37

Angiotensin-converting enzyme activity and cognitiveimpairment during hypoglycaemia in healthy humans

Ulrik Pedersen-Bjergaard,* Carsten E Thomsen,† Hans Høgenhaven,# Annelise Smed,^

Troels W Kjær, # Jens J Holst, § Flemming Dela,¥ Linda Hilsted,∞ Erik Frandsen,+ Stig Pramming,±

Birger Thorsteinsson*

Key words: angiotensin-converting enzyme,hypoglycaemia,cognition,counterregulation

* Endocrinology Section,Division of InternalMedicine I, HillerødHospital, Hillerød.

† Department of OralPathology, CopenhagenSchool of Dentistry,Copenhagen.

# Department ofNeurophysiology.

^ Department ofNeurology, Rigshospitalet,Copenhagen.

§ Department of MedicalPhysiology.

¥ Copenhagen MuscleResearch Centre,Department of MedicalPhysiology, University ofCopenhagen.

∞ Department of ClinicalBiochemistry,Rigshospitalet,Copenhagen.

+ Department of ClinicalPhysiology, GlostrupUniversity Hospital,Glostrup.

± Novo Nordisk A/S,Bagsværd, Denmark.

Paper

AbstractIntroduction: In type 1 diabetes increased riskof severe hypoglycaemia is associated with highangiotensin-converting enzyme (ACE) activity.We tested in healthy humans the hypothesis thatthis association is explained by the reducedability of subjects with high ACE activity tomaintain normal cognitive function duringhypoglycaemia.Methods: Sixteen healthy volunteers selected byeither particularly high or low serum ACE activitywere subjected to hypoglycaemia (plasmaglucose 2.7 mmol/L). Cognitive function wasassessed by choice reaction tests.Results: Despite a similar hypoglycaemicstimulus in the two groups, only the group withhigh ACE activity showed significantdeterioration in cognitive performance duringhypoglycaemia. In the high ACE group meanreaction time (MRT) in the most complex choicereaction task was prolonged and error rate (ER)was increased in contrast to the low ACE group.The total hypoglycaemic symptom response wasgreater in the high ACE group than in the lowACE group (p=0.031).There were no differencesin responses of counterregulatory hormones or inconcentrations of substrates between the groups.Conclusion: Healthy humans with high ACEactivity are more susceptible to cognitivedysfunction and report higher symptom scoresduring mild hypoglycaemia than subjects withlow ACE activity.

IntroductionHypoglycaemia is the dominant side effect toinsulin therapy in type 1 diabetes and representsa major barrier to achievement of optimalglycaemic control in many patients.1 While someepisodes are recognised and corrected by thepatients many hypoglycaemic episodes remainunrecognised due to impaired awareness ofhypoglycaemic warning symptoms.2 Duringunrecognised hypoglycaemic episodes the abilityto maintain cognitive function is important for thepatients to continue their activities until theepisodes are corrected by planned meals, waningof the insulin effect, or hormonalcounterregulation. Impaired cognition may leadto severe hypoglycaemia with need for assistance

from other persons to restore glucose levels. Thehighly variable interindividual risk of severehypoglycaemia in type 1 diabetes that is onlypartly explained by hypoglycaemia awarenessand hormonal counterregulatory capacity3,4

suggests clinically significant differences betweensubjects in the ability to maintain cognitivefunction during hypoglycaemia. The sparseknowledge about factors influencing cognitionduring hypoglycaemia is a major obstacle to theunderstanding and prevention of severehypoglycaemia.

We and others previously reported an associationbetween high spontaneous serum angiotensin-converting enzyme (ACE) activity and occurrenceof severe hypoglycaemia in type 1 diabetesindependent of hypoglycaemia awareness andother previously known risk factors.5-8 Thisfinding led us to hypothesise that subjects withhigh ACE activity may have a reduced capabilityto maintain normal cognitive function duringhypoglycaemia.

The aim of this study was to assess therelationship between ACE activity and cognitivefunction during hypoglycaemia and specificallyto test the hypothesis that healthy subjects withhigh spontaneous ACE activity are more affectedby hypoglycaemia in terms of cognitivedysfunction than subjects with low ACE activity.Moreover, we aimed at exploring associationsbetween ACE activity and symptomatic andhormonal responses.

Research, design and methodsSubjectsIn a cohort of 53 healthy subjects recruited to ascreening phase a venous blood sample wasdrawn for measurement of serum ACE activityand determination of ACE Insertion/Deletiongenotype as described previously.9 Those eightsubjects with the highest and those eight with thelowest screening values of serum ACE activity(mean [range]: 52.3 [35–67] and 20.4 [15–26] U/L)were selected to participate in the experiment.The characteristics of the subjects are shown intable 1. Subjects were instructed to live and eat asnormally as possible and to avoid vigorous

doi:10.3317/jraas.2008.001

Journal ofthe Renin-Angiotensin-AldosteroneSystem(Including other

Peptidergic Systems)

March 2008Volume 9Number 1

COPYRIG

HT JRA

AS LIM

ITED

REPR

ODUCTION PR

OHIBITE

D

38

exercise, as well as the use of alcohol and drugs,in the week preceding the experiments. The studywas approved by the regional Ethics Committee.

DesignThe study was a randomised counterbalancedcross-over study between insulin and placebowith at least three weeks between sessions.

Subjects were blind to treatment andexperimenters were blind to subjects’ ACE status.The purpose was to assess differences duringexperimentally induced hypoglycaemia betweensubjects with high or low ACE activity and suchdifferences could potentially occur in everyaspect of response to, or protection against,hypoglycaemia (degree of hypoglycaemia,

Journal ofthe Renin-Angiotensin-AldosteroneSystem(Including other

Peptidergic Systems)

March 2008Volume 9Number 1

Paper

Correspondence to:Dr Ulrik Pedersen-BjergaardEndocrinology Section0652,Division of InternalMedicine I, HillerødHospital, Helsevej 2,DK-3400 Hillerød,Denmark.Tel: +45 4829 6403Fax: +45 4829 4783E-mail: ulpebj@noh.regionh.dk

Accepted for publication21st January 2008

JRAAS 2008;9:37–48

Table 1Baseline characteristics of subjects according to ACE activity. Mean (SD) or n when indicated.

ACE activity

0 2.55

Low Highn=8 n=8 p *

Age (years) 24.8 (2.0) 24.6 (1.9) 0.90

Gender (F/M) (n) 4 / 4 2 / 6 0.31

BMI (kg/m2) 21.1 (2.2) 22.1 (1.5) 0.32

Renin-angiotensin system (RAS) characteristics

serum ACE activity (U/l) 20.4 (4.0) 52.3 (11.7)ACE genotype (II; ID; DD)† (n) 6; 2; 0 0; 4; 4plasma angiotensinogen (nmol/L) 1664 (1064) 1342 (944) 0.53plasma renin (mIU/L) 26.2 (18.3) 18.8 (8.5) 0.32plasma angiotensin II (pmol/L) 21.6 (9.3) 16.6 (7.4) 0.25plasma aldosterone (pmol/L) 203 (63) 232 (186) 0.68

Plasma concentrations of substrates

glucose (mmol/L) 5.17 (0.12) 5.14 (0.14) 0.89lactate (mmol/L) 0.95 (0.13) 0.99 (0.23) 0.70β-hydroxybutyrate (mmol/L) 0.065 (0.021) 0.091 (0.134) 0.59free fatty acids (micromol/L) 415 (115) 353 (99) 0.27glycerol (micromol/L) 348 (65) 271 (58) 0.025alanine (micromol/L) 309 (40) 324 (59) 0.56

Plasma concentrations of counterregulatory hormones

glucagon (pmol/L) 8.8 (3.0) 11.2 (4.3) 0.23adrenaline (nmol/L) 0.43 (0.16) 0.37 (0.13) 0.44noradrenaline (nmol/L) 1.20 (0.42) 1.15 (0.33) 0.79cortisol (nmol/L) 462 (201) 527 (272) 0.60growth hormone (mU/L) 1.9 (1.9) 5.3 (7.8) 0.25pancreatic polypeptide (pmol/L) 12.4 (4.1) 12.3 (3.8) 0.96

CalCAP performance

MRT 1 (msec) 334 (67) 323 (47) 0.70MRT 2 (msec) 419 (18) 416 (35) 0.84MRT 3 (msec) 471 (58) 486 (83) 0.68MRT 4 (msec) 489 (79) 504 (84) 0.72ER 2 (per cent) 0.25 (0.23) 0.16 (0.27) 0.46ER 3 (per cent) 0.91 (0.74) 1.97 (2.18) 0.23ER 4 (per cent) 4.9 (3.2) 5.3 (2.6) 0.82

Hypoglycaemic symptom scores‡

autonomic 6.4 (0.7) 7.0 (2.1) 0.47neuroglycopenic 9.5 (3.1) 9.3 (3.5) 0.88total§ 24.7 (5.5) 25.2 (7.7) 0.88

Cardiovascular characteristics

heart rate (bpm) 62 (9) 59 (7) 0.48SBP (mmHg) 122 (6) 124 (8) 0.58DBP (mmHg) 72 (6) 75 (6) 0.36pulse pressure (mmHg)ıı 50 (7) 49 (6) 0.79

Key: ACE = angiotensin-converting enzyme; BMI = body mass index; CalCAP = California Computerised AssessmentPackage; D = Deletion; DBP = diastolic blood pressure; I = Insertion; SBP = systolic blood pressure.* = p-values for testsbetween the two groups (unpaired t-test [except for gender and ACE genotype: Chi2 test]); † = ACE genotypes;‡ = Edinburgh Hypoglycaemia Symptom Score Questionnaire (10) § = autonomic + neuroglycopenic + nonspecific; ıı = SBP minus DBP.

COPYRIG

HT JRA

AS LIM

ITED

REPR

ODUCTION PR

OHIBITE

D

39

magnitudes of hormonal responses, per-formance during hypoglycaemia). Therefore, wepreferred a standardised subcutaneous insulinchallenge to the experimental setting ofhypoglycaemic glucose clamping methods. Mostimportantly, we avoided any exogenous nutrientsupply during the hypoglycaemic phase of theexperiments in order not to interfere with thenatural mobilisation and fluxes of substrates.

Experimental protocolOn the day of experiment, subjects arrived at7.30 a.m. in the laboratory after an overnight fast.Following careful introduction to theexperimental setting and scheme, subjects wereequipped as follows: 1) Electroencephalogram(EEG) cap and two precordial electrocardiogram(EKG) leads connected to a digital EEG recorder(Cadwell, Kennewick, Washington, USA); 2) headphones (in-ear type) connected to a computerdelivering auditory stimuli; 3) ambulatory bloodpressure (BP) monitoring unit (Bosch + Sohn,Jungingen, Germany); 4) intravenous cannula inan ante-cubital vein in the non-dominant forearm.Subsequently, subjects were introduced to thecomputerised cognitive function test, theCalifornia Computerised Assessment Package(CalCAP) (Miller and Norland Software, LosAngeles, California, USA);10 the test wasperformed once to accustom subjects to theprocedure. Following this, the experiment wascarried out in a cyclic manner according to thescheme presented in figure 1. Each cycle wasinitiated by a bedside blood glucose measurementfollowed by three minutes of rest for basal EEGmonitoring. Subsequently, 30 ml of venous bloodwas drawn for measurement of plasma glucose,substrates, counterregulatory hormones, andrenin-angiotensin system (RAS) hormones.Thereafter, auditory evoked potentials (AEP) wererecorded, another bedside blood glucose wasmeasured, and the CalCAP test was performed.

Finally, hypoglycaemic symptoms were ratedaccording to the Edinburgh HypoglycaemiaSymptom Score Questionnaire.11

A total of seven cycles were performed. Aftereach of the first three cycles regular humaninsulin (Actrapid, Novo Nordisk, Bagsværd,Denmark) 0.075 U/kg + 0.075 U/kg + 0.15 U/kgor placebo injections were given subcutaneouslyin the abdominal wall (figure 1). In pilot studiesthis dosage scheme resulted in stablehypoglycaemia with nadir in the range between2.5–3.0 mmol/L, which was the hypoglycaemictarget in this study. After the sixth cycle, subjectswere given 0.5 litre of apple juice and rested for20 minutes before the last cycle was carried out.

Cognitive function testsThe Danish version of the abbreviated CalCAP10

test battery was used in the study. This testconsists of four different reaction time tasks withdifferent complexity based on presentations ofnumbers on the computer screen: 1) a simplereaction time test where subjects are instructed tostrike a key whenever a number is presented; 2) achoice reaction task with reaction only to a specificnumber; 3) a choice reaction task with reaction topresentation of two identical numbers in asequence; and 4) a similar test with reaction topresentation of two consecutive numbers in asequence. Performance of the two latter testsincludes the use of working memory. From the testresults the individual mean reaction times (MRT) foreach test (MRT 1, MRT 2, MRT 3, and MRT 4) andthe error rates (ER = percentage of false positive +false negative reactions) for the three choicereaction tests (ER 2, ER 3 and ER 4) were extracted.

EEG and AEPContinuous 20-channel EEG was recordeddigitally (200 Hz sampling rate) throughout theexperiment. AEP was obtained by an auditory

Journal ofthe Renin-Angiotensin-AldosteroneSystem(Including other

Peptidergic Systems)

March 2008Volume 9Number 1

Paper

Experimental protocolTreatment

Insulin* Insulin* Insulin* Apple juice

Cycle:

Baseline Induction Hypoglycaemia Recovery

Sequence of single cycle:

0 30 60 90 120 150 180 230Minutes

Time:

AEPBlood samples B-glucose Symptom scoreB-glucose

200

CalCAP

Cycle 1 Cycle 2 Cycle 3 Cycle 4 Cycle 5 Cycle 6 Cycle 7

EEG

Figure 1Schematic design of experimental sessions. AEP = auditory evoked potentials; CalCAP = California Computerised AssessmentPackage; EEG = electronencephalogram; * = or placebo.

COPYRIG

HT JRA

AS LIM

ITED

REPR

ODUCTION PR

OHIBITE

D

40

stimulation paradigm consisting of more than 500stimuli delivered with an attended: unattendedear ratio of 1:4. In order to maintain attentionsubjects were asked to count the number ofstimuli given to the attended ear. Two differenttones were used: 700 Hz (attended ear) and350 Hz (unattended ear). The duration of eachstimulus was 100 ms and the inter-stimulusinterval was 280 ms. The AEP sweeps wereextracted from the continuous EEG recordingsbased on trigger pulses stored on auxiliarychannels on the EEG recorder. The difference inamplitude between time-averaged (230–255 ms)potentials of attended and unattended sweeps(odd-ball response) was calculated.

Biochemical analysesBlood samples were collected on ice, centrifugedat 5oC within 20 minutes and specimens werestored immediately after at -80oC until analysis.Bedside blood glucose used for monitoringduring the experiments was carried out by aHemoCue Analyser every 15 minutes. Plasmaglucose used for evaluation of the hypo-glycaemic challenge was measured enzymatically(COBAS INTEGRA, Roche, Basel, Switzerland).

Glucagon concentrations in plasma weremeasured with a radioimmunoassay directedagainst the C-terminus of the glucagon molecule(antibody code no. 4305), which mainlymeasures glucagon of pancreatic origin.12 Themeasurement of plasma catecholamines wasbased on high-pressure liquid chromatographicseparation of radioenzymatically labeledcatecholamines.13,14 Plasma cortisol was measuredusing a commercial kit (DSL-2100, DiagnosticSystems Laboratories Inc., Webster, Texas). Serumgrowth hormone concentration was determinedwith a commercial fluoroimmunometric method(Delfia, Wallac, Finland) using two monoclonalantibodies against two separate antigenicdeterminants on the growth hormone molecule.Pancreatic polypeptide was measured by asensitive and specific radioimmunoassaydescribed previously15 using antibody 146-5. Thecalibrator was synthetic human PP (Sigma, StLouis, MO, USA) and the tracer was synthetichuman PP labelled with 125I using iodogen.

Plasma concentration of ketone bodies (D-β-hydroxybutyrate) was measured by afluorometric method (Ascent analyser,Labsystems, Helsinki, Finland) and free fattyacids, glycerol, lactate and alanine weredetermined spectrophotometrically (Hitachi 912autoanalyzer, Roche, Glostrup, Denmark). Plasmaconcentrations of renin and angiotensinogenwere determined using the principle of antibodytrapping16 as modified by Millar et al.17 Plasmaangiotensin (Ang) II concentration was measuredaccording to the method of Kappelgaard et al.18

with the modification that Sep-Pac C18 (MilliporeWaters) was used for plasma extraction. Plasmaaldosterone concentration was measured using acommercial kit (DSL-8600, Diagnostic SystemsLaboratories Inc., Webster, Texas).

StatisticsResponses are shown as mean (±SE) netresponses (=placebo-corrected Δ-values) unlessotherwise stated. The effect of hypoglycaemiawas assessed by one-sample t-test of the areaunder curve (AUC) and differences betweengroups were analysed by unpaired t-tests andtwo-sided p-values <0.05 were consideredsignificant. The SPSS software package (Version11.0) was used for statistical calculations.

ResultsPlasma glucoseInsulin administration resulted in a similarhypoglycaemic stimulus in both of the ACE groupswith plasma glucose nadir of 2.7 (0.1) mmol/L(figure 2). This was also confirmed by thefrequent bedside measurements (not shown).Recovery of blood glucose following ingestion ofapple juice was similar in the two ACE groupswith prompt rises to 4.8 (0.3) mmol/L. No declinein plasma glucose was observed followingplacebo injections in any of the groups.

Cognitive functionThe baseline performance in all tests both interms of reaction time and error rate was similarin the two ACE groups (table 1). Duringhypoglycaemia the group with high ACE activitydeteriorated significantly in performance of themost complicated CalCAP tasks both with regardto speed and precision (figure 3). Thus, MRT 3and ER 3 deteriorated significantly (p=0.027 andp=0.020). A similar trend was observed for MRT4 and ER 4 (p=0.088 and p=0.065). The maximumprolongation of MRT 3 and MRT 4 was 120 ms

Journal ofthe Renin-Angiotensin-AldosteroneSystem(Including other

Peptidergic Systems)

March 2008Volume 9Number 1

Paper

8

7

6

5

4

3

2

1

00 50 100 150 200 250

Time (minutes)

InductionBaseline Hypoglycaemia Recovery

mM

Low ACE - insulinHigh ACE - insulin

Low ACE - placeboHigh ACE - placebo

Figure 2Plasma glucose concentrations during the experiments (mean[SEM]). ACE = angiotensin-converting enzyme.

COPYRIG

HT JRA

AS LIM

ITED

REPR

ODUCTION PR

OHIBITE

D

41

Journal ofthe Renin-Angiotensin-AldosteroneSystem(Including other

Peptidergic Systems)

March 2008Volume 9Number 1

Paper

150

100

50

0

-50

-1000 50 100 150 200 250

Time (minutes)

Low ACE (n.s.)High ACE (p=0.027)

c

150

100

50

0

-50

-1000 50 100 150 200 250

Time (minutes)

Low ACE (n.s.)High ACE (n.s.)

a

150

100

50

0

-50

-1000 50 100 150 200 250

Time (minutes)

b

Low ACE (n.s.)High ACE (n.s.)

150

100

50

0

-50

-1000 50 100 150 200 250

Time (minutes)

Low ACE (n.s.)High ACE (p=0.068)

d

10

8

6

4

2

0

-2

-40 50 100 150 200 250

Time (minutes)

e

10

8

6

4

2

0

-2

-40 50 100 150 200 250

Time (minutes)

f

Low ACE (n.s.)High ACE (n.s.)

Low ACE (n.s.)High ACE (p=0.020)

Mill

isec

on

ds

Mill

isec

on

ds

Mill

isec

on

ds

Mill

isec

on

ds

% %

g

10

8

6

4

2

0

-2

-40 50 100 150 200 250

Time (minutes)

Low ACE (n.s.)High ACE (p=0.065)

%

Figure 3Cognitive function assessed by the abbreviated CalCAP test battery. Shown are mean (SEM) net-responses (placebo-correctedΔ- values) of mean reaction times (MRT) and error rates (ER) during experiments. a: MRT 1; b: MRT 2; c: MRT 3; d: MRT 4; e: ER 2;f: ER 3; g: ER 4. p-values: one-sample t-test of AUC. ACE = angiotensin-converting enzyme; AUC = area under curve;CalCAP = California Computerised Assessment Package; n.s. = non-significant.

COPYRIG

HT JRA

AS LIM

ITED

REPR

ODUCTION PR

OHIBITE

D

42

and 110 ms corresponding to a 25% and 22%increment (figure 3c and d). The maximumabsolute increment in ER 3 and ER 4 was 4.8%and 4.3% corresponding to relative increments of244% and 81% (figure 3e and f). The cognitivedysfunction in the high ACE group persistedeven after normalisation of plasma glucose. Inthe group with low ACE activity, in contrast,cognitive function was not affected significantlyby hypoglycaemia. The high ACE groupdeteriorated significantly more in performance ofthe most complex task compared to the low ACEgroup. Thus, in the high ACE group the ER 3response was greater (p=0.016) and the MRT 3response tended to be greater (p=0.054). Therewere no alterations during the experiments inMRT 1, MRT 2 or ER 2 in any of the two groups.

Auditive evoked potentialsThere was no effect of hypoglycaemia on thedifference between response to attended andunattended sweeps (odd-ball response) in any ofthe groups (not shown).

Hypoglycaemic symptomsThere was no difference in baseline symptomscores between the groups (table 1). Both groupsreported significant autonomic symptom responsesduring hypoglycaemia compared to placebo (figure

4; low ACE group: p=0.036, high ACE group:p=0.003). The neuroglycopaenic and totalsymptom responses were significantly differentfrom placebo in the high ACE group (p=0.02 andp=0.004) but only tended to be so in the low ACEgroup (p=0.074 and p=0.091). The autonomic andtotal symptom response were greater in the highACE group than in the low ACE group (p=0.050and p=0.031). Following normalisation of plasmaglucose, autonomic and total symptom scoresremained elevated in the high ACE group.

Hormonal counterregulationSignificant responses were observed for allcounterregulatory hormones during hypogly-caemia (figure 5). There were no differences inthe catecholamine, glucagon or cortisolresponses between the groups. The responses ofgrowth hormone and pancreatic polypeptide inthe high ACE group tended to be greater than inthe low ACE group (p=0.065 and p=0.060).

SubstratesA significant decline in plasma-free fatty acidswas observed during and after hypoglycaemia inboth groups (figure 6a), whereas glycerolconcentrations remained unchanged during theexperiments (figure 6b). Concentrations of β-hydroxybutyrate were reduced during hypo-

Journal ofthe Renin-Angiotensin-AldosteroneSystem(Including other

Peptidergic Systems)

March 2008Volume 9Number 1

Paper

0 50 100 150 200 250 0 50 100 150 200 250Time (minutes)

b

Low ACE (p=0.036)High ACE (p=0.003)

15

10

5

0

-5

15

10

5

0

-5

Time (minutes)

Low ACE (p=0.074)High ACE (p=0.020)

a

Sym

pto

m s

core

c20

10

0

-100 50 100 150 200 250

Low ACE (p=0.091)High ACE (p=0.004)

Time (minutes)

Sym

pto

m s

core

Sym

pto

m s

core

Figure 4Symptomatic responses during experiments. Shown are mean (SEM) net-responses (placebo-corrected Δ-values). a: autonomicsymptom score; b: neuroglycopaenic symptom score; c: total symptom score. p-values: one-sample t-test of AUC.ACE = angiotensin-converting enzyme; AUC = area under curve.

COPYRIG

HT JRA

AS LIM

ITED

REPR

ODUCTION PR

OHIBITE

D

43

glycaemia but the response was only significantin the low ACE group (figure 6c). Lactateconcentrations rose in both groups and tendedto be greater in the low ACE group (p=0.10)(figure 6d). The alanine response to hypo-glycaemia did not differ from that to placebo(figure 6e). There were no significant differencesbetween the two groups in magnitude ofsubstrate responses.

RAS activationThere were no significant alterations in plasmaangiotensinogen concentrations during hypogly-caemia in any of the groups (figure 7a). Plasmarenin activity rose in both groups (figure 7b),but the response was only significant in the high

ACE group (p=0.034). No significant responses ofserum Ang II or plasma aldosterone con-centrations were observed (figure 7c and d).

Cardiovascular responsesThe pulse pressure tended to increase in bothgroups, but no significant alterations in heart ratewere recorded.

DiscussionThis study demonstrates an association betweenhigh spontaneous serum ACE activity andcognitive dysfunction during hypoglycaemia inhealthy humans. Subjects selected by high ACEactivity developed significant cognitive impair-ment lasting even after restoration of euglycaemia.

Journal ofthe Renin-Angiotensin-AldosteroneSystem(Including other

Peptidergic Systems)

March 2008Volume 9Number 1

Paper

0 50 100 150 200 250

1.0

0.8

0.6

0.4

0.2

0.0

-0.20 50 100 150 200 250

Time (minutes)

b

Low ACE (p=0.041)High ACE (p=0.025)

Time (minutes)

0 50 100 150 200 250

400

300

200

100

0

0 50 100 150 200 250Time (minutes)

f

Low ACE (p=0.013)High ACE (p=0.004)

0 50 100 150 200 250

800

600

400

200

0

0 50 100 150 200 250

Low ACE (p=0.051)High ACE (p=0.015)

d

nM

nM

40

30

20

10

0

Low ACE (p=0.003)High ACE (p=0.011)

c

Time (minutes)

6

5

4

3

2

1

0

Time (minutes)

Low ACE (p=0.002)High ACE (p=0.004)

a

100

80

60

40

20

0

Time (minutes)

e

Low ACE (p=0.021)High ACE (p=0.012)

nM

pM

mU

/L

pM

Figure 5Responses of counterregulatory hormones during experiments. Data shown are mean (SEM) net-responses (placebo-correctedΔ-values). a: adrenaline; b: noradrenaline; c: glucagon; d: cortisol; e: growth hormone; f: pancreatic polypeptide. p-values:one-sample t-test of AUC. ACE = angiotensin-converting enyzme; AUC = area under curve.

COPYRIG

HT JRA

AS LIM

ITED

REPR

ODUCTION PR

OHIBITE

D

Journal ofthe Renin-Angiotensin-AldosteroneSystem(Including other

Peptidergic Systems)

March 2008Volume 9Number 1

44

In contrast those selected by low ACE activityshowed no detectable cognitive dysfunctionduring or after hypoglycaemia despite a similarhypoglycaemic stimulus that was significant asevaluated by the symptomatic and counter-regulatory responses. The result is in accordancewith the association of high rate of severehypoglycaemia with elevated serum ACE activityconsistently reported in independent type 1diabetic cohorts5-8 and the association with the D-allele of the ACE genotype conferring high ACEactivity reported in some5,6 but not in all19 studies.We have recently identified two additional RAS-related risk factors for severe hypoglycaemia intype 1 diabetes: high plasma angiotensinogen

concentration and presence of the A-allele of the1675AT polymorphism of the Ang II subtype 2receptor conferring low expression of thisreceptor.20 This suggests that the effect of high ACEactivity is mediated through increased RAS activity.

A difference between the groups with high orlow ACE activity could potentially occur in everyaspect of the response to insulin or protectionagainst hypoglycaemia: level of hypoglycaemia,magnitudes of hormonal responses, mobilisationof substrates, and cognitive performance duringhypoglycaemia. To avoid interference with thespontaneous metabolism and substrate fluxesduring hypoglycaemia that may be important for

400

200

0

-200

-4000 50 100 150 200 25

Time (minutes)

b

Low ACE (n.s.)High ACE (n.s.)

1.0

0.8

0.6

0.4

0.2

0.0

0 50 100 150 200 25Time (minutes)

Low ACE (p=0.001)High ACE (p=0.035)

d

0 50 100 150 200 250

0 50 100 150 200 250

0.0

-0.1

-0.2

-0.3

-0.4

Time (minutes)

c

0

-100

-200

-300

-400

-500

Time (minutes)

Low ACE (p=0.001)High ACE (p=0.002)

a

Low ACE (p=0.007)High ACE (n.s.)

40

20

0

-20

-40

-60

-80

-1000 50 100 150 200 250

Time (minutes)

e

Low ACE (n.s.)High ACE (n.s.)

mM

μM

μM

mM

μM

Figure 6Plasma concentrations of circulating substrates during experiments. Data shown are mean (SEM) net-responses (placebo-correctedΔ-values). a: free fatty acids; b: glycerol; c: β-hydroxybutyrate; d: lactate; e: alanine. p-values: one-sample t-test of AUC.ACE = angiotensin-converting enzyme; AUC = area under curve.

Paper

COPYRIG

HT JRA

AS LIM

ITED

REPR

ODUCTION PR

OHIBITE

D

Journal ofthe Renin-Angiotensin-AldosteroneSystem(Including other

Peptidergic Systems)

March 2008Volume 9Number 1

45

maintenance of cognitive function we applied astandardised subcutaneous insulin challengewithout exogenous glucose supply instead ofhypoglycaemic clamping methods with variableglucose infusion rates to maintain stablehypoglycaemia as used by most groups.21-24 If ourapproach had resulted in variable depth ofhypoglycaemia it may have been regarded as alimitation of the study. However, we achieved asimilar hypoglycaemic stimulus in the twogroups with both standardised insulin dosageand glucose support (zero). Overall counter-regulatory responses were comparable to thoseobtained in studies using clamp techniques.21-25

Cognitive function during hypoglycaemia can beassessed by a variety of different tests and thereis no generally accepted gold standard.26

Although they do not provide a comprehensiveevaluation of cognition, choice reaction tasks(but not simple reaction tasks) are established assensitive and valid markers of cognitive declineduring hypoglycaemia.26 In accordance with this,there was no effect of hypoglycaemia onperformance of the simple reaction time task ineither ACE group, whereas in the most complexchoice reaction task requiring the use of workingmemory, reaction time was prolonged by 25 percent and error rate was more than doubled in the

group with high ACE activity. We did not find anychanges in the odd-ball response in the AEPalthough we studied the 230–255 ms intervalreported by other groups to change duringhypoglycaemia.27

There are no previous studies addressing therelationship between spontaneous ACE activityand cognitive function during hypoglycaemia.However, data from some human studies supportan association between high ACE or general RASactivity and poor function during hypoglycaemia.Thus, pharmacological blockade of ACE activityby ACE-inhibitor treatment has been reported toattenuate counterregulatory responses in termsof catecholamines28 and adrenocorticotropichormone (ACTH)29 and to reduce the glucoseinfusion rate necessary to maintain glucose levelin hypoglycaemic clamp studies in healthysubjects.30 Together, these studies suggest abeneficial effect of ACE-inhibition onmaintenance of metabolism during hypo-glycaemia. The finding that Ang II receptorblockade results in reduced symptoms andcounterregulatory responses in healthy subjects31

suggests that the effect may be due to inhibitionof RAS activity in general rather than ofACE activity. This is supported by a studyreporting more intense symptoms and greater

400

200

0

-200

-400

0 50 100 150 200 250Time (minutes)

Low ACE (n.s.)High ACE (n.s.)

a

15

10

5

0

-5

-100 50 100 150 200 250

Time (minutes)

c

Low ACE (n.s.)High ACE (n.s.)

pM

nM

30

20

10

0

-10

Time (minutes)

b

0 50 100 150 200 250

Low ACE (p=0.084)High ACE (p=0.034)

400

200

0

-200

-4000 50 100 150 200 250

Time (minutes)

Low ACE (n.s.)High ACE (n.s.)

d

mIU

/Lp

M

Figure 7Responses of renin-angiotensin-aldosterone system hormones during experiments. Data shown are mean (SEM) net-responses(placebo-corrected Δ-values). a: angiotensinogen; b: renin; c: Ang II; d: aldosterone. p-values: one-sample t-test of area undercurve (AUC).

Paper

COPYRIG

HT JRA

AS LIM

ITED

REPR

ODUCTION PR

OHIBITE

D

Journal ofthe Renin-Angiotensin-AldosteroneSystem(Including other

Peptidergic Systems)

March 2008Volume 9Number 1

46

counterregulatory responses during a hypogly-caemic clamp in the standing position (in whichthe RAS is activated) compared with the supineposition in subjects with type 1 diabetes.32 Thesuggestion of both of these latter studies that ablocked31 or inactive32 RAS is associated withblunted symptomatic responses to hypo-glycaemia is in accordance with the lesspronounced symptomatic responses in the lowACE group in this study.

Experimental animal data show that high activityof the RAS increases brain vulnerability tointerruption of fuel supply in conditions such asischaemia and hypoxia. Thus, in vivo studies inrodents subjected to experimental stroke haveshown that animals overexpressing angiotensin-ogen33 develop increased infarction volume andthat intracerebroventricular infusion of Ang IIresults in increased brain damage followingexposure to hypoxia.34 Knockout of the AT1-receptor gene or pharmacological blockade ofthe receptor, which like other components of theRAS is abundant throughout the brain, results inreduced infarction size33 indicating that the effectof Ang II is probably mediated by this receptor.

An effect of the renin-angiotensin system onbrain function during hypoglycaemia could becaused by an influence on either supply ordemand of substrate or both. Duringexperimental hypoglycaemia cerebral glucosesupply is maintained by increased cerebral bloodflow.35 Increasing cerebral blood flow furtherpharmacologically during hypoglycaemia inhealthy humans results in blunted epinephrineand hypoglycaemic symptom responses36

probably as a result of enhanced substratesupply. The cerebral vascular response tohypoglycaemia might be blunted in subjects withhigh circulating ACE activity. In our study, lactateconcentrations in the low ACE subjects tended toincrease more than in the high ACE subjectsduring and after hypoglycaemia. As infusion oflactate conserves human brain function duringhypoglycaemia37 increased lactate supply maycontribute to their superior maintenance ofcognitive function during exposure tohypoglycaemia. On the other hand, the paracrineRAS, which is widely distributed within the brain,may influence the cerebral metabolism ofsubstrates.38 Thus, knockout of the Ang IIsubtype 1 receptor gene is associated withconservation of ATP in neurons exposed toischaemia suggesting either reduced or moreefficient metabolism.33

In addition to suggesting a mechanism for theassociation between high ACE activity and risk ofsevere hypoglycaemia in type 1 diabetes, theresult of this study may contribute to theexplanation of the association between the

Insertion allele of the ACE gene and superiorhuman endurance performance capacitysuggesting a beneficial effect of low ACEactivity.39 As our study shows that cognitivefunction declines significantly in healthy subjectswith high ACE activity exposed to a moderatedegree of substrate deficiency for a relativelyshort duration, ACE activity may be an importantdeterminant of development of fatigue duringendurance sports with exposure to severe andlong-lasting fuel deprivation. Thus, the poorlyunderstood phenomenon of central fatigue isregarded as an important limiting factor ofendurance performance and has been attributedto impairment of brain function due toexhaustion or fuel starvation.40 This phenomenonmay well be influenced by ACE activity.

Further studies are warranted to verify ourfindings including studies during more profoundhypoglycaemia and studies with type 1 diabeticsubjects. Moreover, studies of the effect oftreatment with ACE inhibitors and Ang IIreceptor antagonists (ARBs) on cognitivefunction during experimental hypoglycaemia areneeded to evaluate whether these agents mighthave a role in prevention of severe hypogly-caemia. In this context, older registry-basedstudies have reported a controversial associationbetween ACE-inhibitor treatment and severehypoglycaemia41 which was not present in acontrolled trial.42 A more recent cross-sectionalstudy, on the other hand, reported a reduced rateof severe hypoglycaemia in patients treated withACE-inhibitors or ARBs, compared to patientstreated with other antihypertensives.4

Acknowledgements Research nurses Pernille Banck, Tove Larsen andMargrethe Larsen, Endocrinology Section,Division of Internal Medicine I, Hillerød Hospital,and chief EEG technician Anne Bodil Petersenand the staff at Department of Neurophysiology,Rigshospitalet, are thanked for skilful technicalassistance during the experiments. Researchtechnician Kamma Velin and DMSc GeorgSöletormos, Department of Clinical Biochemistry,Hillerød Hospital, are thanked for performanceof plasma glucose measurements; technicianHanne Damm, Department of ClinicalBiochemistry, Herlev University Hospital, for ACErelated analyses; and technician Birgit Vingaard,Department of Clinical Physiology, GlostrupHospital, and Regitze Kraunsøe and Jeppe Bach,Department of Medical Physiology, University ofCopenhagen, for performance of otherbiochemical analyses.

The study was funded by grants from The EFSD/JDRF/Novo Nordisk Programme for Research inType 1 Diabetes, The Danish Medical ResearchCouncil, The Danish Diabetes Association, The

Paper

COPYRIG

HT JRA

AS LIM

ITED

REPR

ODUCTION PR

OHIBITE

D

Journal ofthe Renin-Angiotensin-AldosteroneSystem(Including other

Peptidergic Systems)

March 2008Volume 9Number 1

47

Research Forum of Eastern Denmark, TheFoundation of Harald Jensen and his wife, andThe Foundation of Olga Bryde Nielsen. UPB wassupported by a grant from Novo Nordisk A/S.

References 1. Cryer PE. Hypoglycaemia: The limiting factor in the

glycaemic management of Type I and Type II diabetes.

Diabetologia 2002;45:637-48.

2. Høi-Hansen T, Pedersen-Bjergaard U, Pramming S,

Thorsteinsson B. Risk factors of severe hypoglycemia and

occurrence of silent hypoglycemia in patients with type 1

diabetes assessed by continuous glucose monitoring.

Diabetes 2005;54(suppl 1): A153.

3. The Diabetes Control and Complications Trial Research

Group. Hypoglycemia in the Diabetes Control and

Complications Trial. Diabetes 1997;46:271-86.

4. Pedersen-Bjergaard U, Pramming S, Heller SR et al.

Severe hypoglycaemia in 1076 patients with type 1 diabetes:

influence of risk markers and selection. Diabet Metab Res Rev

2004;20:479-86.

5. Pedersen-Bjergaard U, Agerholm-Larsen B, Pramming S,

Hougaard P, Thorsteinsson B. Activity of angiotensin-

converting enzyme and risk of severe hypoglycaemia in type

1 diabetes mellitus. Lancet 2001;357:1248-53.

6. Pedersen-Bjergaard U, Agerholm-Larsen B, Pramming S,

Hougaard P, Thorsteinsson B. Prediction of severe

hypoglycaemia by angiotensin-converting enzyme activity

and genotype in type 1 diabetes. Diabetologia 2003;46:89-96.

7. Nordfeldt S, Samuelsson U. Serum ACE predicts severe

hypoglycemia in children and adolescents with type 1

diabetes. Diabetes Care 2003;26:274-8.

8. Zammitt NN, Geddes J, Warren RE, Marioni R, Ashby JP,

Frier BM. Serum angiotensin-converting enzyme activity and

frequency of severe hypoglycaemia in type 1 diabetes: does

a relationship exist? Diabet Med 2007;24:1449-54.

9. Agerholm-Larsen B, Nordestgaard BG, Steffensen R,

Sørensen TI, Jensen G, Tybjærg-Hansen A. ACE gene

polymorphism: ischaemic heart disease and longevity in

10,150 individuals. A case-referent and retrospective cohort

study based on the Copenhagen City Heart Study. Circulation

1997;95:2358-67.

10. Chang L, Ernst T, Speck O, Patel H, DeSilva M, Leonido-

Yee M, Miller EN. Perfusion MRI and computerized cognitive

test abnormalities in abstinent methamphetamine users.

Psychiatry Res 2002;114:65-79.

11. Hepburn DA, Deary IJ, MacLeod KM, Frier BM.

Structural equation modeling of symptoms, awareness and

fear of hypoglycemia, and personality in patients with insulin-

treated diabetes. Diabetes Care 1994;17:1273-80.

12. Holst JJ. Molecular heterogeneity of glucagon in normal

subjects and in patients with glucagon producing tumours.

Diabetologia 1983;24:359-65.

13. Appel E, Bayer P, Hajdu P, Palm D, Shofer J, Uihlein M.

Determination of plasma catecholamines by means of

radioenzymatic labeling and high pressure liquid chro-

matographic separation. Naunyn Schmiedebergs Arch

Pharmacol 1981;315:233-9.

14. Klaniecki TS, Corder CN, McDonald RH Jr, Feldman JA.

High-performance liquid chromatographic radioenzymatic assay

for plasma catecholamines. J Lab Clin Med 1977;90:604-12.

15. Schwartz TW, Holst JJ, Fahrenkrug J et al. cholinergic

regulation of pancreatic polypeptide secretion. J Clin Invest

1978;61:781-9.

16. Poulsen K, Jørgensen J. An easy radioimmunological

microassay of renin activity, concentration and substrate in

human and animal plasma and tissues based on angiotensin I

trapping by antibody. J Clin Endocrinol Metab 1974;39:816-25.

17. Millar JA, Leckie BJ, Morton JJ, Jordan J, Tree M. A micro

assay for active and total renin concentration in human

plasma based on antibody trapping. Clinica Chimica Acta

1980;101:5-15.

18. Kappelgaard AM, Nielsen MD, Giese J. Measurement of

angiotensin II in human plasma: Technical modifications and

practical experience. Clinica Chimica Acta 1976;67:299-306.

19. Bulsara MK, Holman CDJ, van Bockxmeer FM, et al. The

relationship between ACE genotype and risk of severe

hypoglycaemia in a large population-based cohort of children

and adolescents with type 1 diabetes. Diabetologia.

2007;50:965-71.

20. Pedersen-Bjergaard U, Dhamrait SS, Sethi AA et al.

Genetic variation and activity of the renin-angiotensin system

and severe hypoglycemia in type 1 Diabetes. Am J Med

2008;121:246E1-E8.

21. DeRosa MA, Cryer PE. Hypoglycemia and the

sympathoadrenal system: neurogenic symptoms are largely

the result of sympathetic neural, rather than adrenomedullary,

activation. Am J Physiol 2004;287:E32-E41.

22. Porcellati F, Pampanelli S, Rossetti P, Cordoni C, Marzotti

S, Scionti L, Bolli GB, Fanelli CG. Counterregulatory hormone

and symptom responses to insulin-induced hypoglycemia in

the postprandial state in humans. Diabetes 2003;52:2774-83.

23. Heller SR, Amiel SA, Evans ML, Kong MF, Macdonald IA,

Peacey SR. Does insulin lispro preserve the physiological

defences to hypoglycaemia during intensive insulin therapy

with a conventional basal bolus regimen? Diabetes Obes

Metab 2002;4:106-12.

24. Sandoval DA, Guy DL, Richardson MA, Ertl AC, Davis

SN. Effects of low and moderate antecedent exercise on

counterregulatory responses to subsequent hypoglycaemia in

type 1 diabetes. Diabetes 2004;53:1798-106.

25. Frier BM, Al-Dujaili EA, Corrall RJ, Pritchard JL, Edwards

CR. Autonomic neural control mechanisms and the release of

adrenal steroids after hypoglycaemia in man. Horm Metab

Res 1984;16(suppl 1):138-41.

26. Amiel SA. Cognitive function testing in studies of acute

hypoglycaemia: rights and wrongs? Diabetologia

1998;41:713-19.

27. Strachan MW, Ewing FM, Frier BM, McCrimmon RJ,

Deary IJ. Effects of acute hypoglycaemia on auditory

information processing in adults with Type I diabetes.

Diabetologia 2003;46:97-105.

28. Madsen BK, Hølmer P, Ibsen H, Christensen NJ. The

influence of captopril on the epinephrine response to insulin-

induced hypoglycemia in humans. The interaction between

the renin-angiotensin system and the sympathetic nervous

system. Am J Hypertens 1992;5:361-5.

29. Glorioso N, Dessi-Fulgheri P, Alagna S et al. Angiotensin-

converting enzyme inhibition reduces ACTH release due to

hypoglycaemia. Clin Exp Hypertens A 1987;9:665-70.

30. Oltmanns KM, Deininger E, Wellhoener P et al. Influence

of captopril on symptomatic and hormonal responses to

hypoglycaemia in humans. B J Clin Pharmacol 2003:55:347-53.

31. Deininger E, Oltmanns KM, Wellhoener P et al. Losartan

Paper

COPYRIG

HT JRA

AS LIM

ITED

REPR

ODUCTION PR

OHIBITE

D

Journal ofthe Renin-Angiotensin-AldosteroneSystem(Including other

Peptidergic Systems)

March 2008Volume 9Number 1

48

attenuates symptomatic and hor-monal responses to

hypoglycemia in humans. Clin Pharmacol Ther 2001;70:362-9.

32. Hirsch IB, Heller SR, Cryer PE. Increased symptoms of

hypoglycaemia in the standing position in insulin-dependent

diabetes mellitus. Clin Sci 1991;80:583-6.

33. Walther T, Olah L, Harms C et al. Ischemic injury in

experimental stroke depends on angiotensin II. FASEB J

2002;16:169-76.

34. Georgiev V, Opitz M. Participation of angiotensin

receptors in acute hypoxia in mice. II. Effects of angiotensin

II nonpeptide receptor ligands losartan (DuP-753) and PD-

123319. Methods Find Exp Clin Pharmacol 1999;21:463-6.

35. Tallroth G, Ryding E, Agardh CD. Regional cerebral

blood flow in normal man during insulin-induced

hypoglycemia and in the recovery period following glucose

infusion. Metabolism 1992;41:717-21.

36. Thomas M, Sherwin RS, Murphy J, Kerr D. Importance of

cerebral blood flow to the recognition of and physiological

responses to hypoglycemia. Diabetes 1997;46:829-33.

37. Maran A, Cranston I, Lomas J, Macdonald I, Amiel SA.

Protection by lactate of cerebral function during

hypoglycaemia. Lancet 1994;343:16-20.

38. Dzau VJ. Tissue renin-angiotensin system: physiologic

and pharmacologic implications. Circulation 1988;77:I1-I3.

39. Montgomery HE, Marshall R, Hemingway H et al. Human

gene for physical performance. Nature 1998;393:221-2.

40. Nybo L, Møller K, Pedersen BK, Nielsen B, Secher NH.

Association between fatigue and failure to preserve cerebral

energy turnover during prolonged exercise. Acta Physiol

Scand 2003;179:67-74.

41. Morris AD, Boyle DIR, McMahon AD et al. ACE inhibitor

use is associated with hospitalisation for severe

hypoglycaemia in patients with diabetes. Diabetes Care

1997;20:1363-7.

42. The EUCLID study group. Randomised placebo-

controlled trial of lisinopril in normotensive patients with

insulin-dependent diabetes and normoalbuminuria or micro-

albuminuria. Lancet 1997;349:1787-92.

COPYRIG

HT JRA

AS LIM

ITED

REPR

ODUCTION PR

OHIBITE

D