Subcutaneous Continuous Glucose Monitoring: Feasibility of a new microdialysis-based glucose sensor...

OBSERVATIONS

Glucose Instability IsAssociated With aHigh Level ofCirculating P-Selectin

W e have previously shown thatglycemic instability, as measuredby the coefficient of variation for

fasting plasma glucose (CV-FPG), is anindependent predictor of cardiovascularmortality in type 2 diabetes (1,2). Themechanisms, if any, underlying the asso-ciation between long-term plasma glu-cose instability and vascular diseases aredifficult to explain.

In vitro studies with retinal capillarypericytes have shown that rapid glucosefluctuations in the culture medium in-duced cellular damage and death by apo-ptosis (3). Moreover, tubulointerstitialcells exposed to intermittent high glucoseconcentrations underwent changes in cel-lular growth, collagen synthesis, and cy-tokine secretion that were more severethan those observed in cells exposed tostable high-glucose concentrations (4).These data extrapolated in vivo are con-sistent with the notion that rapid excur-sions in glycemia may have relevantpathological effects. Recently, it has beendemonstrated that some adhesion mole-cules (P-selectin in particular) and othermarkers of inflammation are importantrisk factors for cardiovascular diseases(5–7).

Therefore, the aim of the presentstudy was to evaluate whether long-termglucose instability is associated with in-creased plasma concentrations of solubleadhesion molecules in type 2 diabetic pa-tients. To test this hypothesis, type 2 dia-betic patients attending the VeronaDiabetes Center with at least three fastingplasma glucose (FPG) determinations peryear over a 6-year period (1995–2000)were identified, and the CV-FPG was cal-culated. A total of 11 patients with unsta-ble FPG (CV-FPG $25%) were selected(group A). To reduce the effects of con-founders known to increase levels of ad-hesion molecules, participants had to benonsmokers with no clinical evidence ofcardiovascular diseases (assessed byphysical examination, electrocardiogram,

ankle-arm index, and intima-mediathickness of the carotid artery). As thecontrol, a group of 14 subjects with stableFPG (CV-FPG #10%), well-matched forsex, age, duration of diabetes, BMI, totalcholesterol, HDL-cholesterol, triglycer-ides, and blood pressure, were selected(group B). The cutoff of CV-FPG was cho-sen based on the results of our previousstudy: a CV-FPG ,15% was associatedwith the lowest cardiovascular mortalityrate, whereas a CV-FPG .25% predictedthe highest cardiovascular mortality rate(1,2).

Subjects with glucose instability hadstatistically significant higher mean FPG(183 6 43 vs. 157 6 17 mg/dl, P 5 0.01)and were more frequently treated with in-sulin. Circulating levels of E-selectin, P-selectin, intracellular adhesion molecule(ICAM)-1, and vascular cell adhesionmolecule (VCAM)-1 were measured in allpatients between November 2000 andDecember 2000 by enzyme-linked im-munoassay (Bender Medsystems), withintra- and interassay CVs ,8%.

In the covariance analysis, after ad-justing for mean FPG, P-selectin valueswere significantly higher in group A(371 6 102 vs. 284 6 66.7 ng/ml, P 50.03). The other adhesion moleculestended to be higher in group A, but noneof them reached statistical significance(ICAM-1 346 6 102 vs. 299 6 58.9 ng/ml, P 5 0.07; VCAM-1 466 6 175 vs.381 6 153 ng/ml, P 5 0.12; E-selectin62 6 31 vs. 52 6 30.3 ng/ml, P 5 0.8). Inconclusion, in this small selected sampleof diabetic subjects, high FPG instabilityseems to be associated with an endothelialdysfunction that may increase the risk ofcardiovascular morbidity and mortality.The present results must be confirmed ina large prospective study.

ELISABETIA BRUN, MD

GIACOMO ZOPPINI, MD

CRISTINA ZAMBONI, MD

ENZO BONORA, MD

MICHELE MUGGEO, MD

From the Division of Endocrinology and MetabolicDiseases, University of Verona, Verona, Italy.

Address correspondence and reprint requests toProf. Michele Muggeo, Divisione di Endocrinologiae Malattie del Metabolismo, Ospedale Civile Mag-giore Borgo Trento, P.le Stefani 1, 37126 Verona,Italy. E-mail: [email protected].

● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●

References1. Muggeo M, Verlato G, Bonora E, Zoppini

G, Corbellini M, de Marco R: Long-term

instability of fasting plasma glucose, anovel predictor of cardiovascular mortal-ity in elderly patients with non-insulin-dependent diabetes mellitus: the VeronaDiabetes Study. Circulation 961:1750–1754, 1997

2. Muggeo M, Zoppini G, Bonora E, Brun E,Bonadonna RC, Moghetti P, Verlato G:Fasting plasma glucose variability pre-dicts 10-year survival of type 2 diabeticpatients: the Verona Diabetes Study. Dia-betes Care 23:45–50, 2000

3. Li W, Liu X, Yanoff M, Cohen S, Ye X:Cultured retinal capillary pericytes die byapoptosis after fluctuation from high tolow glucose levels: a comparative studywith retinal capillary endothelial cells.Diabetologia 39:537–547, 1996

4. Jones SC, Saunders HJ, Qi W, Pollock CA:Intermittent high glucose enhances cellgrowth and collagen synthesis in culturedhuman tubulointerstitial cells. Diabetolo-gia 42:1113–1119, 1999

5. Frishman WH: Biologic markers as pre-dictors of cardiovascular disease. Am JMed 104:18S–27S, 1998

6. Oparil S, Oberman A: Nontraditional car-diovascular risk factors. Am J Med Sci 317:193–207, 1999

7. Ridker PM, Buring JE, Rifai N: Soluble P-Selectin and the risk of future cardiovas-cular events. Circulation 103: 491–495,2001

MaternalTransmission of Type2 Diabetes Varies byEthnic Group

Cross-sectional survey ofEuropeans and South Asians

P eople with ancestral origins in theIndian subcontinent who migrate toindustrialized countries are at high

risk of type 2 diabetes. The relative con-tribution of genetic and environmentalfactors to this difference is not completelyunderstood.

European origin populations withtype 2 diabetes are substantially morelikely to report diabetes in their mothersthan in their fathers (1). This maternalexcess has not been found in African-Americans (1) or Hispanics (2). Clinic-based reports have found no maternalexcess in South Indians with type 2 dia-betes (3), but we are not aware of anypopulation-based data to confirm this.

L E T T E R S

DIABETES CARE, VOLUME 24, NUMBER 9, SEPTEMBER 2001 1685

We used data from the Newcastle HeartProject (4) to look for an excess of mater-nal diabetes in U.K. South Asians.

We studied a stratified populationsample of 1,509 adults aged 25–74 years,of whom 684 were of South Asian origin(259 Indian, 305 Pakistani, and 120 Ban-gladeshi), and 825 were of European ori-gin. All subjects underwent an oralglucose tolerance test, and we defined di-abetes and impaired glucose tolerance us-ing the 1985 World Health Organizationcriteria. A family history of diabetes wasbased on the report of respondents.

Among Europeans, the proportion re-porting maternal diabetes was higher inthose with type 2 diabetes (Table 1).There was a significant excess of maternaldiabetes in respondents with normal glu-cose tolerance and diabetes. By contrast,the proportion of South Asians reportingmaternal diabetes was similar across cat-egories of glucose tolerance. There was asignificant excess of maternal diabetesamong South Asians with normal glucosetolerance but not among those with dia-betes. Among respondents with diabetes,mother-only diabetes was significantlymore common in Europeans than inSouth Asians (P 5 0.003), whereas father-only diabetes was not significantly differ-ent (P 5 0.37).

Possible explanations for excess ma-ternal diabetes include mitochondrial in-heritance, maternal imprinting, or adetrimental effect of diabetes in pregnancy.Alternatively, the finding may reflect in-accurate reporting, better awareness ofthe health of mothers, longer female lifeexpectancy, or a greater influence ofmothers on lifestyle risk factors. The factthat the maternal excess varied across cat-

egories of glucose tolerance for Europeansmay make these explanations less likely.The most serious limitation of our study(in common with most other such stud-ies) is its reliance on respondents’ reportsof family histories. Uncertainty aboutfamily history is reflected in the largenumbers in the missing or “don’t know”category, particularly among Europeans.Nevertheless, our findings in Europeansare consistent with previous clinic- andpopulation-based reports. Our results inSouth Asians, which were derived from apopulation-based study, add weight toprevious reports that maternal excess isnot a feature of type 2 diabetes in theseethnic groups.

Any proposed explanation of the im-balances between maternal and paternaldiabetes must take account of the ethnicdifferences we have described. Such anexplanation will likely make an importantcontribution to understanding the highprevalence of type 2 diabetes in SouthAsians in the U.K. and elsewhere. Weneed further studies based on biochemi-cal information about family members ofpeople with type 2 diabetes.

COLIN M. FISCHBACHER, MFPHM1

RAJ BHOPAL, MD2

NIGEL UNWIN, MFPHM1,3

MARK WALKER, FRCP3

MARTIN WHITE, FFPHM1

K.G.M.M. ALBERTI, FRCP3

From the 1Department of Epidemiology and PublicHealth, University of Newcastle, Newcastle uponTyne; the 2Department of Public Health Sciences,University of Edinburgh, Edinburgh, Scotland; andthe 3Department of Diabetes, University of New-castle, United Kingdom.

Address correspondence and reprint requests toDr. Colin M. Fischbacher, Lecturer in Public Health

Medicine, Department of Epidemiology and PublicHealth, The Medical School, University of New-castle, Newcastle upon Tyne, NE2 4HH, U.K. E-mail: [email protected].

Acknowledgments— We acknowledge fi-nancial support from the Barclay Trust, theBritish Diabetic Association, the NewcastleHealth Authority, the research and develop-ment directorate of the former Northern Re-gional Health Authority, the U.K. Departmentof Health, and the British Heart Foundation.

The authors thank all those mentioned inthe acknowledgments of their previous study(4) and Louise Hayes for help with data man-agement.

Material in this letter was previously pre-sented at the Society for Social Medicine’s 44thAnnual Scientific Meeting in Norwich, U.K., inSeptember 2000.

● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●

References1. Karter AJ, Rowell SE, Ackerson LM,

Mitchell BD, Ferrara A, Selby JV, NewmanB: Excess maternal transmission of type 2diabetes: the Northern California KaiserPermanente Diabetes Registry. DiabetesCare 22:938–943, 1999

2. Mitchell B, Kammerer C, Reinhart L, SternM, MacCluer J: Is there an excess in ma-ternal transmission of NIDDM? Diabetolo-gia 38:314–317, 1995

3. McCarthy M, Hitman G, Shields D, Mor-ton N, Snehalatha C, Mohan V, Ram-achandran A, Viswanathan M: Familystudies of non-insulin-dependent diabe-tes mellitus in South Indians. Diabetologia37:1221–1230, 1994

4. Bhopal R, Unwin N, White M, Yallop J,Walker L, Alberti K, Harland J, Patel S,Ahmad N, Turner C, Watson W, KulkarniA, Laker M, Tavridou A: Heterogeneityof coronary heart disease risk factors inIndian, Pakistani, Bangladeshi and Euro-pean origin populations: cross sectionalstudy. BMJ 319:215–220, 1999

Epidemiology ofNontraumaticLower-ExtremityAmputation in Area7, Madrid, Between1989 and 1999A population-based study

D iabetes remains the main cause oflower-extremity amputations (LEAs)(1–3). Recently (3), the number of

LEAs carried out worldwide has been es-timated to be .162 3 106 each year, with

Table 1—Family history of maternal and paternal diabetes by ethnic group and respondentglucose tolerance status

NormalImpaired glucose

tolerance Diabetes

Europeansn (% don’t know/missing) 649 (27) 129 (29) 46 (33)Maternal diabetes only 6.3 (4.3, 8.9) 8.7 (3.8, 16.4) 32.3 (16.7, 51.4)Paternal diabetes only 2.1 (1.0, 3.8) 3.3 (0.7, 9.2) 3.2 (0.1, 16.7)Difference 4.2 (1.5, 6.3) 5.4 (22.6, 10.5) 29.0 (6.3, 35.3)

South Asiansn (% don’t know/missing) 386 (14) 140 (14) 158 (18)Maternal diabetes only 12.1 (8.8, 16.1) 14.2 (8.5, 21.7) 10.9 (6.1, 17.5)Paternal diabetes only 8.5 (5.7, 12.0) 15.0 (9.1, 22.7) 7.8 (3.8, 13.8)Difference 3.6 (1.5, 8.5) 20.8 (210.9, 9.3) 3.1 (25.0, 10.4)

Data are n (%) or % (95% CI).

Letters

1686 DIABETES CARE, VOLUME 24, NUMBER 9, SEPTEMBER 2001

a high economic and social cost. Morethan 50% of these LEAs are performed insubjects with diabetes. Differences in theincidence of LEAs could reflect variationsin risk factors, diabetes prevalence, healthcare systems, and lifestyle, as well as dif-ferences in population age-structure, thesources from which the case was identi-fied, the levels of ascertainment, and thedefinitions of LEAs and the distinctionsbetween major and minor LEAs. Recent-ly, a standard approach to data collectionhas been described (4,5), and the first ar-ticle (3) about the epidemiology of LEAsin several countries has been published.In this study, area 7 from Madrid presentsthe lowest LEA rates. The purpose of thisarticle is to compare the incidence of LEAsin diabetic and nondiabetic subjects—inrelation to the provision of improved foot-care after the St. Vincent Declaration—byuse of a standardized approach, in thearea 7 of Madrid between 1989 and 1993,between 1994 and 1996, and between1997 and 1999.

The National Health Service has di-vided Madrid into 13 health care areas.The public health care system is serving.99% of the total population from thearea 7 and is provided by a single hospital(Hospital Clınico San Carlos) with a sin-gle specialized service for vascular pa-tients. According to the census in 1991,the total population for the catchmentarea was 569,307 (261,529 men and307,778 women). The population withdiabetes was calculated according to theLejona study (6). In this study, the prev-alence of diabetes was estimated by a ran-domized sample stratified by 5-year age-groups and sex. The data of this studywere used to calculate the number of peo-ple with diabetes according to age and sexdistribution, bearing in mind that thepopulation composition in both Lejona(Vizcaya) and Madrid are comparable.The total diabetic population estimated(diagnosed and unknown) was 37,932(15,505 men and 22,427 women). From1991 to 1996, the date of the last census,the catchment area was reduced by;4,000 inhabitants, which was takeninto account in order to calculate inci-dence rates. According to the GlobalLower Extremity Amputation Study (3),amputation was defined as the completeloss of any part of the lower limb. Theterm “minor amputation” refers to an am-putation distal to the tarsometatarsaljoint, whereas “major amputation” refers

to one through or proximal to the tarso-metatarsal joint. All patients who under-went an LEA between 1 January 1989 and31 December 1999 were identifiedthrough operating theater records. Vascu-lar surgery department and endocrinol-ogy service discharge records were usedas secondary sources. Only patients whohad resided in area 7 for at least the last 6months were included in the study. Thesocial security system in our country cov-ers between 70 and 100% of the cost ofpharmacological treatments when pre-scribed by a social security doctor. Pa-tients suffering from LEAs usually havesome pharmacological treatment. Thus,prescribing physicians (family doctorsand physicians of the endocrinology ser-vice in an outpatient setting) were used asan additional source to identify peoplewho were amputated outside area 7. Thissource was operative since 1994. Date ofbirth, sex, address, date and level of theactual amputation, and potential causesof the actual amputation were registered.Estimations of the level of ascertainmentwere derived by using capture-recapturemethods (5). Because three independentsources were available from 1994, the lev-el of ascertainment was calculated duringthe 1994–1999 period. To have a greaternumber of cases, men and women wereanalyzed together.

Incidence of LEA data are presentedover three periods. From 1989 to 1993 nointervention was done. Since 1994, a se-ries of improvements in therapeutic mea-sures in diabetes management areavailable in our area, including a prophy-lactic foot care teaching and treatmentprogram (7). As the efficacy of these mea-sures have improved across time, we havestudied the incidence of LEAs during two3-year periods (1994–1996 and 1997–1999). From 1989 to 1993, 139 diabeticsubjects and 118 nondiabetic subjectsunderwent 156 and 139 nontraumaticLEAs, respectively, while from 1994 to1999, 91 diabetic and 43 nondiabeticsubjects suffered 111 and 49 nontrau-matic LEAs, respectively. Age-adjustedincidence per 105 people in risk per yearfor first LEAs (95% CI) decreased from1989 –1993 to 1997–1999 in diabeticpeople as follows: major LEAs decreasedfrom 67.1 (60.9–73.3) to 12.3 (14.1–10.5) and from 13.3 (11.6–15.0) to 5.6(4.9–6.3) for men and women, respec-tively; minor LEAs decreased from 52.1(45.0 –59.2) to 22.5 (19.7–25.3) and

from 10.9 (10.3–11.5) to 7.9 (6.8–9.0)for men and women, respectively. Datafor nondiabetic people are as follows: ma-jor LEAs decreased from 2.6 (2.2–3.0) to1.1 (0.4–1.8) and from 1.3 (1.1–1.5) to0.5 (0.4–0.6) in men and women, re-spectively; minor LEAs decreased from2.7 (2.1–3.3) to 0.7 (0.6–0.8) and from1.5 (1.3–1-7) to 0 for men and women,respectively. Crude and age-adjusted in-cidence per 105 people in risk per year forfirst and all (major and minor) LEAs forthe three periods are displayed in Table 1.

The incidence of LEAs rose steeplywith age .40 years and it rose dramati-cally with age .80 years; it was higher inmen than in women. The incidence of ma-jor LEAs was greater than that of minorLEAs in nondiabetic people, whereas dia-betic subjects more frequently sufferedminor LEAs.

For major LEAs, estimations of ascer-tainment were 100 and 98.7%, and forminor LEAs, they were 83.8 and 76.9%.Given that more than one condition waspresent per patient, LEAs in diabetic sub-jects were associated with peripheral vas-cular disease (PVD) (major/minor: 100/62%), neuropathy (major/minor: 78/92%),and infection (major/minor: 24/84%),whereas LEAs in nondiabetic subjectswere associated with PVD (major/minor:100/98%), neuropathy (major/minor: 22/33%), and infection (major/minor:6/18%). According to the St. Vincent Dec-laration, several studies in recent yearshave found a decrease in the incidence ofLEAs in people with diabetes.

However, only a few studies (8–10)have analyzed a 10-year period, as thisone has. Furthermore, to our knowledge,no other study has used the capture-recapture method to estimate the level ofascertainment from more than two sepa-rate data sources and to measure the inci-dence of first and all LEAs in a separateway in the whole population (i.e., thosewith and without diabetes). According toour data, LEA incidence in area 7 of Ma-drid remains the lowest reported inci-dence in a Caucasian population of bothdiabetic and nondiabetic people. Primaryprevention is defined as a reduction in theincidence of a disease, e.g., first LEA.Since 1994, following the provision of im-proved foot care after the diabetes pro-gram was available (7), a reduction in firstLEA of 57 and 81% (minor and major,respectively) in diabetic men and of 28

Letters


and 57% were detected, with a 5- to6-year increase in age at the first LEA.

These data indicate that efforts to de-lay and reduce the incidence of first LEAin people with diabetes succeeded. In asimilar way, a slightly higher reduction infirst LEAs was found in the nondiabeticpopulation also at the end of the study.Several factors may contribute to thesedata. The decline in LEA rates can be par-tially explained by favorable trends insome risk factors and a better control inothers, including dietary factors as well asa tighter control of hypertension, hyper-cholesterolemia, and reduced tobaccoand alcohol consumption (11). Bettercontrol of these risk factors should po-sitively affect the entire population. Inaddition, the increasing availability of re-sources for the usual clinical managementof these patients, including new drugsand surgical procedures (12) covered bythe social security system, should also beoperative.

This study also evaluated the condi-tions associated with LEAs. The propor-tion of major LEAs associated with PVDwas high and similar in both diabetic andnondiabetic population. The proportionof minor LEAs associated with neuropa-thy and infection in diabetic subjects wasclearly greater than that in the nondia-betic population. This could explain theexcess of minor LEAs in diabetic people,suggesting that we should try to cope withthese problems.

In conclusion, our data show that theincidence of LEAs in area 7 of Madrid re-mains as the lowest incidence reported inEuropean countries in both diabetic andnondiabetic people. A substantial de-crease in LEAs and a later presentation inrelation to a series of improvements in di-abetic treatment were detected. Despitethese figures, the incidence of LEAs re-mains higher in diabetic subjects than inthe nondiabetic population, suggestingthat diabetic foot care remains suboptimalin Madrid. Taking into account that dia-betic subjects suffer minor LEAs more fre-quently than nondiabetic people and thatit is associated with neuropathy and infec-tion, a more substantial reduction in LEAsin diabetic people should be achievedwith an earlier neuropathy diagnosis andan adequate antibiotic management.

ALFONSO L. CALLE-PASCUAL, MD1

NURIA GARCIA-TORRE, MD1

INMACULADA MORAGA, MD1

JOSE A. DIAZ, MD1

ALEJANDRA DURAN, MD1

GUILLERMO MONUX, MD2

FRANCISCO J. SERRANO, MD2

PEDRO J. MARTıN-ALVAREZ, PHD3

ANICETO CHARRO, MD1

JUAN P. MARANES, MD1

From the 1Department of Endocrinology Metabo-lism and Nutrition, Hospital Clınico San Carlos, Ma-drid; the 2Department of Vascular Surgery, HospitalClınico San Carlo, Madrid; and 3Juan de la Cierva

College, Consejo Superior Investigaciones Cientifi-cas, Madrid, Spain.

Address correspondence to A.L. Calle-Pascual,MD, Department of Endocrinology Metabolismand Nutrition 1aS. Hospital, Clınico San Carlos,C/Martın Lagos s/n, E-28040 Madrid, Spain. E-mail:[email protected].

● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●

References1. Amos AF, McCarty DJ, Zimmet P: The ris-

ing global burden of diabetes and its com-plications: estimates and projections tothe year 2010. Diabet Med 14 (Suppl. 5):S7–S85, 1997

2. Mayfield JA, Reiber GE, Sander LJ, JanisseD, Pogach LM: Preventive foot care inpeople with diabetes. Diabetes Care 21:2161–2177, 1998

3. The Global Lower Extremity AmputationStudy Group: Epidemiology of lower ex-tremity amputation in centres in Europe,North America and East Asia. Br J Surg87:328–337, 2000

4. The LEA Study Group: Comparing the in-cidence of lower extremity amputationsacross the world: the global lower extrem-ity amputations study. Diabet Med 12:14–18, 1995

5. International Working Group for DiseaseMonitoring and Forecasting: Capture-re-capture and multiple-record systems esti-mation: applications in human diseases.Am J Epidemiol 142:1047–1058, 1995

6. Bayo J, Sola C, Garcia F, Latorre PM,Vazquez JA: Prevalencia de la diabetesmellitus no dependiente de la insulina enLejona (Vizcaya). Med Clin 101:609–612,1993

7. Ebskov LB, Schroeder TV, Holstein P: Ep-

Table 1—Crude and age-adjusted incidence (number per 105 people in risk per year) of first and all LEAs

1989–1993 1994–1996 1997–1999Decrease

(%)

First All First All First All First All

Crude A-A Crude A-A Crude A-A Crude A-A Crude A-A Crude A-A

Diabetic subjectsMajor

Men 69.5 67.1 72.5 70.6 38.2 36.9 42.5 41.4 12.7 12.3 12.7 12.4 81.7 82.5Women 21.3 13.3 24.5 15.3 12.6 7.9 14.4 9.0 9.0 5.6 9.0 5.6 57.7 63.3

MinorMen 54.3 52.1 60.6 58.9 40.4 38.8 59.9 57.8 23.4 22.5 34.0 33.1 56.9 43.9Women 17.5 10.9 19.0 11.9 14.4 9.0 18.0 11.3 12.6 7.9 18.0 11.3 28.1 5.3

Nondiabetic subjectsMajor

Men 2.7 2.6 2.8 2.7 1.1 1.1 1.1 1.1 1.1 1.1 1.3 1.3 59.2 48.1Women 2.0 1.3 2.4 1.5 0.6 0.4 0.6 0.4 0.8 0.5 0.9 0.6 60.0 62.5

MinorMen 1.1 1.1 1.4 1.4 0.5 0.5 0.5 0.5 0.7 0.7 0.7 0.7 36.4 50.0Women 0.8 0.5 0.9 0.6 0.1 0.1 0.1 0.1 0 0 0.1 0.1 100 88.9

A-A, age-adjusted.

Letters


idemiology of leg amputations: the influ-ence of vascular surgery. Br J Surg 81:1600–1604, 1994

8. Ebskov B, Ebskov LB: Major lower limbamputations in diabetic patients: devel-opment during 1982–1993. Diabetologia39:1607–1610, 1996

9. Larsson J, Apelqvist J, Agardh CD, Sten-strom A: Decreasing incidence of majoramputation in diabetic patients—a conse-quence of a multidisciplinary foot careteam approach? Diabet Med 12:770–776,1995

10. Calle-Pascual, Duran A, Benedi A, CalvoMI, Charro A, Diaz JA, Calle JR, Gil E,Ibarra J, Maranes JP, Cabezas-Cerrato J:Reduction in foot ulcer incidence: relationto compliance with a prophylactic foot-care programme. Diabetes Care 24:405–407, 2001

11. The Diabetes and Nutrition Study Groupof the Spanish Diabetes Association: Dia-betes Nutrition and Complications Trial(DNCT): Food intake and targets of dia-betes treatment in a sample of Spanishpeople with diabetes. Diabetes Care 20:1078–1080, 1997

12. Calle-Pascual AL, Duran A, Diaz A,Monux G, Serrano FJ,: Garcia de la TorreN, Moraga I, Calle JR, Charro A, MaranesJP: Comparison of peripheral arterial re-construction in diabetic and nondiabeticpatients: a prospective clinic-based studyDiabetes Res Clin Pract 53:129–136, 2001

Increased Risk ofLower-ExtremityAmputation AmongCaucasian DiabeticPatients on Dialysis

In two recent issues of Diabetes Care,high amputation rates among Maoripeople from New Zealand (1) and Na-

tive Americans (2) with diabetes on dial-ysis were reported. We would like to addsome information concerning Caucasianswith diabetic foot syndrome and ad-vanced renal disease from a prospectivesample of 400 consecutive patientstreated at our diabetic foot clinic. Similarto chronic renal insufficiency (defined byserum creatinine concentration 1.5–3.0mg/dl), chronic renal failure and end-stage renal disease (ESRD) should be re-garded as a continuum (3). We analyzedthe frequency of these conditions amongthe entire study population and com-pared presentation features and outcomes

among the concerned patients and amongdiabetic foot patients without obvious re-nal impairment.

At the time of presentation for the in-dex foot lesion, 14 patients (4%, group A)were on dialysis, 72 patients (18%, groupB) suffered from chronic renal insuffi-ciency (serum creatinine $1.5 mg/dl),and 314 patients (78%, group C) had nosigns of advanced renal impairment.

There were no significant differencesamong the groups regarding age, diabetesduration, or quality of metabolic control.The proportions of underlying neuropa-thy (86, 91, and 79%, respectively) andperipheral vascular disease (71, 56, and50%, respectively) were higher amongpatients with renal impairment but notsignificantly different from the propor-tions of patients with normal renal func-tion. However, the coincidence ofperipheral vascular disease (PVD) andneuropathy was far more common in di-abetic patients on dialysis than in patientswithout renal impairment (71% in groupA vs. 36% in group C, P 5 0.015); there-fore, patients on dialysis are at evenhigher risk for unrecognized nonhealingfoot lesions. Coexistence of PVD and me-diasclerosis was also far more frequent inthe dialysis group than in the other twogroups (80 vs. 26 and 34%, P 5 0.01 andP 5 0.003).

Total amputation rate was 57% for di-alysis patients and was therefore signifi-cantly higher than that among patientswith chronic renal insufficiency (25%,P 5 0.006) or patients without obviousrenal impairment (16%, P , 0.001). Thiswas primarily caused by a sixfold in-creased risk of major amputation of dial-ysis patients compared with patients withnormal renal function (29 vs. 4.5%, P ,0.001). One-year survival was pooramong diabetic patients on dialysis withfoot lesions compared with their counter-parts without renal impairment (71 vs.94%, P 5 0.007). However, a higher pro-portion of patients in the dialysis groupwere suitable for vascular interventioncompared with the other two groups (46vs. 14 and 13%, P 5 0.02). After success-ful revascularization, the risk of consecu-tive major amputation was similar in allthree groups (17, 20, and 16%, respec-tively). This last finding is in contrast toprevious publications (4 –7) that indi-cated an increased risk of amputationeven in the presence of patent grafts afterrevascularization among ESRD patients,

caused by nonhealing large foot ulcersand uncontrolled infection. The betteroutcome in our series of diabetic patientson dialysis with foot lesions might be ex-plained by the structured interdisciplinaryinpatient care after surgery, as well as thecontinued care by the diabetic foot clinicafter hospital discharge. This interpreta-tion is supported by the observations ofFoster et al. (8), who showed that care bya special diabetic foot clinic reduces epi-sodes of digital gangrene and major am-putation in diabetic renal transplantpatients.

In conclusion, Caucasian patientswith diabetes on dialysis are at high riskfor amputation once a foot lesion occurs.However, vascular interventions shouldnot be refused to those patients if struc-tured interdisciplinary inpatient care aftersurgery and long-term follow up by a spe-cial diabetic foot clinic are available. Pre-vention of foot complications amongdiabetic patients with ESRD and progno-sis improvement in the case of an estab-lished foot lesion should be targeted byintense collaboration of dialysis units andspecialized diabetic foot centers.

STEPHAN MORBACH, MD1

CHRISTOPH QUANTE, MD1

HERMANN RUDOLF OCHS, MD1

FRANZ GASCHLER, MD2

JEAN-MARIO PALLAST, MD3

UDO KNEVELS, MD3

From the Marienkrankenhaus Departments of 1In-ternal Medicine, Diabetic Foot Clinic, and 2VascularSurgery, Soest, Germany, and the 3Limited Care Di-alysis Unit, Soest, Germany.

Address correspondence to Stephan Morbach,MD, Department of Internal Medicine, Marienkran-kenhaus, Widumgasse 5, 59494 Soest, Germany.E-mail: [email protected].

● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●

References1. McGrath NM, Curran BA: Recent com-

mencement of dialysis is a risk factor forlower-extremity amputation in a high-risk diabetic population. Diabetes Care 23:432–433, 2000

2. Rith-Najarian S, Gohdes D: Preventingamputations among patients with diabe-tes on dialysis. Diabetes Care 23:1445–1446, 2000

3. Luke G: Chronic renal failure: a vasculo-pathic state. N Engl J Med 339: 841–843,1998

4. Edwards JM, Taylor LM, Porter JM: Limbsalvage in end-stage renal disease (ESRD).Arch Surg 123:1164–1168, 1988

5. Sanchez LA, Goldsmith J, Rivers SP, Pan-

Letters


etta TF, Wengerter KR, Veith FJ: Limb sal-vage surgery in end stage renal disease: isit worthwhile? J Cardiovasc Surg 33:344–348, 1992

6. Johnson BL, Glickman MH, Bennik DF,Esses GE: Failure of foot salvage in pa-tients with end-stage renal disease aftersurgical revascularization. J Vasc Surg 22:280–286, 1995

7. Leers SA, Reifsnyder T, Delmonte R, Ca-ron M: Realistic expectations for pedal by-pass grafts in patients with end-stage renaldisease. J Vasc Surg 28: 976–983: 1998

8. Foster AVM, Snowden S, Grenfell A,Watkins PJ, Edmonds ME: Reduction ofgangrene and amputations in diabetic re-nal transplant patients: the role of a spe-cial foot clinic. Diabet Med 12:632–635,1995

Acute Charcot’sArthropathy Despite11 Years ofNormoglycemiaAfter SuccessfulKidney and PancreasTransplantation

Charcot’s arthropathy is one of themost debilitating orthopedic se-quelae of diabetes. Regarding its

pathogenesis, the current theory suggestsa role of diabetic peripheral and auto-nomic neuropathy (1).

Kidney-pancreas transplantation isan accepted therapeutic approach in ure-mic type 1 diabetic patients, providingnormalization of glycometabolic controland improvement of secondary diabeticcomplications, including neuropathy(2– 4).

We report a case of a diabetic uremicpatient who developed acute Charcot’sarthropathy despite 11 years of tight met-abolic control after successful kidney andpancreas transplantation. A 41-year-oldtype 1 diabetic uremic male patient un-derwent kidney and pancreas transplan-tation. His immunosuppression regimenconsisted of steroids, cyclosporin, andazathioprine. During the entire follow-upperiod, the patient was insulin-free, ingood metabolic control (HbA1c 5.2 60.6%), and had good renal function (cre-atinine 1.3 6 0.1 mg/dl). At baseline,neurological examination and electromi-ography documented symmetrical sen-sory loss in the distal lower extremitiesand motor deficits mostly on the right

side, with footdrop. A test of autonomicfunction showed a near normal bloodpressure response to standing. Duringdeep breathing, the heart rate was com-promised, whereas the response to stand-ing was normal.

After transplantation, an improve-ment of motor and sensory nerve conduc-tion velocity was observed during thefollow-up period, and the median nervereached normal values at 5 years, whereassural and peroneal nerves remained se-verely impaired. Bone mineral density bydual-energy X-ray absorptiometry withHologic was evaluated regularly, and theresults showed stable osteopenia.

Eleven years after transplantation, thepatient showed a painless profound swell-ing of the right ankle joint with locallyincreased skin temperature, joint effu-sion, and erythema. X-ray, computerizedtomography, magnetic resonance, andTechnetium-99 bone scan showed an ero-sive area at the third distal segment of thetibia, fibula, and astragalus, with involve-ment of the surrounding soft tissue andjoint effusion. The patient was submittedto surgical tibiotalar fusion. During sur-gery, samples were collected for histolog-ical test and microbiological evaluation.Histological examination showed brokendetached osteocartilagineus connectionwithout evidence of phlogistic infiltra-tion, and cultures were negative.

This patient, despite long-term func-tioning kidney and pancreas transplanta-tion, developed Charcot’s arthropathy.The clinical diagnosis was confirmed byhistological analysis via bone and synovialbiopsy—negative cultures ruled out thehypothesis of an infective arthropathy.The development of Charcot’s arthropa-thy in this patient was probably multifac-torial. Among the factors that could haveinfluenced the development of this ar-tropathy, steroids did not seem to haveplayed a fundamental role because theywere administered for a short period aftertransplantation and because a progres-sion of osteopenia was not observed. Cy-closporin was described to induceperipheral neuropathy (5). This couldcounterbalance the positive effects of tightmetabolic control on diabetic neuropa-thy, thus leading to impairment of nervefunction. In fact, in this patient, diabeticneuropathy was partially reversed.

We can conclude that in patients al-ready affected by severe diabetic compli-cations, tight metabolic control achieved

with pancreas transplantation does notprevent the development of Charcot’s ar-thropathy.

ROSSANA CALDARA, MD1

CRISPINO GRISPIGNI, MD2

ENNIO LA ROCCA, MD1

PAOLA MAFFI, MD1

ELENA ORSENIGO, MD3

CARLO SOCCI, MD3

GIANFRANCO FRASCHINI, MD2

VALERIO DI CARLO, MD3

GUIDO POZZA, MD1

ANTONIO SECCHI, MD1

From the Departments of 1Internal Medicine, 2Or-thopedics, and 3General Surgery, San Raffaele Sci-entific Institute, University of Milan, Milan, Italy.

Address correspondence to Antonio Secchi, As-sociate Professor of Medicine, University of Milan,San Raffaele Scientific Institute, Via Olgettina 60,20132 Milan, Italy. E-mail: [email protected].

● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●

References1. Armstrong DG, Todd WF, Lavery LA,

Harkless LB, Bushman TR: The naturalhistory of acute Charcot’s arthropathy in adiabetic foot specialty clinic. Diabet Med14:357–367, 1997

2. Caldara R, La Rocca E, Maffi P, Secchi A:Effects of pancreas transplantation on latecomplications of diabetes and metaboliceffects of pancreas and islet transplanta-tion. J Pediatr Endocrinol Metab 12:777–787, 1999

3. Landgraf R: Impact of pancreas trans-plantation on diabetic complications andquality of life. Diabetologia 39:1415–1424,1996

4. Martinenghi S, Comi G, Galardi G, DiCarlo V, Pozza G, Secchi A: Ameliorationof nerve conduction velocity followingsimultaneous kidney/pancreas trans-plantation is due to the glycaemic controlprovided by the pancreas. Diabetologia 40:1110–1112, 1997

5. Gijtenbeek JM, van den Bent MJ, VechtCJ: Ciclosporine neurotoxicity: a review.J Neurol 246:339–346, 1999

Homogeneity ofMetabolic Control inNew South Walesand the AustralianCapital Territory,Australia

Socioeconomic disadvantage is asso-ciated with higher morbidity andpoorer health behavior, despite

higher rates of health service utilization(1). In a study of American children with

Letters


diabetes, lower socioeconomic status wasassociated with poorer glycemic controland more hypoglycemic episodes (2). Inan audit of 1,190 children with type 1diabetes aged ,15 years, we examinedwhether socioeconomic disadvantage orurban/rural location affect metabolic con-trol in Australia, which has universalhealth insurance and subsidies for diabe-tes supplies.

Case attainment, based on the Diabe-tes Register, was 67% of all eligible chil-dren living in New South Wales (NSW)and the Australian Capital Territory(ACT). There were 649 girls and 541 boysrecruited from three multidisciplinaryhospital-based pediatric diabetes units,four private city-based pediatric diabetespractices, and 18 rural diabetes outreachclinics. Social disadvantage was assessedusing a graded postcode-based social dis-advantage risk score (SDRS) derived fromnine risk indicators (3). HbA1c was as-sayed in a central laboratory using theBio-Rad Variant HPLC Capillary Collec-tion System (Bio-Rad Laboratories, Mu-nich, Germany). Level II LaboratoryCertification of Traceability to the Diabe-tes Control and Complications Trial(DCCT) reference method was obtainedfrom the University of Missouri Second-ary Reference Laboratory no. 1.

The median HbA1c was 8.2% (inter-quartile range 7.6–9.1%). The significantpredictors of HbA1c in a multiple regres-sion model were longer duration (b 50.04, P 5 0.0001) and insulin dose perkilogram (b 5 0.44, P 5 0.0001). Afterexcluding subjects with duration ,2years, a more complex model resulted,with a significant interaction between ageand duration: HbA1c 5 8.74 1 0.28 (in-sulin per kilogram) 2 0.15 (duration) 20.05 (age) 1 0.01 (duration 3 age).

Rural children lived in areas of lowersocial advantage (SDRS 20.4 6 0.6) thancity-based children cared for in either thepublic sector (0.4 6 0.7) or the privatesector (1.0 6 0.7) (P , 0.0001). Therewas no significant difference in the me-dian HbA1c levels among the 25 differentcenters (range 7.7–8.7%), and SDRS wasnot significant in the multiple regressionmodel. There was no difference in the met-abolic control of children living in rural lo-cations compared with those from urbanlocations (median HbA1c 8.2% in both).

The incidence of severe hypoglyce-mia (unconsciousness or seizures) was 36per 100 patient-years. HbA1c, urban/rural

location, or SDRS were not significantpredictors of hypoglycemia in a Poissonregression model. No significant differ-ences were found for either metaboliccontrol or rate of hypoglycemia betweenchildren cared for in the public or privatesystem, despite significantly greater socialadvantage in the latter group.

The median HbA1c of 8.2% in NSW/ACT is comparable with results from themultinational Hvidore study (meanHbA1c 8.6%, DCCT equivalent 8.3%) (4).In that study, there was significant varia-tion in metabolic control among 22 cen-ters from 18 countries. In contrast, theaudit from NSW/ACT shows remarkablehomogeneity in metabolic control amongcenters, regardless of location, socioeco-nomic status, or system of care.

PENNY HANDELSMAN, MPH1

MARIA E. CRAIG, FRACP1

KIM C. DONAGHUE, FRACP1

ALBERT CHAN, M APP STAT1

BARBARA BLADES, PHD1

ROSETTA LAINA, MSC1

DARNA BRADFORD, BSC (HONS)1

ANGELA MIDDLEHURST, RN1

GEOFFREY AMBLER, FRACP, MD1

CHARLES F. VERGE, FRACP, PHD2

PATRICIA CROCK, FRACP3

PATRICK MOORE, FRACP4

MARTIN SILINK, FRACP, MD1

FOR THE NSW/ACT HBA1C

STUDY GROUP

From the 1The Children’s Hospital at Westmead,Sydney, New South Wales; the 2Sydney Children’sHospital, Randwick, New South Wales; the 3JohnHunter Children’s Hospital, Newcastle, New SouthWales; and 4The Canberra Hospital, Canberra, Aus-tralia Capital Territory, Australia.

Address correspondence to Dr. Maria Craig, RayWilliams Institute of Paediatric Endocrinology, Di-abetes and Metabolism, The Children’s Hospital atWestmead, Locked Bag 4001, Westmead, NSW2145, Australia. E-mail: [email protected].

Acknowledgments— This study was sup-ported by an educational grant from NovoNordisk and Bio-Rad Australia.

● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●

References1. Turrell G, Mathers CD: Socioeconomic

status and health in Australia. Med J Aust172:434–438, 2000

2. Overstreet S, Holmes CS, Dunlap WP,Frentz J: Sociodemographic risk factors todisease control in children with diabetes.Diabet Med 14:153–157, 1997

3. Unequal in Life: The Distribution of Social

Disadvantage in Victoria and New SouthWales. Victoria, Australia, The IgnatiusCentre for Social Policy and Research,1999

4. Mortensen HB, Hougaard P: Comparisonof metabolic control in a cross-sectionalstudy of 2,873 children and adolescentswith IDDM from 18 countries: the Hvi-dore Study Group on Childhood Diabe-tes. Diabetes Care 20:714–720, 1997

Remnant-LikeParticle Cholesteroland InsulinResistance inNonobeseNonhypertensiveJapanese Glucose-Tolerant Relatives ofType 2 DiabeticPatients

P atients with type 2 diabetes and im-paired glucose tolerance (IGT) haveincreased incidence of coronary

heart disease (CHD) (1,2). The risk ofCHD appears to be similar in patientswith type 2 diabetes and IGT (3,4). Thus,factors other than the level of glycemiaseem to accelerate the development ofCHD in type 2 diabetes. This idea is sup-ported by the notion that the duration ofdiabetes and the level of glycemia are riskfactors for microvascular disease but notfor CHD in type 2 diabetic patients (5,6).Alternatively, it may be hypothesized thatthe relation between type 2 diabetes andCHD may not be causal and that theevents preceding the onset of diabetes(i.e., the prediabetic state) may contributeto CHD.

Whereas insulin resistance is consid-ered to be associated with CHD, Kugi-yama et al. (7) showed that fastingremnant lipoprotein levels predict coro-nary events in patients with CHD. Ai et al.(8) and our group (9) recently disclosedthat remnant lipoprotein is associatedwith insulin resistance in IGT subjectsand type 2 diabetic patients. Thus, rem-nant lipoprotein, in conjunction with in-sulin resistance, is considered one of theimportant risk factors for the develop-ment of CHD in type 2 diabetes.

Although glucose-tolerant relatives oftype 2 diabetic patients are considered to

Letters


be insulin resistant (10–12), it is unclearwhether remnant lipoprotein is increasedin glucose-tolerant relatives of type 2 dia-betic patients. Only one study containsdata on the lipid profile in nondiabeticrelatives of type 2 diabetic patients. Lawset al. (13) documented that nondiabeticrelatives of patients with type 2 diabeteshad insulin resistance and hypertriglycer-idemia. However, they did not study rem-nant lipoprotein, and the patients studiedwere all obese. It is well known that thedegree of being overweight per se affectsinsulin resistance and lipid profile (14). Inaddition, hypertension itself is associat-ed with insulin resistance and lipid abnor-malities in humans (15). We thereforerecruited nonobese nonhypertensive glu-cose-tolerant relatives of type 2 diabeticpatients who were carefully matched forBMI, blood pressure, and fasting glucoseto glucose-tolerant subjects without anyfamily history of type 2 diabetes. This isthe first report showing that nonobesenonhypertensive Japanese glucose-tolerantrelatives of type 2 diabetic patients havealready high remnant lipoprotein and di-minished insulin sensitivity before theonset of diabetes.

A total of 45 healthy glucose-tolerantrelatives with type 2 diabetes (offspring)participated in the study (34 men and 11women). They all had a parent who de-veloped type 2 diabetes. The controlgroup consisted of 65 healthy glucose-tolerant subjects without a family historyof type 2 diabetes (52 men and 13 wom-en). All subjects studied had a BMI ,25kg/m2 and a blood pressure measurement,140/90 mmHg. They all had normalglucose tolerance on the basis of at leasttwo fasting plasma glucose concentra-tions ,110 mg/dl (16,17).

All subjects studied were Japaneseand had ingested at least 150 g carbohy-drate for the 3 days preceding the study.They did not perform heavy exercise for atleast 1 week before the study. None of thesubjects had significant renal, hepatic, orcardiovascular disease or were taking anymedication. Blood was drawn the morn-ing after a 12-h fast. Plasma glucose wasmeasured with the glucose oxidase method,and serum insulin was measured using atwo-site immunoradiometric assay (Insu-lin Riabead II; Dainabot, Osaka, Japan).Coefficients of variation were 4% for in-sulin .25 mU/ml and 7% for insulin ,25mU/ml, respectively. The total choles-terol, HDL cholesterol, LDL cholesterol,

triglycerides, and remnant-like particlecholesterol (RLP-C) were also measured.The LDL cholesterol level was calculatedusing the Friedewald formula (18). TheRLP-C level was measured by the methodreported by Nakajima et al. (19). The es-timate of insulin resistance by homeosta-sis model assessment (HOMA-IR) wascalculated with the following formula:fasting serum insulin (mU/ml) 3 fastingplasma glucose (mmol/l)/22.5, as de-scribed by Matthews et al. (20).

The statistical analysis was performedwith the StatView 5.0 system (Statview,Berkeley, CA). The differences of the meanswere determined by Student’s t test. Dataare expressed as means 6 SEM.

The clinical characteristics and clini-cal profile between offspring (n 5 45) andcontrol subjects (n 5 65) were compared.No significant difference was observed inage (36.5 6 1.3 vs. 37.6 6 1.0 years, P 50.281), systolic (117 6 2 vs. 114 6 1mmHg, P 5 0.158) and diastolic (68 6 1vs. 66 6 1 mmHg, P 5 0.226) blood pres-sure, and fasting glucose (91 6 1 vs. 89 61 mg/dl, P 5 0.493) levels between thetwo groups. The offspring subjects hadhigher BMI (22.2 6 0.3 vs. 22.0 6 0.2kg/m2, P 5 0.252) and total (190 6 4 vs.184 6 2 mg/dl, P 5 0.092) cholesterollevels than normal control subjects, butthe difference was not statistically signifi-cant. In contrast, the offspring subjectshad significantly higher serum triglycer-ide (98 6 9 vs. 76 6 3 mg/dl, P 5 0.005),RLP-C (4.0 6 0.3 vs. 2.8 6 0.1, P 50.006), serum insulin (5.6 6 0.3 vs.4.7 6 0.1 mU/ml, P 5 0.008), andHOMA-IR (1.25 6 0.07 vs. 1.05 6 0.03,P 5 0.008) levels than normal controlsubjects. No significant difference was ob-served in HDL (60 6 2 vs. 59 6 1 mg/dl,P 5 0.359) and LDL (110 6 4 vs. 110 62 mg/dl, P 5 0.455) cholesterol levelsbetween the two groups.

In the present study, we demon-strated that nonobese healthy glucose-tolerant relatives of type 2 diabeticpatients had higher HOMA-IR levels com-pared with normal control subjects. Thisis compatible with some previous reportsthat showed insulin insensitivity in rela-tives of type 2 diabetic patients (10–12).However, which factor contributes toinsulin resistance in diabetic offspringsubjects has yet to be clarified. BMI,blood pressure, and fasting glucose lev-els are known to be associated with insu-lin resistance in humans (14). Hence, we

recruited nonobese nonhypertensive glu-cose-tolerant relatives of type 2 diabeticpatients and normal control subjects whohad no family history of diabetes, takinginto account BMI, blood pressure, andfasting glucose levels. All subjects studiedhad a BMI ,25 kg/m2, blood pressure,140/90 mmHg, and a fasting glucoselevel ,110 mg/dl.

Interestingly, our offspring subjectsnot only had higher HOMA-IR, but alsohad higher serum triglycerides and RLP-Clevels compared with normal control sub-jects. However, there was no significantdifference in age, BMI, total cholesterol,HDL cholesterol, and LDL cholesterollevels between the two groups. Thesefindings, therefore, suggest that triglycer-ide or RLP-C, but not BMI, is associatedwith insulin resistance in nonobese non-hypertensive glucose-tolerant relatives oftype 2 diabetic patients.

The mechanisms by which serum tri-glyceride or RLP-C level is associated withinsulin resistance in nonobese nonhyper-tensive glucose-tolerant relatives of type 2diabetic patients are presently not known.Although insulin resistance has been sug-gested to be an underlying defect leadingto the development of endogenous hyper-triglyceridemia (21), there are some datasuggesting that elevated triglyceride levelsare preceding factors for the developmentof insulin resistance (22,23). In familieswith multiple cases of hypertriglyceride-mia, increased serum triglycerides levelsserve as a risk marker for subsequent de-velopment of type 2 diabetes (22). Min-grone et al. (23) reported on two sisterswith extreme hypertriglyceridemia anddiabetes in whom the normalization of se-rum triglycerides by operation improvedinsulin resistance and glucose tolerance.We recently disclosed that Japanese type2 diabetic patients with insulin resistancehad significantly higher triglyceride levelsthan those with normal insulin sensitivity(9,24,25). We later demonstrated thatphysical exercise lowers triglyceride, fast-ing glucose, and HOMA-IR levels in Jap-anese type 2 diabetic patients withoutaffecting BMI (26). Furthermore, weshowed that bezafibrate not only reducesserum triglyceride levels but also im-proves insulin sensitivity and glycemiccontrol in type 2 diabetic patients (27).Paolisso et al. (28) recently demonstratedthat simvastatin and atorvastatin, a lipid-lowering drug, reduce serum triglycer-ides, insulin resistance, and HbA1c in type

Letters


2 diabetic patients. An in vitro study (29)showed that incubating IM-9 lympho-cytes with VLDL downregulates the cell’sinsulin receptor. Insulin binding to eryth-rocytes in the blood of patients with hy-pertriglyceridemia is reported to be low(30).

Regardless of the mechanism, ourpresent study showing that both highRLP-C and insulin resistance exist in glu-cose-tolerant relatives of type 2 diabeticpatients might suggest that the events pre-ceding the onset of diabetes contribute tothe evolution of CHD. Alternatively, ourpresent study might support the view thatthe duration of diabetes and the level ofglycemia, the risk factors for microvascu-lar disease, are not risk factors for CHD intype 2 diabetic patients (5,6).

In summary, nonobese nonhyperten-sive Japanese glucose-tolerant relatives oftype 2 diabetic patients are characterizedby impairments in insulin sensitivity,high serum triglyceride levels, and highRLP-C levels. Further study should be un-dertaken to clarify whether diabetic off-spring in association with high RLP-C andinsulin resistance is a risk factor for coro-nary atherosclerosis.

Acknowledgments— We thank the Depart-ment of Biochemistry, Kansai-Denryoku Hos-pital, Shionogi Biomedical Laboratory, andOotsuka Pharmaceutical Company, Japan, fortheir help in our research.

MITSUO FUKUSHIMA, MD1

ATARU TANIGUCHI, MD2

YOSHIKATSU NAKAI, MD3

MASAHIKO SAKAI, MD2

KENTARO DOI, MD4

KAZUKO NIN, MD5

TSUYOSHI OGUMA, MD2

SHOICHIRO NAGASAKA, MD6

KUMPEI TOKUYAMA, PHD7

YUTAKA SEINO, MD1

From the 1Department of Metabolism and ClinicalNutrition, Graduate School of Medicine, Kyoto Uni-versity, Kyoto; the 2Division of Diabetes, Kansai-Denryoku Hospital, Osaka; the 3College of MedicalTechnology, Kyoto University, Kyoto; 4GraduateSchool of Medicine, Kyoto University, Kyoto;5Nagoya University School of Health Sciences,Aichi; the 6Division of Endocrinology and Metabo-lism, Jichi Medical School, Tochigi; and 7Laboratoryof Biochemistry of Exercise and Nutrition, TsukubaUniversity, Tsukuba, Japan.

Address correspondence to Ataru Taniguchi,M.D., First Department of Internal Medicine, Kansai-Denryoku Hospital, 2-1-7 Fukushima, Fukushima-ku, Osaka-city, Osaka 553-0003 Japan. E-mail:[email protected].

● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●

References1. Bierman EL: Atherogenesis in diabetes.

Arterioscler Thromb 12:647–656, 19922. Piedrola G, Novo E, Serrano-Gotarredona

J, Escobar-Morreale HF, Villa E, Luna JD,Garcia-Robles R: Insulin resistance in pa-tients with a recent diagnosis of coronaryartery disease. J Hypertens 14:1477–1482,1996

3. Fuller JH, Shipley MJ, Rose G, Jarrett RJ,Keen H: Coronary heart disease risk andimpaired glucose tolerance. Lancet 1:1373–1376, 1980

4. Yano K, Kagan A, McGee D, Rhoads GG:Glucose intolerance and nine-year mor-tality in Japanese men in Hawaii. Am J Med72:71–80, 1982

5. Diabetes Drafting Group: Prevalence ofsmall vessel and large vessel disease in di-abetic persons from 14 centers: the WorldHealth Organization multinational studyof vascular disease in diabetics. Diabetolo-gia 28:615–640, 1985

6. Herman JB, Medalie JH, Goldbourt U:Differences in cardiovascular morbidityand mortality between previously knownand newly diagnosed adult diabetics. Dia-betologia 13:229–234, 1977

7. Kugiyama K, Doi H, Takazoe K, KawanoH, Soejima H, Mizuno Y, Tsunoda R,Sakamoto T, Nakano T, Nakajima K,Ogawa H, Sugiyama S, Yoshimura M,Yasue H: Remnant lipoprotein levels infasting serum predict coronary events inpatients with coronary artery disease. Cir-culation 99:2858–2860, 1999

8. Ai M, Tanaka A, Ogita K, Sekine M, Nu-mano F, Numano F, Reaven GM: Rela-tionship between hyperinsulinemia andremnant lipoprotein concentrations inpatients with impaired glucose tolerance.J Clin Endocrinol Metab 85:3557–3560,2000

9. Taniguchi A, Fukushima M, Sakai M,Miwa K, Makita T, Nagata I, Nagasaka S,Doi K, Okumura T, Fukuda A, KishimotoH, Fukuda T, Nakaishi S, Tokuyama K,Nakai Y: Remnant-like particle choles-terol, triglycerides, and insulin resistancein nonobese Japanese type 2 diabetic pa-tients. Diabetes Care 23:1766–1769, 2000

10. Eriksson J, Kallunki AF, Saloranta EC,Widen E, Schalin C, Groop L: Early met-abolic defects in persons at increasing riskfor non-insulin-dependent diabetes mel-litus. N Engl J Med 321:337–343, 1989

11. Osei K, Cottrell KDA, Orabella MM: Insu-lin sensitivity, glucose effectiveness, andbody fat distribution pattern in nondia-betic offspring of patients with NIDDM.Diabetes Care 14:890–896, 1991

12. Ramachandran A, Snehalatha AC, MohanV, Bhattacharyya PK, Viswanathan M:Decreased insulin sensitivity in offspringwhose parents both have type 2 diabetes.

Diabet Med 7:331–334, 199013. Laws A, Stefanick ML, Reaven GM: Insu-

lin resistance and hypertriglyceridemiain nondiabetic relatives of patients withnoninsulin-dependent diabetes mellitus.J Clin Endocrinol Metab 69:343–347, 1989

14. Reaven GM: Role of insulin resistance inhuman disease. Diabetes 37:1595–1607,1998

15. Lind L, Berne C, Lithell H: Prevalence ofinsulin resistance in essential hyperten-sion. J Hypertens 13:1457–1462, 1995

16. The Expert Committee on the Diagnosisand Classification of Diabetes Mellitus:Report of the Expert Committee on theDiagnosis and Classification of DiabetesMellitus. Diabetes Care 21 (Suppl. 1):S5–S19, 1998

17. Abbasi F, McLaughlin T, Lamendola C,Yeni-Komshian H, Tanaka A, Wang T,Nakajima K, Reaven GM: Fasting remnantlipoprotein cholesterol and triglycerideconcentrations are elevated in nondia-betic, insulin-resistant, female volunteers.J Clin Endocrinol Metab 84:3903–3906,1999

18. Friedewald WT, Levy RI, FredericksonDS: Estimation of the concentration oflow-density lipoprotein cholesterol inplasma without use of the preparative ul-tracentrifuge. Clin Chem 18:499–508,1972

19. Nakajima K, Saito T, Tamura A, SuzukiM, Nakano T, Adachi M, Tanaka A, TadaT, Nakamura H, Campos E, Havel RJ:Cholesterol in remnant-like lipoproteinsin human serum monoclonal anti-apoB-100 and anti apo-A-I immunoaffinitymixed gels. Clin Chim Acta 223:53–71,1993

20. Matthews D, Hosker J, Rudenski A, Nay-lor B, Treacher D, Turner R: Homeostasismodel assessment: insulin resistance andB-cell function from fasting plasma glu-cose and insulin concentrations in man.Diabetologia 28:412–419, 1985

21. Reaven GM, Lerner RL, Stern MP, Far-quhar JW: Role of insulin in endogenoushypertriglyceridemia. J Clin Invest 46:1756–1767, 1967.

22. Sane T, Taskinen MR: Does familiar hy-pertriglyceridemia predispose to NIDDM?Diabetes Care 16:1494–1501, 1993

23. Mingrone G, Henriksen FL, Greco AV,Krogh LN, Capristo E, Gastaldelli A, Cast-agneto M, Ferrannini E, Gasbarrini G,Beck-Nielsen H: Triglyceride-induced di-abetes associated with familiar lipopro-tein lipase deficiency. Diabetes 48:1258–1263, 1999

24. Taniguchi A, Fukushima M, Sakai M,Kataoka K, Miwa K, Nagata I, Doi K, To-kuyama K, Nakai Y: Insulin-sensitive andinsulin-resistant variants in nonobese Jap-anese type 2 diabetic patients (Letter). Di-abetes Care 22:2100–2101, 1999

Letters


25. Taniguchi A, Fukushima M, Sakai M,Kataoka K, Nagata I, Doi K, Arakawa H,Nagasaka S, Tokuyama K, Nakai Y: Therole of the body mass index and triglycer-ide levels in indentifying insulin-sensitiveand insulin-resistant varaints in Japanesenon-insulin-dependent diabetic patients.Metabolism 49:1001–1005, 2000

26. Taniguchi A, Fukushima M, Sakai M, Na-gasaka S, Doi K, Nagata I, Matsushita K,Ooyama Y, Kawamoto A, Nakasone M,Tokuyama K, Nakai Y: Effect of physicaltraining on insulin sensitivity in Japanesetype 2 diabetic patients (Letter). DiabetesCare 23:857–858, 2000

27. Taniguchi A, Fukushima M, Sakai Y, To-kuyama K, Nagata I, Fukunaga A, Kishi-moto H, Doi K, Yamashita Y, Matsuura T,Kitatani N, Okumura T, Nagasaka S, Na-kaishi S, Nakai Y: Effects of bezafibrate oninsulin sensitivity and insulin secretion innon-obese Japanese type 2 diabetic pa-tients. Metabolism 50:477–480, 2001

28. Paolisso G, Barbagallo M, Petrella G, et al:Effects of simvastatin and atorvastatin ad-ministration on insulin resistance andrespiratory quotient in aged dyslipidemicnon-insulin dependent diabetic patients.Atherosclerosis 150: 121–127. 2000

29. Steiner G: Insulin and hypertriglyceride-mia.Diabetes.WaldhauslWK,Ed.Amster-dam, the Netherlands, Excerpta Medica,1979, p. 590–593

30. Bieger WP, Michel G, Barwich D, Biehl K,Wirth A: Diminished insulin receptors onmonocytes and erythrocytes in hypertri-glyceridemia. Metabolism 33:982–987, 1984

Glucodynamics andPharmacokinetics of70/30 vs. 50/50NPH/Regular InsulinMixtures AfterSubcutaneousInjection

Optimal insulin therapy for many pa-tients with type 1 and type 2 diabe-tes requires the use of combinations

of short- and intermediate-acting insulinpreparations. A significant number of pa-tients experience difficulties with mixingintermediate- and short-acting insulinpreparations because of an inability tomix insulin preparations in correct ratios(1–2) or confusion between the differenttypes of insulins (3–5). The fixed ratio of70% NPH to 30% regular (70/30) insulinpreparation does not fulfill the needs of

many of these patients because of signifi-cant postprandial blood glucose excur-sions. A premix insulin preparation, suchas 50% NPH to 50% regular (50/50) in-sulin, could benefit those individuals withpostprandial hyperglycemia. To investi-gate differences in pharmacodynamic andpharmacokinetic properties between50/50 and 70/30 premix insulins, we per-formed a randomized single-dose two-period crossover study using a 12-heuglycemic clamp technique (7) inhealthy subjects after double-blindedsubcutaneous injections of each insulinpreparation.

The study was approved by the Insti-tutional Review Board of the University ofPittsburgh. Written informed consentwas obtained before initiation of screen-ing procedures. All subjects had a fastingblood glucose #108 mg/dl at screening,did not ingest alcohol or caffeine-containing beverages for 48 h before re-ceiving study medication, and refrainedfrom exercise for 24 h before the study. Ina double-blind crossover study, singlesubcutaneous doses (0.2 units/kg) of50/50 or 70/30 insulin were administeredto 12 healthy male subjects. Glucose, in-sulin, and C-peptide levels were mea-sured at baseline, every 10 min for 1 h,and every 15 min for 11 h after injection.Baseline euglycemia was maintained witha variable-rate glucose infusion (GIR). Se-rum insulin concentrations were analyzedusing an enzyme-linked immunosorbentassay (ELISA) with a reference interval of1.5–15.6 mIU/l, a detection limit of ,0.5mIU/ml, and cross-reactivity to C-peptideand proinsulin ,0.3%. Serum C-peptideconcentrations were analyzed by ELISA(reference interval 0.7–1.02 nmol/l, pre-cision 8%). Glucose determinations werecarried out immediately on site using aYellow Spring Instruments 2300 STATglucose analyzer. The 50/50 insulin pre-mix (50% NPH in human insulinisophane suspension 1 50% regular hu-man insulin injection [rDNA], 100 units/ml, 3 ml PenFill cartridge) and the 70/30insulin premix (Novolin 70/30, 3 ml Pen-Fill cartridge) were supplied by NovoNordisk Pharmaceutical Industries (Clay-ton, NC). The insulin premixes were ad-ministered using a NovoPen 3.0 with aNovoFine 30G needle (Novo Nordisk,Bagsvaerd, Denmark).

Glucodynamic parameters includ-ed the area under the curve (AUC) asfollows: AUCGIR(0 – 4), AUCGIR(4 – 8),

AUCGIR(8–12), maximum GIR (GIRMAX),and time to GIRMAX [TGIR(MAX)]. The fol-lowing pharmacokinetic measures werecalculated from free insulin concentra-tion–time curves adjusted for C-peptide:AUCINS(0–4), AUCINS(4–8), AUCINS(8–12),and maximum insulin concentration[CMAX and TINS(MAX)]. Adjusted insulinwas derived according to the followingformula: adjusted insulin 5 actual insu-lin 2 (baseline insulin / baseline C-pep-tide) 3 actual C-peptide. Analysis ofvariance was used to test the sequence,subject (within sequence), period, andtreatment effect on AUC and CMAX forglucose, insulin, and GIRMAX. The analy-sis was performed on raw or logarithmictransformed data, depending on the dis-tribution of the data.

A total of 12 subjects completed thestudy (age 29.7 6 9.1 years, weight78.2 6 11.6 kg, BMI 25.1 6 1.6 kg/m2,and 100% Caucasian). Mean basal glu-cose values of 90.8 6 4.9 mmol/l and90.5 6 5.8 mmol/l during 50/50 and70/30 periods of the study were main-tained for 12 h after the insulin bolus,with coefficients of variability of 5.4 and6.4%, respectively. LogAUCGIR(0–4) was6.6 6 0.6 and 6.1 6 0.6 with 50/50 and70/30 injections, respectively (P 50.0087). No significant differences wereobserved for AUCGIR(4–8), AUCGIR(8–12),AUCGIR(0–12), TGIR(MAX), or GIRMAX be-tween the two insulin preparations.LogAUCINS(0–4) was 10.2 6 0.5 after ad-ministration of 50/50 insulin and 9.6 60.5 after an injection of 70/30 insulin(P 5 0.018). No significant differenceswere obse rved in AUC I N S ( 4 – 8 ) ,AUCINS(8–12), and TINS(MAX) between thetwo premixes.

The major differences observed in thepharmacokinetic and glucodynamicproperties of 50/50 and 70/30 insulinpremixes occurred during the first 4 h af-ter injection, with significantly greaterAUCGIR(0–4), AUCINS(0–4), and CMAX after50/50 vs. 70/30 insulin. Differences inpharmacokinetics and glucodynamics be-tween NPH and regular insulins havebeen described previously (9–12). How-ever, there are few published studiesavailable that directly compare differentconcentrations of premix insulin prepara-tions. In one open-label study directlycomparing single injections of 50/50 and70/30 insulin (6), significant differencesbetween AUC for GIR and insulin concen-trations adjusted for C-peptide were ob-

Letters


served during the first 4 h after a higherdose of insulin (0.3 units/kg) than wasused in this trial. In this previous trial, theinvestigators were not blinded to the in-sulin being used, which could have con-tributed to higher GIR with the 50/50preparation because of anticipation ofpeak insulin action. In another study,postprandial hyperglycemic responses toa standard breakfast were measured in el-derly individuals with insulin-requiringtype 2 diabetes after a single injection of50/50 or 70/30 insulin mixtures (0.33units/kg) (8). No differences were ob-served in nadir and peak glucose concen-trations, time to nadir and peak glucoses,or plasma insulin concentrations duringthe first 4 h after insulin injection. Theabsence of C-peptide measurements, theuse of uncorrected insulin measures, andthe low insulin dose used during thestudy period relative to the usual insulinrequirements of the subjects (55.3 6 24.0units/day) may have contributed to theabsence of an observed difference in thetwo insulin premix preparations.

No significant differences were ob-served in any glucodynamic or pharma-cokinetic parameters during hours 4–12after the insulin injections. This suggeststhat the risk for either late hypo- or hy-perglycemia is unaffected by the type ofinsulin premix used. An overlap in actionprofiles of regular and NPH insulins dur-ing hours 3 and 5 and a relatively flat NPHaction profile during hours 6–12 couldexplain these observations (12–13).

In summary, there are statistically sig-nificant differences in glucodynamic andpharmacokinetic parameters between50/50 and 70/30 insulin mixtures duringthe first 4 h after a single injection inhealthy volunteers. This suggests that50/50 insulin premix preparations mayoffer the advantage of reducing postpran-dial glycemic excursions in patients withinsulin-requiring diabetes.

ESTHER I. KRUG, MD

LOUISE DERISO, CRNP

MARY BETH TEDESCO, CRNP

HARSHA RAO, MD

MARY T. KORYTKOWSKI, MD

From the University of Pittsburgh School of Medi-cine, Pittsburgh, Pennsylvania.

Address correspondence to Mary T. Koryt-kowski, MD, Falk, Room 581, University of Pitts-burgh Medical Center, 3601 Fifth Avenue,Pittsburgh, PA 15213. E-mail: [email protected].

Acknowledgments— This work was sup-ported by a research grant from Novo Nordiskand funds received from the National Insti-tutes of Health/National Center for ResearchResources/General Clinical Research CenterGrant no. 5M01-RR-00056.

M.T.K. serves on an advisory panel for NovoNordisk.

● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●

References1. Puxty JAN, Hunter DH, Burr WA: Accu-

racy of insulin injection in elderly pa-tients. Br Med J 287:1762, 1983

2. Bell D, Clements RS, Perentesis G, Rod-dam R, Wagenknecht L: Dose accuracy ofself-mixed vs. premixed insulin. Arch IntMed 151:2265–2269, 1991

3. Coscelli C, Calabrese G, Fedele D, Pisu E,Calderini C: Use of premixed insulinamong the elderly: reduction of errors inpatient preparation of mixtures. DiabetesCare 15:1628–1630, 1992

4. Bell D, Bode B, Clements RS, Jerron J,Hollander PA: Premixed vs. self-mixed in-sulin in the treatment of type 2 diabetesmellitus: a randomized trial. Today’s TherTrends 9:63–73, 1991

5. Klauser R, Schernthaner G, Prager R: Mix-tures of human intermediate and humanregular insulin in type 1 diabetic patients:evaluation of free insulin levels and insu-lin action on glucose metabolism aftercombined and separate subcutaneous ad-ministration. Diabetes Res Clin Pract 5:185–190, 1988

6. Woodworth JR, Howey DC, Bowsher RR:Comparative pharmacokinetics and glu-codynamics of two human insulin mix-tures (70/30 and 50/50 insulin mixtures).Diabetes Care 17:366–371, 1994

7. DeFronzo RA, Tobin JD, Andres R: Glu-cose clamp technique: a method of quan-tifying insulin secretion and resistance.Am J Physiol 237:E214–E223, 1979

8. Brodows R, Chessor R: A comparison ofpremixed insulin preparations in elderlypatients: efficacy of 70/30 and 50/50 hu-man insulin mixtures. Diabetes Care 18:855–857, 1995

9. Nijs HGT, Radder JK, Frolich M, KransHMJ: In vivo relationship between insulinclearance and action in healthy subjectsand IDDM patients. Diabetes 39:333–339, 1990

10. Bilo HJG, Heine RJ, Sikkenk AC, van derMeer J, van der Veen EA: Absorption ki-netics and action profiles of intermediate-acting human insulins. Diabetes Res 4:39–43, 1987

11. Starke AAR, Heinemann L, Hohmann A,Berger M: The action profiles of humanNPH insulin preparations. Diabet Med6:239–244, 1989

12. Gardner DF, Arakaki RF, Podet EJ, NellLJ, Thomas JW, Field JB: The pharmaco-kinetics of subcutaneous regular insulinin type 1 diabetic patients: assessment us-ing a glucose clamp technique. J Clin En-docrinol Metab 63:689–694, 1986

13. Woodworth JR, Howey DC, Bowsher RR:Establishment of time-action profiles forregular and NPH insulin using pharmaco-dynamic modeling. Diabetes Care 17:64–69, 1994

Presence ofAutoantibodies toCarbonic AnhidraseII and Lactoferrin inType 1 Diabetes

Proposal of the concept ofautoimmune exocrinopathy andendocrinopathy of the pancreas

Impairment of exocrine as well as endo-crine function of the pancreas has beenreported in type 1 diabetes (1). Histo-

logically, lymphocytic infiltration of theexocrine gland has been reported in abouthalf of Japanese type 1 diabetic patients(2). Fibrosis and atrophy of the exocrinegland have also been reported.

On the other hand, the concept of au-toimmune pancreatitis has been recentlyproposed (3). This new clinical entity ischaracterized by swelling of the pancreaswith extensive fibrosis and lymphocyteinfiltration in the exocrine pancreas (3–5). It has been suggested that carbonicanhidrase II, an antigen of the duct cells ofthe pancreas, or lactoferrin in the pancre-atic acinus may be the candidates for thetarget antigen (6).We have reported thefirst case of association of autoimmunepancreatitis and type 1 diabetes as auto-immune exocrinopathy and endocrinop-athy of the pancreas (7).

To test the hypothesis that autoim-munity against the exocrine pancreas isinvolved in the pathological process oftype 1 diabetes, we investigated the pres-ence of autoantibodies against carbonicanhidrase II and lactoferrin in type 1 dia-betes.

Eighteen patients with type 1 diabetes(2 men and 16 women, aged 45.4 619.5years, range 23–72) and 18 patients withtype 2 diabetes (7 men and 11 women,aged 59.1 6 10.9 years) as control sub-

Letters


jects were recruited for the study after in-formed consent was obtained. Diabeteswas diagnosed according to the new cri-teria of the American Diabetes Associa-tion. Diagnosis of type 1A (immune-mediated) diabetes was based on thepresence of at least one of the two immu-nological markers: anti-GAD antibody(measured by radioimmunoassay [GADAb Cosmic; RSR, Cardif, U.K.]) or anti–IA-2 (tyrosine phosphatase–like protein)antibody (measured by radioimmunoas-say [RSR]). Diagnosis of type 1B (idio-pathic) diabetes was based on insulindeficiency and the absence of these im-munological markers. Among 18 type 1diabetic patients, 12 had type 1A diabetesand 6 had type 1B diabetes. One patienthad chronic thyroiditis and one hadGraves’ disease. There was no history ofother autoimmune diseases in the otherpatients. No patient had a history of pan-creatitis.

Serum levels of autoantibodiesagainst carbonic anhidrase II (ACA) andagainst human lactoferrin (ALF) weremeasured using the solid-phase enzyme-linked immunosorbent assay, as previ-ously described (8).

Of 18 patients with type 1 diabetes,ACA was detected in 9 patients and ALFwas detected in 8 patients. A total of 12(67%) patients tested positive for either ofthe two antibodies. In contrast, neitherACA nor ALF were detected in any pa-tients with type 2 diabetes. The preva-lences of both ACA and ALF in type 1diabetic patients were significantly higherthan those of type 2 diabetic patients (P ,0.01, x2 test). Our results suggest the in-volvement of autoimmunity against theexocrine as well as endocrine pancreas inthe pathogenesis of type 1 diabetes. Theseconditions could be referred to as autoim-mune exocrinopathy and endocrinopathyof the pancreas.

In addition, it is noteworthy that ei-ther of these antibodies were detected inthree of six type 1B diabetic patients.These observations suggest that ALF andACA may be novel immunological mark-ers for type 1 diabetes.

TAKAO TANIGUCHI, MD, PHD1

KAZUICHI OKAZAKI, MD, PHD2

MOTOZUMI OKAMOTO, MD, PHD1

SHUJI SEKO, MD, PHD1

KAZUSHIGE UCHIDA, MD2

YUTAKA SEINO, MD, PHD3

From the 1Department of Internal Medicine, OhtsuRed Cross Hospital, Shiga, Japan; the 2Departmentof Endoscopic Medicine and Gastroenterology,Kyoto University Faculty of Medicine, Kyoto, Japan;and the 3Department of Metabolism and ClinicalNutrition, Kyoto University Faculty of Medicine,Kyoto, Japan.

Address correspondence to Takao Taniguchi,MD, PhD, Department of Internal Medicine, OhtsuRed Cross Hospital, 1-1-35, Nagara, Ohtsu, Shiga,Japan.

● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●

References1. Lankisch PG, Manthey G, Otto J, Koop H,

Talaulicar M, Willms B, Creutzfeldt W:Exocrine pancreatic function in insulin-dependent diabetes mellitus. Digestion 25:211–216, 1982

2. Waguri M, Hanafusa T, Itoh N, MiyagawaJ, Imagawa A, Kuwajima M, Kono N, Mat-suzawa Y: Histopathologic study of thepancreas shows a characteristic lmpho-cytic infiltration in Japanese patients withIDDM. Endocrine J 44: 23–33, 1997

3. Yoshida K, Toki F, Takeuchi T, WatanabeS, Shiratori K, Hayashi N: Chronic pan-creatitis caused by an autoimmune abnor-mality: proposal of the concept ofautoimmune pancreatitis. Dig Dis Sci 40:1561–1568, 1995

4. Taniguchi T, Seko S, Azuma K, TamegaiM, Nishida O, Inoue F, Okamoto M, Mi-zumoto T, Kobayashi H: Autoimmunepancreatitis detected as a mass in the tailof the pancreas. J Gastroenterol Hepatol 15:461–464, 2000

5. Taniguchi T, Seko S, Azuma K, Asagoe K,Tamegai M, Nishida O, Inoue F, OkamotoM, Mizumoto T, Kobayashi H: Autoim-mune pancreatitis detected as a mass inthe head of the pancreas with contiguousfibrosis around the superior mesentericartery. Dis Dis Sci 46:187–191, 2001

6. Okazaki K, Uchida K, Ohana M, NakaseH, Uose S, Inai M, Matsushima Y,Katakura K, Ohmori K, Chiba T: Autoim-mune-related pancreatitis is associatedwith autoantibodies and a Th1/Th2-typecellular immune response. Gastroenterol-ogy 118:573–581, 2000

7. Taniguchi T, Seko S, Okamoto M, Ha-masaki A, Ueno H, Inoue F, Nishida O,Miyake N, Mizumoto T: Association ofautoimmune pancreatitis and type 1 dia-betes: autoimmune exocrinopathy andendocrinopathy of the pancreas (Letter).Diabetes Care 23:1592–1594, 2000

8. Kino-Ohsaki J, Nishimori I, Morita M,Okazaki K, Yamamoto Y, Onishi S, Holl-ingsworth MA: Serum antibodies to car-bonic anhydrase I and II in patients withidiopathic chronic pancreatitis andSjogren’s syndrome. Gastrenterology 110:1579–1586, 1996

SubcutaneousContinuous GlucoseMonitoring

Feasibility of a new microdialysis-based glucose sensor system

D evices for continuous glucose mon-itoring (1,2) should display the datain real time rather than retrospec-

tively. This is possible by means of themicrodialysis technique (3). Therefore,we examined the feasibility of continuousglucose monitoring with functional mod-els of a new microdialysis-based subcuta-neous continuous glucose monitoring(SCGM) system (Roche Diagnostics,Mannheim, Germany) for up to 72 h. Thissystem was calibrated by a one-point cal-ibration shortly after the beginning of therecording period.

The SCGM system consists of a mi-crodialysis catheter and a portable extra-corporal electrochemical glucose sensor.It displays a new glucose value everyminute. The system was tested in 23 am-bulatory inpatients with insulin-treatedtype 1 and type 2 diabetes. For compari-son, up to 75 capillary blood glucose sam-ples were taken in each patient. TheSCGM signal was corrected for the timethat is needed for fluid transportationfrom the catheter to the extracorporal sen-sor (31 min) and calibrated by a linearone-point calibration model using thefirst capillary blood glucose value after aninitial equilibration period (mean 4.7 h).

The SCGM system was well toleratedin all patients over the 3-day study. It didnot limit the patients’ daily activities. Noadverse event occurred, except for a mildskin irritation caused by dressing tape.The sensor signal of the SCGM systemshowed no system-inborn drift. The per-cent difference between the SCGM sys-tem and the capillary blood glucose wascomparable over the observed bloodglucose range, with an intra-individualmean absolute difference of 14.8 6 9.9%(mean 6 SD). A representative recordingis shown in Fig. 1.

Furthermore, the ability of two glu-cose spot-measurement schemes to detecthypoglycemic episodes was analyzed ina standardized manner. Glucose spotmeasurements were retrospectively de-rived on the basis of continuous glucose

Letters


profiles according to two different mea-surement schemes: four times per day(three times preprandially and bedtime)and seven times per day (adding threetimes 2-h postprandially). Hypoglycemiawas arbitrarily defined as glucose ,70mg/dl.

Spot measurement of glucose, per-formed four or seven times per day, wouldnot detect 71% (22 of 31) or 58% (18 of31), respectively, of all hypoglycemic epi-sodes registered by continuous glucosemonitoring.

Continuous glucose monitoring bythis new SCGM system in diabetic pa-tients was feasible and safe for at least 3days. In contrast to intracorporally placedglucose electrodes (4), the SCGM sys-tem showed no time-dependent declinein glucose-sensing sensitivity. Therefore,calibration of the system by just one cap-illary blood glucose measurement at thebeginning of the monitoring period wassufficient.

This study further supports the obser-vation that hypoglycemic episodes ininsulin-treated patients are, to a large ex-tent, not detected by spot measurements,even if they are performed seven-timesper day (3,5).

In conclusion, the SCGM system of-fers real-time continuous glucose moni-toring with sufficient precision in diabeticpatients over 72 h. Detecting asymptom-atic hypoglycemic episodes by continu-

ous glucose monitoring will facilitate themanagement of patients with diabetes.

KARSTEN JUNGHEIM, MD1

KLAAS-JAN WIENTJES, MD2

LUTZ HEINEMANN, PHD3

VOLKER LODWIG, PHD4

THEODOR KOSCHINSKY, MD, PHD1

ADELBERT J. SCHOONEN, PHD2

FOR THE GLUCOSE MONITORING STUDY

GROUP

From the 1German Diabetes Research Institute at theHeinrich-Heine University of Duesseldorf, Duessel-dorf, Germany; the 2Department of PharmaceuticalTechnology and Biopharmacy, University Center ofPharmacy, University of Groningen, Groningen, theNetherlands; the 3Profil Institute for Metabolic Re-search, Neuss, Germany; and 4Roche Diagnostics,Mannheim, Germany.

Address correspondence to Karsten Jungheim,MD, German Diabetes Research Institute, Auf’mHennekamp 65, D-40225 Duesseldorf, Germany. E-mail: [email protected].

L.H. and T.K. are members of the advisory boardon SCGM systems to Roche Diagnostics, and V.L. isemployed full-time at Roche Diagnostics.

Acknowledgments— This study was sup-ported by Roche Diagnostics, Mannheim, Ger-many.

K.J., K-J.W., L.H., T.K., and A.J.S. have re-ceived unrestricted research grant supportfrom Roche Diagnostics, Mannheim, Ger-many.

The authors thank Elisabeth Moll for excel-lent technical assistance.

APPENDIX

GLUCOSE MONITORING STUDYGROUPJens Sandahl Christiansen, Hans Orskov(Aarhus, Denmark); Michael Berger, MaxSpraul, Theodor Koschinsky (Duessel-dorf, Germany); Lutz Heinemann (Neuss,Germany); Volker Lodwig (Mannheim,Germany); Cornelia Haug (Ulm, Germa-ny); Adelbert J. Schoonen (Groningen,Netherlands); Urban Ungerstedt (Stock-holm, Sweden).

● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●References1. Bode BW, Gross TM, Thornton KR, Mas-

trototaro JJ: Continuous glucose moni-toring used to adjust diabetes therapyimproves glycosylated hemoglobin: a pi-lot study. Diabetes Res Clin Pract 46:183–190, 1999

2. Maran A, Crepaldi C, Tiengo A, GrassiG, Vitali E, Pagano G, Bistoni S, CalabreseG, Santeusanio F, Leonetti F, Ribaudo M,Di Mario U, Annuzzi G, Genovese S, Ric-cardi G, Previti M, Cucinotta D, GiorginoR, Poscia A, Varalli M: Continuous subcu-taneous glucose monitoring in diabetic pa-tients: a multicenter evaluation (Abstract).Diabetes Res Clin Pract 50:S333, 2000

3. Bolinder J, Hagstrom-Toft E, UngerstedtU, Arner P: Self-monitoring of bloodglucoseintype1diabeticpatients:compar-ison with continuous microdialysis mea-surements of glucose in subcutaneousadipose tissue during ordinary life condi-tions. Diabetes Care 20:64–70, 1997

4. Gerritsen M: Problems associated with sub-cutaneously implanted glucose sensors(Editorial). Diabetes Care 23:143–145, 2000

5. Gross TM, Mastrototaro JJ, FredricksonLP: Detection of unseen hypoglycemia us-ing continuous glucose monitoring (Ab-stract). Diabetologia 43:A5, 2000

Serum Concentrationsof Soluble VascularCell AdhesionMolecule-1 andE-Selectin AreElevated in Insulin-Resistant PatientsWith Type 2 Diabetes

I t is well known that soluble intercellu-lar adhesion molecule-1 (sICAM-1),soluble vascular cell adhesion mole-

cule-1 (sVCAM-1), and soluble E-selectin(sE-selectin) levels are elevated in patients

Figure 1—Continuous glucose recording (—) in a patient with type 1 diabetes by means of theSCGM system. One capillary blood glucose value was used for one-point calibration (Œ), andsucceeding capillary blood glucose values (‚) were used for comparison.

Letters


with type 2 diabetes (1–3). Previous stud-ies suggest that hyperglycemia, hyperin-sulinemia, or insulin resistance may beresponsible for the elevation of adhesionmolecules (4–6). However, to our knowl-edge, the relation between soluble adhe-sion molecules and directly measuredinsulin sensitivity has not been studied inpatients with type 2 diabetes. To investi-gate the direct association between insu-lin resistance and the elevation of solubleadhesion molecules, we performed acase-control study. Serum concentrationsof sICAM-1, sVCAM-1, and sE-selectinwere compared in insulin-sensitive andinsulin-resistant patients who werematched for age, sex, BMI, duration ofdiabetes, and glycemic control (HbA1c).Insulin sensitivity was assessed by the Kindex of short insulin tolerance test (KITT)(7). Briefly, a bolus of regular insulin (0.1unit/kg) was infused, and blood sampleswere collected at 3, 6, 9, 12, and 15 minafter infusion. KITT was calculated usingthe following formula: KITT 5 0.693/t1/2(8). Patients whose KITT values were,2.00% per min were diagnosed as hav-ing insulin resistance, and those with KITTvalues .3.00% per min were defined asinsulin-sensitive. Serum concentrationsof sICAM-1, sVCAM-1, and sE-selectinwere measured by enzyme-linked immu-nosorbent assay kits (R&D Systems, Min-neapolis, MN). Data were compared withStudent’s t test or contingency table anal-ysis.

A total of 37 insulin-sensitive patients(24 men and 13 women, mean 6 SEM ofKITT 5 3.82 6 0.12% per min) and 37insulin-resistant patients (24 men and 13women, mean KITT 5 1.47 6 0.06% permin) were compared. The age (61 6 2 vs.62 6 2 years, P 5 0.51, respectively), BMI(23.2 6 0.5 vs. 24.3 6 0.7 kg/m2, P 50.22, respectively), duration of diabetes(6 6 1 vs. 7 6 1 years, P 5 0.59, respec-tively), and HbA1c levels (8.9 6 0.2 vs.9.1 6 0.3%, P 5 0.66, respectively) werecomparable between the two groups. Thepercentage of insulin therapy (21.6 vs.37.8%, P 5 0.20, respectively), presenceof retinopathy (18.9 vs. 10.8%, P 5 0.52,respectively), and presence of coronaryartery disease (8.1 vs. 18.9%, P 5 0.31,respectively) did not differ between thetwo groups. Concerning the lipid profile,insulin-sensitive patients had signifi-cantly lower serum triglyceride levelsthan insulin-resistant patients (1.0 6 0.1vs. 1.5 6 0.1 mmol/l, P , 0.01, respec-

tively). sICAM-1 levels were comparablebetween the two groups (166.5 6 11.9 vs.186.2 6 1 0.5 ng/ml, P 5 0.22, respec-tively). In contrast, levels of sVCAM-1and sE-selectin in insulin-sensitive pa-tients were significantly lower than thoseof insulin-resistant patients (sVCAM-1683 6 23 vs. 813 6 44 ng/ml, P 5 0.01;and sE-selectin 48.9 6 4.0 vs. 81.5 65.7 ng/ml, P , 0.01, respectively). Be-cause HbA1c levels in insulin-sensitiveand -resistant patients were comparable,the significant elevation of sVCAM-1and sE-selectin in insulin-resistant pa-tients may be independent of the plas-ma glucose level. Thus, in type 2 diabeticpatients, insulin resistance is associatedwith an elevation in sVCAM-1 and sE-selectin. Increased levels of sVCAM-1 andsE-selectin may partly explain the pre-disposition of insulin-resistant type 2diabetic patients to atherosclerotic vas-cular disease.

KAZUNARI MATSUMOTO, MD1

YASUNORI SERA, MD1

YASUYO ABE, MD2

YUKITAKA UEKI, MD3

SEIBEI MIYAKE, MD1

From the Departments of 1Diabetes and Metabo-lism, 2Endocrinology, and 3Rheumatology, SaseboChuo Hospital, Nagasaki, Japan.

Address correspondence to Kazunari Matsu-moto, MD, Department of Diabetes and Metabolism,Sasebo Chuo Hospital, 15 Yamato-cho, Sasebo,Nagasaki 857-1195, Japan. E-mail: [email protected].

● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●

References1. Fasching P, Waldhausl W, Wagner OF:

Elevated circulating adhesion moleculesin NIDDM; potential mediators in dia-betic macroangiopathy. Diabetologia 39:1242–1244, 1996

2. Carter AM, Grant PJ: Vascular homeosta-sis, adhesion molecules, and macrovas-cular disease in non-insulin-dependentdiabetes mellitus. Diabet Med 14:423–432, 1997

3. Albertini JP, Valensi P, Lormeau B,Aurousseau MH, Ferriere F, Attali JR,Gattegno L: Elevated concentrations ofsoluble E-selectin and vascular cell adhe-sion molecule-1 in NIDDM: effect of in-tensive insulin treatment. Diabetes Care21:1008–1013, 1998

4. Morigi M, Angioletti S, Imberti B, Don-adelli R, Micheletti G, Figliuzzi M, RemuzziA, Zoja C, Remuzzi G: Leukocyte-endothe-lial interaction is augmented by high glu-cose concentrations and hyperglycemia in aNF-kB-dependent fashion. J Clin Invest 101:

1905–1915, 19985. Chen NG, Holmes M, Reaven GM: Re-

lationship between insulin resistance,soluble adhesion molecules, and mono-nuclear cell binding in healthy volunteers.J Clin Endocrinol Metab 84:3485–3489,1999

6. Matsumoto K, Miyake S, Yano M, Ueki Y,Tominaga Y: High serum concentrationsof soluble E-selectin in patients with im-paired glucose tolerance with hyperin-sulinemia. Atherosclerosis 152:415–420,2000

7. Matsumoto K, Miyake S, Yano M, Ueki Y,Yamaguchi Y, Akazawa S, Tominaga Y:Insulin resistance and arteriosclerosis ob-literans in patients with NIDDM. DiabetesCare 20:1738–1743, 1997

8. Bonora E, Moghetti P, Zancanaro C, Cigo-lini M, Querena M, Cacciatori V, CorgnatiA, Muggeo M: Estimates of in vivo insulinaction in man: comparison of insulin tol-erance tests with euglycemic and hyper-glycemic clamp studies. J Clin EndocrinolMetab 68:374–378, 1989

COMMENTS ANDRESPONSES

Response to Maloneet al.

The article on early retinopathy byMalone et al. (1) suggested that allnewly diagnosed people with type 1

diabetes should have annual retinal exam-inations through a dilated pupil. The datato support this suggestion were takenfrom the Diabetes Control and Complica-tions Trial (DCCT). The DCCT recruitedsubjects $13 years of age who were atleast Tanner II stage for pubertal develop-ment. Therefore, the data do not apply toprepubertal children. Unfortunately, nosuch qualification was noted in their arti-cle or the accompanying editorial (2). Ret-inal changes are almost never present inchildren before puberty. I am concernedthat Malone’s article will result in evenmore children than at present having reg-ular ophthalmologic assessments that areentirely unnecessary. The American Dia-betes Association recommends that chil-dren $10 years of age only need to haveretinal examinations after 3–5 years, atime when most will have entered puberty(3). Perhaps a qualification should benoted in a future issue of Diabetes Care.

ROBERT M. EHRLICH, MD, FRCP

Letters


From Markham Stouffville Hospital, Markham, On-tario, Canada.

Address correspondence to Robert M. Ehrlich,381 Church St., Markham, Ontario L3P 7P3, Can-ada. E-mail: [email protected].

● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●

References1. Malone JI, Morrison AD, Pavan PR, Cuth-

bertson DD: Prevalence and significanceof retinopathy in subjects with type 1 di-abetes of less than 5 years’ durationscreened for the Diabetes Control andComplications Trial. Diabetes Care 24:522–526, 2001

2. Palmberg P: Screening for diabetic reti-nopathy (Editorial). Diabetes Care 24:419–420, 2001

3. American Diabetes Association: Diabeticretinopathy (Position Statement). Diabe-tes Care 24 (Suppl. 1):S73–S76, 2001

Response to Ehrlich

D r. Ehrlich’s (1) point is well taken.The Diabetes Control and Compli-cations Trial data did not include

subjects under the age of 13 years. Ourstudy (2) was not intended to deal with

the subject of retinal exams for prepuber-tal children. Most physicians who care forchildren with diabetes agree with Dr. Ehr-lich, who stated “retinal changes are al-most never present in children beforepuberty.” This claim seems to overlookthe Wisconsin Epidemiologic Study ofDiabetic Retinopathy (3), a population-based study that reported that 4 of 26children (15.4%) aged 0–9 years and 23of 42 children (54.8%) aged 10–12 years,or 27 of 68 children (40%) aged 0–12years, had retinopathy on retinal colorphotographs. This widely accepted studysuggests that retinopathy does occur be-fore puberty. If knowledge of the pres-ence of microvascular disease in a patientis not important before it becomes vision-threatening, as suggested by Dr. Palmberg(4), then exams before puberty may beunnecessary. One should not believe,however, that microvascular disease doesnot exist in prepubertal children whohave diabetes.

JOHN I. MALONE, MD

From the Diabetes Center, Health Sciences Center,University of South Florida, Tampa, Florida.

Address correspondence and reprint requests toJohn I. Malone, MD, Diabetes Center, Health Sci-ences Center, University of South Florida, 12901Bruce B. Downs Blvd., MDC 45, Tampa, FL 33612.E-mail: [email protected].

● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●

References1. Erlich RM: Response to Malone et al. (Let-

ter). Diabetes Care 24:1698, 20012. Malone JI, Morrison AD, Pavan PR, Cuth-

bertson DD: Prevalence and significance ofretinopathy in subjects with type 1 diabetesof less than 5 years’ duration screened forthe Diabetes Control and ComplicationsTrial. Diabetes Care 24:522–526, 2001

3. Kline R, Klein BEK, Moss SE, Davis MD,DeMets DL: The Wisconsin EpidemiologicStudy of Diabetic Retinopathy. IX. Four-year incidence and progression of diabeticretinopathy when age at diagnosis is lessthan30years.ArchOphthalmology107:237–243, 1989

4. Palmberg P: Screening for diabetic retino-pathy (Editorial). Diabetes Care 24:419–420,2001

Letters


Subcutaneous Continuous Glucose Monitoring: Feasibility of a new microdialysis-based glucose sensor...

Documents

Transcript of Subcutaneous Continuous Glucose Monitoring: Feasibility of a new microdialysis-based glucose sensor...