients with sepsis

syndrome or septic shock treated with catecholamines. Intensive Care Med 1996; 22:213-21931 Girbes ARJ, Patten MT, McCloskey BV, Van Twuyver P, Zwaveling JH, Smit AJ. The renal and

neurohumoral effects of the addition of low-dose dopamine septic critically-ill patients. IntensiveCare Med 1996; 22 [Suppl 3]: S391

32 Flancbaum L, Choban P, Dasta J. Quantitative effects of low-dose dopamine on urine output inoliguric surgical intensive care unit patients. Crit Care Med 1994; 22: 61-66

33 Duke GJ, Briedis JH, Weaver RA. Renal support in critically ill patients: Low-dose dopamine orlow-dose dobutamine? Crit Care Med 1994; 22:1919-1925

34 Baldwin L, Henderson A, Hickman P. Effect of postoperative low-dose dopamine on renal func-tion after elective major vascular surgery. Ann Intern Med 1994; 120: 744-747

35 Chertow GM, Sayegh MH, Allgren RL, Lazarus JM. Is the administration of dopamine associatedwith adverse or favorable outcomes in acute renal failure? Am J Med 1996; 101: 49-53

Acknowledgments

We are indebted to mrs. A. Drent-Bremer and mrs. M. van Kammen from thedepartment of nephrology, for their skilful assistance while performing renal haemo-dynamic assessments. Dr. Hoogenberg is supported by a grant from the Dutch DiabetesFoundation.

Department of Endocrinology, State University, Groningen, The Netherlands1

Acta Endocrinologica (Copenhagen) 1993; 129: 151-157

CHAPTER 7

EFFECT OF GROWTH HORMONE (GH) AND INSULIN-LIKE GROWTH FACTOR-I ON URINARY ALBUMIN

EXCRETION: STUDIES IN ACROMEGALY AND GH DEFICIENCY

K. Hoogenberg , W.J. Sluiter and R.P.F. Dullaart1 1 1

Glomerular hyperfiltration is a characteristic feature of acromegaly, but it is uncertain whetheralbuminuria is elevated in this disease. To investigate the role of abnormal growth hormone (GH) andinsulin-like growth factor 1 (IGF-1) levels on urinary protein excretion we measured overnight urinaryalbumin excretion rate (Ualb.V) and creatinine clearance in 14 acromegalic patients with metabolicallyactive disease (fasting GH>5 µg/l) and IGF-1>2.2 kU/l), 8 GH deficient patients and 20 controlsubjects. Ualb.V was higher in the acromegalic patients (8.4 (4.2-68.2) µg/min; median (range)) thanin the GH deficient patients (2.0 (0.9-5.9) µg/min, p<0.001) and control subjects (3.3 (1.0-7.8) µg/min,p<0.01). Five acromegalic patients had Ualb.V levels above the normal upper normal limit of 10µg/min. Only one patient with concomitant untreated hypertension had persistent microalbuminuria.Creatinine clearance was also higher in the acromegalic patients (p<0.05) and lower in the GHdeficient patients (p<0.05) than in the control subjects. In 11 of these acromegalic cases the loweringof GH by 63% and of IGF-1 by 48%, following treatment with the somatostatin analogue (n=10) orspontaneous pituitary infarction (n=1), reduced Ualb.V by 29% to 4.9 (3.1-45.2) µg/min (p<0.01).Among the acromegalic patients (25 observations) Ualb.V was related to GH (r=0.61, p<0.01), IGF-1(r=0.57, p<0.01) and creatinine clearance (r=0.54, p<0.01). In conclusion, circulatory GH and IGF-1levels influence albuminuria. Since persistent microalbuminuria is uncommon in acromegaly, it isunlikely that GH elevations alone predispose to clinically important glomerular damage.

Introduction

The appearance of microalbuminuria is an early and characteristic sign of renalinvolvement in patients with (insulin-dependent) diabetes mellitus (IDDM) [1]. Renalhaemodynamic factors are thought to play an important role in the development ofglomerular damage in experimental diabetes mellitus, renal ablation and intrinsic renaldisease [2,3]. Although not consistently reported, an increased glomerular filtration rate(GFR) has also been found to precede the development of microalbuminuria, overtproteinuria and the subsequent decline in renal function in IDDM patients [4-6]. Glo-merular hyperfiltration is a feature of only few other human diseases. In acromegaly anincreased renal function is well recognized [7-10]. Despite similarities in renalhaemodynamic changes between acromegaly and IDDM [10,11], renal insufficiency

Chapter 7102

Table 1. Clinical and biochemical characteristics.

Acromegalic patients GH deficient Control patients subjects

pretreatment post-treatment change (n=14) (n=11) (n=11) (n=8) (n=20)

Age (years) 42(24-67) 42(25-67) 40(26-56) 39(19-63)Female/male 6/8 6/5 4/4 8/12MAP (mmHg) 93(72-130) 92(77-20) -4(-10 to 3) 96(92-107) 99(73-110)Hypertension(n) 2 2 0 0CrCl (ml/min 125(85-182) 112(87-138) -8(-16 to -1) 86(56-100) 100(83-148)

a,c b d a

per 1.73m )2

GH (µg/l) 16.8(6.5-69.5) 5.2(0.5-13.9) -12(-26 to -7) <0.5 ndc b e

IGF-1 (kU/l) 4.8(2.2-13.8) 2.1(0.2-6.7) -2(-7 to -1) 0.15(0.1-0.2) ndc b e

Data in median(range); changes within acromegalic group in median(95%confidence interval).GH:growth hormone; IGF-1:insulin-like growth factor-I; MAP:mean arterial pressure; CrCl:creatinine clearance, nd: not determined; p<0.05 from control subjects; p<0.01 and

a b

p<0.001 from GH deficient patients; p<0.05 and p<0.01 from pretreatment.c d e

seems uncommon in acromegaly [12], whereas in contrast 30 to 40% of diabetic patientsdevelop nephropathy [13].

Growth hormone (GH) stimulates renal haemodynamics indirectly through an effecton insulin-like growth factor 1 (IGF-1) [14,15]. The effects of circulatory GH and IGF-1on urinary albumin excretion are not well studied. In acromegaly no correlation could bedemonstrated between urinary GH and albumin excretion [9], but recent observationsshow that IGF-1 infusion increases albuminuria in healthy subjects [16].

To investigate the role of abnormal GH and IGF-1 levels in urinary protein excretionwe measured albuminuria in patients with active acromegaly, severe GH deficiency andcontrol subjects. In acromegalic patients the influence of effective GH lowering therapyon albuminuria was also studied.

Subjects and methods

Subjects

All subjects consented to participate in the study which was approved by the localmedical ethics committee. Fourteen adult patients with active acromegaly, 8 adult patientswith severe GH deficiency and 20 control subjects were included. The diagnosis of activeacromegaly was based on a fasting serum GH concentration >5µg/l that did not decreaseto <2µg/l after 100 g glucose, and a plasma IGF-1 level >2.2 kU/l [17]. Two potentiallyeligible acromegalic patients were not studied because of renal tract abnormalities. Thepretreatment acromegalic group was composed of 6 patients who had not undergone anyprevious treatment for acromegaly and 8 patients in whom surgery, irradiation,bromocriptine or a combination of them had failed to normalize GH and IGF-1 levels.None of the participants used GH lowering medical treatment at the start of the study.

Growth hormone, insulin-like growth factor 1 and albuminuria 103

Eleven of these 14 patients could be prospectively studied and were included in the post-treatment acromegalic group. They were reevaluated 3 to 4 months after effectivereduction of GH and IGF-1 levels. In 10 patients this was accomplished by subcutaneousinjections of the somatostatin analogue, octreotide (Sandostatin®, Sandoz, Basel,Switzerland), in doses of 300 to 600 µg daily. In one patient pituitary infarction shortlyprior to planned therapy obviated the need for octreotide treatment. Severe GH deficiencywas considered to be present if peak GH after insulin-induced hypoglycaemia was <1.5µg/l and if IGF-1 was <0.34 kU/l [18]. GH deficiency was due to primary hypopituitarismin 6 patients and to non-functioning pituitary adenoma in 2 patients. The acromegalic andGH deficient patients received supplementation doses of thyroxine, cortisone acetate andsex steroids if necessary.

Clinical and laboratory measurements

In the acromegalic and GH deficient patients fasting venous blood was obtained toassay GH (2 occasions) and IGF-1. Precisely timed overnight urine samples were collectedunder sterile conditions to measure urinary albumin excretion rate (Ualb.V) and creatinineclearance. Three urine samples were obtained from each acromegalic and GH deficientpatient. The control subjects collected one specimen. Creatinine clearance was used as aparameter of renal function. Blood pressure was measured using the auscultatory method.Mean arterial pressure (MAP) was calculated as a systolic pressure+ b diastolic pressure.In the post-treatment acromegalic group the procedures were repeated at follow-up.Overnight Ualb.V has been previously established to range from 1.6 to 10.0 µg/min (2.5to 97.5 percentile) in 50 healthy subjects in our clinic. Micro-albuminuria was definedth

as Ualb.V>20 µg/min [1]. Hypertension was defined as systolic blood pressure >160mmHg and/or diastolic blood pressure >95 mmHg.

Serum GH was measured by radioimmunoassay (Farmos Diagnostica, Turku,Finland). The lower detection limit was 0.5 µg/l. Plasma IGF-1 was assayed with a kitpurchased from Nichols Institute of Diagnostics (San Juan Capistrano, CA, USA).Reference values for adults range from 0.34 to 2.2 kU/l. Urinary albumin was measuredby radioimmunoassay (Diagnostics Products Corporation, Apeldoorn, The Netherlands).The intra- and interassay coefficients of variation were 2.3% and 7.7% for urinary albuminconcentrations between 0.07 and 60 mg/l [19]. The lower detection limit was 0.07 mg/l.Serum and urinary creatinine and serum albumin were measured on a SMAC IIautoanalyzer (Technicon Instruments Inc., Tarrytown, NY, USA).

Statistical analysis.

Data are expressed as median (range). Within group changes in variables are givenas median (95% confidence interval (CI)). Parameters from each individual were averagedfor analysis. Between group differences in parameters were assessed by Kruskal-Wallisanalysis of variance. Adjustment for multiple comparisons was carried out using Duncan'smethod. Paired Wilcoxon tests were used to analyse within group changes. Spearman'srank correlation was performed to calculate correlations. Multiple regression analysis wasapplied to evaluate the independent contribution of parameters. A two-sided p-value<0.05was taken as significant.

100

50

Ual

bV (

g/m

in)

10

5

1

0.5

20

2

,

x ,

acromegalic patie

nts

acromegalic patie

nts

pretreatm

ent

posttreatm

ent contro

l

subjects

GH deficient

patients

Chapter 7104

Table 2. Parameters of urinary albumin excretion.

Acromegalic patients GH deficient Controlpatients subjects

pretreatment post-treatment change (%)(n=14) (n=11) (n=11) (n=8) (n=20)

Ualb.V 8.4(4.2-68.2) 4.9(3.1-45.2) -29(-46 to -14) 2.0(0.9-5.9) 3.3(1.0-7.8)b,d c f

(µg/min)Ualb.V 7.1(3.2-57.8) 4.4(2.8-50.3) -23(-44 to -11) 2.4(0.8-7.1) 3.1(0.9-7.5)

b,d c f

(µg/min/1.73m )2

FalbCl (10 ) 1.3(0.6-10.8) 0.9(0.5-6.7) -27(-39 to -3) 0.8(0.2-1.6) 0.7(0.2-1.8)-6 a,c e

Data in median (range); % change within acromegalic group in median (95% confidence interval).Ualb.V denotes urinary albumin excretion rate. FalbCl denotes fractional albumin clearance.

a

p<0.05 and p<0.01 from control subjects; p<0.05 and p<0.001 from GH deficient patients; b c d e

p<0.05 and p<0.01 from pretreatment.f

Figure 1. Urinary albumin ex-cretion rate (Ualb.V) in G acro-megalic patients pretreatment(n=14), G acromegalic patientspost-treatment (n=11), F controlsubjects (n=20) and M growthhormone (GH) deficient patients(n=8). The bars represent themedian levels. The shaded arearepresents the normal range. †p<0.01 from control subjects; ’p<0.05 and ‡ p<0.001 from GH deficientpatients; § p<0.01 from pre-treatment.

Results

Growth hormone, insulin-like growth factor 1 and albuminuria 105

The clinical and biochemical characteristics of the study participants are given inTable 1. The groups were comparable regarding age and sex distribution. Median bloodpressure levels were comparable, although 2 acromegalic patients had hypertension. Dur-ing the study hypertension was treated with a calcium-entry blocker in one patient, mildhypertension being untreated in the other. One acromegalic patient had concomitant diabe-tes mellitus that was adequately controlled with insulin. Creatinine clearance was higherin the untreated acromegalic patients and lower in the GH deficient patients as comparedwith the control subjects. In the post-treatment acromegalic patients serum GH wasreduced by 63(58-77)% and IGF-1 was reduced by 48(32-69)% (p<0.01 for both). Cir-culatory GH and IGF-1 levels decreased to <5 µg/l and to <2.2 kU/l in 5 and 6 patients,respectively. The reduction in GH and IGF-1 levels was accompanied by a fall in creatini-ne clearance (Table 1). Blood pressure (Table 1) and body weight (change -1.0 (-3.8 to 1.5) kg) were unaltered at follow-up.

Ualb.V was elevated in the group of untreated acromegalic patients and tended to belower in the group of GH deficient subjects as compared with the control subjects (Figure1, Table 2). Ualb.V was above the upper normal limit of 10µg/min in 5 acromegalicpatients (36%). One acromegalic patient with concomitant untreated hypertension hadpersistent microalbuminuria (Ualb.V of 68.2µg/min), whereas Ualb.V was 7.6 and9.1µg/min in the treated hypertensive and the diabetic patient, respectively. In theacromegalic patients urinary albumin excretion was still elevated after correction for body-surface area and when expressed as fractional albumin clearance (Table 2). At follow upUalb.V was lower in 10 of 11 patients. Ualb.V was also reduced in the patient with pitu-itary infarction. In this post-treatment acromegalic group both Ualb.V and fractional al-bumin clearance were significantly decreased to a median level comparable with that inthe control subjects (Figure 1, Table 2).

Among the acromegalic patients (25 observations) Ualb.V was significantly corre-lated with serum GH (r=0.61, p<0.01, Figure 2A), plasma IGF-1 (r=0.57, p<0.01, Figure2B) and creatinine clearance (r=0.54, p<0.01, Figure 3A). Multiple regression analysisconfirmed that Ualb.V was independently related to serum GH (p<0.02), or alternativelyto plasma IGF-1 (p<0.03), and creatinine clearance (p=0.05). Ualb.V was not related toblood pressure (Figure 3B), age and body-surface area. Fractional albumin clearance wasalso independently related to circulatory GH (p=0.01) and IGF-1 (p<0.01).

Discussion

This study demonstrates that urinary albumin excretion rate is elevated in acromega-ly and tends to be reduced in GH deficiency. Furthermore, urinary albumin excretion wasconsistently reduced after GH and IGF-1 lowering. The positive relations between GH,IGF-1 and albuminuria supports that these hormonal factors are involved in urinaryprotein excretion. However, the elevations in albuminuria that occurred with GH excesswere minor and only 1 of 14 patient had microalbuminuria according to the cut-off levelof 20µg/min, proposed to define incipient nephropathy in IDDM [1].

100 B

50

Ual

bV (

g/m

in)

10

5

1

0.5

20

2

IGF 1 (kU/l)

1 10 5 2 150.1 0.50.2

100 A

50

Ual

bV (

g/m

in)

10

5

1

0.5

20

2

0.5 1 10 5 100 50GH ( g/l)

2 20<0.5

Chapter 7106

Figure 2.Relationships between serum growth hormone (GH), plasma insulin-like growth factor I(IGF-1) and urinary albumin excretion rate (Ualb.V) in G acromegalic patients pretreatment (n=14),G acromegalic patients post-treatment (n=11) and M growth hormone (GH) deficient patients (n=8).The shaded area represents the normal range. A: GH and Ualb.V in acromegalic patients (25 observa-tions) r=0.61, p<0.01; B: IGF-1 and Ualb.V in acromegalic patients (25 observations) r=0.57, p<0.01.

100 B

50

Ual

bV (

g/m

in)

10

5

1

0.5

20

2

80 90 100 110

MAP (mmHg)

70 120 130

N.S.

100 A

50

Ual

bV (

g/m

in)

10

5

1

0.5

20

2

75 100 125 150creatinine clearance

50 175 200(m m m )2in per 1 73.l /

Growth hormone, insulin-like growth factor 1 and albuminuria 107

Figure 3.Relationships between creatinine clearance, mean arterial blood pressure (MAP) and urinaryalbumin excretion rate (Ualb.V) in G acromegalic patients pretreatment (n=14),G acromegalic patients post-treatment (n=11) and M growth hormone (GH) deficient patients (n=8).A: creatinine clearance and Ualb.V in acromegalic patients (25 observations) r=0.54, p<0.01 and inGH deficient patients r=0.26, NS; B: MAP and Ualb.V in acromegalic patients (25 observations)r=0.07, NS and in GH deficient patients r=-0.18, NS.

Chapter 7108

Previous studies have also shown that microalbuminuria is infrequent in acromegaly[9,10]. These observations would agree with the contention that the elevated urinaryalbumin excretion in acromegaly represents a functional rather than a structural phenom-enon.

The mechanisms responsible for the stimulatory effects of GH and/or IGF-1 onalbuminuria are not precisely understood. Haemodynamic factors and permselectivityproperties of the glomerular filtration barrier govern the glomerular passage of albumin[20]. The presence of IGF-1 receptors on all type of glomerular cells represents astructural basis for the in vivo effects of IGF-1 on glomerular function [21]. IGF-1infusion rapidly increases GFR, renal plasma flow and albuminuria in human subjects[16]. Glomerular vasodilation and a rise in the ultrafiltration coefficient are probablyresponsible for these stimulatory effects of IGF-1 on renal haemodynamics [22]. Theimpaired renal vasodilatory response to amino acid infusion in already hyperfilteringpatients corroborates the presence of abnormal glomerular vasodilation in acromegaly[10]. In this study the expected increase in renal function could be demonstrated usingcreatinine clearance as a parameter of glomerular filtration. The positive relation betweencreatinine clearance and albuminuria supports that renal haemodynamic factors play acontributory role in urinary albumin excretion. The present observations also show thatfractional albumin excretion was elevated in acromegaly. In accordance, IGF-1 infusionenhances fractional albumin clearance as well [16]. Thus, the elevated urinary albuminexcretion cannot be fully explained by an increase in renal function per se. Experimentalevidence indicating a direct effect of IGF-1 on glomerular protein handling is not yetavailable, but IGF-1 could alter glomerular permselectivity properties by relaxing themesangium [21]. It cannot be excluded that the observed reduction in albuminuria thatfollowed GH and IGF-1 lowering was to some extent directly caused by the somastatinanalogue, octreotide. However, albuminuria was also decreased after spontaneous GH andIGF-1 normalization and octreotide does not acutely reduce albuminuria in IDDMpatients, despite a rapid decrease in renal function [23]. It is evident that systemic bloodpressure can influence albuminuria. In patients with IDDM a rise in blood pressure andprogression of albuminuria develop concomitantly [24]. Microalbuminuria has also beenobserved in essential hypertension, albeit at higher blood pressure levels [25]. In our studysystemic blood pressure did not have an independent effect on albuminuria in acromegaly,which could have been due to the fact that most studied patients were normotensive. Inthis respect it is noteworthy that the only patient with persistent microalbuminuria haduntreated hypertension. Moreover, we cannot exclude that alterations in tubular proteinreabsorption contributed to the increased albuminuria, but the normal $2-microglobulinexcretion suggests that tubular function is not impaired in acromegaly [9].

In a retrospective histopathological survey no severe glomerular lesions weredocumented in 20 acromegalic cases studied at autopsy [26], consistent with the presentlyobserved minor elevations in albuminuria. Besides acromegaly and IDDM long-standingglomerular hyperfiltration is a feature of subjects with one kidney and patients with type1 glycogen storage disease [27,28]. Microalbuminuria is rare in subjects with a singlekidney [27], but glomerulosclerosis, overt proteinuria and renal insufficiency are commoncomplications of type 1 glycogen storage disease [28]. Taken together, these observationssuggest that glomerular hyperfiltration does not result in important glomerular injury in

Growth hormone, insulin-like growth factor 1 and albuminuria 109

humans, unless additional, yet to be defined factors, are present. In contrast, experimentalanimal studies have shown that mice transgenic for bovine GH and rats bearing GH-producing tumours develop considerable glomerular enlargement, glomerulosclerosis andproteinuria, whereas less severe abnormalities have been observed in experimental modelsof IGF-1 hypersecretion [29,30]. In such animals, GH is extremely elevated to levels notencountered in human subjects with acromegaly. Between species differences insusceptibility to glomerular lesions may possibly also account for these apparentlycontradictory findings.

In conclusion, circulatory GH and IGF-1 levels influence urinary albumin excretion.In active acromegaly slightly elevated levels of albuminuria are observed that can bereduced with GH-lowering measures. Since persistent microalbuminuria is uncommon inacromegaly, it is unlikely that GH elevations alone predispose to clinically importantglomerular damage.

References

1 Mogensen CE. Prediction of clinical diabetic nephropathy in IDDM patients. Alternatives tomicroalbuminuria? Diabetes 1990; 39: 761-767

2 Hostetter TH, Rennke HG, Brenner BM. The case for intrarenal hypertension in the initiation andprogression of diabetic and other glomerulopathies. Am J Med 1982; 72: 375-380

3 Brenner BM, Meyer TW, Hostetter TH. Dietary protein intake and the progressive nature ofkidney disease: the role of hemodynamically mediated glomerular injury in the pathogenesis ofprogressive glomerular sclerosis in aging, renal ablation and intrinsic renal disease. N Engl J Med1982; 307: 652-659

4 Mogensen CE. Early glomerular hyperfiltration in insulin-dependent diabetics and latenephropathy. Scand J Clin Lab Invest 1986; 46: 201-206

5 Lervang HH, Jensen S, Brøchner-Mortensen J, Ditzel J. Early glomerular hyperfiltration and thedevelopment of late nephropathy in Type 1 (insulin-dependent) diabetes mellitus. Diabetologia1988; 31: 723-729

6 Rudberg S, Persson B, Dahlquist G. Increased glomerular filtration rate as a preditor of diabeticnephropathy - an 8 year prospective study. Kidney Int 1992; 41: 822-828

7 Ikkos D, Ljunggren H, Luft R. Glomerular filtration rate and renal plasma flow in acromegaly.Acta Endocrinol (Copenh) 1956; 21: 226-236

8 Corvilain J, Abramow M, Bergans A. Some effects of human growth hormone on renalhemodynamics and on tubular phosphate transport in man. J Clin Invest 1962; 41: 1230-1235.

9 Eskildsen PC, Parving H-H, Mogensen CE, Christiansen JS. Kidney function in acromegaly. ActaMed Scand 1979; Suppl 624: 79-82

10 Dullaart RPF, Meijer S, Marbach P, Sluiter WJ. Effect of a somatostatin analogue, octreotide, onrenal haemodynamics and albuminuria in acromegalic patients. Eur J Clin Invest 1992; 23: 494-502

11 Ter Wee PM, Van Ballegooie E, Roman JB, Meijer S, Donker AJM. Renal reserve filtrationcapacity in patients with Type 1 (insulin-dependent) diabetes mellitus. Nephrol Dial Transplant1987; 2: 504-509

12 Molitch ME. Clinical manifestations of acromegaly. Endocrinol Metab Clin North Am 1992; 21:597-614

13 Kofoed-Enevoldsen A, Borch-Johnson K, Kreiner S, Nerup J, Deckert T. Declining incidence ofpersistent proteinuria in Type 1 (insulin-dependent) diabetic patients in Denmark. Diabetes 1987;36: 205-209

14 Hirschberg R, Rabb H, Bergamo R, Kopple JD. The delayed effect of growth hormone on renal

Chapter 7110

function in humans. Kidney Int 1989; 35: 865-87015 Hirschberg R, Kopple JD. Evidence that insulin-like growth factor I increases renal plasma flow

and glomerular filtration rate in fasted rats. J Clin Invest 1989; 83: 326-33016 Hirschberg R, Brunori G, Kopple JD, Guler H-P. The effects of insulin-like growth factor I on

renal function in normal man. Kidney Int 1993; 43: 387-397 17 Barkan AL. Acromegaly: diagnosis and therapy. Endocrinol Metab Clin North Am 1989; 18: 227-

31018 Salomon F, Cuneo RC, Hesp R, Sönksen PH. The effects of treatment with recombinant human

growth hormone on body composition and metabolism in adults with growth hormone deficiency.N Engl J Med 1989; 321: 1797-1803

19 Dullaart RPF, Roelse H, Sluiter WJ, Doorenbos H. Variability of albumin excretion in diabetes.Neth J Med 1989; 34: 287-296

20 Kanwar YS. Biology of disease. Biophysiology of glomerular filtration and proteinuria. LabInvest 1984; 51: 7-21

21 Hirschberg R, Kopple JD. The effects of growth hormone and insulin-like growth factor I on renalglomerular and tubular function. In: Flyvbjerg AF, Ørskov H, Alberti KGMM, eds. Growthhormone and insulin-like growth factor I in human and experimental diabetes. Chichester, UK:John Wiley & Sons 1993; 229-254

22 Hirschberg R, Kopple JD, Blantz RC, Tucker BJ. Effects of recombinant human insulin-likegrowth factor I on glomerular dynamics in the rat. J Clin Invest 1991; 87: 1200-1206

23 Stegeman CA, Dullaart RPF, Meijer S, Marbach P, Sluiter WJ. Acute renal effects of somatostatinanalogue, octreotide, in insulin-depedent diabetic patients: antagonism by low dose glucagon.Diab Nutr Metab 1993; 6: 87-95

24 Mogensen CE, Christensen CK. Blood pressure changes and renal function in incipient nephropat-hy and overt diabetic nephropathy. Hypertension 1985; 7 [Suppl 2]: 64-73

25 Christensen CK, Krusell LR, Mogensen CE. Increased blood pressure in diabetes: essentialhypertension or diabetic nephropathy. Scand J Clin Lab Invest 1987; 47: 363-370

26 Newbold KM, Howie AJ, Girling AJ, Kizaki T, Bryan RL, Carey MP. A simple method forassessment of glomerular size and its use in the study of kidneys in acromegaly and compensatoryrenal enlargement. J Pathol 1989; 158: 139-146

27 Schmitz A, Christensen CK, Christensen T, Sølling K. No microalbuminuria or other adverseeffects of long standing hyperfiltration in humans with one kidney Am J Kidney Dis 1989; 13:131-136.

28 Chen Y-T, Coleman RA, Scheinman JI, Kalbeck PC, Sidbury JB. Renal disease in Type 1glycogen storage disease. N Engl J Med 1988; 318: 7-11

29 Doi T, Striker LJ, Gibson CC, Agodoa LYC, Brinster RL, Striker GE. Glomerular lesions in micetransgenic for growth hormone and insulinlike growth factor-I. Am J Pathol 1990;137: 541-552

30 Kawaguchi H, Itoh K, Mori H, Hayashi Y, Makino S. Renal pathology in rats bearing tumour-secreting growth hormone. Pediatr Nephrol 1991; 5: 533-538

Departments of Internal Medicine, State University Hospital, Groningen; Free University1 2

of Amsterdam, The Netherlands. Transplant Proc 1994; 26: 505-507

CHAPTER 8

INSULIN-LIKE GROWTH FACTOR I AND ALTEREDRENAL HAEMODYNAMICS IN GROWTH HORMONE

DEFICIENCY, ACROMEGALY AND DIABETES MELLITUS

K. Hoogenberg , P.M. ter Wee , A.G. Lieverse , 1 2 1

W.J. Sluiter and R.P.F. Dullaart 1 1

Growth hormone (GH) influences renal haemodynamics indirectly by enhancing the synthesisof insulin-like growth factor I (IGF-I). In patients with acromegaly increases in glomerular filtrationrate (GFR) are well correlated with IGF-I levels. Circulatory GH-levels have also been implicated inthe elevated GFR of IDDM patients, but IGF-I levels are often low in these patients. In the presentstudy, we found IGF-I levels to be highly correlated with renal haemodynamics parameters inGH-deficient, acromegalic and healthy subjects, whereas this relationship was not present in normo-hyperfiltering IDDM groups. In contrast, in all groups amino acid stimulated renal function wasinversely correlated with baseline renal haemodynamics, indicating similar renal haemodynamicabnormalities in hyperfiltering IDDM and acromegalic patients, despite large differences in plasmaIGF-I levels. It is suggested, therefore, that the lack of a relationship between plasma IGF-I and renalfunction in IDDM does not exclude the role of elevated GH levels in diabetic glomerularhyperfiltration.

Introduction

Evidence has accumulated that glomerular hyperfiltration plays an important role inthe development of glomerular damage in experimental diabetes mellitus [1] and possiblyin human diabetes as well [2]. An increased glomerular filtration rate (GFR) is also acharacteristic feature of acromegaly [3]. Growth hormone (GH) influences renalhaemodynamics indirectly by enhancing the synthesis of insulin-like growth factor I (IGF-I), which has a direct stimulatory effect on renal function [4]. Since circulatory levels ofGH are frequently elevated in patients with insulin-dependent diabetes mellitus (IDDM),especially during poor metabolic control, it is possible that circulatory GH plays apathogenetic role in diabetic glomerular hyperfiltration. However, no differences indiurnal GH-profile could be demonstrated between normo- and hyperfiltering diabeticpatients [5]. In contrast, recent studies show a positive correlation between GH-releasinghormone-stimulated [6] or exercise-stimulated GH levels [7] and renal haemodynamics inType 1 diabetes mellitus, independently of metabolic control.

In the present study we therefore investigated the relationship between plasma IGF-I

Chapter 8112

and GFR as well as effective renal plasma flow (ERPF) among GH-deficient, acromegalic,normo- and hyperfiltering IDDM patients and healthy control subjects. As the assessmentof renal reserve filtration capacity using amino acids (AA) infusion has been demonstratedto provide insight into the presence of glomerular vasodilation associated with abnormalGH secretion [3,8] and diabetes mellitus [9], the AA-induced changes in renalhaemodynamics in these patients were compared.

Subjects and methods

The study was approved by the local medical ethics committee and all participantsconsented to the procedure. Fasting GFR and ERPF were determined simultaneously,using I-iothalamate and I-hippuran, respectively [9]. On the second day, renal reserve125 131

filtration capacity was measured as the % increment of GFR and ERPF after 17-hours ofamino acids infusion (Vamin-N®, 7% weight/volume, Kabi Vitrum, Limoges, France; in-fusion rate 83 ml/h). All subjects had an age between 21 and 55 years. Only subjectswithout hypertension (systolic blood pressure <160 mmHg and diastolic blood pressure<95 mmHg) and without blood pressure lowering or non-steroidal anti-inflammatory med-ication participated. The IDDM patients did not have clinical proteinuria and had a diseaseduration of at least 5 years.

The relationships between plasma IGF-I and baseline GFR and ERPF were deter-mined in 8 GH-deficient patients [8], 7 patients with active acromegaly who were studiedbefore and after treatment with the somatostatin analogue octreotide [3], 17 normo- and9 hyperfiltering Type 1 diabetic patients (GFR >130 ml/min per 1.73m ) and 6 healthy2

controls. Renal reserve filtration capacity was compared among these GH deficient andacromegalic patients, another 6 normo- and 6 hyperfiltering diabetic patients and 12healthy controls [9]. Serum GH and plasma IGF-I concentrations were measured byradioimmunoassays. Data are given as mean ±SEM. Correlation coefficients were cal-culated with linear regression analysis. A p-value <0.05 was considered significant.

Results

Serum GH levels were higher in the normo- and hyperfiltering IDDM patients thanin the control subjects (4.1±1.2 and 5.9±2.9 µg/l vs 0.8±0.2 µg/l, p<0.001 for both). Incontrast, plasma IGF-I was significantly lower in both IDDM groups as compared tocontrols (0.37±0.02 and 0.36±0.03 kU/l vs 1.08±0.20 kU/l, p<0.02 for both). In the com-bined groups of GH-deficient, acromegalic and control subjects, positive correlations wereobserved between plasma IGF-I, GFR (r=0.87, p<0.001, Figure 1A) and ERPF (r=0.77,p<0.001, Figure 1B). In the combined IDDM groups, these renal haemodynamicsparameters were not related to plasma IGF-I (r=0.08 for GFR and r=0.11 for ERPF).

AA infusion significantly increased GFR in GH-deficient patients (p<0.001), healthysubjects (p<0.01), normofiltering IDDM (p<0.01) and treated acromegalic patients(p<0.02) but not in patients with active acromegaly and hyperfiltering IDDM patients(Figure 2A). ERPF was stimulated in GH-deficient patients (p<0.001), healthy subjects(p<0.05) and treated acromegalic patients (p<0.02), whereas there was no

180

140

100

60

0.1

GF

R

0.5 5.0

80

IGF-I (kU/l)

160

120

A

900

600

400

200

0.1 0.5 5.0

300

800

500

B

700

IGF-I (kU/l)

1.00.05

0.05 1.0

(mm

m)

2in

per

173.

l/E

RP

F(m

mm

)2

inpe

r1

73.l/

IGF and renal reserve filtration capacity 113

Figure 1.Relationships of plasma IGF-I with GFR (A) and ERPF (B) in healthy sub-jects ‘, GH-deficient patients F, acromegalic patients before M and after F treat-ment, normofiltering G and hyperfiltering G diabetic patients.

25%

15%

5%

-5%

80 170

A

AA

-indu

ced

incr

ease

in G

FR

from

bas

elin

e (%

)

25%

15%

-5%

-15%

B

AA

-indu

ced

incr

ease

in E

RP

F fr

om b

asel

ine

(%)

5%

90 100 110 120 130 140 150 160

300 350 400 450 500 550 600 650 700

0%

0%

hyperfiltering

diabetic

untreated acromegalic

normofiltering diabetic

treated acromegalic

controlGH deficient

hyperfiltering

diabetic

untreated acromegalic

normofiltering diabetic

treated acromegalic

control

GH deficient

baseline GFR (m m m )2in per 1 73.l /

baseline ERPF (m m m )2in per 1 73.l /

Chapter 8114

Figure 2. Relationships of baseline renal haemodynamics with AA-induced increments inGFR (A) and ERPF (B) in healthy subjects ‘, GH-deficient patients F, acromegalicpatients before M and after F treatment, normofiltering G and hyperfiltering G diabeticpatients.

increase in patients with active acromegaly as well as in normo- and hyperfiltering IDDM

IGF and renal reserve filtration capacity 115

patients (Figure 2B). Taken all groups together, significant inverse relationships could bedemonstrated between mean baseline renal haemodynamics and the mean AA-induced re-sponses (GFR: r=-0.95, p<0.005; ERPF: r=-0.90, p<0.02).

Discussion

This study showed that baseline GFR and ERPF are positively related to the plasmalevel of IGF-I in GH-deficient, acromegalic and control subjects, but not in diabeticpatients. Both hyperfiltering diabetic and acromegalic patients had an abolished responseto AA infusion. Thus it appears that hyperfiltering IDDM and acromegalic patients sharecomparable renal haemodynamic abnormalities, despite large differences in plasma IGF-Ilevels.

The stimulatory effects of IGF-I on renal haemodynamics are ascribed to glo-merular vasodilation and a rise in the ultrafiltration coefficient [4]. The enhanced re-sponse of GFR and ERPF to AA in GH deficiency, in contrast to acromegaly, supportsthe notion that IGF-I is an important determinant of these renal haemodynamic parame-ters. The renal vasodilatory effects of IGF-I could be mediated by nitric oxide, whereasprostaglandins may have a permissive role [4]. These factors are also thought to play arole in the multifactorial process that leads to diabetic glomerular hyperfiltration [10].

Plasma IGF-I was low in the diabetic patients, despite elevated GH levels. Thisparadox is in accordance with other observations [11], but does not exclude the potentialcontributory role of abnormalities in the GH-IGF-I-system in diabetic glomerularhyperfiltration. For example, increases in IGF binding protein I facilitate the action ofIGF-I on aortic smooth muscle cells [12]. Thus alterations in IGF-I binding proteins indiabetes mellitus may influence its biological activity. Furthermore, augmented accumu-lation of IGF-I [11] and increased expression of kidney IGF-I receptors [13] have beenfound in the diabetic kidney, possibly resulting in an enhanced renal haemodynamiceffect of IGF-I in the diabetic state. Therefore, the lack of a relationship between plasmaIGF-I and renal function in IDDM does not contradict that elevated circulatory GH levelscould play a contributory role in diabetic glomerular hyperfiltration. Obviously, furtherexperiments are needed to clarify the impact of abnormalities in the GH-IGF-I-axis ondiabetes-related changes in renal function.

References

1 Hostetter TH, Rennke HG, Brenner BM. The case for intrarenal hypertension in the initiationand progression of diabetic and other glomerulopathies. Am J Med 1982; 72: 375-380

2 Rudberg S, Persson B, Dahlquist G. Increased glomerular filtration rate as a predictor of diabeticnephropathy. An 8-year prospective study. Kidney Int 1992; 41: 822-828

3 Dullaart RPF, Meijer S, Marbach P, Sluiter WJ. Effect of a somatostatin analogue, octreotide, onrenal haemodynamics and albuminuria in acromegalic patients. Eur J Clin Invest 1992; 23: 494-502

4 Hirschberg R, Kopple JD. The effects of growth hormone and insulin-like growth factor I onrenal glomerular and tubular function. In: Flyvbjerg AF, Ørskov H, Alberti KGMM, eds. Growthhormone and insulin-like growth factor I in human and experimental diabetes. Chichester, UK:John Wiley & Sons 1993; 229-254

5 Wiseman MJ, Redmond S, House F, Keen H, Viberti GC. The glomerular hyperfiltration of

Chapter 8116

diabetes is not associated with elevated plasma levels of glucagon and growth hormone.Diabetologia 1985; 28: 718-721

6 Blankestijn PJ, Derkx FHM, Birkenhäger JC, Lamberts SWJ, Mulder P, Verschoor L,Schalenkamp MADH, Weber RFA. Glomerular hyperfiltration in insulin-dependent diabetesmellitus is correlated with enhanced growth hormone secretion. J Clin Endocrinol Metab1993;77: 498-502

7 Hoogenberg K, Dullaart RPF, Freling NJM, Meijer S, Sluiter WJ. Contributory roles of glycatedhaemoglobin, circulatory glucagon and growth hormone to increased renal haemodynamics inType 1 (insulin-dependent) diabetes mellitus. Scand J Clin Lab Invest 1993; 53: 821-828

8 Dullaart RPF, Meijer S, Marbach P, Sluiter WJ. Renal reserve capacity in growth hormonedeficient subjects. Eur J Clin Invest 1992; 22: 562-568

9 Ter Wee PM, Van Ballegooie E, Rosman JB, Meijer S, Donker AJM. Renal reserve filtrationcapacity in patients with Type 1 (insulin-dependent) diabetes mellitus. Nephrol Dial Transplant1987; 2: 504-509

10 Bank N. Mechanisms of diabetic hyperfiltration. Kidney Int 1991; 40: 792-80711 Flyvbjerg A. Growth factors and diabetic complications. Diab Med 1990; 7: 387 -39912 Busby WH, Klapper DG, Clemmons DR. Purification of a 31,000-dalton insulin-like growth

factor binding protein from human amniotic fluid. J Biol Chem 1988; 263:14203-1420913 Werner H, Shen-Orr Z, Stannard B, Burguera B, Roberts CT, LeRoith D. Experimental diabetes

increases insulinlike growth factor I and II receptor concentration and gene expression in kidney.Diabetes 1990; 39:1490-1497

Department of Endocrinology and Nephrology , and Department of Radiology , Groningen1 1 2 3

State University Hospital, Groningen, The Netherlands. Scand J Clin Lab Invest 1993; 53: 821-828

CHAPTER 9

CONTRIBUTORY ROLES OF CIRCULATORY GLUCAGONAND GROWTH HORMONE TO INCREASED RENAL

HAEMODYNAMICS IN IDDM PATIENTS

K. Hoogenberg , R.P.F. Dullaart , N.J.M. Freling , 1 1 3

S. Meijer and W.J. Sluiter2 1

The stimulatory effects of growth hormone (GH) and glucagon on renal function are wellknown, but it is uncertain whether these hormones are involved in the increase in renal function,characteristic of IDDM patients. Therefore, the circulatory levels of GH and glucagon were measuredin 10 IDDM patients with an elevated glomerular filtration rate (GFR >130 ml/min per 1.73m ) and2

in 20 age and sex matched normofiltering patients (GFR ranging from 90 to 130 ml/min per 1.73m ).2

In the patients, fasting glomerular filtration rate and effective renal plasma flow (ERPF) weredetermined using I-iothalamate and I-hippuran, respectively, during near-normoglycaemia. On125 131

a separate day, the levels of glucagon and GH were determined in the fasting basal state and afterexercise. Multiple regression analysis disclosed that GFR was positively correlated with HbA1c(r =0.18, p<0.01), glucagon (r =0.14, p<0.03) as well as exercise-stimulated GH (r =0.10, p<0.05).2 2 2

ERPF was independently associated with HbA1c (r =0.24, p<0.005) and glucagon (r =0.18,2 2

p<0.01), whereas renal vascular resistance (RVR) was negatively correlated with stimulated GH(r =0.18, p<0.02). Kidney volume was positively correlated with HbA (r =0.26, p<0.001) and2 2

1

inversely with RVR (r =0.16, p<0.01), but not with glucagon or stimulated GH. The present study2

suggests that circulatory GH and glucagon play a contributory role in the renal haemodynamic changesin IDDM.

Introduction

Elevations in glomerular filtration rate (GFR) are well recognized in insulin-dependent diabetes mellitus (IDDM) and can persist for many years after onset of thedisease [1-4]. In experimental diabetes mellitus glomerular hyperfiltration appears to beinvolved in the pathogenesis of glomerulosclerosis and the progressive loss of kidneyfunction [5]. Some clinical observations also support the suggestion that an elevated GFRmight contribute to the subsequent development of diabetic nephropathy [6], but in otherstudies no such role could be assigned to glomerular hyperfiltration [7].

Moderate hyperglycaemia has been found to increase GFR [8,9], whereas improvedmetabolic control reduces renal haemodynamics as well as kidney volume [2,10-12].Plasma levels of growth hormone (GH) and glucagon, which have a renal vasodilatoryeffect [13,14], are frequently elevated under circumstances of suboptimal metaboliccontrol [15-18]. However, no differences in the diurnal profiles of GH and glucagon could

Chapter 9118

Table 1. Clinical characteristics.

Group H Group N(n=10) (n=20)

Age (years) 39 ± 14 40 ± 12

Sex (M/F) 9 / 1 18 / 2

Duration of disease (years) 24 ± 13 21 ± 10

BMI (kg/m ) 24.0 ± 2.4 23.8 ± 2.32

Retinopathy (O/B/P) 2 / 4 / 4 6 / 8 / 6a

Ualb.V (µg/min) 25 (13 - 44) 15 (12 - 44)b

Ualb.V>20 µg/min 6 6

Mean arterial blood pressure (mmHg) 93 ± 9 95 ± 9

HbA (%) 8.0 ± 1.5 7.2 ± 0.91

Blood glucose (mmol/l) 9.3 ± 2.3 7.7 ± 1.7c

H: IDDM patiens with glomerular filtration rate >130 ml/min per 1.73m ; N: IDDM patients with2

glomerular filtration rate ranging from 90 to 130 ml/min per 1.73m . Data are given as mean±SD,2

except for Ualb.V which is given in median and interquartile ranges. Retinopathy: O: absent; B:a

background; P: proliferative; Ualb.V: urinary albumin excretion rate; Blood glucose: mean ofb c

three 8-sample-based 24-h blood glucose profiles

be demonstrated between normo- and hyperfiltering IDDM patients [19], althoughrecently an increased responsiveness to GH-releasing hormone has been observed inIDDM patients with glomerular hyperfiltration [20]. Thus, the contributory roles of thesehormones in the maintenance of diabetic glomerular hyperfiltration are still controversial.Therefore, we employed an exercise test to stimulate GH physiologically and related GHas well as glucagon levels to renal function and kidney size in IDDM patients withoutclinical nephropathy.

Patients and methods

Patients

The study was approved by the local medical ethics committee and all patientsconsented to the procedure. Inclusion criteria to participate were an onset of disease before30 years of age, glucagon-stimulated plasma C-peptide concentration <0.2 nmol/l, diabetesduration for at least 5 years, no arterial hypertension (systolic blood pressure $160 mmHgand/or diastolic blood pressure $95 mmHg), no orthostatic hypotension, serum creatinine#120 µmol/l, urinary albumin excretion rate (Ualb.V) #200 µg/min on three consecutiveovernight urine samples, and no use of anti-inflammatory and antihypertensive drugs.

The study groups were comprised of 10 patients with glomerular hyperfiltration,defined as a GFR>130 ml/min per 1.73m [21], and 20 normofiltering patients (GFR2

between 90 and 130 ml/min per 1.73m ) (Table 1). The subjects from these groups,2

designated H for the hyperfiltering and N for the normofiltering patients, were individu-ally matched with respect to sex, age and diabetes duration (within 5 years) to exclude any

Glucagon, growth hormone and kidney function 119

influence due to differences in clinical parameters. The proprotion of patients withmicroalbuminuria, as well as retinopathy was not significantly different (Table 1). HbA1cvalues and mean blood glucose concentrations (obtained from the three 8-sample-based24-h blood glucose profiles) tended to be higher in H than in N, but this difference did notreach statistical significance (p=0.14 and p=0.08, respectively; Table 1). None of thepatients consumed a protein or sodium-restricted diet. Besides insulin, no other medicationwas used.

Renal haemodynamics and kidney volume

The patients were studied after an overnight fast, both when measuring kidneyfunction and during the exercise procedure to eliminate the confounding effects of recentnutrient intake. Supine GFR and effective renal plasma flow (ERPF) were determinedsimultaneously using I-iothalamate and I-hippuran, respectively [22]. To minimize125 131

the possible effect of actual glycaemia on GFR and ERPF the patients were studied duringnear-normoglycaemia (blood glucose between 4.4 and 8.3 mmol/l [9]. Starting at 0800 h,a 5% glucose solution was infused intravenously at a rate 1 ml (0.28 mmol)/kg per h towhich regular acting insulin (Velosulin H.M., Novo-Nordisk, Bagsvaerd, Denmark) wasadded in a dose of 1% of the total daily requirements per hour. After a 4 h period tonormalize and stabilize blood glucose levels no further corrections of blood glucose weremade and GFR and ERPF were measured over a 2 h clearance period from 1200 to 1400h. The renal vascular resistance (RVR, in dynes.s/cm per 1.73m ) was calculated by5 2

dividing the mean arterial blood pressure by the renal blood flow (RBF) multiplied by 80,where RBF was estimated from the ERPF and the haematocrit (HTC) using the formula:RBF(l/min per 1.73m ) = ERPF(l/min per 1.73m )/(1-HTC). Kidney volume was2 2

determined by summation of the volumes of both kidneys, obtained by ultrasonography[23].

Exercise test

On a separate day following the renal haemodynamic studies, the patients performeda bicycle-exercise test with a fixed workload of 600 kpm for 20 min. The morning insulindose was withheld. After 2 h rest, blood samples were taken at 1000, 1100, 1120 (i.e.directly at the end of exercise) and 1200 h from a cannula inserted into an antecubital vein.

Laboratory methods

Blood glucose was measured using a Yellow Springs glucose Analyzer (Model 23A,Yellow Springs Inc,. Yellow Springs, Ohio, USA). HbA was determined by col-orimetry1

[24].

300

175

50

40

25

10

15

12

9

gluc

agon

(ng/

l)in

sulin

(mU

/l)gl

ucos

e (m

mol

/l)

10.00 11.00 11.20 12.00h

exercise

*

*15.0

7.5

0grow

th h

orm

one

(g/

l)

*

*

**

A

B

C

D

Chapter 9120

Figure 1. Metabolic assessments during theexercise. Hyperfiltering F, normofiltering MIDDM patients: see text and Table 1. Mea-surements at 1000, 1100, 1120 and 1200 h: 1h before, at the start, directly after and 40min after the exercise. Data are given inmean ± SEM. Differences in both IDDM -groups: * denotes p<0.005 for growth hor-mone from preexercise, ** denotes p<0.01for insulin from baseline.

Urinary albumin was measured byradioimmunoassay (Diagnostic Prod-ucts Corporation, Apeldoorn, TheNetherlands). Serum growth hormone(GH) concentrations were measuredby radioimmunoassay (FarmosDiagnostica, Turku, Finland). Insulin-like growth factor-I (IGF-I) was mea-sured by radio-immunoassay (NicholsInstitute of Diagnostics, San JuanCapistrano, Calif., USA). Plasmaglucagon was measured by radio-immunoassay, using antibody 30K(obtained from Professor RH Unger,Dallas, Tex., USA [25]). Plasma freeinsulin and plasma C-peptide con-centrations were measured by radio-immunoassays.

Statistical analysis

Results are expressed as mean±-SD or mean±SEM (Figures) forparametrically distributed data and asmedian (interquartile ranges) for non-parametrically distributed data. Datawere compared with unpaired t-tests.Within group changes of parameterswere evaluated using analysis of vari-ance. Duncan's method [26] was usedto adjustment for multiplecomparisons. Multiple regressionanalysis was used to disclose the independent contribution of the various metabolicparameters to GFR, ERPF, RVR and kidney volume. P-values less than 0.05 wereconsidered to be significant.

Glucagon, growth hormone and kidney function 121

Table 2. Renal haemodynamics, corresponding plasma glucose and insulin levels, and kidney volume.

Group H Group N (n=10) (n=20)

Glomerular filtration rate (ml/min per 1.73m ) 160 ± 22 109 ± 122 a

Effective renal plasma flow (ml/min1 per 1.73m ) 670 ± 142 470 ± 742 a

Filtration fraction 0.24 ± 0.04 0.24 ± 0.03

Renal vascular resistance (dynes.s/cm per 1.73m ) 6654 ± 1422 9195 ± 16975 2 a

Kidney volume (ml/1.73m ) 332 ± 99 291 ± 542

Plasma glucose (mmol/l) 6.7 ± 1.4 6.1 ± 1.1

Insulin (mU/l) 24.4 ± 11.5 23.9 ± 11.1

Group H: IDDM patients with glomerular filtration rate >130 ml/min per 1.73m ; group N: IDDM2

patients with glomerular filtration rate ranging from 90 to 130 ml/min per 1.73m . Data are given2

as mean±SD. denotes P<0.001 from group N.a

Results

Renal haemodynamic parameters and kidney volume

The renal haemodynamics and kidney volume measurements are given in Table 2.By selection, GFR was higher in H than in N, whereas ERPF was also clearly elevated ingroup H compared with N (p<0.001). The FF was similar in the groups H and N, but RVRwas lower in group H than in group N (p<0.001). The difference in kidney volume wasnot significant. The mean blood glucose and insulin concentrations during the renalhaemodynamic assessments were comparable between the groups (Table 2).

Metabolic assessments during exercise

As expected, the circulatory level of GH increased in response to the exercise(p<0.005, Figure 1A). Plasma glucagon did not significantly change during the procedure(Figure 1B). Plasma insulin levels decreased (p<0.01, Figure 1C). Blood glucose rangedbetween 4.9 and 15.4 mmol/l and tended to increase at the end of the exercise (Figure 1D).No significant differences were found between the two groups in the absolute levels ofgrowth hormone, glucagon, insulin and glucose neither at the separate measurements norin the averaged values of the test (p>0.10 for all comparisons). Baseline plasma IGF-Ilevels were 0.35 ± 0.08 kU/l and 0.36 ± 0.07 kU/l in the groups H and N, respectively(NS).

Correlation analyses

For GH the averaged baseline and exercise-stimulated levels were used separatelyin the regression analysis. For glucagon the averaged values of the whole procedure weretaken since glucagon was not stimulated by exercise.

Simple regression analysis showed significant correlations between stimulated GHas well as glucagon and GFR (Figure 2A and B). The correlation between GFR andHbA1c did not reach statistical significance (r=0.32, p=0.054). ERPF was correlated withHbA1c and glucagon (Figure 2C and D), but not with GH (r=0.28, p=0.13).

200

160

120

14000

10000

6000

2000

1000

800

600

400

200

1000

800

600

400

200

A

C D

E

0.0 12.5 25.0

4.0 6.0 10.08.0 12.0 0 150 300 450HbA (%) glucagon (ng/l)

stimulated GH ( g/l)

GF

RE

RP

FR

VR

80

600

450

300

150

F

4.0 6.0 10.08.0 12.0HbA (%)

kidn

ey v

olum

e

200

160

120

B

0 150 300 450glucagon (ng/l)

80

stimulated GH ( g/l)

1

1

0.25 2.5 25.0

(mm

m)

2in

per

173.

l/(m

mm

)2

inpe

r1

73.l/

ER

PF

(mm

m)

2in

per

173.

l/(m

m)

2pe

r1

73.l

GF

R(m

mm

)2

inpe

r1

73.l/

(m

dyn

ec

ss

m)

25

per

173.

./

Chapter 9122

Figure 2. Relationships between renal haemodynamics and metabolic parameters. Hyperfiltering F,normofiltering M IDDM patients. A: exercise stimulated-serum growth hormone and GFR, r=0.40, p<0.05;B: mean plasma glucagon and GFR, r=0.40, p<0.05; C: HbA1c and ERPF, r=0.42, p<0.02; D: meanplasma glucagon and ERPF, r=0.36, p<0.05; E: exercise-stimulated serum growth hormone and RVR, r=-0.43, p<0.01; F: HbA and kidney volume, r=0.61, p<0.001.1

Glucagon, growth hormone and kidney function 123

Table 3. Determinants of renal haemodynamics and kidney volume as assessed by stepwise multipleregression analyses.

Variable Partial correlation Contribution Significancecoefficient to variance

Determinants of glomerular filtration rate.

HbA1c 0.42 18% p<0.01Glucagon 0.38 14% p<0.03Stimulated GH 0.32 10% p<0.05Multiple-r 0.67 44% p<0.002

Determinants of effective renal plasma flow.

HbA1c 0.48 24% p<0.005Glucagon 0.43 18% p<0.01Multiple-r 0.62 39% p<0.002

Determinants of renal vascular resistance. Stimulated GH 0.43 18% p<0.01

Determinants of kidney volume.HbA 0.51 26% p<0.0011

Renal vascular resistance -0.42 18% p<0.01Multiple-r 0.72 52% p<0.001

RVR was negatively related to GH (Figure 2E). Basal GH levels were not correlated withrenal haemodynamic parameters. Kidney volume was correlated with HbA1c (Figure 2F),but not with glucagon or GH.

Multiple regression analysis disclosed that HbA (r =0.18, p<0.01), glucagon12

(r =0.14, p<0.03) and stimulated GH (r =0.10, p<0.05) independently contributed to the2 2

GFR (multiple r=0.67, p<0.002, Table 3). Also, HbA (r =0.24, p<0.005) and glucagon12

(r =0.18, p<0.01) showed independent relations with the ERPF (multiple r=0.62, p<0.002,2

Table 3). RVR only showed a negative relation with stimulated GH (r =0.18, p<0.01)2

Table 3). Kidney volume was positively related HbA1c (r =0.26, p<0.001) independent2

from the inverse relation with RVR (r =0.16, p<0.01) (multiple r=0.72, p<0.001, Table2

3). Blood glucose levels, IGF-I concentrations and clinical parameters such as age,diabetes duration, body mass index, retinopathy and Ualb.V were not related to renalfunction or kidney size.

Discussion

In this study an exercise test was performed to stimulate GH levels physiologicallyand on both study days the patients were kept fasting to eliminate the effects of recentnutrient intake on hormone levels and renal function. Under these circumstances renalhaemodynamic parameters were significantly related to circulatory levels of plasmaglucagon, exercise-stimulated GH levels as well as HbA1c, whereas the absolute levels ofthese hormonal and metabolic factors were not significantly different between the hyper-and normofiltering IDDM patients. Possible between group differences in these parameters

Chapter 9124

could have been obscured because of the arbitrarily employed definition of glomerularhyperfiltration. A GFR above the upper normal limit of 130 ml/min per 1.73m was2

designated as glomerular hyperfiltration, comparable to a cut-off value of 135 ml/min per1.73m used by others [10]. However, such a ‘control population-based’ definition has the2

limitation that subtle renal haemodynamic alterations in apparently normofiltering patientscannot be discriminated and is likely to represent a minimal estimation of thehyperfiltration phenomenon [27].

In the multiple regression analysis with the renal haemodynamic parameters asdependent variables, glucagon contributed to 14% and 18% of the variances in GFR andERPF, respectively. The contributory effect of stimulated GH to GFR was 10% andamounted to 18% of the variance in RVR. In other studies plasma glucagon and GHconcentrations were not different in normo- and hyperfiltering patients and were notcorrelated with renal haemodynamic parameters [1,4,19]. These negative results couldpossibly be explained by differences in study design, since stimulatory tests were not used[1,4,19] or single basal hormone levels were obtained [4]. Recently, an augmented GHresponse to GH-releasing hormone has been documented in hyperfiltering patients [20]supporting an association between an abnormal GH release and renal haemodynamicchanges in IDDM. Long-term metabolic control, as determined by HbA1c levels, contrib-uted to 18, 24 and 26% of the variances in GFR, ERPF and kidney volume, respectively.In accordance, other cross-sectional and intervention studies showed a relation ofmetabolic control with renal function and kidney size [4,10,11].

Several lines of evidence support that glucagon and GH are important humoraldeterminants of renal function in man. Infusion of glucagon to reach levels as encounteredin poorly controlled IDDM acutely increases both GFR and ERPF in IDDM patients [14].Conversely, the rapid decrease in renal function during the administration of thesomatostatin analogue octreotide is closely related to reductions in plasma glucagon [28]and can be completely prevented by concomitant low-dose glucagon administration [29].Furthermore, renal haemodynamics have been shown to be more responsive to incrementsin plasma glucagon in well-controlled IDDM patients than in normal subjects [14]. Therole of circulatory GH in the maintenance of renal function is illustrated by the fact thatacromegalic and GH-deficient patients show differences in GFR and ERPF of about 50%[30,31]. The abolished renal-stimulatory effects of amino acids in acromegaly as opposedto the enhanced response in GH-deficiency further supports that GH is involved in renalvasodilation [30,31]. GH increases GFR and ERPF indirectly by stimulating IGF-Isynthesis [32,33]. In spite of the high concentrations of GH, plasma IGF-I levels arefrequently found to be reduced in IDDM [34,35]. In agreement, plasma IGF-I levels werelow in both groups of IDDM patients. In the interpretation of these data it should be notedthat plasma IGF-binding proteins could have interfered with the presently usedradioimmunoassay [36] and that free IGF-I concentrations were not measured.Disturbances in IGF-I binding proteins, modulating its biological activity [36,37], oralterations in renal IGF-I accumulation [35] might possibly explain why, in accordancewith other observations [4,38], the low circulatory IGF-I levels were not significantlycorrelated with renal haemodynamics in these IDDM patients. Moreover, an altered orvariable renal haemodynamic responsiveness to circulatory IGF-I in IDDM cannot beexcluded. Indeed, in vitro studies have shown increased expression of IGF-I receptors in

Glucagon, growth hormone and kidney function 125

cultured mesangial cells from diabetic mice [39].In conclusion, the present results suggests that circulatory glucagon as well as GH

play a contributory role in the increase in renal haemodynamics in IDDM patients.

References

1 Ditzel J, Junker K. Abnormal glomerular filtration rate, renal plasma flow, and renal protein excretionin recent an short-term diabetics. Br Med J 1972; 2: 13-19

2 Christiansen JS, Gammelgaard J, Frandsen M, Parving HH. Increased kidney size, glomerularfiltration rate and renal plasma flow in short-term insulin-dependent diabetics. Diabetologia 1981; 20:451-456

3 Mogensen CE, Christensen CK. Predicting diabetic nephropathy in insulin-dependent patients. N EnglJ Med 1984; 311: 89-93

4 Mogensen CE, Christensen CK, Pedersen MM, Alberti KGMM, Boye N, Christensen T, ChristiansenJS, Flyvbjerg A, Ingerslev J, Schmitz A, Ørskov H. Renal and glycemic determinants of glomerularhyperfiltration in normoalbuminuric diabetics. J Diab Comp 1990; 4: 159-165

5 Hostetter TH, Rennke HG, Brenner BM. The case for intrarenal hypertension in the initiation andprogression of diabetic and other glomerulopathies. Am J Med 1982; 72: 375-380

6 Rudberg S, Persson B, Dahlquist G. Increased glomerular filtration rate as a predictor of diabeticnephropathy. An 8-year prospective study. Kidney Int 1992; 41: 822-828

7 Lervang HH, Jensen S, Brøchner-Mortensen J. Early glomerular hyperfiltration and the developmentof late nephropathy in Type 1 (insulin-dependent) diabetes mellitus. Diabetologia 1988; 31: 723-729

8 Wiseman MJ, Mangili R, Alberetto M, Keen H, Viberti GC. Glomerular response mechanisms toglycemic changes in insulin-dependent diabetics. Kidney Int 1987; 31: 1012-1018

9 Dullaart RPF, Meijer S, Sluiter WJ, Doorenbos H. Renal haemodynamic changes in response tomoderate hyperglycaemia in Type 1 diabetes mellitus. Eur J Clin Invest 1990; 20: 208-213

10 Wiseman MJ, Saunders AJ, Keen H, Viberti GC. Effect of blood glucose control on increasedglomerular filtration rate and kidney size in Type diabetes. N Engl J Med 1985; 312: 617-621

11 Feldt-Rasmussen B, Hegedüs L, Mathiesen ER, Deckert T. Kidney volume in Type 1 (insulin-dependent) diabetic patients with normal or increased urinary albumin excretion: effect of long-termimproved metabolic control. Scand J Clin Lab Invest 1991; 51: 31-36

12 Tuttle KR, Bruton JL, Perusek MC, Lancaster JL, Kopp DT, DeFronzo RA. Effect of strict glycemiccontrol on renal hemodynamic response to amino acids and renal enlargement in insulin-dependentdiabetes mellitus. N Engl J Med 1991; 324: 1626-1632

13 Christiansen JS, Gammelgaard J, Frandsen M, Ørskov H, Parving H-H. Kidney function and size inType 1 (insulin-dependent) diabetic patients before and during growth hormone administration for oneweek. Diabetologia 1982; 22: 333-337

14 Parving H-H, Christiansen JS, Noer I, Tronier B, Mogensen CE. The effect of glucagon infusionkidney function in short-term insulin-dependent juvenile diabetics. Diabetologia 1980; 19: 350-354

15 Gerich JE, Tsalikian E, Lorenzi M, Bohanon NV, Gustavson G, Karam JH. Normalization of fastinghyperglucagonemia and excessive glucagon responses to intravenous arginine in human diabetesmellitus by prolonged infusion of insulin. J Clin Endocrinol Metab 1975; 41: 1178-1180

16 Hansen AP. Abnormal serum growth hormone response to exercise in juvenile diabetics. J Clin Invest1970; 49: 1467-1472

17 Tamborlane WV, Sherwin RS, Koivisto V, Hendler R, Genel M, Felig P. Normalization of the growthhormone and catecholamine response to exercise in juvenile-onset diabetic subjects treated with aportable insulin infusion pump. Diabetes 1979; 28: 785-788

18 Asplin CM, Faria ACS, Carlsen EC, Vaccaro VA, Barn RE, Iranmanesh A, Lee MM, Veldhuis JD,Evans WS. Alterations in the pulsatile mode of growth hormome release in men and women withinsulin-dependent diabetes mellitus. J Clin Endocrinol Metab 1989; 69: 239-245

19 Wiseman MJ, Redmond S, House R, Keen H, Viberti GC. The glomerular hyperfiltration of diabetes

Chapter 9126

is not associated with elevated plasma levels of glucagon and growth hormone. Diabetologia 1985;28: 718-721

20 Blankestijn PJ, Derkx FHM, Birkenhäger JC, Lamberts SWJ, Schalekamp MADH, Mulder P,Peperkamp E, Verschoor L, Weber RFA. Glomerular hyperfiltration is associated with enhancedgrowth hormone secretion and increased renal plasma flow. Kidney Int 1991; 40: 342 A

21 Ter Wee PM, Van Ballegooie E, Roman JB, Meijer S, Donker AJM. Renal reserve filtration capacityin patients with Type 1 diabetes mellitus. Nephrol Dial Transplant 1987; 2: 504-509

22 Donker AJM, Van der Hem GK, Sluiter WJ, Beekhuis H. A radioisotope method for simultaneousdetermination of the glomerular filtration rate and effective renal plasma flow. Neth J Med 1977; 20:97-103

23 Jones TB, Riddick LR, Harpen MD, Dubuisson RL, Samuels D. Ultrasonographic determination ofrenal mass and renal volume. J Ultrasound Med 1983; 2: 151-154

24 Flückiger R, Winterhalter KH. In vitro synthesis of haemoglobin A1C. FEBS Lett 1976;71: 356-36025 Harris V, Faloona GR, Unger RH. Glucagon. In: Jaffe B, Behrman H, eds. Methods of hormone

radioimmunoassay, 2nd edn. New York: Academic Press, 1979; 643-65626 Duncan DB. Multiple range and multiple F-test. Biometrics 1955; 11: 1-4227 Viberti GC, Wiseman MJ. The kidney in diabetes: significance of the early abnormalities. Clin

Endocrinol Metab 1986; 15: 753-78228 Pedersen MM, Christensen SE, Christiansen JS, Pedersen EB, Mogensen CE, Ørskov H. Acute effects

of a somatostatin analogue on kidney function in Type 1 diabetic patients. Diab Med 1990; 7: 304-309

29 Stegeman CA, Dullaart RPF, Meijer S, Marbach P, Sluiter WJ. Acute renal effects of somatostatinanalogue, octreotide, in insulin-depedent diabetic patients: antagonism by low dose glucagon. DiabNutr Metab: 1993; 6: 87-95

30 Dullaart RPF, Meijer S, Marbach P, Sluiter WJ. Effect of a somatostatin analogue, octreotide, onrenal haemodynamics and albuminuria in acromegalic patients. Eur J Clin Invest 1992; 23: 494-502

31 Dullaart RPF, Meijer S, Marbach P, Sluiter WJ. Renal reserve capacity in growth hormone deficientsubjects. Eur J Clin Invest 1992; 22: 562-568

32 Hirschberg R, Kopple JD. Evidence that insulin-like growth factor I increases renal plasma flow andglomerular filtration rate in fasted rats. J Clin Invest 1989; 83: 326-330

33 Hirschberg R, Brunori G, Kopple JD, Guler H-P. The effects of insulin-like growth factor I on renalfunction in normal man. Kidney Int 1993; 43: 387-397

34 Schaper NC. Growth hormone in Type 1 diabetic and healthy man. Acta Endocrinol (Copenh) 1990;122 [Suppl 2]: 7-44

35 Flyvbjerg A. Growth factors and diabetic complications. Diab Med 1990; 7: 387-39936 Baxter RC, Martin JL. Radioimmunoassay of growth hormone-dependent insulinlike growth factor

binding protein in human plasma. J Clin Invest 1986; 78: 1504-151237 Suikkari A-M, Koivisto VA, Rutanen E-M, Yki-Järvinen H, Karonen S-L, Seppälä M. Insulin

regulates the serum levels of low molecular weight insulin-like growth factor-binding protein. J ClinEndocrinol Metab 1988; 66: 266-272

38 Johansson B-L, Berg U, Freyschuss U, Hall K, Troel S. Factors related to renal haemodynamics inyoung type-1 diabetes mellitus patients. Pediatr Nephrol 1990; 4: 589-592

39 Oemar BS, Foellner HG, Hodgdon-Anandant L, Rosenzweig SA. Regulation of insulin-like growth factor I receptors in diabetic mesangial cells. J Biol Chem 1991; 266: 2369-2374

AcknowledgmentThe present study was in part supported by a grant (88.744) from the Dutch Kidney

Foundation. We are indebted to M. van Kammen, F. Kranenburg-Nienhuis and A. Drent-Bremer for their skilful assistance.

Department of Endocrinology and Immunochemistry, State University Hospital, Groningen,1 2

The Netherlands. Diabetic Medicine 1996; 13: 651-655

CHAPTER 10

INCREASED URINARY IGG/ALBUMIN INDEX INNORMOALBUMINURIC IDDM PATIENTS:

A LABORATORY ARTEFACT

K. Hoogenberg , P. Visser , J. Marrink , W.J. Sluiter and R.P.F. Dullaart1 2 2 1 1

The previous observation that urinary IgG excretion is increased in normoalbuminuric IDDMpatients is unexplained and could possibly be related to a laboratory phenomenon. When untreatedurine samples were stored at -20 C for 2 to 4 weeks, the IgG/albumin index (IgG clearance divided byo

albumin clearance) was higher in normoalbuminuric IDDM patients than in control subjects(0.91(0.68-1.54), n=27 vs 0.72(0.55-0.79), n=15 (median(interquartile range)), p<0.05). In normo- andmicroalbuminuric IDDM patients the IgG/albumin index was higher in urine samples with glucosethan without glucose (1.16(0.93-1.68), n=11 vs 0.73(0.50-0.91), n=16; p<0.05, and 0.33(0.23-0.60),n=17 vs 0.15(0.10-0.26), n=14, p<0.02 for normo- and microalbuminuric patients, respectively). We,therefore, evaluated the preserving effects of glucose and bovine serum albumin (BSA) on urinary IgGafter 1 hour to 16 weeks of freezing at -20 C in 4 non-diabetic subjects (proteinuria ranging from 0.05o

to 8.0 g/day). Urine samples were either stored without precautions or treated with addition ofphosphate buffer, BSA (1%) and glucose (100 and 300 mM). The weekly decline from 1 to 16 weeksof IgG in the urine aliquots diluted 1:1 with buffered glucose 300 mM and glucose 300 mM + BSA 1%was insignificant, whereas urinary IgG declined with all other storage regimes (p<0.05). These resultssuggest that glucose in urinary specimens of IDDM patients prevents at least in part the loss of urinaryIgG and may thus explain the higher urinary IgG/albumin index when unprocessed urine is storedfrozen before assay. Laboratory precautions are necessary when urinary IgG cannot be measuredimmediately.

Introduction

The glomerular filtration of macromolecules is dependent on their charge, size andstructure, on the properties of the glomerular filtration barrier which govern charge andsize selectivity, as well as on renal haemodynamic factors [1]. In the microalbuminuricstage of diabetic renal disease an important role is attributed to a loss of negative chargeof the glomerular basement membrane, which is probably due to a decrease in thenegatively charged proteoglycan, heparansulphate [2-5]. This impaired charge selectivityleads to enhanced glomerular passage and urinary excretion of negatively chargedmolecules like albumin and IgG4 [1,6,7]. In more advanced stages of diabetic nephropathya concomitant defect in glomerular size selectivity is thought to result in a grosslyincreased urinary excretion of neutrally charged IgG [1]. Therefore, the observation thatthe urinary IgG excretion [3,8] or the IgG/albumin clearance ratio [9] is elevated in

Chapter 10128

normoalbuminuric diabetic patients compared to healthy subjects is puzzling and awaitsfurther explanation. This phenomenon could possibly be related to laboratory conditionsof urine storage before IgG assay. In this respect it is important to note that urinary IgGconcentration decreases within several weeks after freezing of samples at -20 C and thato

addition of bovine serum albumin (BSA) to the urine specimens and buffering urine toneutral pH diminishes this decline in IgG concentration [10].

In the present study we describe the apparent increasing effect of glucosuria onurinary IgG measurement when aliquots were stored at -20 C without precautions.o

Laboratory experiments were carried out to determine the effect of different storageconditions, including addition of glucose and BSA, on urinary IgG measurement.

Patients and methods

Study A: Urinary IgG (UIgG.V) and albumin (Ualb.V) excretion were determinedin overnight and in day-time urine collections of 15 healthy controls (age 25±5 years,group C), in 27 normoalbuminuric (age 30±9 years, diabetes duration 15±10 years, groupD1) and 31 microalbuminuric insulin-dependent diabetic patients (IDDM) (age 43±11years, diabetes duration 22±9 years, group D2). Microalbuminuria was defined as anUalb.V>20 µg/min in an exactly timed overnight urine collection. On the followingmorning, directly after completion of the overnight collection, a one hour urine collectionwas made and venous blood was taken for measurements of blood glucose, serum albuminand IgG at the beginning of this period. In the overnight samples, the IgG/albumin indexwas calculated as the IgG clearance divided by the albumin clearance.

Study B: The effects of laboratory storage conditions on urinary IgG measurementwas evaluated in 4 non-diabetic subjects with a total urinary protein excretion of 0.05, 0.2,2.0 and 8.0 g/day.

In all participants of both studies urinary tract infection was excluded by Dipslidetests. Urine samples for IgG measurement were stored at -20 C without precautions for 2o

to 4 weeks in study A. Freshly collected samples were processed in study B (see below).Urinary IgG was measured by an in-house developed enzyme linked immunosorbent

assay (ELISA), essentially according to Fomsgaard et al. [11]. The following materialswere used: coating: polyclonal goat-antihuman-IgG (gammachain) (Tago, Burlingame,CA, cat no 4100), as conjugate: goat-antihuman-IgG peroxidase (GAHu/IgG (H+L)/PO,Nordic, Tilburg, The Netherlands), enzyme substrate: o-phenylenediamine di-HCl(Eastman Kodak, Rochester, NY, USA). The IgG standard (standard human serum ORDT06/07, Behringwerke AG, Marburg, FRG) was diluted providing a range from 1.5 to 200µg/l. Incubations were carried out at 37 C. The colour development was stopped by addingo

100 µl H SO (0.5 mol/l). The absorbance was measured at 492 nm using a Titertek2 4

Multiscan (Flow Laboratories, Irvine, UK). Results were computed according to Rodbardet al. [12]. The intra- and interassay coefficients of variation (CV) were 5% and 15%,respectively. The lower detection limit was 1.5 µg/l.

To evaluate the effect of laboratory storage conditions on urine IgG measurement(study B) the following experiments were carried out: within 4 hours after voiding urinealiquots were treated (A) undiluted; (B) diluted 1:1 with 40 mM phosphate buffer, 0.1 MNaCl adjusted to pH 7.4 and sodium azide (NaN 0.2%); (C) diluted 1:1 with buffer and2

Glucose and urinary IgG in IDDM. 129

BSA 1% (10 g/l); (D) diluted 1:1 with buffer and glucose 100 mM; (E) diluted 1:1 withbuffer and glucose 300 mM; (F) diluted 1:1 with buffer, glucose 100 mM and BSA 1%;(G) diluted 1:1 with buffer, glucose 300 mM and BSA 1%. IgG was measured in thesesamples before freezing and after 1 hour, and 1, 2, 4, 8 and 16 weeks of storage at -20 C.o

All IgG measurements were performed in quadruplicate. In both studies urine aliquots for albumin measurement were stored at 4 C ando

analyzed within 24 hours. Urinary albumin was measured in duplicate using a commer-cially available double antibody radioimmunoassay (Diagnostic Products Corporation,Apeldoorn, The Netherlands, cat no KHAD ). The intra- and interassay CV's were 2.3%2

and 7.7%. respectively. The lower detection limit was 0.07 mg/l. Serum IgG wasmeasured by nephelometry (Behring Nephelometer Analyzer , Behringwerke AG,TM

Marburg, FRG). The intra- and interassay CV's were 2.4 to 3.6% and 2.5% respectively.Serum albumin was measured on a SMAC-II autoanalyser (Technicon Instruments Inc.,Tarry Town, NY, USA). The intra- and interassay CV were 1.9% and 2.5%, respectively.Blood glucose was measured using a Yellow Springs glucose Analyzer (Model 23A,Yellow Springs Inc,. Yellow Springs, Ohio, USA). Urinary glucose was measured usinga Polarimeter Type 243 (Thorn Automation, Nottingham, UK).

Statistical analysis

Results are expressed as mean±SD or as median (interquartile range) forparametrically and non-parametrically analysed data, respectively. In study A, differencesin urinary albumin excretion, urinary IgG excretion and the albumin/IgG index wereevaluated by Kruskal-Wallis oneway analysis of variance with Duncan's method to correctfor multiple comparisons. Correlation coefficients between urinary and blood glucose andprotein excretion were calculated using Spearman's rank analysis. In study B, mean (±SD)recovery of IgG at each interval was calculated as percentage of the freshly analysedsamples for each of the four urine collections. The overall means and their common SD'sare given in the Table, and the data are compared by Student t-tests. Using linearregression analysis, the weekly percentage decline (±SD) from 1 to 16 weeks wascalculated for each collection. The overall mean declines (± SD) are given in the Table. Atwo-sided p-value less than 0.05 was considered to be significant.

Results

Urinary albumin and IgG excretion (study A)

Overnight Ualb.V and urinary IgG excretion (UIgG.V) were significantly higher ingroup D2 (54.6(41.6-109.5) and 3.10(2.35-8.81) µg/min, respectively) than in groups D1and C (7.1(4.6-10.1) and 4.6(3.0-7.2) for Ualb.V; 1.64(1.02-2.37) and 0.46(0.35-0.66)µg/min for UIgG.V, p<0.01 for all comparisons). Ualb.V was similar in group D1 andgroup C, but UIgG.V was higher in group D1 compared to group C (p<0.05). Since nodifferences were present in serum albumin and IgG concentrations, the overnightIgG/albumin index was higher in group D1 (0.91(0.68-1.54)) than in group C (0.72(0.55-0.79), p<0.05). This index was lower in group D2 (0.27(0.14-0.40), p<0.01) than in groupC. As shown in Figure 1, in both diabetic groups the IgG/albumin index was higher inurine samples with glucose as compared to those without glucose (1.16(0.93-1.68) (n=11)

2.5

2.0

1.5

1.0

0.5

0.0

IgG

/alb

umin

inde

x

no glucosuria

normoalbuminuric IDDM

control subjects microalbuminuric IDDM

glucosuria no glucosuria glucosuria

*

Chapter 10130

vs 0.73(0.50-0.91) (n=16), p<0.05, for group D1; 0.33(0.23-0.60) (n=17) vs 0.15(0.10-0.26) (n=14), p<0.02, for group D2). In group D1, the IgG/albumin index in sampleswithout glucose was not different from group C (Figure 1). In the day-time urine collec-tions similar differences were present (not shown). In the combined diabetic patients(n=58), the day-time IgG/albumin index was positively correlated with urinary glucoseconcentration (r = 0.33, p<0.05), urinary glucose excretion (r =0.44, p<0.01) and bloods s

glucose (r = 0.56, p<0.01). s

F i g u r e1 . T h eovernight IgG/albumin index in non-diabetic control subjects (group C,�), in normoalbuminuric IDDMpatients (group D1) without (Q) and with glucosuria (#), and in microalbuminuric IDDM patients(group D2) without (Î) and with glucosuria (>). Bars represent median values. * denotes p<0.05 fromgroup C and group D1 without glucosuria; † denotes p<0.01 from groups C and D1 with and withoutglucosuria ‡ denotes p<0.02 from group D2 without glucosuria.

Effects of laboratory storage condition on urinary IgG measurement (study B)

In the 4 non-diabetic patients the urinary IgG concentration was 0.68, 1.52, 12.1 and109 mg/l in the immediately processed samples. The IgG concentrations after storage at

Glucose and urinary IgG in IDDM. 131

Table 1. Recovery of urinary IgG after different storage conditions during variable intervals at -20 C in 4 non-diabetic subjects with proteinuria. o

Storage time weeklyfresh 1h 1wk 2wk 4wk 8wk 16wk decline %

Dilution (1:1) % % % % % % %

A undiluted urine mean 100 98 83* 86* 81* 66* 39* 3.18*SD 6.2 7.2 7.9 6.5 8.5 14.6 24.9 0.92

B buffer alone mean 100 98 73* 73* 65* 60* 49* 1.65*SD 8.3 10.7 10.8 7.7 4.6 5.8 8.9 0.25

C +BSA 1% mean 100 97 83* 85* 80* 73* 65* 1.37*SD 3.8 7.4 7.4 7.7 10.0 7.1 8.7 0.34

D +glucose mean 100 95 86* 86* 77* 78* 65* 1.35* 100 mM SD 3.6 6.6 13.2 8.1 6.6 8.8 14.7 0.37

E +glucose mean 100 99 95 96 95 97 86* 0.61 300 mM SD 3.8 6.1 8.2 8.5 12.9 11.5 8.4 0.43

F +BSA 1% mean 100 97 94 96 95 91* 86* 0.66* +glucose 100mM SD 3.1 6.2 6.9 8.8 13.8 6.7 7.2 0.28

G +BSA 1% mean 100 101 95 100 96 95 91* 0.39 +glucose 300mM SD 4.3 5.1 6.4 12.5 10.7 11.3 7.1 0.29

Data are given as overall mean and common SD in % of measurements in fresh urine andpercentage decline per week (from 1 to 16 weeks) and common SD. * denotes p<0.05 as comparedto fresh urine or as compared to zero-decline.

-20 C are given as percentage recovery of the initial value (Table 1). No immediate effecto

of 1 hour of freezing could be demonstrated. In contrast, the IgG content after 1 week ofstorage declined significantly in the untreated (p<0.01, A), buffered (p<0.005, B),buffered and BSA 1% diluted (p<0.01, C), buffered and glucose 100 mM diluted (p<0.05,D) urine specimens (Table 1). Moreover, this decline was progressive, since the weeklydecline from 1 to 16 weeks of storage was significant in these storage regimes (p<0.01 forA,B,C and D). In glucose 300 mM diluted (E), glucose 100mM and BSA 1% diluted (F)and glucose 300 mM and BSA 1% diluted (G) urine specimens a small but significantdecline in IgG was noted after 16 weeks. The weekly decline in IgG was not significant inthe urine specimens diluted with glucose 300 mM (p<0.20, E) or glucose 300 mM andBSA 1% (p<0.20, G), but significant if diluted with glucose 100 mM and BSA 1%(p<0.05, F).

Discussion

In addition to documentation of microalbuminuria, measurement of IgG withsensitive assays will gain insight in the glomerular permelectivity defects underlying thedevelopment and evolution of diabetic renal disease. In a cross-sectional study weconfirmed recent unexplained findings showing that normoalbuminuric IDDM patients

Chapter 10132

have a higher urinary IgG excretion and a higher IgG/albumin index compared to controls[3,8,9]. Of importance, urine samples for IgG measurement were stored at -20 C withouto

precautions in the study by Deckert et al [3] and in the present study, whereas in the otherreports [8,9] the urinary storage conditions were not mentioned. The present data showedthat among both normo- and microalbuminuric IDDM patients the IgG/albumin index washigher in samples with glucose compared to samples without glucose. Moreover, theIgG/albumin index was positively correlated with glucosuria. This unexpected findingprompted us to evaluate the effect of laboratory storage conditions on IgG measurement.In agreement with other data, storage at -20 C without precautions resulted in ao

pronounced loss of IgG [10], but in contrast with that study, addition of BSA alone did notprevent the decline in IgG. Only addition of 300 mM glucose (preferably in combinationwith BSA 1%) to the specimens prevented a rapid decline in IgG. Moreover, such highurinary glucose concentrations can be encountered in vivo during periods of severehyperglycaemia. It is, therefore, plausible that the apparently increasing effect of glucoseon IgG excretion in IDDM is attributable to a diminished or absent loss of IgG duringstorage without precautions. An alternative explanation, such as a general abnormality inglomerular or tubular IgG handling in diabetes mellitus, is difficult to envisage, since theglomerular permselectivity is considered to be intact [13] and $ -microglobulin excretion2is unchanged [1] in normo-albuminuric IDDM patients. Indeed, we were unable toreproduce an elevated IgG/albumin index in normoalbuminuric IDDM patients when usingadequate storage procedures [unpublished observations]. Deckert et al. were also unableto find a difference in the IgG/albumin index in normoalbuminuric IDDM patientscompared to controls if urine samples were stored at -40 C after predilution with 5% BSAo

in phosphate buffer [14, personal communication B. Feldt-Rasmussen].A gradual loss in urinary IgG during storage at -20 C was observed, without ano

immediate effect of freezing and thawing. Denaturation or aggregation of globulins intomultimeres of 2 to 4 molecules at low temperatures, which might interfere with epitoperecognition in the enzyme-linked immunosorbent assay, can possibly explain this loss inIgG. The employed glucose concentrations of 100 to 300 mM in the specimens were veryhigh compared with the measured IgG concentrations that ranged from 0.0045 to 0.716mM. This excess of glucose molecules could prevent IgG denaturaton or aggregationduring prolonged freezing. In contrast to IgG, urinary albumin does not decline afterfreezing for 1 week [15], although a period of longer storage is also associated with a lossof albumin [16]. Of note, addition of glucose does not influence the albumin measurementin samples assayed within one week after collection [15].

In conclusion, the presence of glucose in urine samples of IDDM patients mostlikely prevents, at least in part, the loss of urinary IgG when aliquots are frozen withoutprecautions. This may explain the apparently higher IgG/albumin index innormoalbuminuric IDDM patients as compared to healthy subjects. Addition of glucose300 mM plus BSA 1% (1:1) to the urine samples is an appropriate measure to prevent asubstantial decline in IgG. It is, therefore, necessary to use such precautions when it is notpossible to assay IgG immediately.

References

Glucose and urinary IgG in IDDM. 133

1 Viberti CG, Keen H. The patterns of proteinuria in diabetes mellitus. Relevance to pathogenesisand prevention of diabetic nephropathy. Diabetes 1984; 33: 686-692

2 Deckert T, Kofoed-Enevoldsen A, Nøgaard K, Borch-Johnsen K, Feldt-Rasmussen B, Jensen T.Microalbuminuria. Implications for micro- and macrovascular disease. Diabetes Care 1992; 15:1181-1191

3 Deckert T, Feldt-Rasmussen B, Djurup R, Deckert M. Glomerular size and charge selectivity ininsulin-dependent diabetes mellitus. Kidney Int 1988; 33: 100-106.

4 Nakamura Y, Myers DB. Charge selectivity of proteinuria in diabetic glomerulopahty. Diabetes1988; 37: 1202-1211

5 Tamsma JTM, Van den Born J, Bruijn JA, Assmann KJM, Weening JJ, Berden JHM, et al.Expression of glomerular extracellular matrix components in human diabetic nephropathy:decrease of heparan sulphate in the glomerular basement membrane. Diabetologia 1994; 37: 313-320

6 Bangstad H-J, Kofoed-Enevoldsen A, Dahl-Jørgensen, Hanssen KF. Glomerular charge selectiv-ity and the influence of improved blood glucose control in Type 1 (insulin-depedent) diabeticpatients with microalbuminuria. Diabetologia 1992; 35: 1165-1169

7 Pietravalle P, Morano S, Cristina G, Grazia De Rossi M, Mariani G, Cotroneo P et al. Chargeselectivity of proteinuria in Type 1 diabetes explored by Ig subclass clearance. Diabetes 1991; 40:1685-1690

8 Jerums G, Allen TJ, Cooper ME. Triphasic changes in charge selectivity with increasingproteinuria in Type 1 and Type 2 diabetes. Diab Med 1989; 6: 772-779

9 Fox JG, Quin JD, Paterson KR, O'Reilly DStJ, Smith MP, Boulton-Jones JM. Glomerular chargeselectivity in Type 1 (insulin-dependent) diabetes mellitus. Diabetic Med 1995; 12: 387-391

10 Kofoed-Enevoldsen A, Jensen K. Measuring urinary IgG and IgG4 excretion. Clin Chem 1992;37:1136-1137

11 Fomsgaard A, Feldt-Rasmussen B, Deckert M, Dinesen B. Micro-Elisa for the quantification ofhuman urinary IgG. Scand J Clin Lab Invest 1987; 47: 195-198.

12 Rodbard D, Bridson W, Rayford PL. Rapid calculation of radioimmmunoassay results. J Lab ClinMed 1969; 74: 770-781.

13 Mogensen CE. Kidney function and glomerular permeability to macromolecules in early juvenilediabetes. Scand J Clin Lab Invest 1971; 28: 79-90.

14 Deckert T, Kofoed-Enevoldsen A, Vidal P, Nørgaard K, Andreasen HB, Feldt-Rasmussen B.Size- and charge selectivity of glomerular filtration in Type 1 (insulin-dependent) diabeticpatients with without albuminuria. Diabetologia 1993; 36: 244-251

15 Dullaart RPF, Roelse H, Sluiter WJ, Doorenbos H. Variability of albumin excretion in diabetes.Neth J Med 1989; 34: 287-296.

16 Elving LD, Bakkeren JAJM, Jansen MJH, De Kat-Angelino CM, De Nobel E, Van Munster PJJ.Screening for microalbuminuria in patients with diabetes mellitus: Frozen storage of urinesamples decreases their albumin content. Clin Chem 1989; 35: 308-310

Departments of Endocrinology, General Internal Medicine, Physiology and Clinical1 2 3 4

Chemistry, and Isotope Laboratory, Groningen State University, The Netherlands5

J Clin Endocrinol Metab 1995; 80: 3002-3008

CHAPTER 11

ALTERATIONS IN CORTISOL METABOLISM IN IDDMPATIENTS: RELATIONSHIP WITH METABOLIC CONTROL

AND ESTIMATED BLOOD VOLUME AND EFFECT OFANGIOTENSIN CONVERTING ENZYME-INHIBITION

R.P.F. Dullaart , F.L. Ubels , K. Hoogenberg , A.J. Smit , J.J. Pratt , 1 2 1 2 5

J.H.J. Muntinga , W.J. Sluiter and B.G. Wolthers3 1 4

11ß-hydroxysteroid dehydrogenase (11ß-HSD) catalyses the interconversion of cortisol and itsinactive metabolite, cortisone, and protects the mineralocorticoid receptor from activation by cortisol.Sodium and fluid retention is a well documented phenomenon in IDDM patients, but it is unknownwhether alterations in cortisol metabolism contribute to its pathogenesis. Therefore, we evaluated someaspects of cortisol metabolism by measuring urinary metabolites of cortisol and cortisone in 8microalbuminuric and 8 normoalbuminuric IDDM patients and 8 matched control subjects. In bothIDDM groups, the overnight excretion of tetrahydrocortisol (THF), allo-tetrahydrocortisol (allo-THF)and tetrahydrocortisone (THE) was lower than in the control group (p<0.05 to p<0.01). Both the allo-THF/THF ratio, a parameter of 5"/5ß reduction of cortisol, and the cortisol to cortisone metaboliteratio (THF+allo-THF/THE), which reflects the overall direction of the cortisol to cortisoneinterconversion, were lower in the IDDM groups (p<0.05 to p<0.01). In the combined subjects (n=24),allo-THF, allo-THF/THF and THF+allo-THF/THE were inversely correlated with HbA1c (r=-0.69,p<0.001, r=-0.61, p<0.01 and r=-0.58, p<0.01, respectively). Upper arm segmental blood volume,estimated by an electrical impedance technique, was positively correlated with the cortisol to cortisonemetabolite ratio both in the control subjects (r=0.77, p<0.05) and in the IDDM patients in whom it wasmeasured (r=0.56, p<0.05, n=13), whereas the regression line was shifted leftwards in IDDM (i.e. alower ratio at the same blood volume, p<0.03 by analysis of covariance). In 7 microalbuminuric IDDMpatients, the angiotensin converting enzyme-inhibitor, enalapril (10 mg daily for 6 to 12 weeks)resulted in a moderate further lowering of the cortisol to cortisone metabolite ratio (p<0.05). Thepresent data suggest a chronic hyperglycemia-related impairment in the reduction of corticoids totetrahydro metabolites and an imbalance in 11ß-HSD. Altered 11ß-HSD activity is unlikely to beprimarily responsible for the sodium and fluid retention in IDDM. Moreover, an additional mechanismof action of angiotensin-converting enzyme inhibition might be provided by an effect on 11ß-HSDactivity.

Introduction

Corticosteroid hormones play a central role in extracellular sodium and fluidhomeostasis. The mineralocorticoid receptor has an equal affinity for cortisol andaldosterone in vitro, but the distal renal tubule binds almost exclusively mineralocorti-coi-

Chapter 11136

ds in vivo [1-4]. Recently, it has become clear that the mineralocorticoid receptor isprotected from excess circulating cortisol by the enzyme 11ß-hydroxysteroiddehydrogenase (11ß-HSD) [3]. This enzyme, of which several isoforms have beenidentified, catalyses the interconversion of cortisol and its inactive metabolite, cortisone[4]. These observations have provided a link between cortisol metabolism and the regula-tion of volume and sodium homeostasis, and there is increasing evidence to underscore theclinical importance of 11ß-HSD. The type 1 syndrome of apparent mineralocorticoidexcess (AME), characterized by sodium retention and hypokalaemic hypertension despitesuppressed plasma aldosterone, has been shown to be caused by impaired 11ß-HSDactivity [5,6]. Similarly, glycyrrhethinic acid, the main metabolite of glycyrrhizic acidpresent in licorice, inhibits 11ß-HSD, thus allowing more cortisol to gain access to themineralocorticoid receptor [7]. In these situations defective dehydrogenase activity of 11ß-HSD is reflected by an abnormal proportion of urinary cortisol to cortisone metabolites [5-8].

In diabetes mellitus an increase in exchangeable sodium accompanied by extracellul-ar volume expansion has been well documented and might contribute to the blood pressurerise associated with microalbuminuria [9-15]. The mechanisms responsible for thisabnormal sodium retention are still incompletely understood, although enhanced renaltubular sodium reabsorption most probably plays an important role [15-16]. Plasmaaldosterone levels are usually found to be within normal limits in insulin-dependentdiabetic (IDDM) patients [14,15]. Mild hypercortisolism, possibly associated with thepresence of complications, has been found in diabetic patients [17-19], but it is unclear inhow far the diabetic state leads to alterations in cortisol metabolism and whether possiblechanges in the cortisol to cortisone conversion influence volume homeostasis.

Against this background we measured urinary cortisol and cortisone metabolites innormo- and microalbuminuric IDDM patients and in control subjects. Parameters ofcortisol metabolism were related to metabolic control and upper arm segmental bloodvolume, estimated by an electrical impedance technique. Since recent experimental dataindicate that 11ß-HSD can be stimulated by angiotensin converting enzyme (ACE)-inhibition [20], the effect of this treatment was studied in microalbuminuric patients.

Subjects and methods

Subjects

The protocol was approved by the local medical ethics committee and writteninformed consent was obtained from all participants. The subjects also participated inanother study (investigating the effects of catecholamines on renal function) which will bereported elsewhere. That study was started after completion of the data collection and doesnot interfere with the present results. The IDDM patients had ketosis-prone diabetesmellitus and glucagon-stimulated C-peptide levels were <0.2 nmol/l in all of them. Noneof the participants had autonomic neuropathy. Microalbuminuria was defined as urinaryalbumin excretion rate (Ualb.V) between 20 and 200 µg/min in at least 2 of 3 overnighturine collections obtained over a one year period [21]. Ualb.V was assessed in the absenceof urinary tract infection as indicated by a sterile urine culture. Three groups of subjectswere studied: 8 IDDM patients with normoalbuminuria, 8 IDDM patients with

Cortisol metabolism in IDDM 137

microalbuminuria and 8 healthy control subjects, individually matched for sex and age.Except for IDDM the participants did not have other diseases. In particular, all subjectswere clinically and biochemically euthyroid, as indicated by a normal TSH level (0.3 to5.0 mU/l). None of the participants had severe obesity, defined as body mass index (BMI)>30 kg/m . The mean±SD BMI was 23.2±2.9, 24.5±3.0 and 25.6±2.6 kg/m for the2 2

control subjects, the normo- and the microalbuminuric IDDM patients, respectively(p>0.10). There were 7 men and one woman in each group. Mean age was 44±9, 46±10and 47±10 yr in these three groups, respectively (p>0.40). The mean diabetes duration was22±6 and 28±6 yr in the normo- and the microalbuminuric IDDM patients (p<0.08). In thenormoalbuminuric group 5 patients had no retinopathy, 2 had background retinopathy andone had proliferative retinopathy, whereas in the microalbuminuric group 5 patients hadbackground and 3 patients had proliferative retinopathy (p<0.05). None of the control andnormoalbuminuric IDDM subjects had arterial hypertension, defined as systolic bloodpressure $160 mm Hg and/or diastolic blood pressure $95 mm Hg. An ACE-inhibitor wasused by 7 microalbuminuric IDDM patients. This treatment was withheld for a 6 weekperiod before the study.

Clinical procedures

The participants were instructed by a dietician to adhere to a diet containing 100mmol sodium/day starting 1 week before the study to avoid possible confounding effectsof differences in sodium intake on gluco- and mineralocorticoid status. At the end of thisperiod urinary sodium excretion was 109±36, 103±26 and 109±19 mmol/24 h in thecontrol subjects and the normo- and microalbuminuric IDDM patients (p>0.40), indicatingacceptable diet compliance. An exactly timed overnight urine collection was then obtainedfor measurement of steroid metabolites, free cortisol, aldosterone, Ualb.V and creatinineclearance. None of the IDDM patients had symptomatic nocturnal hypoglycaemia andblood glucose at 0300 h varied between 4- 9 mmol/l. On the following morning, fastingvenous blood was drawn at 0900 h, while the subjects were in the supine position. Supineblood pressure was measured at 5 min intervals during one hour using an automateddevice (Dinamap ) and the results were averaged for analysis. Initial blood volume,R

estimated by a previously described electrical impedance technique [22] was used as anindex of upper arm segmental blood volume. In brief, a blood pressure cuff is placed overa 2 cm wide segment of the left upper arm. With a computer-assisted model the segment'sinitial blood volume (expressed in ml/cm) is calculated from the increase in impedanceduring rapid inflation of the cuff at a period of constant suprasystolic cuff pressure. Inhealthy subjects the coefficient of variation of the initial blood volume measurementdetermination is 6% (derived from Ref. 23). This technique can also be used to calculateother vascular parameters, like arterial blood volume and static arterial compliance. Thevalidity of the mathematical model and its derived parameters has been demonstrated bya close correlation of the diameter of the brachial artery as estimated by this techniquewith its diameter as assessed by high-resolution echography (r=0.93 in 11 healthy subjects,Van Leeuwen JTM, unpublished results). The impedance technique was available to allcontrol subjects, to 7 normoalbuminuric and to 6 microalbuminuric IDDM patients.

The 7 microalbuminuric IDDM patients in whom ACE-inhibition treatment wasstopped before the study consented to be reevaluated 6-12 weeks after resumption of ACE

Chapter 11138

inhibition therapy (enalapril, 10 mg/day). Venous blood samples were then taken 30-60min after enalapril administration. On this second evaluation the electrical impedance

technique was not available.

Laboratory measurements

Blood samples for measurement of cortisol, aldosterone, ACTH and renin werecollected in EDTA-anticoagulated tubes and placed on ice immediately. The plasmaspecimens and aliquots of overnight urine collections were stored at -20 C until assay.Plasma cortisol and aldosterone were measured by RIA's as described [24,25]. Plasma ACTH was assayed with a commercially available IRMA (Nichols Institute of Diagnos-tics, Wijchen, The Netherlands). PRA was measured by RIA (Du Pont, Billerica, MA,USA). Urinary steroid metabolites were determined by gas chromatography [26]. ß-glycyrrhetinic acid in the urine specimens was analysed by gaschromatography/massspectrometry (GC/MS) using "-glycyrrhetinic acid as internal standard. The detectionlimit of the assay is approximately 5µg/l. Urinary free cortisol and free aldosterone weremeasured after extraction and chromatography by RIA's. Aldosterone-18-glucuronide wasassayed by RIA after hydrolysis at pH 2.0 followed by chromatography. Serum andurinary electrolytes and creatinine were measured on SMAC and SMA-II autoanalyzers,respectively (Technicon Instruments, Tarrytown NY, USA). HbA1c was measured byhigh-performance liquid chromato-graphy (HPLC, Bio-Rad, Veenendaal, The Nether-lands; normal values 4.6 to 6.1%). Blood glucose was determined on a YSA glucoseanalyser (model 23A, Yellow Springs, Yellow Springs OH, USA). Urinary albumin wasmeasured by RIA (Diagnostic Products Corporation, Apeldoorn, The Netherlands).

Evaluation of cortisol metabolism

5ß tetrahydrocortisol (THF), 5" tetrahydrocortisol (allo-THF) and 5ß tetrahydro-cortisone (THE) are major metabolites of cortisol and cortisone which are excreted in theurine [27]. These metabolites are primarily produced in the liver by the irreversible ringA reduction of glucocorticoids to dihydro metabolites and the subsequent keto reductionto tetrahydro compounds. The first step is catalysed by 5" and 5ß reductases. The urinaryallo-THF/THF and androsterone (A, 5"-androstan-3"-ol-17-on)/etiocholanolone (E, 5ß-androstan-3"-ol-17-on) ratios reflect the proportion of 5"/5ß reduced C-21 and C-19steroids, respectively [27,28]. According to Ulick et al [5,8], the urinary cortisol tocortisone metabolite ratio (THF+allo-THF/THE) is an index of 11ß-HSD activity and thisratio reflects the set-point of the overall direction of the cortisol to cortisoneinterconversion. Since 11ß-HSD activity may be influenced by food intake [20], we usedovernight instead of 24-h urine collections. It has also been suggested that one calculatesthe urinary THF+allo-THF/free cortisol ratio as a measure of the efficiency of the ring Areduction of cortisol [8]. As urine was collected overnight, we did not use this parameterbut instead calculated the excretion of cortisol metabolites and free cortisol separately.

Statistical analysis

Parameters are expressed as mean±SD, or as median and range. Differences inparameters between the control group, the normoalbuminuric IDDM group and the

Cortisol metabolism in IDDM 139

microalbuminuric IDDM group were evaluated by Kruskal-Wallis analysis of varianceusing Duncan's method to correct for multiple comparisons. Bivariate correlations wereassessed by linear regression analysis. Analysis of covariance was used to evaluateindependent relationships between variables. In the microalbuminuric IDDM group theeffect of ACE-inhibition therapy on the various parameters was expressed as mean ormedian change with 95% Confidence Interval and was analysed by the paired Wilcoxonrank test. A two-sided p-value < 0.05 was taken as significant.

Results

Table 1 shows some clinical and laboratory data of the study groups. Systolic bloodpressure was higher in the untreated microalbuminuric IDDM patients compared to thenormoalbuminuric IDDM and the control subjects. No differences in diastolic bloodpressure were observed. Overnight Ualb.V was >20 µg/min in each micro-albuminuricpatient and <20 µg/min in all other subjects. Overnight creatinine clearance and serumpotassium concentration were not different among the groups. Initial blood volume washigher in the normoalbuminuric IDDM than in the control group, but the differencebetween the microalbuminuric IDDM and the control subjects was not significant. Meanfasting blood glucose and HbA1c levels were similarly elevated in the IDDM groups.

As shown in Table 2 the urinary excretion of the major metabolites of cortisol, THFand allo-THF, as well as of cortisone, THE, were lower in both IDDM groups than in thecontrol group. When expressed per mmol of creatinine, the urinary excretion of theseglucocorticoid metabolites was similarly decreased in the IDDM groups (not shown). Ofimportance, the excretion of these metabolites was not equally decreased. Consequently,the THF+allo-THF/THE ratio, i.e. the cortisol to cortisone metabolite ratio, was lower inboth IDDM groups compared to controls. The allo-THF/THF ratio, which reflects 5"/5ßreduction of cortisol, was also lower in the IDDM groups. This change in 5"/5ß reductionwas not restricted to cortisol, since the A/E ratio, representing 5"/5ß reduction of C-19steroids, was again lower, especially in the microalbuminuric IDDM group. ß-glycyrrhetinic acid was detected in the urine samples of 2 control subjects and 1normoalbuminuric IDDM patient. Exclusion of the steroid data of these subjects from theanalysis did not change the statistical results.

The overnight urinary excretion of cortisol and aldosterone, as well as the fastingplasma levels of these corticosteroids and of ACTH and PRA are given in Table 3.Urinary free cortisol, but not plasma cortisol, was lower in the normoalbuminuric IDDMgroup than in the other groups. In the microalbuminuric IDDM group the plasma level ofcortisol was higher, but urinary free cortisol was not different compared to the controlgroup. No differences in plasma ACTH were observed. Urinary free aldosterone,aldosterone-glucuronide and plasma aldosterone were not different in the diabetic groupscompared to the control group, whereas urinary free aldosterone was higher in the

Chapter 11140

Table 1. Clinical and laboratory data of study subjects

Control IDDM patients Change with Subjects Normal Ualb.V Micro Ualb.V ACE inhibition (n=8) (n=8) (n=8) (n=7)

Systolic BP (mmHg) 127±9 127±12 144±11 -12(-22 to -2)a,b d

Diastolic BP (mmHg) 79±7 75±8 80±7 -8(-11 to -4)d

Ualb.V (µg/min) 4(1-11) 5(2-14) 56(32-174) -8(-55 to 25)a,b

Creatinine clearance 1.68±0.50 1.84±0.28 1.75±0.50 0.02(-0.44 to 0.46)(ml/sec per 1.73m )2

Initial blood volume 3.53±1.14 4.81±1.07 4.25±1.48 NDc

(ml/cm) (n=8) (n=7) (n=6)

Blood glucose (mmol/l) 5.0±0.5 9.7±4.1 8.9±2.5 1.1(-4.3 to 6.5)a a

HbA1c (%) 5.7±0.5 8.3±1.0 8.2±0.9 0.2(-0.5 to 0.9)a a

Serum potassium (mmol/l) 4.2±0.2 4.4±0.4 4.2±0.3 0.1(-0.3 to 0.5)

Data are the mean±SDS or the median range (in parenthesis) changes in the mean or median (95%confidence interval). BP: blood pressure; Ualb.V: albumin excretion rate; ND: not determined

a

p<0.01 vs. control subjects; p<0.01 vs. normoalbuminuric IDDM patients (by selection); p<0.05b c

vs. control subjects; p<0.02 vs without ACE inhibition d

microalbuminuric than in the normoalbuminuric IDDM group. PRA was not differentamong the groups.

Significant relationships were present between the HbA1c level and the urinaryexcretion of steroid metabolites (Figure 1). In the combined groups (n=24), urinary allo-THF (Figure 1A, r=-0.69, p<0.001), THF+allo-THF/THE (Figure 1B, r=-0.58, p<0.01),allo-THF/THF (Figure 1C, r=-0.61, P<0.01), A/E (Figure 1D, r=-0.59, p<0.01), THF (r=-0.40, p=0.05, not shown) and THF+allo-THF+THE (r=-0.48, p<0.05, not shown) wereinversely correlated with HbA1c. When analysing the data from the IDDM groups (n=16)separately, the relationships between HbA1c, THF + allo-THF/THE (r=-0.52, p<0.05) andA/E (r=-0.54, p<0.05) remained significant. No significant correlations were foundbetween fasting blood glucose and urinary steroid parameters. Figure 2 shows the positiverelationship between the urinary cortisol to cortisone metabolite ratio and initial bloodvolume in the control subjects (n=8, r=0.77, P<0.05) and in the IDDM patients (n=13,r=0.56, p<0.05). Remarkably, the bivariate relationship between the cortisol to cortisonemetabolite ratio and initial blood volume was different between IDDM patients andcontrol subjects, as depicted by the leftward shift of the regression line in the IDDMpatients (i.e. a lower ratio at the same blood volume). Analysis of covariance disclosedthat initial blood volume was independently related to the cortisol to cortisone metaboliteratio (p<0.01) and to the presence of IDDM (as a categorical covariate, p<0.03), but notto the HbA1c level (p=0.76). Thus, the leftward shift in this relationship in the IDDMpatients remained significant after controlling for the HbA1c level. Except for possiblebivariate correlations of diastolic

Cortisol metabolism in IDDM 141

Table 3. Urinary and plasma measurements of cortisol, aldosterone, ACTH and PRA.

Control IDDM patients Change withSubjects Normal Ualb.V Micro Ualb.V ACE inhibition(n=8) (n=8) (n=8) (n=7)

Urine Free cortisol 68.3±29.1 35.1±11.1 61.0±30.4 8.6(-53.8 to 71.1)

a b

(pmol/min) Free aldosterone 0.83±0.44 0.72±0.42 1.26±0.65 -0.72(-1.28 to -0.17)

b c

(pmol/min) Aldosterone 21.5±15.9 17.3±7.6 26.5±15.5 -16.5(-30.7 to -23)

c

glucuronide (pmol/min)

Plasma Cortisol (nmol/l) 381±119 401±154 551±117 39(-73 to 152)

a,b

ACTH (pmol/l) 5.5±4.2 7.5±3.5 7.5±4.2 7.1(-5.5 to 19.7) Aldosterone 0.45±0.27 0.64±0.42 0.55±0.32 -0.17(-0.58 to -0.01)

d

(nmol/l)PRA (ng/l.s) 0.30±0.22 0.32±0.20 0.40±0.38 0.52(0.02 to 1.82)

d

Data are the mean±SD or changes in the mean (95% confidence interval). Ualb.V: urinary albuminexcretion. p<0.01 vs. control subjects and p<0.05 vs. normoalbuminuric IDDM; p<0.02 and

a b c

p<0.05 vs without ACE inhibition d

Table 2. Urinary excretion of steroid metabolites in the study subjects

Control IDDM patients Change withSubjects Normal Ualb.V Micro Ualb.V ACE inhibition(n=8) (n=8) (n=8) (n=7)

THF (nmol/min) 4.2±1.4 3.1±1.0 3.0±1.3 0.5(-1.3 to 2.2)a a

Allo-THF (nmol/l) 3.4±1.0 1.7±0.9 1.3±0.7 0.3(-0.4 to 0.9)b b

THE (nmol/min) 7.5±2.2 5.8±1.6 5.3±2.7 1.3(-1.4 to 4.0)a a

THF+allo-THF+THE 15.1±4.1 10.6±1.6 9.6±4.4 2.0(-2.9 to 7.0)a b

(nmol/min)

THF+allo-THF/THE 1.02±0.14 0.82±0.19 0.86±0.15 -0.08(-0.19 to 0.0)a a c

(µmol/µmol)

Allo-THF/THF 0.86±0.26 0.55±0.19 0.46±0.21 0.04(-0.08 to 0.16)a b

(µmol/µmol)

A/E (µmol/µmol) 1.35±0.39 1.04±0.34 0.91±0.36 -0.04(-0.23 to 0.14)a

Data are the mean±SD or changes in the mean (95% confidence interval). Ualb.V: uirnary albuminexcretion rate. Allo-THF/THF and A/E reflect 5"/5$-reduction of C-21 and C-19 steroids,respectively. p<0.01 and p<0.01 vs. control subjects; p<0.05 vs without ACE inhibition

a b c

4 6 8 105 7 9HbA (%)

0.5

0.9

1.5

0.7

1.1

1.3

BA

C D

THF+

allo

-TH

F/TH

E (m

ol/m

ol)

4 6 8 105 7 9

0

1

2

3

4

5

6

allo

-TH

F (n

mol

/min

)

4 6 8 105 7 9

0

0.3

0.6

0.9

1.2

1.5

allo

-TH

F/TH

F (m

ol/m

ol)

4 6 8 105 7 9

0

0.5

1.0

1.5

2.0

A/E

(mol

/mol

)1c HbA (%)1c

HbA (%)1c HbA (%)1c

Chapter 11142

Figure 1.Relationships among HbA1c, urinary allo-THF (A), urinary THF+allo-THF/THE (B), urinaryallo-THF/THF (C), and urinary A/E (D). ! Control subjects, Q normoalbuminuric and Qmicroalbuminuric IDDM patients. A: r=-069, p<0.001: B: r=-0.58, p<0.01; C: r=-0.61, p<0.01; D: r=-0.59, p<0.01 (in the combined groups).

blood pressure with initial blood volume and with the urinary cortisol to cortisone me-tabolite ratio in the control group (n=8, r=0.63 and r=0.63, p<0.10 for both), no furtherinterrelationships between blood pressure, initial blood volume, BMI, HbA1c, steroidmetabolites and aldosterone could be demonstrated.

In the microalbuminuric IDDM group ACE-inhibition treatment resulted in a fall insystolic and diastolic blood pressure, whereas overnight Ualb.V and creatinine clearancedid not change significantly (Table 1). Mean HbA1c and fasting blood glucose levelsremained unaltered (Table 1). Overnight urinary sodium excretion rose by 42 (95%CI, 18to 66) µmol/min from 69±41 to 111±47µmol/min at this second evaluation (p<0.02). As

0.5 0.7 1.1 1.50.9 1.3

THF+aTHF/THE (mol/mol)

3

5

8

4

6

7

2

initi

al b

lood

volu

me (

mL

/cm

)

Cortisol metabolism in IDDM 143

presented in Table 2 the absolute urinary excretion of THF, allo-THF and THE did notchange significantly, but there was a modest further decrease in the cortisol to cortisonemetabolite ratio. No changes were observed with respect to the allo-THF/THF ratio andthe A/E ratio. Urinary and plasma cortisol, as well as plasma ACTH did not changesignificantly following ACE-inhibition treatment (Table 3). As expected, urinary andplasma aldosterone decreased and PRA increased after administration of the ACE-inhibi-tor (Table 3).

Figure 2.Relationships between the urinary cortisol to cortisone metabolite ratio(THF+allo-THF/THE) and initial blood volume an indexof blood volume.! Controlsubjects (r=0.77,p<0.05;n=8), Q normo- and Q micro-albuminuric IDDM patients (r=0.56, p<0.05; n=13).The regression line is shifted to the left in the IDDM patients (p<0.03, by analysis of covariance).

Discussion

The lower urinary excretion of tetrahydro metabolites of cortisol and cortisone innormo- and microalbuminuric IDDM patients strongly suggest that the diabetic stateinfluences cortisol metabolism via an impaired reduction of glucocorticoids. It was alsofound that the urinary allo-THF/THF ratio as well as the A/E ratio was lower,indicating that the proportion of 5"/5ß reduced corticoids of both the C-21(glucocorticoids) and the C-19 (adrenal androgens) series was decreased. In addition, theurinary cortisol to cortisone metabolite ratio was lower in IDDM. Since this metaboliteratio reflects the set-point of overall direction of the 11ß-HSD-catalysed cortisol tocortisone interconversion [5-8], the present findings support the notion that IDDM isassociated with an imbalance in 11ß-HSD activity. Opposite changes in the urinary allo-THF/THF ratio and in the cortisol to cortisone metabolite ratio have been demonstratedin type 1 AME syndrome and after ingestion of licorice, conditions associated withimpaired dehydrogenase activity of 11ß-HSD [5-8,28,29].

It is uncertain whether the presently observed lower urinary excretion oftetrahydro metabolites of glucocorticoids reflects decreased cortisol secretion in IDDM,since urine was collected overnight and cortisol production rate was not measured.Previous studies have shown mild hypercortisolaemia in IDDM and a diminished

Chapter 11144

activation of the hypothalamic-pituitary adrenal axis is unlikely [17-19]. Fasting plasmacortisol was indeed higher in the microalbuminuric IDDM patients. Overnight urinaryfree cortisol excretion was within normal limits in both IDDM groups, albeit that itslevel was lower in normoalbuminuric patients compared to controls. Obviously, diabetesmellitus is associated with multiple alterations in the regulation of cortisol metabolismand further study is indicated to document their exact nature.

5" and 5ß reductases as well as 11ß-HSD are widely distributed enzymes[4,27,30]. In rat liver 5" reductase is found in the endoplasmatic reticulum, whereas 5ßreductase is a cytosolic enzyme [30]. Both reductases require NADPH as a cofactor andare capable of metabolising different classes of steroids, including C-21 and C-19 com-pounds [27]. 11ß-HSD is predominantly concentrated in the microsomal cell fraction andthus far two functional isoforms have been identified [4,31]. 11ß-HSD , isolated from1a variety of tissues including liver and vascular smooth muscle, exhibits bothdehydrogenase and reductase activity and catalyses the dehydrogenation of cortisol tocortisone as well as the reduction of cortisone to cortisol. The direction of the reactionis dependent on the availability of its cofactor NADP /NADPH, the ambient+

cortisol/cortisone concentration and the glycosylation status of the protein [4]. 11ß-HSD is present in placenta and in the distal nephron. This isoform only shows2dehydrogenase activity and requires NAD as a cofactor [4,32-34]. In rat kidney both+

NADP and NAD availability affect dehydrogenase activity of 11ß-HSD, suggesting+ +

the expression of both isoforms [31]. In human renal tissue the 11ß-HSD isoform is2preferentially expressed and 11ß-HSD activity is largely NAD -dependent [33,34].+

Experimental studies have demonstrated a diabetes-induced accumulation of NADP and+

NADH in the cytosolic cell compartment as a result of increased reduction of glucose tosorbitol and oxidation of sorbitol to fructose via the sorbitol pathway [35]. The presentstudy showed inverse correlations of urinary allo-THF, the allo-THF/THF and the A/Eratio as well as THF+allo-THF/ THE ratio with the HbA1c level. These relationshipswould fit the hypothesis that chronic hyperglycemia induces a decrease in intracellularNADPH which impairs apparent 5"/5ß reductase activity. The data might also supportthe concept that a higher NADP /NADPH ratio enhances dehydrogenase activity and+

inhibits reductase activity of 11ß-HSD , but it seems very improbable that a lower1NAD /NADH stimulates dehydrogenase activity of 11ß-HSD . In the interpretation of+

2the current results it should be noted that the urinary THF+allo-THF/THE ratio is onlyan index of the overall in vivo direction of the cortisol to cortisone interconversion.Therefore, it is not possible to direct the imbalance in 11ß-HSD activity to a specifictissue or to a specific isoform.

In IDDM patients without nephropathy extracellular volume tends to be increased,whereas plasma volume is normal or elevated [9-15]. We found that initial blood volumewas elevated in the normoalbuminuric IDDM group, but our technique which estimatesupper arm segmental blood volume cannot be easily compared with other methods tomeasure whole body extravascular and blood volume. In both the control group and thecombined diabetic groups a positive correlation was found between the urinary cortisolto cortisone metabolite ratio and initial blood volume. Of note, the regression line wassignificantly shifted leftwards in the diabetic patients, thus essentially excluding that the

Cortisol metabolism in IDDM 145

overall change in 11ß-HSD activity is primarily responsible for an abnormal fluid andsodium retention in IDDM. 11ß-HSD could also play a role in blood pressure regulation[36]. Inhibition of dehydrogenase activity of 11ß-HSD enhances the vasoconstrictorpotency of cortisol and impaired cortisol to cortisone conversion is a salient feature ina subgroup of patients with essential hypertension [37,38]. The similar decrease in theurinary cortisol to cortisone metabolite ratio in the microalbuminuric compared to thenormoalbuminuric IDDM group makes it unlikely, however, that the blood pressure riseassociated with incipient nephropathy is related to altered dehydrogenase activity of 11ß-HSD.

Plasma aldosterone as well as urinary free aldosterone and aldosterone-glucuronideexcretion were unchanged in the IDDM groups compared to control subjects. Normalto low levels of circulating aldosterone have been previously shown in heterogeneousgroups of diabetic patients [15,39] and in IDDM patients irrespective of the presenceof microalbuminuria [14], whereas discrepant data have been reported with respect toPRA and plasma active renin [11-15,40,41]. Our data support the contention thatcirculating aldosterone might be insufficiently suppressed in relation to the alteredvolume homeostasis, although it is not likely that the sodium and fluid retention inIDDM is caused by a stimulated plasma aldosterone per se [14,15].

During ACE-inhibition therapy urinary and plasma cortisol remained unchanged,in agreement with data in healthy subjects [42]. ACE-inhibition resulted in a modestfurther lowering of the cortisol to cortisone metabolite ratio, without an effect on theproportion of 5"/5ß reduced corticoids. Since the HbA1c level was unaltered, this effectwas not attributable to a change in metabolic control. In the rat, ACE-inhibitor adminis-tration stimulates 11ß-HSD activity in renal tissue and this influence is partly antago-nized by angiotensin II [20]. The presently observed moderate decrease in urinarycortisol to cortisone metabolite ratio is in keeping with enhanced dehydrogenase activityof 11ß-HSD in vivo. It is possible that, in addition to inhibition of angiotensin II andaldosterone, stimulation of 11ß-HSD could contribute to the saliuretic effect of ACE-inhibition treatment [20]. Indeed, urinary sodium excretion was increased during treat-ment, but our study design does not allow a conclusion with respect to the responsiblemechanism.

In conclusion, IDDM patients show multiple abnormalities in urinary corticoidmetabolite excretion, suggestive of impaired corticoid reduction and altered 11ß-HSDactivity. It seems very unlikely that a change in 11ß-HSD activity is primarily responsi-ble for the sodium and fluid retention in IDDM. Since ACE-inhibition resulted in afurther decrease in the urinary cortisol to cortisone metabolite ratio, an additionalmechanism of action of this treatment might be provided by an effect on 11ß-HSDactivity.

References

1 Arriza JL, Weinberger C, Cerelli G, Glaser TM, Handeling BL, Housman DE, Evans RM.Cloning of human mineralocorticoid receptor complementary DNA: structural and functionalkinship with the glucocorticoid receptor. Science 1987; 237: 268-275

2 Sheppard K, Funder JW. Mineralocorticoid specificity of renal type I receptors: in vivo binding

Chapter 11146

studies. Am J Physiol 1987; 252: E224-E2293 Walker BR, Edwards CRW. 11ß-hydroxysteroid dehydrogenase and enzyme-mediated receptor

protection: Life after liquorice? Clin Endocrinol. 1991; 35: 281-2894 Seckl JR. 11ß-hydroxysteroid dehydrogenase isoforms and their implications for blood pressure

regulation. Eur J Clin Invest 1993; 22: 589-6015 Ulick S, Levine LS, Gunczler P, Ganconato G, Ramirez LC, Rauh W, Rösler A, bradlow HL,

New MI. A syndrome of apparent mineralocorticoid excess associated with defects in theperipheral metabolism of cortisol. J Clin Endocrinol Metab 1979; 49: 757-764

6 Stewart PM, Corrie JET, Shackleton CHL, Edwards CRW. Syndrome of apparent mineralo-corticoid excess. A defect in the cortisol-cortisone shuttle. J Clin Invest 1988; 82: 340-349

7 Stewart PM, Valentino R, Wallace AM, Burt D, Shackleton CHL, Edwards CRW.Mineralocorticoid activity of liquorice 11ß-hydroxysteroid dehydrogenase activity comes ofage. Lancet 1987; ii: 821-824

8 Ulick S, Tedde R, Wang JZ. Defective ring A reduction of cortisol as the major error in thesyndrome of apparent mineralocorticoid excess. J Clin Endocrinol Metab 1992; 74: 593-599

9 Brøchner-Mortensen J. Glomerular filtration rate and extracellular fluid volumes during normo-glycemia and moderate hyperglycemia in diabetics. Scand J Clin Lab Invest 1973; 32: 311-316

10 Brøchner-Mortensen J, Ditzel J. Glomerular filtration rate and extracellular fluid volume ininsulin-dependent patients with diabetes mellitus. Kidney Int 1982; 21: 696-698

11 O'Hare JA, Ferris JB, Twomey BM, Brady D, O'Sullivan DJ. Essential hypertension andhypertension in diabetic patients without nephropathy. J Hypertens 1983; 1 [Suppl 2]: 200-203

12 O'Hare JA, Ferris JB, Brady D, Twomey B, O'Sullivan DJ. Exchangeable sodium and reninin hypertensive diabetic patients with and without nephropathy. Hypertension 1985; 7 [Suppl2]: 43-48

13 Weidmann D, Beretta-Piccoli C, Trost BW. Pressor factors and responsiveness in hypertensionaccompanying diabetes mellitus. Hypertension 1985; 7 [Suppl 2]: 33-42

14 Feldt-Rasmussen B, Mathiesen ER, Deckert T, Giese J, Christensen NJ, Bent-Hansen, NielsenMD. Central role for sodium in the pathogenesis of blood pressure changes independent ofangiotensin, aldosterone and catecholamines in Type 1 (insulin-dependent) diabetes mellitus.Diabetologia 1987; 30: 610-617

15 Weidmann D, Ferrari D. Central role of sodium in hypertension in diabetic subjects. DiabetesCare 1991; 14: 220-232

16 De Fronzo RA. The effect of insulin on renal sodium metabolism. Diabetologia 1981;21:65-17117 Roy M, Collier B, Roy A. Dysregulation of the hypothalamo-pituitary-adrenal axis and duration

of diabetes. J Diabetic Complications 1991; 5: 218-22018 Roy MS, Roy A, Gallucci WT. The ovine corticotropin-releasing-hormone-stimulation test in

Type I diabetic patients and controls: suggestion of mild chronic hypercortisolism. Metabolism1993; 42: 696-700

19 Tsigos C, Young RJ, White A. Diabetic neuropathy is associated with increased activity of thehypothalamic-pituitary-adrenal axis. J Clin Endocrinol Metab 1993; 76: 554-558

20 Riddle MC, Mc Daniel PA. Renal 11ß-hydroxysteroid dehydrogenase activity is enhanced byramipril and captopril. J Clin Endocrinol Metab 1994; 78: 830-834

21 Mogensen CE. Prediction of clinical diabetic nephropathy in IDDM patients. Alternatives tomicroalbuminuria? Diabetes 1990; 39: 761-767

22 Muntinga JHJ, Visser KR. Estimation of blood pressure-related parameters by electrical imped-ance measurement. J Appl Physiol 1992; 73: 1946-1957

23 Muntinga JHJ, Gelt ME, Terpstra WF, Visser KR. Investigation of the arterial and venousupperarm vascular bed. Medical Engineering and Physics 1995; 17: 264-272

24 Pratt JJ. Steroid immunoassay in clinical chemistry. Clin Chem. 1978; 84: 329-33725 Pratt JJ, Boonman R, Woldring MG, Donker AJM. Special problems in radioimmunoassay of

Cortisol metabolism in IDDM 147

plasma aldosterone without prior extraction and purification. Clin Chim Acta 1978; 84: 329-33726 Wolthers BG, De Vries IJ, Volmer M, Nagel GT. Detection of 3ß-hydroxysteroid

dehydrogenase deficiency in a newborn by means of urinary steroid analysis. Clin Chim Acta1987; 109: 109-116

27 Brownie AL. The metabolism of adrenal cortical steroids. In: James VHT, ed. The adrenalgland. 2nd ed. New York: Raven Press Ltd; 1992, pp 209-224

28 Monder C, Shackleton CHL, Bradlow HL, New MI, Stoner E, Iohan F, Lakshmi V. Thesyndrome of apparent mineralocorticoid excess: its association with 11ß-dehydrogenase and 5ß-reductase deficiency and some consequences for corticosteroid metabolism. J Clin EndocrinolMetab. 1986; 63: 550-557

29 Farese R, Biglieri EG, Shackleton CHL, Irony I, Gomez-Fontes R. Liquorice-inducedhypermineralocorticoidism. N Engl J Med 1991; 1223-1227

30 Orth DN, Kovacs WJ, Debold CR. The adrenal cortex. In:Wilson JD and Foster DW, eds.Williams textbook of Endocrinology. 8th ed. Philadelphia: WB Saunders 1992, pp 489-619

31 Stewart PM. 11ß-hydroxysteroid dehydrogenase. Clin Endocrinol Metab 1994; 8: 357-37832 Albiston AL, Obeyesekere VR, Smith RE, Krozowski ZS. Cloning and tissue distribution of the

human 11ß-HSD type 2 enzyme. Mol Cell Endocrinol 1994; 105: R11-R1733 Stewart PM, Murray BA, Mason JI. Human kidney 11ß-hydrosteroid dehydrogenase is a high

affinity nicotinamide adenine dinucleotide-dependent enzyme and differs form the cloned Type1 isoform. J Clin Endocrinol Metab. 1994; 79: 480-484

34 Agarwal AK, Mune T, Monder C, White PC. NAD -dependent isoform of 11ß-hydroxysteroid+

dehydrogenase. Cloning and characterization of cDNA from sheep kidney. J Biol Chem 1994;269: 25959-25962

35 Williamson JR, Chang K, Frangos M, Hasan KS, Ido Y, Kawamura T, Nyengaard JR, Van denEnden M, Kilo C, Tilton RG. Hyperglycemic pseudohypoxia and diabetic complications.Diabetes 1993; 42: 801-813

36 Seckl JR, Brown RW. 11-beta-hydroxysteroid dehydrogenase: on several roads to hypertension.J Hypertens 1994 12:105-112

37 Walker BR, Connacher AA, Webb DJ, Edwards CRW. Glucocorticoids and blood pressure:a role for the cortisol/cortisone shuttle in the control of vascular tone in man. Clin Sci 1992;82:171-178.

38. Walker BR, Stewart PM, Shackleton CHL, Padfield PL, Edwards CRW. Deficient inactivationof cortisol by 11ß-hydroxysteroid dehydrogenase in essential hypertension. Clin Endocrinol1993; 29: 221-227

39 Ferris JB, Sullivan PA, Gonggrijp H, Cole M, O'Sullivan D. Plasma angiotensin II andaldosterone in unselected diabetic patients. Clin Endocrinol 1982; 17: 261-269

40 Drury PL, Bodansky HJ. The relationship of the renin-angiotensin system in Type I diabetes tomicrovascular disease. Hypertension 1985; 7 [Suppl 2]: 84-89

41 Franken AAM, Derkx FHM, Blankestijn PJ, Janssen JAMJL, Mannesse CK, Hop W,Boomsma F, Weber R, De Jong PTVM, Schalenkamp MADH. Plasma prorenin as an earlymarker of microvascular disease in patients with diabetes mellitus. Diabète Metab. 1992;18:137-143

42 Winer LM, Molteni A, Molitch ME. Effect of angiotensin-converting enzyme inhibition onpituitary hormone responses to insulin-induced hypoglycemia in humans. J Clin EndocrinolMetab. 1990; 71: 256-259

Acknowledgment

We appreciate the skilful technical assistance of Jan Van der Molen, Wouter Bosman andHans Pol. Berta Beusekamp provided dietary advice.

NADP NAD

NADPH NADH

++

dehydrogenation

11β-HSD

reduction

NADP NADPH+

ring A reduction

5α−reductase

3-keto reduction

5β−reductase

C=O

CH OH2

HO

O

OH

C=O

CH OH2

O

O

OH

allo-tetrahydrocortisol

5β−dihydrocortisol dihydrocortisone5α−dihydrocortisol

cortisol cortisone

allo-THFtetrahydrocortisol tetrahydrocortisone

THF THE

Chapter 11148

Addendum

Scheme representing metabolic pathways of cortisol.

CHAPTER 12

SUMMARY AND CONCLUSIONS

Chapter 1 outlines the epidemiology, the functional stages, the pathogenesis andtherapeutic aspects of renal disease in patients with insulin-dependent diabetes mellitus(IDDM). Several aspects of the pathogenesis of diabetic nephropathy (DN) are moreextensively overviewed in sections on the influence of norepinephrine (NE) and thegrowth hormone-insulin-like growth factor-I (GH-IGF-I) axis on renal function. Bothsubstances belong to hormonal systems that control renal haemodynamics in oppositeways: NE causes renal vasoconstriction, the GH-IGF-I-axis induces renal vasodilatation.Since the early stages of diabetic renal involvement are characterised by an imbalance inglomerular vasodilation and vasoconstriction, the possible role of these humoral systemsin diabetic nephropathy (DN) is discussed. The role of 11$-hydroxysteroid dehydrogena-se (11$-HSD) in protecting the mineralocorticoid receptor from activation by cortisol isbriefly recapitulated in the context of abnormalities in sodium and volume homeostasis inIDDM.

Effects of NE on renal protein handling and renal haemodynamics.

In chapter 2, the relationships between plasma NE and the rise in albuminuria aftera fixed exercise test is evaluated in normo-and microalbuminuric IDDM patients andhealthy subjects. Physical exercise provides a strong sympathetic stimulus and can lead toan increase in urinary albumin excretion. NE is both a neurotransmitter and a hormonallyactive substance spilled over from the sympathetic nervous system. Plasma NE levelsmay influence renal function by vasoconstriction mediated via "-adrenoceptors, whichhave been located along glomerular arterioles. The albuminuric response after exercise isthought to result from an enhanced glomerular passage of macromolecules inconjunction with rises in systemic blood pressure and alterations in renalhaemodynamics.

Moderate strenuous exercise was found to induce an exaggerated rise inalbuminuria in both IDDM groups, in keeping with earlier reports. Blood pressure roseto higher levels in the microalbuminuric IDDM patients. The rise in plasma NE levelswas significantly greater in normo- and microalbuminuric IDDM patients than inhealthy subjects. Multiple regression analyses revealed that both elevations in bloodpressure and stimulated plasma NE levels independently contributed to the albuminuricresponse.

Exercise and exogenous NE induce comparable renal haemodynamic changes inhumans. Both are associated with a decrease in effective renal plasma flow (ERPF)without much change in glomerular filtration rate (GFR). Consequently, filtrationfraction rises which suggests a change in pressure profile along glomerular vessels, infavour of an increase in intraglomerular pressure. Indeed, in experimental studies NEincreases intraglomerular pressure. Thus, the relationship between changes in plasmaNE concentrations and changes in albuminuria during exercise support the hypothesis

Chapter 12150

that NE is involved in the albuminuric response by a renal haemodynamic mechanism,like a rise intraglomerular pressure. The observations also suggest that an enhancedcatecholamine response may contribute to an exaggerated rise in albuminuria in micro-albuminuric IDDM patients. From these experiments no conclusion can be drawn wheth-er an altered renal vascular responsiveness to NE in IDDM is involved in thisphenomenon.

In chapter 3, the possible relationships between circulatory NE levels and renalhaemodynamic parameters are investigated in normo- and microalbuminuric IDDMpatients and in healthy subjects. Both GFR and ERPF were higher in IDDM patientscompared to healthy subjects. GFR and ERPF were found to be inversely correlated withvenous plasma NE levels. No differences were observed in the relationships betweenplasma NE and ERPF between the IDDM and healthy subjects. The slightly lower plasmaNE levels in the IDDM patients could thus contribute to the elevations in ERPF. GFRwas negatively related to plasma NE and positively with the presence of IDDM. This sup-ports the notion that concomitant vasodilating mechanisms play a role in the elevations ofGFR in IDDM patients. These results suggest that circulatory NE is a determinant ofrenal haemodynamics both in IDDM patients and healthy subjects.

In chapter 4, a randomized, placebo controlled NE infusion experiment is under-taken in matched groups of normo- and microalbuminuric IDDM patients and healthysubjects. Microproteinuria, renal and systemic haemodynamic responses were measuredduring stepwise exogenous NE infusions at individually determined NE threshold, 20%pressor and pressor doses. The study addressed the following questions. First, does -exogenous NE induce a microproteinuric response? Second, is there a difference inmicroproteinuric response in normo- and microalbuminuric IDDM patients and healthysubjects? Third, what are the determinants of such a microproteinuric response? Fourth,are there differences in renal haemodynamic NE- responsiveness between these groups?

Exogenous NE was found to increase microproteinuria in conjunction with a rise insystemic blood pressure and renal vasoconstriction in all groups. NE increased urinaryalbumin and IgG excretion, but no effect was seen on urinary $2-microglobulinexcretion. This indicates that NE increases glomerular protein leakage. Furthermore, theabsolute microproteinuric response was more pronounced in microalbuminuric IDDM pa-tients than in normoalbuminuric IDDM patients and healthy subjects. Multiple regressionanalysis showed that the increase in microproteinuria was not only related to the rise insystemic blood pressure induced by the NE infusions, but also to the increase in plasmaNE level itself. The renal haemodynamic NE responsiveness (i.e. a fall in ERPF and risein filtration fraction) was similar in the groups.

This study is the first to demonstrate that a vasopressor agent causes an increase inmicroproteinuria. These results disagree with previous studies using angiotensin II. Ourfindings support the hypothesis that an intrarenal mechanism contributes to the NE-induced increase in microproteinuria, and demonstrate that a low dose of NE causesglomerular vasoconstriction. This is in accord with the role of circulatory NE in thealbuminuric response following exercise, and with the relationship of plasma NE withrenal haemodynamics as outlined in the preceding chapters. The enhanced micro-proteinuric response in microalbuminuric IDDM is likely to be the result of a glomerular

Summary and conclusions 151

permselectivity defect. Plasma NE rises during strenuous daily life activities and this maycontribute to the perpetuation of microproteinuria. Proteinuria itself is a determinant offuture loss of renal function, although it is still unknown if this also holds true for themicroalbuminuric phase of diabetic renal disease. It can, therefore, be argued thatprotection against the renal NE effects may be of clinical benefit.

In chapter 5, the possibility that treatment with the ACE-inhibitor, enalapril,attenuates systemic and renal haemodynamic NE responsiveness in microalbuminuricIDDM is investigated. Such an effect would be of particular relevance in micro-albuminuric IDDM patients, since systemic NE responsiveness has been found to beexaggerated, and strict blood pressure control has been shown to prevent or delay pro-gression of albuminuria in these patients.

Enalapril was found to lower systemic blood pressure and overnight urinaryalbumin excretion, and to increase ERPF. The blood pressure lowering effect of enalaprildisappeared with NE pressor infusion. The overall mean increase in blood pressure inresponse to NE was even higher with than without enalapril. ERPF remained elevatedduring NE infusion with enalapril treatment, and the NE-induced fall in ERPF wasunaltered by enalapril. Urinary albumin excretion was similar during the NE infusionsbefore and after enalapril treatment. These results are in keeping with earlier reports inpatients with non insulin-dependent diabetes mellitus, but contrast with findings inpatients with essential hypertension. The lack of effect of ACE inhibition treatment onthe systemic and renal effects of NE may have clinical implications for thedesign of renoprotective strategies in IDDM patients.

In chapter 6, the possibility that low dose dopamine infusion counteracts NE-induced renal vasoconstriction is investigated. Although low dose dopamine is widelyused to attenuate the decrease in renal haemodynamics during NE infusion therapy, thiseffect has not been proven in humans. Dopamine (4 µg/kg per min) was added toincremental doses of NE in normotensive healthy subjects. This dose of dopamine wasshown to prevent the fall in ERPF during NE infusion. Dopamine also attenuated the risein blood pressure, enlarged pulse pressure, blunted the fall in heart rate, and induced alarge natriuretic response. Thus, dopamine is indeed able to oppose NE-induced renalvasoconstriction. They also indicate that dopamine influences systemic haemodynamicsduring NE infusion. Further studies are warranted to confirm these findings in criticallyill patients, and to establish whether low dose dopamine infusion is able to improve theirclinical outcome. The latter is of particular relevance since a recent study reporteddisappointing effects of low dose dopamine on prevention of renal failure in this patientcategory.

Renal effect of the GH-IGF-I-system

In chapter 7, it is investigated whether abnormal GH and IGF-I levels influenceurinary albumin excretion. GH deficient patients, patients with (un)treated acromegalyand healthy subjects were compared. Urinary albumin excretion rate was shown to beelevated in acromegalic patients and tended to be reduced in GH deficient patients ascompared to healthy subjects. Moreover, GH and IGF-1 lowering by treatment with the

Chapter 12152

somatostatin analogue, octreotide, reduced albuminuria in the acromegalic patients. Thelevel of albuminuria was positively correlated with the GH and IGF-I level. Thesefindings support the notion that the GH-IGF-I system is involved in urinary proteinexcretion. Since renal insufficiency is uncommon in acromegaly, it is unlikely that GHand IGF-I elevations alone predispose to clinically important glomerular damage.

In chapter 8, baseline and amino acid-stimulated GFR and ERPF are compared inGH deficient, acromegalic and normo- and hyperfiltering IDDM patients as well as inhealthy subjects. Moreover, the possible relationship between plasma IGF-I levels andrenal haemodynamics were evaluated across these groups.

Baseline GFR and ERPF were shown to covary with GH status: the lowest valueswere found in the GH deficient patients followed by higher levels in the healthy subjects,treated and untreated acromegalic patients. The amino acid-induced increase in GFR andERPF was enhanced in the GH deficient patients and was abolished in the acromegalicand hyperfiltering IDDM patients. Taken all groups together, an inverse relationship wasfound between baseline GFR and ERPF and the amino acid-induced increment in GFRand ERPF. This indicates the renal reserve filtration capacity is exhausted in glomerularhyperfiltration, and suggest that hyperfiltering IDDM and acromegalic patients sharecomparable renal haemodynamic abnormalities.

The plasma level of IGF-I was a determinant of GFR and ERPF across the GHdeficient, acromegalic and healthy subjects, but not in the IDDM groups. The latter doesnot exclude a role of abnormalities in the GH-IGF-I system in glomerular hyperfiltrationassociated with IDDM. Enhanced glomerular IGF-I accumulation due to increased IGF-Ireceptor expression, alterations in local production of IGF-binding proteins and in IGF-binding protein 3 protease activity, could affect renal haemodynamics in IDDM.Obviously, IGF-I infusion experiments are required to evaluate whether renalhaemodynamic sensitivity to IGF-I is enhanced in IDDM. The very limited availabilityto clinical use of IGF-I currently does not enable us to carry out such experiments.

In chapter 9, an exercise test is used to stimulate GH physiologically in IDDMpatients with a normal and elevated GFR. GFR and ERPF were measured under stand-ardised conditions, and IDDM patients with glomerular hyperfiltration (GFR>130 ml/minper 1.73m ) were individually matched with IDDM patients with a normal GFR (90 to2

130 ml/min per 1.73m ). Kidney size was measured by ultrasonography. The circulatory2

levels of glucagon and GH were determined on a separate day in the fasting state andafter exercise.

The mean levels of these hormones were not significantly different in the hyper-and normofiltering IDDM patients. However, multiple regression analysis showed thatexercise-stimulated GH levels, circulatory plasma glucagon as well as HbA1c weresignificantly related to renal haemodynamic parameters and kidney size. These findingssupport the hypothesis that stimulated levels of GH and circulating glucagon contribute toglomerular hyperfiltration in IDDM. In contrast, previous studies showed that diurnal GHand glucagon profiles were not different in normo- and hyperfiltering IDDM patients.These discrepancies may due to the lack of use of stimulated GH levels in those studies,and to the definition of glomerular hyperfiltration.

Summary and conclusions 153

From the studies described in the chapters 7,8 and 9, we conclude that the GH-IGF-I system plays a role in renal haemodynamics and in glomerular protein handling invarious disease states in humans, including GH deficiency and GH excess, as well as inIDDM. From a therapeutic point of view, it will be of interest to determine whethersomatostatin analogues prevent progression of albuminuria and loss of renal function inIDDM patients. Such intervention could have a place as an adjunct to or as an alternativefor blood pressure lowering therapy. The availability to clinical use of long actingsomatostatin analogues will facilitate the evaluation of such treatment.

Urinary IgG excretion in normoalbuminuric IDDM

Chapter 10 describes the artefacts that can be encountered when urinary IgG ismeasured at low concentrations. It has been reported that urinary IgG excretion is in-creased in normoalbuminuric IDDM patients, but this phenomenon in not well under-stood. In normo- and microalbuminuric IDDM patients and healthy subjects, urinary IgGwas measured in samples that were kept frozen for 2 to 4 weeks. Urinary IgG excretionwas higher both in normo- and microalbuminuric IDDM patients compared to healthysubjects. Furthermore in IDDM patients, the IgG clearance divided by the albumin clear-ance was found to be higher in urine collections that contained glucose as compared tosamples without glucose. This raised the possibility that glucose influences urinary IgGconcentration, possibly by a preserving effect of glucose during storage.

In a laboratory experiment, the effects various storage procedures were evaluated inurine samples with different amounts of protein. Urinary IgG declined when sampleswere frozen for several weeks without precautions. The best results were obtained whenurine was stored frozen with addition of bovine serum albumin, phosphate buffer andhigh concentrations of glucose. These results indicate that glucose in urinary specimensof IDDM patients can in fact prevent the decrease in IgG, and may thus explain theapparently higher urinary IgG excretion in normoalbuminuric IDDM patients whenunprocessed urine is stored frozen before assay. This study indicates that precautionsshould be taken when urinary IgG cannot be measured immediately.

Sodium and volume homeostasis in IDDM and the role of 11$-HSD

In chapter 11, urinary cortisol and cortisone metabolites are evaluated in normo-and microalbuminuric IDDM patients and in healthy subjects. The primary objective wasto establish whether possible abnormalities in cortisol metabolism as a consequence of analtered 11$-HSD enzyme activity, are involved in the abnormal sodium and fluidretention in IDDM patients. 11$-HDS catalyses the interconversion of cortisol and itsinactive metabolite, cortisone, and thereby protects the mineralocorticoid receptor frombeing activated by cortisol. A change in the so-called cortisol-cortisone shuttle towardscortisol could lead to sodium retention and volume expansion in IDDM. Alternatively, achange towards cortisone could attenuate sodium retention.

Lower urinary excretion rates of cortisol and cortisone metabolites were found innormo- and microalbuminuric IDDM patients suggesting that the diabetic state influ-ences cortisol metabolism via an impaired reduction of glucocorticoids. Furthermore,the urinary cortisol to cortisone metabolite ratio was lower in IDDM. This indicates

Chapter 12154

that the set-point of overall direction of the 11ß-HSD catalysed cortisol to cortisoneinter-conversion is shifted towards cortisone. In the IDDM patients, the urinary cortisolto cortisone metabolite ratio was inversely related to the HbA1c level. A positiverelation between the urinary cortisol to cortisone metabolite ratio and initial bloodvolume, as a measure of whole body extravascular and blood volume, was found bothin healthy subjects and in IDDM patients. Interestingly, the regression line was betweenthese parameters was shifted leftwards in the IDDM patients. This indeed suggests thatthe cortisol to cortisone ratio is a determinant of volume homeostasis, but essentiallyexcludes the possibility that the an abnormal 11ß-HSD activity is primarily responsiblefor an abnormal fluid and sodium retention in IDDM. Finally in microalbuminuricIDDM patients, ACE-inhibition treatment was shown to induce a modest further lower-ing of the cortisol to cortisone metabolite ratio.

These results raise the possibility that altered cofactor availability as consequenceof chronic hyperglycaemia influences glucocorticoid reduction and 11$-HSD activity inhumans. Improved metabolic control could induce a backward shift of the cortisol-cortisone shuttle towards cortisol in IDDM patients, which would accentuate volumeand sodium homeostasis. Moreover, this study suggests that stimulation of 11$-HDSactivity may be an additional mechanism whereby ACE inhibitors promote saliuresis.

SAMENVATTING

Inleiding.

In hoofdstuk 1 worden het voorkomen, de stadia, de theorieën over het ontstaan en debehandelingsmethoden van diabetische nierziekte (nefropathie) besproken. Naar schattingtreedt nefropathie op bij 30% van de patiënten met van insuline-afhankelijke diabetischemellitus (IADM). Diabetische nefropathie wordt gekenmerkt door een verhoogde eiwi-tuitscheiding in de urine (proteïnurie) van meer dan 0.5 gram per dag en een geleidelijkverlies van nierfunctie. Nefropathie is een ernstige complicatie van diabetes mellitus,omdat naar verloop van tijd terminaal nierfunctieverlies kan optreden, zodat nierfunctievervangende behandeling met dialyse of transplantatie noodzakelijk wordt. Daarnaast isgebleken dat IADM patiënten met proteïnurie een sterk verhoogd risico hebben op hart-en vaatziekten. Verbeterde behandelingsmethoden kunnen het beloop van de nefropathiegunstig beïnvloeden. Met name agressieve bloeddrukverlaging vermindert de snelheidvan het nierfunctieverlies en heeft naar alle waarschijnlijkheid de vermindering van hetoverlijden ten gevolge van hart- en vaatziekten tot gevolg.

Het identificeren van patiënten met een verhoogde kans op het ontwikkelen van eendiabetische nefropathie is mogelijk gemaakt door de introductie van gevoeligelaboratoriummethoden waarmee lage concentraties van eiwit, albumine, in de urinegemeten kunnen worden. Onder normale omstandigheden passeert een kleine hoeveelheidalbumine het glomerulaire filter in de nieren. De glomerulus is de kleinste functioneleeenheid van de nier. Hier wordt de urine gevormd uit het bloed dat onder druk gefilterdwordt over een poreuze membraan, de glomerulaire basaal membraan. De glomerulairebasaal membraan vormt een effectieve barrière tegen verlies van grote eiwitten, zoalsalbumine. Gebleken is dat bij IADM patiënten een geringe verhoging van de albumineuitscheiding in de urine (microalbuminurie) van prognostische betekenis is voor het laterontwikkelen van nefropathie. Deze verhoogde doorlaatbaarheid van albumine kan wijzenop vroege veranderingen in het filtratie proces ten gevolge van een hogere filtratie drukin de glomeruli en veranderingen in de glomerulaire basaal membraan.

Het ontstaan van diabetische nefropathie is waarschijnlijk multifactorieel bepaald.Men veronderstelt dat veranderingen in de bloedsomloop door het lichaam en de nieren(haemodynamische factoren) en veranderingen in de stofwisseling (metabole factoren) bijIADM patiënten bijdragen aan het ontstaan van diabetische nefropathie. Immers, zowelhoge bloeddruk als onvoldoende diabetes regulatie zijn risicofactoren voor het ontstaanvan deze complicatie. Daarnaast spelen erfelijke factoren een rol bij de predispositie voordiabetische nefropathie.

Het is thans nog niet mogelijk om het ontwikkelen van diabetische nefropathie tevoorkomen. Onderzoek naar factoren, die een rol spelen bij het ontstaan van diabetischenefropathie, is van belang om de behandelingsmogelijkheden te kunnen optimaliseren. Indit proefschrift is getracht meer inzicht te krijgen in de werking van noradrenaline (NA)en de groeihormoon (GH)-insulin-like growth factor-I (IGF-I)-as op de nieren. NA iseen niervaatvernauwende stof (renale vasoconstrictor) die de doorstroming in de niervermindert en de filtratiedruk kan verhogen. GH veroorzaakt indirect niervaatverwijding(renale vasodilatatie) door stimulatie van IGF-I synthese. De vroege veranderingen in denierfunctie bij IADM patiënten worden gekenmerkt door een dysbalans tussen vasodilatie

156

van de vaten die de glomerulus van bloed voorzien (preglomerulaire arteriolen) envasoconstrictie van de vaten die het bloed afvoeren (postglomerulaire arteriolen). Dezeveranderingen leiden waarschijnlijk tot een toegenomen filtratiedruk. Een apart hoofdstukis gewijd aan methoden om urine ingevroren te bewaren alvorens kleine hoeveelhedeneiwit, in casu immunoglobuline-G, te bepalen. In een ander hoofdstuk wordt de rol vande balans tussen de hormonen, cortisol en cortison, bij zoutretentie en volume expansiebij IADM patiënten belicht.

Effecten van noradrenaline op de nierdoorbloeding en het eiwit verlies in de urine.

NA is de neurotransmitter van het sympathische zenuwstelsel, die vrijkomt uit dezenuwvezels. NA lekt voor een deel naar de circulatie. Dit circulerende NA is als vaso-actieve stof werkzaam. NA veroorzaakt vasoconstrictie na binding aan zogenaamde "-adrenerge receptoren . In de nieren zijn "-adrenerge receptoren gelokaliseerd in de vatenvan de glomerulaire arteriolen. Deze bepalen in belangrijke mate de nierdoorstroming enreguleren de filtratiedruk. Fysieke inspanning stimuleert het sympathische zenuwstelselen kan een stijging van de albuminurie veroorzaken. Deze toename van albuminurie nafysieke inspanning wordt toegeschreven aan een verhoogde glomerulaire passage voorgrote moleculen, in combinatie met een stijging van de systemische bloeddruk en glome-rulaire filtratiedruk.

In hoofstuk 2 is onderzocht of een stijging van de plasma NA spiegel bijdraagt aan eentoename van het eiwitverlies in de urine tijdens een gestandaardiseerde fietstest. Bestu-deerd werden IADM patiënten met normo- en microalbuminurie (normale en licht ver-hoogde albumine excretie in de urine) en controle personen. Fysieke inspanning indu-ceerde een duidelijke en meer uitgesproken toename van de albuminurie bij de microal-buminurische IADM patiënten dan de controle personen. In de micro-albuminurischeIADM patiënten werden hogere bloeddrukken bereikt. In beide IADM groepen steeg deplasma NA spiegel tot hogere waarden dan bij de controle personen. Zowel de stijging inde bloeddruk als de toename van de NA spiegels correleerden met de toename van albu-minurie. Deze bevinding suggereert dat NA via een renaal haemo-dynamisch mechanis-me bijdraagt aan de stijging van albuminurie als gevolg van inspanning. De hogere NAspiegels bij de IADM patiënten zouden op deze wijze de versterkte albuminurischerespons in IADM patiënten kunnen verklaren. Dit wordt gesteund door eerder onderzoekwaaruit bleek dat NA infusie en fysieke inspanning een gelijkwaardige daling van de nierdoorstroming veroorzaken. Een dergelijke daling kan gepaard gaan met een hogereglomerulaire filtratiedruk en kan leiden tot een stijging van albuminurie. Een NA infusieexperiment zou deze hypothese kunnen bevestigen.

In hoofdstuk 3 werd de mogelijke relatie tussen het in het bloed aanwezig NA enrenale haemodynamische parameters, zoals de klaring (glomerulaire filtratie rate: GFR)en de nierdoorbloeding (effectieve renale plasma flow: ERPF), geëvalueerd in normo- enmicroalbuminurische IADM patiënten en controle personen. De GFR en de ERPF werdenbepaald met standaard infusie methoden. De in rust afgenomen plasma NA spiegel wasiets lager bij de IADM patiënten dan bij de controle personen. De IADM patiëntenhadden een hogere GFR en ERPF dan de controle personen, wijzend op glomerulairevasodilatatie. De ERPF bleek in alle groepen negatief gecorreleerd met te zijn met de

Nederlandstalige samenvatting 157

plasma NA spiegel. Tussen de IADM en de controle personen bleek geen verschil in dezerelatie te bestaan. Dus de lagere plasma NA spiegels zouden kunnen bijdragen aan eenverhoogde ERPF bij IADM patiënten. De GFR was eveneens negatief gecorreleerd metde plasma NA concentratie. Bovendien bleek dat IADM patiënten een hogere GFRhadden, onafhankelijk van de plasma NA spiegel. Dit impliceert dat additionele vasodila-terende mechanismen, naast de invloed van een lagere NA spiegel, een rol spelen bij deverhoogde GFR bij IADM patiënten.

Hoofdstuk 4 geeft de resultaten weer van een gerandomiseerde, placebo gecontroleerdeNA infusie studie bij gematchte normo- en microalbuminurische IADM patiënten en niet-diabetische controle personen. Dit onderzoek werd uitgevoerd om de volgende vragen tebeantwoorden. Ten eerste, kan NA toediening een toename van de eiwituitscheiding in deurine (microproteïnurie) induceren bij de mens? Ten tweede, zijn er verschillen in eendoor NA-geïnduceerde microproteïnurische respons tussen normo- en microalbuminuri-sche IADM patiënten en controle personen? Ten derde, wat zijn de determinanten vaneen door NA-geïnduceerde toename van microproteïnurie? Ten vierde, bestaan er ver-schillen in de renale haemodynamiek tijdens NA infusie tussen de onderzochte groepen?In dit onderzoek werden de microproteïnurie (uitscheiding van albumine enimmunoglobuline-G), de systemische en de renale haemodynamiek gemeten tijdens 3opeenvolgende NA infusie doses, die respectievelijk een bloeddrukstijging van 0, 4 en 20mmHg veroorzaakten. De NA infusie dosis werd per deelnemer van te voren bepaald.

De systemische bloeddruk gevoeligheid voor NA was toegenomen bij beide groepenIADM patiënten, zoals eerder is gerapporteerd voor heterogene groepen diabetespatiënten. NA, gegeven in een dosis die de bloeddruk met 20 mmHg deed stijgen, veroor-zaakte een toename van de microproteïnurie in combinatie met renale vasoconstrictie inalle onderzochte groepen. NA verhoogde zowel de uitscheiding van albumine alsimmunoglobuline-G, maar niet van $2-microglobuline. Dit impliceert dat NA de glome-rulaire passage van eiwitten doet toenemen. In absolute zin was de toename van de mi-croproteïnurie groter in de microalbuminurische IADM patiënten ten opzichte van denormoalbuminurische IADM patiënten en controle personen. Multipele regressie analysetoonde aan dat de stijging van de bloeddruk en de toename van de plasma NA spiegelsonafhankelijk bijdroegen aan de door NA geïnduceerde toename van microproteïnurie.NA veroorzaakte een uitgesproken daling van de ERPF en stijging van de filtratie fractie(het quotiënt van GFR en ERPF), maar had geen effect op de GFR. De relatie tussen deplasma NA spiegel en de veranderingen in renale haemodynamiek verschilden niet tussende onderzochte groepen.

Bovengenoemde studie is het eerste acute infusie experiment dat een toename vanmicroproteïnurie als reactie op de vasopressor, NA, aantoont. In eerdere, niet met place-bo gecontroleerde, studies met de vasopressor, angiotensine II, werd een dergelijk effectop de microproteïnurie niet gevonden. Onze resultaten geven steun aan de hypothese datNA via een renaal mechanisme de glomerulaire eiwitlekkage in de nieren doet toenemen.Dit is in overeenstemming met de in de voorgaande hoofdstukken veronderstelde rol vancirculerende plasma NA op de toename van albuminurie bij fysieke inspanning, en debeschreven relatie met renaal haemodynamische parameters. De versterktemicroproteïnurische respons bij de microalbuminurische IADM patiënten iswaarschijnlijk het gevolg van een preëxistent permeabiliteitsdefect van de glomerulaire

158

basaal membraan. Plasma NA spiegels stijgen tijdens dagelijkse inspanningen en zoudenaldus kunnen bijdragen aan eiwitlekkage. De aanwezigheid van proteïnurie wordt als eenongunstige indicator beschouwd voor toekomstig nierfunctieverlies, alhoewel het nietzeker is of dit ook geldt voor de microalbuminurische fase van diabetische nefropathie.Desalniettemin zou bescherming van de nier tegen renale NA effecten van klinischebelang kunnen zijn bij microalbuminurische IADM patiënten.

In hoofdstuk 5 is onderzocht of behandeling met een angiotensine-converting enzyme(ACE) remmer de NA-geïnduceerde stijging van de bloeddruk en renale vasoconstrictiebeïnvloed bij IADM patiënten met microalbuminurie. Bij patiënten met essentiële hyper-tensie is gebleken dat ACE-remmers inderdaad de door NA-geïnduceerde stijging van debloeddruk verminderen, hetgeen als een additioneel werkingmechanisme van deze medi-camenten kan worden gezien. Een studie bij gezonde vrijwilligers heeft bovendienaangetoond dat dit ook geldt voor de door NA-geïnduceerde ERPF daling. Zeven micro-albuminurische IADM patiënten werden voor en 6 weken na behandeling met de ACE-remmer, enalapril, onderzocht. Conform de verwachting trad er tijdens enalapril behande-ling een daling op van de bloeddruk en de albuminurie, en bleek de ERPF te zijn toege-nomen. De bloeddrukstijging onder invloed van NA was echter meer uitgesproken tijdensenalapril behandeling. De ERPF daling en de albuminurie tijdens NA infusie werd nietdoor enalapril beïnvloed. Deze bevindingen suggereren dat enalapril onvoldoende be-scherming biedt tegen systemische en renale NA effecten.

In hoofdstuk 6 is onderzocht in hoeverre intraveneus toegediend dopamine in staat ishet renale vasoconstrictieve effect van NA infusie te antagoneren. Dopamine wordt veeltoegepast om bij haemodynamisch instabiele patiënten de nierfunctie te ondersteunentijdens NA behandeling. Bij 7 gezonde vrijwilligers werd in gerandomiseerde volgordedopamine en placebo toegevoegd aan 3 verschillende NA infusie doses. Dopamine bleekde door NA-geïnduceerde ERPF daling zeer effectief te antagoneren. Er was een minderuitgesproken bloeddruk stijging en een sterke toename van de zout uitscheiding na toe-voeging van dopamine aan de NA infusen. Deze studie bevestigt de veronderstelling datlage dosis dopamine bij de mens de vasoconstrictieve effecten van NA in de nier tegeng-aat. Of dit ook geldt voor instabiele patiënten zal nader onderzocht moeten worden. Deverminderde door NA-geïnduceerde bloeddrukstijging en de toegenomen zoutexcretieonder invloed van dopamine kunnen bij haemodynamisch instabiele patiënten echterminder gewenst zijn.

Effecten van groeihormoon en insulin-like growth factor-I op de nier.

De GH-IGF-I-as wordt bij de mens beschouwd als een determinant van de nierfunctie.Exogeen toegediend GH en IGF-I stimuleren zowel de GFR als de ERPF. Het effect vande GH-IGF-I-as op microproteïnurie is niet goed bekend. In hoofdstuk 7 is onderzocht ofde albumine uitscheidingssnelheid in de urine gerelateerd is aan GH en IGF-I spiegels inhet bloed. In deze studie werden patiënten met GH deficiëntie, patiënten met(on)behandelde acromegalie ten gevolge van een GH producerend hypofyse adenoom engezonde controle personen met elkaar vergeleken. De albumine uitscheiding bleek hoger

Nederlandstalige samenvatting 159

bij de patiënten met acromegalie en was iets lager bij de patiënten met GH deficiëntie tenopzichte van de controle personen. Bij de acromegaliepatiënten resulteerde medicamenteuze verlaging van de GH en IGF-I spiegels met octreo-tide in een daling van de albumine uitscheidingssnelheid. Er bestond een positieve relatietussen de GH en IGF-I spiegels en de albumine uitscheidingssnelheid, onafhankelijk vande kreatinine klaring. Deze bevindingen bevestigen dier experimentele studies dat ver-hoogde GH en IGF-I spiegels microproteïnurie kunnen induceren, en zijn in overstem-ming met de toename van albuminurie na IGF-I infusie bij gezonde vrijwilligers.

In hoofdstuk 8 werden de parameters voor de nierfunctie (GFR en ERPF) basaal en nastimulatie door middel van aminozuren infusie gemeten. Vergeleken werden GH defici-ente patiënten, (on)behandelde acromegalie patiënten, IADM patiënten met een normaleen een verhoogde GFR (normo- versus hyperfiltratie) en gezonde controle personen. Ookwerden de IGF-I spiegels in relatie tot de gemeten nierfunctieparameters vergeleken.

De basale GFR en ERPF waarden correspondeerden met de mate van GH secretie: delaagste GFR en ERPF waarden werden gevonden bij de GH deficiënte patiënten, gevolgddoor een hogere GFR en ERPF bij de controle personen, behandelde en onbehandeldeacromegalie patiënten. Er bleek in alle groepen een inverse relatie te bestaan tussen dehoogte van de basale GFR en ERPF en de toename van GFR en ERPF na stimulatie metaminozuren infusie. De GFR en ERPF waren bij de acromegalie en hyperfiltrerendeIADM patiënten niet meer stimuleerbaar. De GFR en ERPF correleerden met de plasmaIGF-I spiegel bij de GH deficiënte patiënten, (on)behandelde acromegalie patiënten en decontrole personen, maar niet bij de IADM patiënten.

Een verhoogde GFR (glomerulaire hyperfiltratie) is een fenomeen dat bij IADM enacromegalie wordt gevonden. Bij patiënten met acromegalie is deze het gevolg van dehoge GH en IGF-I spiegels, maar de rol van de GH-IGF-I-as bij diabetische glomerulairehyperfiltratie is niet duidelijk. De afwezige GFR en ERPF respons na aminozuren infusiesuggereert dat de renale reserve capaciteit bij glomerulaire hyperfiltratie maximaal benutis. Deze opmerkelijke overeenkomst tussen de acromegalie en hyperfiltrerende IADMpatiënten werd echter niet weerspiegeld in een verhoogde plasma IGF-I concentratie bijde hyperfiltrerende IADM patiënten. Bij diabetische ratten is accumulatie van IGF-I in denier aangetoond. Men veronderstelt dat toegenomen renale IGF-I receptor activiteit enverandering in lokale productie van IGF-I bindende eiwitten hieraan ten grondslag liggen.Op een dergelijke wijze zou de GH-IGF-I-as een rol kunnen spelen bij glomerulairehyperfiltratie bij IADM patiënten, ondanks het feit dat de plasma IGF-I spiegel nietverhoogd is. Deze veronderstelling zou bij IADM patiënten met een IGF-I infusie experi-ment onderzocht kunnen worden. De beperkte beschikbaarheid van IGF-I maakt eendergelijke studie momenteel niet mogelijk.

In hoofdstuk 9 werden de door inspanning gestimuleerde GH en glucagon spiegelsgeëvalueerd in IADM patiënten met normale en verhoogde GFR. Glucagon heeft evenalsGH een vasodilaterende werking op de nier. Glucagon spiegels zijn dikwijls verhoogd bijIADM patiënten. Onderzocht werd of deze hormonen zouden kunnen bijdragen aan deverhoging van GFR, ERPF en niergrootte bij IADM patiënten. Door middel van gestand-aardiseerde GFR en ERPF metingen werden IADM patiënten met glomerulaire hyperfil-tratie geselecteerd. Deze werden individueel gematcht met normo-filtrerende IADM

160

patiënten.De gemiddelde concentratie van deze hormonen was niet significant verschillend

tussen de normo- en hyperfiltrerende IADM patiënten. Echter, multipele regressie analy-se toonde dat de door inspanning gestimuleerde GH spiegel, de glucagon concentratie enhet HbA1c gehalte als maat voor diabetes regulatie, positief correleerden met de GFR,ERPF en niergrootte.

Een eerdere studie toonde geen verschillen aan in dag profielen van deze hormonen bijnormo- en hyperfiltrerende IADM patiënten. Overeenkomstig onze studie is ook gevon-den dat de GH afgifte na GH-releasing hormoon toediening verhoogd is bij hyperfiltre-rende ten opzichte van normofiltrerende IADM patiënten. Deze bevindingen suggererendat GH en glucagon een rol kunnen spelen bij glomerulaire hyperfiltratie in IADMpatiënten.

Uit de hoofdstukken 7, 8 en 9 wordt geconcludeerd dat de GH-IGF-I-as de renalehaemodynamiek en proteïnurie beïnvloedt bij diverse patiënten categorieën, waaronderGH deficiëntie, acromegalie en IADM. Vanuit een therapeutisch perspectief, bestaat devraag of somatostatine analogen de progressie van albuminurie en nierfunctieverlies bijIADM patiënten zouden kunnen tegengaan. Een dergelijke interventie met GH en IGF-Iconcentratie verlagende middelen zou mogelijk een waardevolle aanvulling kunnen zijnop bloeddruk verlagende behandeling. Met het beschikbaar komen van langwerkendesomatostatine analogen zal dit binnen afzienbare tijd onderzocht kunnen worden.

Het effect van glucose op de immunoglobuline-G concentratie in de urine.

Het immunoglobuline-G is een groter eiwit dan albumine. Het meten van deimmunoglobuline-G uitscheidingsnelheid in de urine verschaft informatie over de veran-deringen in de poriën grootte van het glomerulaire basaal membraan. Bij normoalbumi-nurische IADM patiënten is gerapporteerd dat ook immunoglobuline-G in ver-hoogdemate wordt uitgescheiden. Deze bevinding is onbegrepen omdat de poriën-grootte van deglomerulaire basaal membraan bij deze patiënten geacht wordt onveranderd te zijn.

In hoofdstuk 10 vonden ook wij dat normoalbuminurische IADM patiënten een ver-hoogde mate van immunoglobuline-G excretie in de urine hadden ten opzichte vancontrole personen. Tevens bleek de immunoglobuline-G uitscheiding hoger te zijn inurine verzamelingen van IADM patiënten die glucose bevatten dan in die van patiëntenzonder glucosurie. Aangezien het immunoglobuline-G gehalte in de urine kanverminderen bij het bewaren van de monsters bij -20 C gedurende enkele weken, veron-0

derstelden wij dat de aanwezigheid van glucose mogelijk een conserverend effect op hetimmunoglobuline-G had. In een laboratorium experiment werd deze hypothese getoetst.Getest werden diverse bewaarregimes op de houdbaarheid van immuno-globuline-G inurine monsters van niet-diabetes patiënten met een verschillende mate van proteïnurie. Deimmunoglobuline-G concentratie daalde in de monsters die onbehandeld bij -20 C waren0

ingevroren. Daarentegen bleek een voorbehandeling met fosfaatbuffer, koeienalbumineen glucose te beschermen tegen het verlies van immuno-globuline-G. Het minste verliesna 12 weken invriezen bij -20 C werd gevonden bij die urine monsters, waaraan een hoge0

concentratie van glucose was toegevoegd. Deze bevinding impliceert dat de aanwezigheid

Nederlandstalige samenvatting 161

van glucose in de urine van IADM patiënten preserverend werkt op het urineimmunoglobuline-G gehalte, wanneer urine verzamelingen zonder voorzorgsmaatregelenworden bewaard bij -20 C. Een dergelijk effect leidt tot de een schijnbare verhoging van0

de immunoglobuline-G uitscheiding bij normoalbuminurische IADM patiënten tenopzichte van niet diabetische personen. Er dienen dus voorzorgsmaatregelen getroffen teworden wanneer de urine immunoglobuline-G bepaling niet onmiddellijk kan wordenuitgevoerd.

De rol van 11$$-hydroxysteroid dehydrogenase bij de water- en zouthuishouding bijIADM patiënten.

In hoofdstuk 11 is onderzocht of veranderingen in het metabolisme vanglucocorticoïden, met name van de activiteit van het enzym 11$-hydroxysteroiddehydrogenase (11$-HSD), een rol spelen in de toegenomen water en zout retentie bijIADM patiënten. Hiertoe werden de metabole afbraakproducten van cortisol en cortisonin de urine gemeten bij normo- en microalbuminurische IADM patiënten en controlepersonen. De verhouding in de metaboliet uitscheiding werd als maat gebruikt voor de11$-HSD enzym activiteit. Dit enzym catalyseert de inter conversie tussen cortisol en hetmetabool inactieve, cortison, en beschermt de mineralocorticoïde receptor tegen active-ring door cortisol. Een verschuiving in het cortisol-cortison evenwicht richting cortisolzou bij IADM patiënten kunnen bijdragen aan de toegenomen zoutretentie en volumebel-asting. Omgekeerd zou een verandering in de richting van cortison hiertegen beschermen.

Bij normo- en microalbuminurische IADM patiënten werd een lagere excretie vancortisol en cortison metabolieten in de urine gevonden in vergelijking met de controlepersonen. Dit wijst erop dat de reductie van glucocorticoïden tot tetrahydrometabolietenbij IADM patiënten gestoord is. De urine cortisol/cortison metaboliet ratio was lager inde IADM groepen. Dit impliceert een verschuiving in het evenwicht tussen cortisol encortison ten gunste van cortison. Bij IADM patiënten bleek een inverse relatie tussen hetHbA1c gehalte en de urine cortisol/cortison metaboliet ratio te bestaan. Zowel bij IADMpatiënten als bij de controle personen was de urine cortisol/cortison metaboliet ratiopositief gerelateerd met het via een impedantie techniek gemeten extravasculaire- enbloedvolume. Deze relatie was in de IADM groepen naar links verschoven. Ditsuggereert dat de 11$HSD activiteit zoals weerspiegelt in de urine cortisol/cortisonmetaboliet ratio, een determinant is van volume homeostase. De linksverschuiving indeze relatie bij de IADM patiënten maakt het echter onwaarschijnlijk dat een veranderde11$-HSD activiteit verantwoordelijk is voor de toegenomen zoutretentie en volumeexpansie. De behandeling van de microalbuminurische IADM patiënten met een ACE-remmer induceerde een verdere verlaging van de urine cortisol/cortison metaboliet ratio.

Op basis van deze resultaten wordt de hypothese naar voren gebracht dat de gestoordereductie van glucocorticoïden en de 11$-HSD gemediëerde verschuiving in het corti-sol/cortison evenwicht bij IADM patiënten het gevolg kan zijn van een veranderdecofactor beschikbaarheid ten gevolge van chronische hyperglycaemie. De bevindingensuggereren ook dat verbeterde metabole controle tot een verandering in het evenwicht van cortisol/cortison ten gunste van cortisol kan leiden en aldus de zoutretentieen volume expansie bij IADM patiënten kan doen toenemen. De stimulatie van 11$-HSD

Nederlandstalige samenvatting 163

door ACE-remmers suggereert een additioneel werkingsmechanisme waardoor dezefarmaca zoutexcretie kunnen bevorderen.

Information on the author 163

CURRICULUM VITAE

Name Klaas Hoogenberg Address Middelhorsterweg 34

9751 TG Haren (Gn), The Netherlands

Date of birth 14 february 1962Place of birth The Hague Gender MaleCivil state Married, 1 son 2 years of age Nationality Dutch

Education Highschool Atheneum B, graduated june 1980Study Psychology for one year, 1981Medical School, accomplished august 1987Graduated as general physician, august 1989Educated in Internal Medicine 1990-1996

Previous position 1990 General physician Diabetes Center, Beatrixoord, Haren (Gn)1990-1996 Training in Internal Medicine in Univerisity Hospital,Groningen

Present position 1996 October assigned as internist for training in endocrinologyand diabetes (supported by a grant from the Dutch DiabetesFoundation) at the department of Edocrinology, UniversityHospital Groningen, P.O. Box 30.001, 9700 RB Groningen, The Netherlands

Publications

Dullaart RPF, Hoogenberg K, Groener JEM, Dikkeschei LD, Erkelens DW, DoorenbosH. The activity of cholesteryl ester transfer protein is decreased in hypothyroidism: apossible contribution to alterations in high-density lipoproteins. Eur J Clin Invest 1990; 20: 581-587

Dullaart RPF, Beusekamp BJ, Meijer S, Hoogenberg K, Van Doormaal JJ, Sluiter WJ.Long-term effects of linoleic-acid-enriched diet on albuminuria and lipid levels in Type1 (insulin-dependent) diabetic patients with elevated urinary albumin excretion.Diabetologia 1992; 35: 165-172

Hoogenberg K, Tupker RA, Van Essen LH, Smit AJ, Kallenberg CGM. Successfultreatment of ulcerating livedo reticularis with infusions of prostacyclin. Br J Dermatol 1992; 127: 64-66

164

Hoogenberg K, Dullaart-RPF. Abnormal plasma noradrenaline response and exerciseinduced albuminuria in type 1 (insulin-dependent) diabetes mellitus.Scand J Clin Lab Invest 1992; 52: 803-811

Hoogenberg K, Sluiter WJ, Dullaart RPF. Effect of growth hormone and insulin-likegrowth factor I on urinary albumin excretion: studies in acromegaly and growth hormonedeficiency.Acta Endocrinol (Copenhagen) 1993; 129: 151-157

Hoogenberg K, Dullaart-RPF, Freling NJM, Meijer-S, Sluiter WJ. Contributory roles ofcirculatory glucagon and growth hormone to increased renal haemodynamics in type 1(insulin-dependent) diabetes mellitus.Scand J Clin Lab Invest 1993; 53: 821-828

Hoogenberg K, Ter Wee PM, Lieverse AG, Sluiter WJ, Dullaart RPF. Insulin-likegrowth factor I and altered renal hemodynamics in growth hormone deficiency,acromegaly, and type I diabetes mellitus.Transplant Proc 1994; 26: 505-507

Dullaart RPF, Sluiter WJ, Dikkeschei LD, Hoogenberg K, Van Tol A. Effect of adiposityon plasma lipid transfer protein activities: a possible link between insulin resistance andhigh density lipoprotein metabolism.Eur J Clin Invest 1994; 24: 188-194

Dullaart-RPF, Hoogenberg K, Dikkeschei BD, Van Tol A. Higher plasma lipid transferprotein activities and unfavorable lipoprotein changes in cigarette-smoking men.Arterioscler Thromb 1994; 14: 1581-1585

Van Haeften TW, Hoogenberg K; Van Essen LH. Acute pancreatitis in a patient withfamilial benign hypercalcaemia.Neth J Med 1994; 45: 110-113

Hoogenberg K, Van Essen LH, Van den Dungen JJ, Limburg AJ, Boeve WJ, KleibeukerJH. Chronic mesenteric ischaemia: diagnostic challenges and treatment options.J Intern Med 1995; 237: 293-299

Dullaart RPF, Van Doormaal JJ, Hoogenberg K, Sluiter WJ. Triiodothyronine rapidlylowers plasma lipoprotein (a) in hypothyroid subjects.Neth J Med 1995; 46: 179-184

Hoogenberg K, Girbes ARJ, Stegeman CA, Sluiter WJ, Reitsma WD, Dullaart RPF.Influence of ambient plasma noradrenaline on renal haemodynamics in type 1(insulin-dependent) diabetic patients and healthy subjects.Scand J Clin Lab Invest 1995; 55: 15-22

Dullaart RPF, Ubels FL, Hoogenberg K, Smit AJ, Pratt JJ, Muntinga JHJ, Sluiter WJ,Wolthers BG. Alterations in cortisol metabolism in insulin-dependent diabetes mellitus:

Information on the author 165

relationship with metabolic control and estimated blood volume and effect ofangiotensin-converting enzyme inhibition.J Clin Endocrinol Metab 1995; 80: 3002-3008

Hoogenberg K, Visser P, Marrink J, Sluiter WJ, Dullaart RPF. Increased urinaryIgG/albumin index in normoalbuminuric insulin-dependent diabetic patients: a laboratoryartefact.Diabetic Medine 1996; 13: 651-655

Riemens SC, Van Tol A, Hoogenberg K, Van Gent T, Scheek LM, Sluiter WJ, DullaartRPF. Higher high density lipoprotein cholesterol associated with moderate alcoholconsumption is not related to altered plasma lecithin:cholesterol acyltransferase and lipidtransfer protein activity levels.Clin Chim Acta 1997; 258: 105-115

Jessurun GA, Hoogenberg K, Crijns HJ. Bronchiolitis obliterans organizing pneumoniaduring low-dose amiodarone therapy.Clin Cardiol 1997; 20: 300-302

Hoogenberg K, Smit AJ, Girbes ARJ. Effects of low dose dopamine on renal andsystemic haemodynamics during incremental norepinephrine infusion in healthyvolunteers. Crit Care Med 1998 (in press)

Girbes ARJ, Hoogenberg K, Smit AJ. ‘Renal’ dopamine. Br J Anaesthesia 1997 (in press)

Girbes ARJ, Hoogenberg K. The use of dopamine and norepinephrine in the ICU. Focuson renal function: hope and hype of dopamine versus fear and facts of norepinephrine.Year Book of Intensive Care Medicine 1998 (in press)

Dullaart RPF, Hoogenberg K, Riemens SC, Groener JEM, Van Tol A, Sluiter WJ, StulpBK. Cholesteryl ester transfer protein gene polymorphism is a determinant of highdensity lipoprotein cholesterol and of the lipoprotein response to a lipid lowering diet inIDDM. Diabetes (in press)

Hoogenberg K, Sluiter WJ, Navis G, Van Haeften TW, Smit AJ, Reitsma WD, DullaartRPF Exogenous norepinephrine induces an enhanced microproteinuric response inmicroalbuminuric IDDM.J Am Soc Nephrol (in press)

Hoogenberg K, Navis G, Dullaart RPF. Norepinephrine-induced blood pressure rise andrenal vasoconstriction is not attenuated by enalapril in microalbuminuric IDDM.Nephrol Dial Transplant (in press)