ADA Scientific 2000

7/30/2019 ADA Scientific 2000

1/6

American Diabetes Association 60thScientific Sessions, 2000Glucose tolerance, diabetes and cancer, glycemic control, monitoring, andrelated topics

ZACHARY T. BLOOMGARDEN, MD

This is the sixth in a continuing seriesof articles on the American Diabetes

Association (ADA) 60th ScientificSessions held in San Antonio, TX, in June2000. It covers topics related to use offasting versus 2-h glucose for diagnosis of

diabetes, diabetes and cancer, glycemiccontrol, monitoring, and hypoglycemia.

Postload Glycemia in the Diagnosisof Diabetes

At a symposium on diagnostic criteria fordiabetes James Gavin, Chevy Chase, MD,discussed the dilemma that fasting andpostglucose load glucose levels are dif-ferently regulated and that diagnosisbased on the two criteria therefore definesomewhat different populations. The useof fasting glucose has several advantages.

It is more convenient to perform and lim-its a missed diagnosis of diabetes. Gavinnoted that microvascular complicationsbegin to occur above a fasting glucose of126 mg/dl. A concern, however, is thatmore than half of patients with diabetesdiagnosed by a glucose tolerancetest havea fasting glucose below 126 mg/dl. Givennew data showing that the fasting glucoselevel is less predictive of cardiovasculardisease (CVD) risk than postchallengeglucose level, Gavin concluded that fun-damentally, we are back at square one inasking What is diabetes? Daniel Porte,

San Diego, CA, noted that cause and ef-fect is not necessary when we see associ-ations, an important caveat in the newemphasis on glucose tolerance testing and

postchallenge glycemia for diabetes diag-nosis.

Jaakko Tuomilehto, Helsinki, Fin-land, reviewed data from the Diabetes Ep-idemiology Collaborative Analysis ofDiagnostic Criteria in Europe (DECODE)

Study of 18,048 men and 7,316 womenaged 30 years or older followed for anaverage of 7 years after glucose tolerancetesting, addressing the question of wheth-er fasting glucose or 2-h postload glucosebetter predicts mortality (1) There were atotal of 1,836 deaths. The World HealthOrganization (WHO) criteria for diabetes,based mainly on 2-hglucose200 mg/dl,andthe ADAcriteria, based on fasting glu-cose126 mg/dl, were shown to conveya similar (1.85-fold and 1.75-fold) in-crease in risk among men; but among

women, the WHO criteria for diabetesconveyed a 2.43-fold increase, comparedwith the 1.77-fold increase with the ADAcriteria. Adjusting the fasting for the 2-hglucose greatly decreased the significanceof the increase in mortality risk, whereasadjusting the 2-h for the fasting glucosehad little such influence. In all fasting glu-cose categories (normal, impaired fastingglucose [IFG],and diabetes), mortality in-creased with increasing 2-h glucose, butin the 2-h categories (normal, impairedglucose tolerance [IGT], and diabetes),there was little increase in risk as the fast-

ing glucose level increased. No thresholdwas seen for increasing mortality with in-creasing 2-h glucose, whereas for fastingglucose the threshold was 126 mg/dl.

Finally, Tuomilehto pointed out that thegreatest number of excess deaths oc-curred in the large group of individualswith normal fasting glucose and IGT de-termined on the basis of postload glucose.

David Nathan, Boston, MA, stressedtheneed to clarify theassociation betweenhyperglycemia and risk of complications,listing three questions: What is thecorrectglycemic threshold for diagnosis of diabe-

tes? Do glucose levels in the subdiabeticrange impart risk for diabetes? And isthere therapeutic value in identifying pa-tients in this glucose range? Most diabetesoccursin thesetting of aging. HbA

1clevels

increase with age, even in patients withneither diabetes nor IGT. For diabetes-specific complications, risk increases withglucose above a threshold defined byHbA

1c6.0%. TheDECODE data shows,

however, that subdiabetic degrees of gly-cemia increase the risk of CVD. Similarly,in the Framingham Offspring Study, indi-viduals without diabetes showed an in-crease in CVD risk with increasing levelsof HbA

1c(2). The important question

Nathan raised is whether it is glucose perse that increases the risk of CVD in thispopulation. In the Framingham study,there is a direct and continuous relation-ship between blood glucose and bloodpressure, triglyceride, low HDL, and avariety of other risk factors in patientswithout diabetes. Thus, it may be overlysimplistic to conceptualize individuals aseither having normal glucose tolerance orIGT. Similarly, glycemic treatment may or

maynot be relevantto CVD risk reductionin this population. The Diabetes Preven-tion Program (DPP) is studying patients inthe upper half of IGT, allowing the po-tential to identify patients before theonsetof disease and to prevent CVD complica-tions. Nathan acknowledged that CVDrisk factor treatment is being recom-mended for all patients in the DPP, in-cluding those in the control group. Thus,it may not be possible to learn from thisstudy what are the essential characteris-tics of the increase in CVD risk in patients

Zachary T. Bloomgarden, MD, is a practicing endocrinologist in New York, New York, and is affiliated withthe Division of Endocrinology, Mount Sinai School of Medicine, New York, New York.

Abbreviations:ADA, American Diabetes Association; CHD, coronary heart disease; CVD, cardiovasculardisease; DCCT, Diabetes Control and Complications Trial; DECODE, Diabetes Epidemiology Collaborative

Analysis of Diagnostic Criteria in Europe; DPP, Diabetes Prevention Program; FFA, free fatty acid; IFG,impaired fasting glucose; IGT, impaired glucose tolerance; WHO, World Health Organization.

A table elsewhere in this issue shows conventional Systeme International (SI) units and conversion factorsfor many substances.

R e v i e w s / C o m m e n t a r i e s / P o s i t i o n S t a t e m e n t s

P E R S P E C T I V E S O N T H E N E W S

DIABETES CARE, VOLUME 24, NUMBER 4, APRIL 2001 779


2/6

wit h post loa d hype rglyc e mia a ndwhether efforts directed at glycemia itselfare sufficient to improve their prognosis.

In the Kelley West Lecture, EvelineEschwege, Paris, France, brought addi-tional arguments from the Paris Prospec-tive Study to a discussion of use of the 2-hglucose level in the diagnosis of diabetes.Many different sets of glucose tolerancetest criteria have been suggested over thepast three decades, with a consequent841% variation in the frequency of dia-betes in a group of500 individuals inher center. Use of a 2-h glucose of 11.2mmol/l, an increased fasting glucose re-quirement (from 7.2 to 7.8 mmol/l, assuggested in 1979), and a different fastingglucose criterion (7.0 mmol/l, as sug-gested in 1997) can be reassessed by anal-ysis of the 2-h glucose tolerance tests in

7,000 Paris policemen initially studiedin the 1970s. Mortality follow-up showedthat 8 and 23% had died at 10 and 20years, respectively, 1.6 and 4.7% of coro-nary disease.

About 200 of the Paris policemen had2-h glucose 11.1 mmol/l and fastingglucose 7.0 mmol/l, associated with2.5-fold increases in coronary mortality.The 68 whose fasting glucose was 7.0mmol/l with 2-h level 11.2 mmol/l,however, had lower mortality rates. Theirlipid and insulin levels were similar tothose in the nondiabetic group, theirblood pressure was similar to that in thediabetic group, and they had higher fast-ing free fatty acid (FFA) levels. At 30months, 8% of men with isolated 2-h glu-cose elevation vs. 2.4% of the nondiabeticgroup developed fasting hyperglycemia.

At 10 years, 18 had died, a 2.5-foldgreater mortality than in the nondiabeticgroup, although the coronary heart dis-ease (CHD) mortality was 1.5%, similar tothat in thenondiabetic group. At 20 years,total mortality was again 2.5-fold greater,again with similar CHD mortality to the

nondiabetic group. Cancer mortality ap-peared to explain the increased deaths.

Patients with IGT based on postloadhyperglycemia comprised 10% of theoverall group, with half having normalfasting glucose and half IFG. Men withfasting glucose 6.0 mmol/l had charac-teristics almost identical to the nondia-betic men, whereas those with IFG weresimilar to the group with diabetes amongthe risk factors. At 30 months, 6% hadfasting glucose in the diabetic range,mainly because of progression of those

with IFG. At 10 and 20 years, mortalitywas 1.7- and 1.5-fold increased in thosewho had isolated postload IGT, againmainly reflecting an increase in cancermortality, although CHD mortality wassomewhat greater than in the normal glu-cose tolerance group at 20 years.

Eschwege concluded that, amongParis policemen, isolated postchallengehyperglycemia was not associated withincreased CHD, but rather with increasedrisk of malignancy, the 10-year death ratefrom cancer increasing from 3 to 5 to 8%and the 20 year risk increasing from 10 to16 to 31% in normal versus isolated post-load IGT versus isolated postload diabe-tes, respectively. Insulin levels were notelevated in this group, arguing against thetheory that these patients had insulin re-sistance. FFA levels appeared to be a

marker of tobacco and/or alcohol con-sumption. FFAs were independent riskfactors for cancer and were particularly arisk factor for malignancies usually felt tobe related to alcohol, even excludingthose with death from malignancy duringthe first 5 years of follow-up. The onlyavailable marker of alcohol intake was theerythrocyte mean corpuscular volume,which was higher in theisolated postchal-lenge hyperglycemia groups. It should benoted that other authors have suggestedthat increased alcohol use may be asso-ciated with isolated postchallenge hy-perglycemia. Although more data arerequired, Eschwege suggested that thefasting glucose appears sufficient toscreen for individuals at increased CHDrisk and that, at least among policemen inParis, high postload glucose indicate ex-cess alcohol use.

Diabetes and MalignancyAddressing the topic of whether diabetesis associated with an increased incidenceof cancer, Julie C. Will of the Centers forDisease Control and Prevention, Atlanta,

GA, discussed data pertaining to colorec-tal and prostate cancer. Data from theFirst Cancer Prevention Study, collectedby the American Cancer Society, withmore than1 million participants recruitedbetween 1959 and 1960 who had fol-low-up to ascertain diagnoses of cancerover the subsequent 13 years, included15,000 patients with a history of diabe-tes and 850,000 without diabetes. Ap-proximately 150 and 7,000 individuals inthe two groups developed colorectal can-cer, adjusting for dietary factors including

consumption of fruit, cereal, etc.; BMI;pregnancy history; and cigarette use. Formen, diabetes was associated with a 1.3-fold increase in colon cancer risk. Therewas also an 8% increase for every year ofage. Family history and cigarette smokingwere risk factors, whereas exercising,drinking milk, and taking aspirin wereprotective factors. For women, the 1.16-fold higher risk of colorectal cancer withdiabetes was not statistically significant.Other studies of the association of diabe-tes with colorectal cancer have shownpositive association for both sexes, withsimilar risk increases approximating 25%and peak rates at 11 years of diabetes. Po-tential mechanisms include slower boweltransit time increasing exposure to carcin-ogens, altered bacterial flora affecting bileacid metabolism, and growth-promoting

effects of insulin and IGF-1.The incidence of prostate cancer

among the6,000 men with diabetes in thesame American Cancer Society databaseshowed 65 developing prostate cancer, asopposed to 2,500 cases among the400,000 men without diabetes. After ad-

justment for age, race, dietary variables,cigarettes, and exercise, there was no sig-nificant overall association of diabetes,although newly onset diabetes was asso-ciated with a 1.56-fold increase in risk. Aquestion is whether there was detectionbias for men with diabetes because of in-creased rectal examination and prostate-specific antigen screening or whetherthere were effects of altered sex hormonelevels or growth-promoting effects of in-sulin and IGF-1 in diabetes.

Thomas A. Sellers, Rochester, MN,discussed the association of diabetes withbreast cancer based on the Iowa WomensHealth Study, a prospective cohort studyof nearly 42,000 women that beganaround 1985 (3). Most participants livedin rural areas, reported little alcohol use,and had high school educations. The

study was geared toward assessment ofthe relationship between obesity andbreast cancer and included height,weight, and waist and hip girth measure-ments. Family history of breast cancer,age, and abdominal obesity are known tobe associated with breast cancer; thewaist-to-hip ratio is only associated withbreast cancer in women with a positivefamily history, suggesting an interaction.Potential explanations include familialtendency to abdominal obesity. Sellersnoted that insulin is a growth factor for

Perspectives on the News

780 DIABETES CARE, VOLUME 24, NUMBER 4, APRIL 2001


3/6

breast epithelium, that estrogen-depen-dent breast cancer cell lines require insu-lin or IGF for cell survival, and thataromatization of androgens to estrogenmay play a role.

Sellers reported decreased prognosisfor survival in women with breast cancerand diabetes. For patients without familyhistory of breast cancer, diabetes did notaffectrisk of breastcancer, butwith a fam-ily history of breast cancer, diabetes ap-peared to convey increased risk, most ofwhich was accounted for by obesity. Fur-ther, among 743 individuals who devel-oped diabetes during the follow-upperiod, there was a 21% increase in riskwith a family history of breast cancer,again related to obesity. However, a fam-ily history of breast cancer was associatedwith a doubling of diabetes-related mor-

tality, which was not explained by obe-sity. Having a sister with diabetes wasassociated with having a sister with breastcancer, independent of the obesity level ofthe subject, but other familial associationswere not consistent. Sellers noted thatmeasures of insulin resistance in otherstudies suggest an association with breastcancer as well.

Glycemic ControlA number of studies presented at themeeting gave important additional insightinto questions of theadverse effects of gly-cemia. The concerns from Nathans pre-sentation were underscored by a reporton the baseline characteristics of the3,234 participants in the Diabetes Preven-tion Program (abstract 303). The BMIaveraged 34.0 kg/m2, with 57% of themen and 73% of the women having BMI30 kg/m2. The average age was 51years;67.7% were women (64% postmenopau-sal), 55% were Caucasian, 20% African-

American, 16% Hispanic-American, 5%American Indian, and 4% Asian-Amer-ican. Hypercholesterolemia was present

in 34% and hypertension in 28%, so ifvigorous treatment of these was pursuedin the control group, as is ethically re-quired, it may be difficult to documentbenefit of glycemic treatment. Addressingearly treatment, Komatsu et al. (abstract299) reviewed the outcome of glyburidetherapy in 111 previously untreated pa-tients with type 2 diabetes. Those whofailed to achieve HbA

1c6.5% had mean

diabetes duration of 10.4 years, whilemean duration was 4.6 years for the 69patients whose HbA

1cdecreased to

6.5%. In the latter group, insulin resis-tance and decreased pancreatic -cellfunction correlated with longer durationwithout treatment, but not with the initialHbA

1c, suggesting the benefit of early

treatment.The relationship between glycated

hemoglobin and complications is com-plex. Manley et al. (abstract 742) deriveda regression equation for the relationshipbetween fasting plasma glucose (in milli-moles per liter) and HbA

1c(percentage)

from 1,180 measurements taken in theU.K. Prospective Diabetes Study, similarto the reported relationship in the Dia-betes Control and Complications Trial(DCCT). They noted the wide confidencerange: an HbA

1cof 7% corresponds to a

fasting plasma glucose of 149 mg/dl, butwith 95% confidence limits from 83 to

212 mg/dl. The difference between glyce-miaand hemoglobin glycation maybe im-portant in prognosis. McCarter et al.(abstract 195) analyzed the DCCT data-base, showing that patients whose HbA

1c

exceeded that expected based on theirglucose profile measurements had in-creased development of retinopathy andnephropathy. Rates of protein glycationmay vary for a given degree of glycemiaand may contribute to differing rates ofdevelopment of complications. Alterna-tively, those with higherHbA

1cfor a given

mean glucose may have unrecognizedglycemic variability. Bastyr et al. (abstract389) reported stronger correlation of 2-hpostprandial glucose than of fasting glu-cose with HbA

1cin 131 patients with sec-

ondary sulfonylurea failure 3 monthsafter initiating combination therapy. Les-lie et al. (abstract 304) studied 44 mono-zygotic and 45 dizygotic pairs of twinswithout diabetes, showing that heritablegenetic effects explained 65% of the pop-ulation variance in HbA

1c. Among 33and

45 pairs of monozygotic twins concor-dant and discordant for type 1 diabetes,

HbA1c correlated with r 0.73 and r 0.47, suggesting that even in type 1 di-abetes, factors independent of glycemiamay play a role in HbA

1cvariability. In an

assessment of economic implications,Menzin et al. (abstract 189) followed2,394 patients with diabetes from 1994on, showing a 3-year hospitalization rateof 11% and a tripling of hospitalizationrate as HbA

1cincreased from 6 to 12%.

Average inpatient hospitalization costswere $1,100, $1,700, and $2,500 for in-dividuals with HbA

1cof 8, 10, and 12%,

respectively. Wiedmeyer et al. (abstract386) reported that the intraindividual co-efficient of variation in glycohemoglobinin 126 stable patients with type 2 diabeteswas 4.81%, suggesting that factors otherthan glycemic status, such as assay vari-ability, altered life span of red blood cells,or other biological factors, do not them-selves play an important role in HbA

1c

changes for a given patient.

MonitoringLaffel et al. (abstract 369) and Fineberget al. (abstract 426) assessed a capillaryblood beta-hydroxybutyrate measure-ment technology for home use. Levels ex-ceeded0.5 and 1.0 mmol/l in 4 and 1%of15,893 determinations in 86 children and88 adults performing home glucose mon-itoring over a 4-week period. Of the ele-

vated levels, 87% were in children, halfof the episodes occurred between 5:00and 11:00 A.M., and the risk of levels 1mmol/l was eightfold greater in individu-als with mean glucose 200 mg/dl, sug-gesting that such testing may be useful forselected patients. Compared with urinaryketone testing, this method showed neg-ative or trace urine on 56 and 69%, inchildren and adults, respectively, of oc-casions when blood levels were 0.5mmol/l, suggesting that this may proveuseful for sick day testing. Bohannon(396) and Koschinsky et al. (abstract 461)reported on use of a home glucose moni-toring device using0.3l of blood, lessthan one-tenth of that used for fingersticksample determination, allowing lancingof forearm and thigh for capillary glucosesampling in several hundred individuals.

Accuracy was high, and patients reporteddecreased discomfort with the device.

Use of a device for measurement ofglucose extracted through theskin by ion-tophoresis and measured three times perhour with an electrochemical sensor(GlucoWatch; Cygnus, Redwood City,

CA) was reported on by Garg et al. (ab-stracts 433 and 434) and Edelman et al.(abstract 420) from a home use trial in111 insulin-treated patients. With pat-tern-management and sliding scale in-sulin dose adjustment algorithms, themethod identified five times as many dosechanges as were found with twice dailycapillary glucose testing, with 2351% ofpatients showing postprandial hypergly-cemic episodes that would not be recog-nized with conventional preprandialtesting and for which supplemental insu-

Bloomgarden



4/6

lin could be administered. Sensitivity ofthe device for detection of hypoglycemiawas 75%, in comparison with sensitivitiesof 14 and 39% for twice-daily and four-times-daily capillary glucose testing. In15 adults using the device for 6 weeks,glucose testing performance was stable.

All patients showed skin erythema fromthe initial use, which was mild in two-thirds and moderate in one-third at theconclusion of study. Wilson et al. (ab-stract 524) reported on use of the deviceby 103 individuals, with none developingmore than moderate erythema or morethan mild local edema. Tamada et al. (ab-stract 514) reported on 452 individualstested with 13,676 paired capillary anddevice data points, showing a mean dif-ference of5 mg/dl, correlation of r0.80.85, and 94% of tests in the

AB region on error grid analysis.A number of investigators evaluated a

subcutaneously placed glucose sensorthat allowed measurement of interstitialglucose levels every 5 min during a 72-hperiod (MiniMed, Sylmar, CA). Blumaueret al. (abstract 392) used this approach inrats administered insulin, with glucosefalling from 175 to 80 mg/dl over a 1-hperiod. Venous and interstitial glucoseshowed high correlation (r2 0.740.80). Steil et al. (abstract 510) placed thesensor in five nondiabetic men undergo-ing glucose infusion to increase plasmaglucose from 5.5 to 17.2 mmol/l, then sta-bilized at 10.2 mmol/l, with levels subse-quently falling to 4.4 mmol/l. Afterabruptly increasing glucose levels, a lagwith t

1/2averaging 3 min was seen be-

tween interstitial and venous plasmablood glucose levels, which showedstrong correlation during a 5-h period,with mean difference averaging 8%.Boyne et al. (abstract 398) assessed thetime lag between interstitial and venousglucose after a meal in 11 individuals withtype 1 diabetes who had two sensors in-

serted simultaneously, showing that in-terstitial glucose lags behind venous levelsby a mean of 13 min. It was not possible,however, to statistically distinguish thelag in glucose fall after meals from thesensor-to-sensor variation. Bode et al.(abstract 393) used the system for two1-week periods, changing insulin treat-ment on both occasions. At 5 and 10weeks, HbA

1chad decreased from an ini-

tial mean level of 9.9 to 8.8 and 8.6%,without significant change in total dailyinsulin dose, suggesting that the device

allows fine tuning, which in turn leadsto improvement in glycemia. Boland et al.(abstract 397) studied the sensor in 16children with type 1 diabetes and HbA

1c

averaging 7.6%. Interstitial and capillaryglucose levels showed a mean differenceof 4 mg/dl with a correlation ofr 0.91.Despite only 1 patient having a clinicallysymptomatic episode of hypoglycemia,10 had nocturnal glucose 55 mg/dl,with mean duration of hypoglycemia of220 min, 5 of whom had glucose 40mg/dl for a mean of 188 min.

HypoglycemiaA number of studies assessed aspects ofhypoglycemia in patients with diabetes.Sood et al. (abstract 270) studied 14healthy subjects and 8 subjects with type

2 diabetes to assess whether moderateethanol use (mimicked by continuous in-fusion of placebo or ethanol) delays re-covery from hypoglycemia in people withdiabetes compared with nondiabetic indi-viduals. Insulin was administered to re-duce plasma glucose to 41 mg/dl, withrates of recovery from hypoglycemiaof 33and 34 mg dl1 h1 in the nondiabeticsubjects during the placebo and ethanolstudies, but 26 and 18 mg dl1 h1,respectively, in the patients with diabetes.

Stork et al. (abstract 311) studied 17individuals with type1 diabetes and12con-trol subjects in a state-of-the-art moving-base driving simulator to assess drivingwith blood glucose levels of 5.0 and 2.7mmol/l. At euglycemia, there was more va-riability in lane keeping among the individ-uals with diabetes, whereas this groupreacted to a simulated drive through a built-up area with a reduction in speed. No ad-verse effect of hypoglycemia was observed.

Heptulla et al. (abstract 531) studiedepinephrine, glucagon, and growth hor-mone responses to hypoglycemic insulinclamp studies in five healthy adult men,

showing an increase in peak response ofeach counterregulatory hormone with in-gestion of oral glucose, suggesting po-tentiation by gut factors or by nutriententry into the portal circulation. Segel etal. (abstract 533) compared five insulin-requiring with five oral agenttreatedpatients with type 2 diabetes, showingmarkedly impaired response of epi-nephrine and, particularly, glucagon tohypoglycemia in the former group, sug-gesting that defective glucose counter-regulation, analogous to that occurring in

type 1 diabetes, develops in advanced type2 diabetes.

Davis et al. (abstract 534) treated 16patients with type 2 diabetes for 6 monthswith an intensive regimen using met-formin, glipizide, and acarbose, reducingHbA

1cfrom 11.0 to 6.7%, with hypogly-

cemia occurring at a rate of 0.6 per pa-tient-month. After 2 weeks with completeavoidance of hypoglycemia, there wasgreater suppression of hepatic glucoseproduction during a 2-h hypoglycemicclamp. Muscle sympathetic nerve activity,epinephrine and norepinephrine re-sponses, and symptoms of hypoglycemiawere less pronounced after the improve-ment in glycemic control, suggesting thatthe decrease in counterregulatory re-sponse with improved glycemia requiresneither recurrent hypoglycemia nor insu-

lin treatment. Fanelli et al. (abstract 542)compared 11 patients with type 1 diabe-tes whose glucose was reduced to 42mg/dl over either 30 or 90 min. There wasno difference in symptoms, norepineph-rine, glucagon, or lactate levels, and epi-nephrine levels were greater during theslower reduction in blood glucose, butperformance on psychometric tests dete-riorated more with rapid glucose lower-ing. Spyer et al. (abstract 535) studied thecounterregulatory response to hypoglyce-mia in 8 patients with glucokinase muta-tions, well-controlled type 2 diabeticsubjects, and normal individuals, show-ing a reduction in C-peptide secretion athyperglycemia and higher glucose thresh-olds for glucagon and growth hormonesecretion during hypoglycemia, suggest-ing a roleof the enzymein glucose sensingand counterregulatory hormone release.

Wysocki et al. (abstract 1331) showedno difference in intelligence quotientamong 67 children with type 1 diabeteswith or without a history of severe hypo-glycemia. Intensively treated patientsshowed a 62% greater likelihood of ex-

periencing episodes. Ferguson et al.(abstract 536) showed no effect of thefrequency of hypoglycemia on brainstructure, based on magnetic resonanceimaging, or in cognitive function in 74 in-dividuals with type 1 diabetes. There was,however, an association of both durationof diabetes and microvascular diseasewith cortical atrophy and brain leukoara-iosis, suggesting that while recurrent hy-poglycemia has no demonstrable effect,there may be adverse consequences ofcumulative glycemia. Jacobson et al.




5/6

(abstract 537), however, reported thatamong 16 patients with type 1 diabetes oflonger duration, those with five or moreepisodes of unconsciousness from hypo-glycemia had more abnormal magneticresonance imaging results, were morelikely to have had psychiatric illness, par-ticularly depression, and had lower intel-ligence quotients.

Health Care IssuesReviriego et al. (abstract 922) assessed di-rect and indirect costs of hypoglycemia in100 patients with type 1 diabetes inSpain, showing annual direct costs of $US227, of which $174 was accounted for bythe cost of hospitalization, while indirectcosts from lost work amounted to $120.Patients receiving three or more insulindoses per day showed annual cost of

$281, while those with one or two insulindoses had annual cost of $583, suggestingactual benefit of intensive treatment inthis population. Zheng et al. (abstract780) used statewide emergency roomvisit data to assess determinants of re-peated visits for hypoglycemia. Of 3,150individuals with at least one visit, 480 hadrepeat visits for hypoglycemia; men were57% more likely than women, and Afri-can-Americans 51% were more likelythan Caucasians to have visits. Patientswith Medicare and Medicaid had 76%more visits than those with private insur-ance or self-insurance.

Addressing issues of health care cost,Ettaro et al. (abstract 185) analyzed dataon insurance and the use of health ser-vices for 658 individuals with type 1 dia-betes enrolled in the Pittsburgh Epidemi-ology of Diabetes Complications Studyseen biennially for 10 years between 1986and 1988 and 1996 and 1998. Those pa-tients with health insurance were 77%more likely to inject insulin at least threetimes daily, had lower HbA

1cand albu-

minuria, and had a lower frequency of

neuropathy. Bagust et al. (abstract 892)reported that the lifetime health care costper patient in the U.K. is 2.5 times greaterfor men with diabetes and 1.9 timesgreater for women with diabetes than fortheir nondiabetic counterparts, being 5.9-and 3.8-fold greater for men and women3549 years of age and 1.8-fold more ex-pensive for both sexes for individuals over85 years of age. Benjamin et al. (abstract893) reported that the 19961998 Be-havioral Risk Factor Surveillance Systempopulation-based telephone interview

showed that 20% of individuals with dia-betes were aware of having had an HbA

1c

measurement in the past year, 51% adilated eye exam, and 47% a foot exami-nation and that 37% performed self-monitoring of blood glucose at least daily.

Specialist Versus GeneralistDiabetes Care

A number of studies examined features ofspecialist versus generalist care of patientswith diabetes. Champagne et al. (abstract1357) compared responses to a question-naire about health care in 413 individualswith diabetes who had attended an 8-heducational conference; 49% were treatedby a diabetes specialist vs. 51% by a gen-eral practitioner. The groups reported anaverage of 4.2 and 4.4 visits per year. Ofthe specialists, 41% discussed the pa-

tients diabetes for 15 min versus 17% ofthe general physicians, 83 vs. 54% en-couraged self-care, 95 vs. 82% had or-dered an HbA

1cdetermination, 93 vs.

79% explained the HbA1c

results, 93 vs.63% reviewed home glucose monitoringrecords, and 95 vs. 74% were felt by thepatients to give adequate care.

Zgibor et al. (abstract 186) studiedthe 429 participants in the 10-year fol-low-up exam of the Pittsburgh Epidemi-ology of Diabetes Complications Study,reporting that there were significant de-creases in neuropathy, nephropathy, andcoronary artery disease in those individu-als cared for by a board-certified endocri-nologist, diabetologist, or diabetes clinic,controlling for duration, sex, income,health care practices, and known physio-logic risk factors. Philis-Tsimikas et al.(abstract 190) compared 64 patients withdiabetes in a managed care insurance pro-gram treated in a nurse-managed groupwith 59 patients treated by primary careproviders. All the patients in the nurse-managed group received instruction onself-monitoring of blood glucose and foot

exams. HbA1c and cholesterol decreasedfrom 8.2 to 7.7% and from 194 to 173mg/dl, respectively, in the nurse-man-aged group while increasing from 8.3 to8.6% and from 210 to 224 mg/dl withprimary care providers. Emergency roomand hospitalization costs were 72 and45% lower, but laboratory, medication,and nursing costs were 52, 3, and 300%greater, for a 31% higher total cost in theprimary care group. Ziemer et al.(abstract191) compared 1,925 patients treated in adiabetes clinic staffed by nurse providers

and endocrinologists with 353 patients ina medical clinic staffed by general medi-cine residents and faculty. Despite similardemographics, use of insulin was morefrequent, treatment was more likely to beintensified when glycemic control waspoor, and mean HbA

1cwas 7.6 vs. 8.5%.

Aubert et al. (abstract 192) compared 79patients assigned to a nurse care managerwith 76 receiving usual care in a managedcare organization, showing decreases inHbA

1cfrom 9.5 to 8.0% vs. 9.3 to 8.7%,

with similarly greater fall in fasting glu-cose, without evidence of adverse events.

One approach to improving access tonurse or physician specialists involves thenovel use of telephone-delivered care.Edelman et al. (abstract 421) compared14 patients instructed to perform homeglucose monitoring without further inter-

vention with 11 patients who performedhome glucose monitoring and measuredfructosamine at home using a fingerstickmethod, with weekly medication adjust-ment by telephone for fructosamine350 mol/l. HbA

1cdecreased from 9.4

to 9.1% at 3 months without additionaladvice, but from 8.2 to 8.0% with weeklytelephone adjustment. Allen et al. (ab-stract 889) traced costs and outcomes fortelephone care as an adjunct to office vis-its by three pediatric endocrine nursesand three pediatric endocrinologists in1998. Nurses spent 12.1 h weekly on tele-phone calls and 9.7 h documenting thecalls, and physicians spent 6.4 and 6.1 h.The total weekly personnel cost for tele-phone care was $1,002 for the 355 pa-tients with diabetes, or $202 per patientper year, with evidence of a decrease inhospitalizations and emergency room vis-its. Halvorson et al. (abstract 906) re-ported on 9,892 telephone calls to thenursing team using a 24-h new-patientpager, a diabetes hotline, and voicemail/pagers, lasting 7 min each, ata total cost of$59,352 in 1999. Hospitalizations and

emergency room visits decreased becauseof this service, providing further rationalethat telephone consultation be reim-bursed. Sone et al. (abstract381) reportedresults of the Japan Diabetes Complica-tions Study of 2,547 patients in 40 in-stitutions randomized to telephoneinstruction and other information regard-ing diet andexercise given mainly by pub-lic health nurses or a conventionaltherapy group. HbA

1clevels were 7.8 and

7.9% in the intervention and controlgroups at baseline, 7.7 and 7.8% at 1 year,

Bloomgarden



6/6

7.6 and 7.8% at 2 years, and 7.5 and 7.7%at 3 years, with the latter two differencesbeing statistically significant.

References

1. DECODE Study Group: Glucose toler-ance and mortality: comparison of WHO

and American Diabetes Association diag-nostic criteria. Lancet 354:617621, 1999

2. Singer DE,Nathan DM,Anderson KM,Wil-son PWF, Evans JC: Association of HbA1cwith prevalent cardiovascular disease inthe original cohort of the FraminghamHeart Study. Diabetes 41:202208, 1992

3. Folsom AR, Kushi LH, Anderson KE,Mink PJ, Olson JE, Hong CP, Sellers TA,Lazovich D, Prineas RJ: Associations ofgeneral and abdominal obesity with mul-tiple healthoutcomes in older women: theIowa Womens HealthStudy.Arch IntMed160:21172128, 2000



Documents

ADA Scientific 2000