early presentation of type 2 diabetes in mexican - diabetes care

E p i d e m i o I o g y / H e a 1 1 h S e r v i c e s / P s y c h o s o c i a I R e s e a r c hO R I G I N A L A R T I C L E

Early Presentation of Type 2 Diabetes inMexican-American YouthNAOMI D. NEUFELD, MD

LESLIE J. RAFFEL, MD

CHRIS LANDON, MD

Y.-D. IDA CHEN, PHD

CONSTANCE M. VADHEIM, PHD

OBJECTIVE — To describe features of pediatric-onset type 2 diabetes in the Hispanic pop-ulation.

RESEARCH DESIGN AND METHODS— The medical records of 55 Hispanic subjectswith diabetes who were treated from 1990 to 1994 in a pediatric clinic serving lower incomeMexican-Americans were reviewed to assess the frequency and clinical features of type 2 dia-betes. Additionally, nondiabetic siblings of several patients underwent oral glucose tolerancetesting, and a survey of six high schools in the same county was performed.

RESULTS — Seventeen of 55 (31%) of the diabetic children and adolescents had type 2 dia-betes. An additional 4 Hispanic children with type 2 diabetes treated in other clinics were alsoidentified, yielding a total of 21 subjects who were used to describe the characteristics of child-hood type 2 diabetes. At presentation, all were obese (mean BMI 32.9 ± 6.2 kg/m2), 62% hadno ketonuria, and fasting C-peptide levels were elevated (4.28 ± 3.43 ng/ml). Diabetes was eas-ily controlled with diet, sulfonylureas, or low-dose insulin. No autoantibodies were present inthose tested, and family histories were positive for type 2 diabetes. Compliance was poor, and3 subjects developed diabetic complications. Of the tested siblings, 2 of 8 had impaired glu-cose tolerance and 5 of 8 had stimulated hyperinsulinemia, correlated with BMI (r = 0.80, P< 0.05). The school survey identified 28 diabetic adolescents, 75% more than expected (P <0.01). The Hispanic enrollment at each school was highly correlated with the number of dia-betic students (r = 0.87, P = 0.011).

CONCLUSIONS — Genetic susceptibility to type 2 diabetes, when coupled with obesity,can produce type 2 diabetes in Mexican-American children. This diagnosis should be consid-ered in young Hispanic patients, who might otherwise be assumed to have type 1 diabetes, andalso when caring for overweight Hispanic youth with a family history of type 2 diabetes, inwhom intervention may prevent or delay diabetes onset.

Diabetes is a chronic disorder associ-ated with significant morbidity andmortality in the U.S. Hispanic popu-

lation. In Hispanic adults, as in most adultpopulations, diabetes is primarily type 2,and its incidence is correlated with theoccurrence of obesity (1,2). Hispanicadults, in general, have an increased risk fordiabetes compared with non-HispanicCaucasians (3). In children, however, most

diabetes is type 1, and the prevalence oftype 1 diabetes among Mexican-Americanchildren is low compared with that in non-Hispanic Caucasian children (4).

Type 2 diabetes has generally beenthought to be rare in children and esti-mated to account for <5% of all childhooddiabetes (5). Recently, however, there havebeen reports that type 2 diabetes is increas-ing in adolescents, particularly in African-

From the Pediatric Diagnostic Center (N.D.N., C.L.), Ventura County Medical Center, Ventura; AhmansonPediatrics Department (N.D.N., L.J.R.) and Medical Genetics-Birth Defects Center (L.J.R.), Cedars-Sinai Med-ical Center, Los Angeles; Department of Medicine (Y.-D.I.C), Stanford University, Stanford; and Departmentof Pediatrics and the UCLA Center for Vaccine Research (C.M.V), Harbor-UCLA Medical Center, Torrance,California.

Address correspondence and reprint requests to Naomi D. Neufeld, MD, 8733 Beverly Blvd., Suite 200,Los Angeles, CA 90048.

Received for publication 10 March 1997 and accepted in revised form 24 September 1997.Abbreviations: IGT, impaired glucose tolerance; MODY, maturity-onset diabetes of the young; NDDG,

National Diabetes Data Group; OGTT, oral glucose tolerance test.

Americans, and is associated with growingrates of childhood obesity (6-8). We reportour experience with Mexican-Americanchildren and adolescents with type 2 dia-betes, suggesting that the prevalence ofchildhood-onset type 2 diabetes may bemuch higher than previously suspected inHispanic children as well.

RESEARCH DESIGN ANDMETHODS— The medical records ofdiabetic Mexican-American children <17years of age who presented to the PediatricEndocrinology Service at the Pediatric Diag-nostic Clinic in Ventura, California between1990 and 1994 were reviewed. This clinic isa specialty care center, developed as part ofa comprehensive primary and specialty carenetwork in Ventura County, California, thatserves the predominantly Hispanic popula-tion of southern Ventura County. The clinicprovides care for children, the majority ofwhom are from uninsured or working poorHispanic families in the county. In 1991,there were 12,670 outpatient visits; the eth-nic makeup of the clinic population is 86%Hispanic, predominantly Mexican-Ameri-can. All of the patients seen by the PediatricEndocrinology Service are referred by theirprimary health care provider in the com-munity, the Ventura County Medical CenterEmergency Room, or by one of the foursatellite clinics affiliated with the PediatricDiagnostic Clinic.

Subjects were classified as having type1 diabetes, type 2 diabetes, secondary dia-betes, or impaired glucose tolerance (IGT)following the National Diabetes Data Group(NDDG) criteria (9). Patients with type 2diabetes met the following criteria: 1) oralglucose tolerance test (OGTT) diagnostic fordiabetes or two random glucose values > 11mmol/1, 2) not ketosis-prone under basalconditions, and 3) required no insulin orwere receiving significantly <0.67 U insulin• kg"1 • day"1 1 year after diagnosis.

The medical records of the Mexican-American children with type 2 diabeteswere reviewed and summarized in 1994and again in 1996. Data were collected onfeatures at presentation (including age,height, weight, plasma glucose and insulin,C-peptide, islet cell antibodies, thyroidantibodies, glycosylated hemoglobin, and

80 DIABETES CARE, VOLUME 21, NUMBER 1, JANUARY 1998

Dow

nloaded from http://diabetesjournals.org/care/article-pdf/21/1/80/585295/21-1-80.pdf by guest on 08 January 2022

Neufeld and Associates

Table 1—Characteristics at diagnosis of Hispanic adolescents and preadolescents with type 2 diabetes

Patient

123456789101112131415161718192021Mean ± SDor ratio

Age(years)

101117161411131116131614171012111616161514

13.7 ±2.4

Sex

MMFMMMFMFFMMFFMFFMFMM

9F/12M

Familyhistory

+NA+

NA+

NANANA+++++++++++++

16 of 16 +

BM1(kg/m*)

44.442.630.027.526.235.744.424.232.635.137.337.930.623.632.028.830.727.638.034.627.5

32.9 ±6.2

Glucose(mg/dl)

337240415608331115110131172250334240288461367266294575173113401

296.2 ± 142

Glycosylatedhemoglobin (%)

1612.46.1—

12.45.15.7

13.68

10.67.47.8

11.110.412.511.610.620

7.36.7

11.3 9.9 ±3.0

C-peptide(ng/ml)

2.8—8—1.2

24———7.42.26.42.82.62.23.4———

13.03.1

4.28 ±3.43

Bicarbonate(meq/1)

20————————————202322————12

18.6 ±5.6

Therapy

GlyburideInsulin, 48 UGlyburideDietDietDietDietGlyburideDietInsulin, 12 UInsulin, 35 UDietGlyburideGlyburideGlyburideGlyburideGlyburideInsulin/dietTolbutamideDietInsulin, 11 V

—

NA, not available.

the presence of ketones in blood and/orurine) and on current therapy (insulin ororal hypoglycemic agents, type of diet),compliance, and disease-related complica-tions. Relevant features from past medicalhistory, including birth weight and mater-nal history of gestational diabetes, werenoted. Family histories of diabetes wereobtained, permitting assessment of thegenetic aspects of this disorder. In additionto the 17 children seen at the PediatricDiagnostic Clinic in Ventura, 4 others wereseen by N.D.N. at clinics in Los AngelesCounty. Information on these individuals isincluded for descriptive purposes; theywere not included in estimating type 2 dia-betes incidence rates.

SiblingsThree-hour OGTTs were performed on 8siblings of several of the patients, afterobtaining informed consent. These siblings(4 girls, 4 boys) ranged in age from 8 to 16years. Subjects were given 1.75 g/kg glucoseto a maximum of 75 g. Samples were ana-lyzed for glucose and insulin at each timepoint, using standard laboratory methods(10,11). Informed consent for studies ofsiblings of affected individuals was obtainedaccording to guidelines established and

monitored by the institutional reviewboards of Cedars-Sinai Medical Center andVentura County Medical Center.

School surveyIn an effort to determine whether an excessprevalence of diabetes existed among His-panic youth in Ventura County, schoolnurses serving six high schools in thecounty were surveyed. They were asked forthe following information: total schoolpopulation, ethnic breakdown of the stu-dent population, and the number of dia-betic students, with treatment whenavailable. The relative excess number ofdiabetic students was determined by com-paring the observed number in the schoolpopulation with that expected based ontype 1 diabetes prevalence estimates of 1 of500 in Caucasian children and 1 of 1,000in Hispanic children (4,12).

Data analysisFor purposes of comparison, means andSDs were calculated. Comparisons to pop-ulation standards were obtained by usingunpaired Students t tests, and changes insubjects during follow-up were evaluatedusing paired Students t tests. Linear regres-sion was performed using the method of

least squares. A P value ^0.05 was consid-ered significant for all tests conducted.

RESULTS

Characterization of patientsA total of 80 patients with diabetes, 25 ofwhom were non-Hispanic, were seen inthe Pediatric Diagnostic Clinic from 1990to 1994. Of the 55 patients who were His-panic, 17 met the criteria for type 2 dia-betes; the other 38 patients had type 1diabetes. Thirty-one of these diabeticpatients were newly diagnosed; of these, 14(45%) had type 2 diabetes.

Summary data for the 17 patients withtype 2 diabetes and for the 4 additionalpatients identified in Los Angeles Countyare shown in Table 1. At the time of pre-sentation, subjects ranged in age from 10 to17 years (mean 13.7 ± 2.4 years). Two weresiblings (patients 5 and 14 in Table 1). Roysand girls were equally represented; 9 of 21were girls and 12 of 21 were boys. Thepatients shared several important clinicalfeatures. At diagnosis, all were obese; theBMIs for 18 of 21 patients exceeded the95th percentile for age and exceeded the90th percentile for age in the other threepatients (13). Mean BMI for the entire

DIABETES CARE, VOLUME 21, NUMBER 1, JANUARY 1998 81

Dow


Type 2 diabetes in Mexican-American youth

Figure 1—Example pedigrees for some of the early-onset type 2 diabetic subjects. Note that histories ojtype 2 diabetes in both maternal and paternal rel-atives is common.

group was greatly increased at 32.9 ± 6.2kg/m2. For comparison, the 95th percentilefor BMI, based on National Health andNutrition Examination Survey (NHANES) 1data, ranges from 19.6 (age 10) to 30.5(age 17) (13). In general, these patients hadbeen obese throughout childhood, and atleast 4 of them had been macrosomic (>4kg) at birth.

The majority of children (18 of 21) pre-sented with histories of polyuria, polydipsia,polyphagia, and fatigue, lasting from a fewdays to several weeks. Significant hypergly-cemia was observed in 18 of 21 patients atthe time of presentation (mean glucose326.8 ± 129.9 mg/dl; range 131-608) whopresented symptomatically. Seven hadketones in their urine, and one did havesevere ketoacidosis (bicarbonate 8 meq/1)on admission. Mean glycosylated hemoglo-bin values (9.9 ± 3.0%; normal range4.0-6.4%) at the time of diagnosis weresignificantly elevated for the majority ofpatients. Three other asymptomatic children(patients 6, 7, and 20) were evaluatedbecause of obesity and a strong family his-tory of diabetes and were found to have dia-betes on OGTT. For most children, glycemic

control was easily achieved with diet therapy(7 of 21), glyburide therapy (9 of 21), ordoses of insulin lower than usually requiredby adolescents with type 1 diabetes (5 of 21)(14). Normal to elevated C-peptide levels(mean 4.28 ± 3.43 ng/ml), suggesting hyper-insulinemia, were present in 10 of 11patients for whom data were available. Anti-bodies were undetectable in the 5 tested forislet cell antibodies and the 8 tested for thy-roid antibodies, suggesting an absence ofsignificant endocrine-related autoimmunityin those patients who were assessed.

Family histories were positive for type2 diabetes for all 15 of the 20 index patients(excluding patient 14, the affected siblingof patient 5) for whom pedigree informa-tion was available (Fig. 1). In 9 of these 15cases, the mother had type 2 diabetesand/or a history of gestational diabetes withsubsequent IGT. In 3 of 15 cases, thefathers were noted to have type 2 diabetes.The maternal grandmother was reported tohave diabetes in 7 of 15 cases. Data werealso available on 25 siblings, of whom 9had diabetes or IGT.

In 1996, medical records of the type 2diabetic subjects were again reviewed.

Short-term follow-up (0.5-6 months) wasavailable for 18 of 21, and longer-term (1-6years) follow-up data were located for 11 of21. As shown in Table 2, shortly after diag-nosis, there was a significant reduction inblood glucose levels in almost all of thepatients (P < 0.001). The lack of a con-comitant reduction in glycosylated hemo-globin might reflect the short period oftime that had elapsed since the institutionof diabetes treatment. For the 11 for whomlonger term follow-up data were available,the duration of follow-up averaged 3.3 ±1.8 years. During follow-up, the degree ofobesity as measured by BMI remainedunchanged (P > 0.8). Treatment over thecourse of follow-up changed for 6 of the 11subjects. Of the 7 who had been on oralagents, 3 progressed to insulin; 2 of the 3who had been on insulin were taken off (1to diet alone, 1 to oral agents), and the boyinitially treated with diet alone was subse-quently prescribed an oral agent. Unfortu-nately, 10 of the subjects were lost tofollow-up less than 1 year after initial diag-nosis. This is likely due to the fact thatmany of the patients served by the VenturaCounty clinics come from migrant families.


Dow


Table 2—Follow-up data on diabetic subjects at 0.5-6 months after initial diagnosis

Patient

1234567811121314

161718192021Mean ± SD

Duration(months)

5

5211124847

0.534426

3.5 ±2.2

BMI(kg/m2)

41.242.626.227.526.233.940.123.337.832.230.031.029.832.331.339.334.024.3

32.4 ±5.9

Glucose(mg/dl)

8910612590

200115

5.111080—

272331

902001041586785

138 ±75.3


—6.1—

12.45.1—

13.56.7——————7.4—5.6

7.7 ±3.3

Therapy

GlyburideInsulinGlyburideDiet and exerciseGlyburideDietDiet and exerciseGlyburideInsulinDietGlyburideGlyburideGlyburideDiet and exerciseDiet and exerciseTolazamideDiet and exerciseInsulin

—

Additionally, the passage of CaliforniaProposition 187 in 1995, which threatenedto deny medical coverage to illegal aliens,resulted in a significant decline in the num-bers of undocumented Hispanic individu-als seeking medical services.

The most disconcerting finding wasthe low level of compliance with therapy;as shown in Table 3, on 6 of 11 charts, spe-cific notation of poor compliance wasmade. Although not reaching statistical sig-nificance, in 4 of the 5 cases where follow-up values were available, glycosylated

hemoglobin levels were higher on follow-up (12.9 ± 3.0%) than initially (9.1 ±3.0%). Six patients were followed for morethan 4 years from initial diagnosis. Ofthese, 2 had been hospitalized for diabetesout of control, and 3 had evidence of acuteor chronic diabetic complications; 1 patientdeveloped an abscess, 1 became hyperten-sive, and 1 developed a third nerve palsy

OGTTs of siblingsStandard 3-h OGTTs were performed on 8nondiabetic siblings whose families agreed


to participate (Table 4). Of the 8 siblings, 5were significantly obese (BMI >95th per-centile for age). However, mean BMI was25.7 ± 6.9, which was significantly lowerthan the mean for our patients (P < 0.015).Data for glucose and insulin are shown inFigs. 2 and 3 and compared with valuesderived from normal children (15).Although none had diabetes, 2 of the 8 sib-lings had IGT by NDDG criteria for chil-dren (Fig. 2) (9).

In addition, abnormalities in insulinresponses were noted. As seen in Fig. 3,basal insulin values were below the 95thpercentile in all but one case. The siblingssegregated into two groups based on theirinsulin responses to a glucose challenge,with 5 of the 8 siblings (including the 2 withIGT) exhibiting clinically significant eleva-tions in stimulated values of insulin whencompared with normal standards. There wasa significant correlation between obesity, asdetermined by BMI, and stimulated hyper-insulinemia (r = 0.80, P < 0.05).

One of the siblings with an abnormalinsulin response (sibling 13-A) developedovert diabetes within 1 year, at the age of 17,and is currently being treated with gly-buride. Her BMI at the time of the OGTTwas 33.9, the highest of the 8 siblings tested.

School survey data: prevalence byschool and relative excessIn an effort to determine whether theoccurrence of type 2 diabetes among His-panic youth was having an impact on theoverall prevalence of diabetes in the ado-lescent population, school nurses were sur-veyed at six high schools affiliated with theVentura County School-based clinic pro-

Table 3—Follow-up data on diabetic subjects at 1-6 years after initial diagnosis

Patient

38101112131416171819Mean ± SD

Age atfollow-up

2114202020201513171718

17.7 ±2.8

Duration(years)

414565521.512

3.3 ±1.8

BMI(kg/m*)

30.023.13835.030.028.0—

40.031.027.640.0

29.3 ±11.2


12.715.89.6

16.310.4——————

12.9 ±3.0

Therapy

MetforminGlyburideMetformin, enalaprilInsulinMetforminInsulinInsulinInsulinGlyburideDiet and exerciseTolazamide

—

Other

Poor compliance, readmission X1Poor complianceHypertension beginning at age 19Poor compliance, 2 admissionsThird nerve palsy at age 20Poor compliance, glucose = 320Poor compliance, ER visit, abscess at age 15Poor complianceGlucose 249Glucose 116Glucose 158

—


Dow



Table 4—Characteristics of siblings who underwent OGTTs

Sibling

1-A2-A3-A10-A10-B10-C10-D13-AMean ± SDor ratio

Sex

MMMMFFFF

4F/4M

Age (years)

1216148

11141516

13.25 ±2.76

BMI (kg/m2)

33.219.118.919.419.831.330.033.9

25.7 ±6.9

OGTT 2-h insulin(uU/ml)

229652611

153214136152

123 ±81.7

gram (Table 5). Using estimates based on atype 1 diabetes prevalence rate of 1/500 inCaucasian children and 1/1,000 in His-panic children (4,12), a total of 16 cases ofdiabetes would have been anticipated.However, our survey identified 28 childrenwith known diabetes, 75% more thanexpected (P < 0.01). The Hispanic enroll-ment at each school was highly correlatedwith both the total number of diabetic stu-dents (r = 0.87, P = 0.011) and the excessover the anticipated numbers of type 1 dia-betic cases (r = 0.94, P = 0.003). Nosignificant correlations were found betweenthe numbers of diabetic students and non-Hispanic enrollment rates.

CONCLUSIONS— In the present sur-vey, 22 young Mexican-American subjects(including 1 sibling who developed dia-betes subsequent to an OGTT demonstrat-ing IGT) who appear to have type 2diabetes have been identified. The conclu-sion that these subjects have type 2 dia-betes rather than type 1 diabetes is basedon several features of this disease: 1) at thetime of diagnosis, all patients in this surveywere significantly obese; 2) the majority ofchildren had positive family histories fortype 2 diabetes; 3) at presentation, most didnot have ketosis; 4) their diabetes was eas-ily controlled with diet, oral agents, or lowdoses (<0.2 U • kg"1 • day"1) of insulin; 5)the 6 patients who were tested had elevatedfasting C-peptide levels, suggesting normalto exaggerated insulin secretion; and 6)none of the patients who were tested haddemonstrable autoimmunity against pan-creatic islet or thyroid tissue.

These findings suggest that type 2 dia-betes may be quite common among Mexi-can-American young people, even thoughthis condition is considered to be an adult-

onset disorder. Although our data do notallow precise estimations of incidence orprevalence, type 2 diabetes accounted for~45% of the newly diagnosed cases of dia-betes and 31% of all diabetic children fol-lowed in the Pediatric Diagnostic Clinic inVentura from 1990 to 1994. Based on ourexperience in other clinics in Southern Cal-ifornia, type 2 diabetes appears to accountfor 10-15% of all currently followed dia-betic children and may be responsible foras many as 25% of Hispanic children withdiabetes. These estimates are similar tothose recently reported by Glaser et al. (16)from San Diego.

It is possible that some of the subjectswe report on represent cases of maturity-onset diabetes of the young (MODY), theautosomal dominant form of type 2 dia-betes. As shown in Fig. 1, however, not allof the pedigrees are consistent with autoso-mal dominant inheritance. Additionally, in

some families, the proband is the only indi-vidual in the family in whom diabetes wasdiagnosed at <25 years of age, and type 2diabetes frequently occurs in both maternaland paternal relatives. Because MODY hasbeen estimated to account for only 1-2% ofall diabetes (17) and has not been reportedto occur with any greater frequency in His-panics than in other ethnic groups, it islikely that many of the subjects described inthis report have early-onset "classical" type2 diabetes rather than MODY.

Hispanic children now account formore than half of the school-age popula-tion in Southern California; thus, all physi-cians who care for children need to beaware of the characteristics of this form ofdiabetes (18). Mexican-American childrenappear susceptible to the development ofobesity, which may reflect both genetic pre-disposition and cultural factors, such asmaternal feeding practices, nutritionknowledge, and values (19). Indeed, inHispanic adolescents in Ventura County,we found that close to one-third have BMIs^ 2 5 , although severe obesity (BMI >30)was less common (11%) (20). Our experi-ence with the children described in thisreport suggests that family history of type 2diabetes, when combined with the earlydevelopment of obesity, may result in child-hood development of type 2 diabetes, espe-cially in Hispanic populations. Thepresence of hyperinsulinemia in some (butnot all) of the siblings, even ones who havenormal glucose tolerance, suggests thatthese children come from families with apredisposition to insulin resistance. Whenthe insulin resistance associated with obe-

3 hour OGTT

18Dmln

Figure 2—Glucose values from 3-h OGTTs of siblings. The shaded areas represent data inclusive ofthe 95th percentile for normal children, ages 1-14 years, for purposes of comparison (13).


Dow



3 hr OGTT

180mh

Figure 3—Insulin values from 3-h OGTTs of siblings. Data show segregation into two groups; somesiblings have clearly normal responses and some have responses that fall outside the upper limits of nor-mal at more than one time point.

sity is added to this inherited predisposi-tion, the combination may lead to the earlydevelopment of overt diabetes in theseyoung people (21).

Because type 2 diabetes can so fre-quently have few, if any, early symptoms,adult-onset diabetes often remains unrec-ognized for a long period of time. Because ofthis long period of clinically inapparent dis-ease, patients with type 2 diabetes oftendevelop severe complications, includingblindness, macro vascular disease, andnephropathy within a relatively short timeafter diagnosis (22,23). It has been sug-gested that diabetes in Hispanics is moresevere, since diabetes-related complicationsare often identified within 5-10 years of theinitial diagnosis, even sooner than is gener-ally seen in non-Hispanic patients (24,25).Asymptomatic diabetes occurring duringadolescence, as in the patients reportedabove, but going undiagnosed until later

life, may well contribute to the "early" devel-opment of complications. Even during theshort period of follow-up achieved withthese patients, some serious complicationsof type 2 diabetes (hypertension andperipheral neuropathy) have alreadyoccurred in some of these adolescents.Because those individuals who develop dia-betes during adolescence may be exposed tohyperglycemia for decades longer thanthose who develop diabetes during adultlife, it is likely that even more severe com-plications may occur over their life spans, ifthe disease does not first lead to early death.

The fear that severe complicationsand/or early death may be the outcome ofchildhood-onset diabetes is even greatergiven the evidence of poor compliance inthe patients described above. Poor compli-ance is a well-known problem in adoles-cents with type 1 diabetes (26) and is likelyto be an even greater problem in those

Table 5—Prevalence of diabetes among high school students in Ventura County, California

High school

ABCDEFTotal

Totalenrollment

2,6501,7001,7802,3002,000

75011,180

Hispanicenrollment

1,590850890

1,54140

5635,474

Observeddiabetes

834832

28*

Predicteddiabetes

3.52.52.52.73.51.2

15.9

* Predicted values were calculated based on a type 1 diabetes prevalence of 1/500 in Caucasian and 1/1,000in Hispanic children. *P < 0.01 compared with predicted values.

with type 2 diabetes, since noncompliancewill be unlikely to produce immediatesymptoms. Therefore, without aggressiveintervention, a poor prognosis appearslikely for these young people, who have thecombined risks associated with early-onsetdisease, poor compliance, and obesity.

While the recognition of early-onsettype 2 diabetes in Hispanics is of mostimmediate importance for the care ofknown diabetic children, it is also importantfor disease prevention. There is evidencethat some cases of diabetes can be pre-vented by routine exercise and maintenanceof ideal body weight (27). At the very least,weight loss and dietary management havebeen associated with improvement in glu-cose tolerance, often leading to a reducedneed for medication (24,25). Therefore, thediagnosis of diabetes must be entertainedand investigated in obese Hispanic childrenand adolescents, particularly those with afamily history of diabetes. Once the diag-nosis of childhood-onset type 2 diabetes ismade, all siblings, regardless of weight orthe presence of symptoms, should be testedfor diabetes as well.

In summary, we recommend that thediagnosis of type 2 diabetes be consideredwhen treating young Hispanic patientswho, based on age, might otherwise beassumed to have type 1 diabetes and alsowhen caring for overweight Hispanic youthwith a family history of type 2 diabetes, inwhom intervention may prevent or delaydiabetes onset.

Acknowledgments— This study was sup-ported by National Institute of Diabetes andDigestive and Kidney Disease Grant 1-R21-DK-45154-01.

We thank Drs. JoAnn Brasel, W. Paul Lee,and Jerome I. Rotter for thoughtful commentsupon review of this manuscript.

Information from this study was presentedas a poster at the American Diabetes AssociationConference on Obesity and Diabetes, Boston,Massachusetts, 26 August 1994.

References1. Stern MP: Diabetes in Hispanic Americans.

In Diabetes in America. Harris Ml, HammanRF, Eds. Washington, DC, U.S. Govt. Print-ing Office, 1985 (DHHS publ. no. 85-1468)

2. Stern MP, Gaskill SP, Hazuda HP, GardnerLL, HafTner SM: Does obesity explain theexcess prevalence of diabetes among Mexi-can-Americans? Results of the San AntonioHeart Study. Dkihetdoffa. 24:272-277,1983


Dow



3. Stern MP, Haffner SM: Type II diabetes andits complications in Mexican- Americans.Diabetes Metab Rev 6:29-45, 1990

4. Gay EC, Hamman RF, Carosone-Link PJ,Lezotte DC, Cook M, Stroheker R, Klingen-smith G, Chase HP: Colorado IDDM reg-istry: lower incidence of IDDM inHispanics: comparisons of disease charac-teristics and care patterns in bi-ethnic pop-ulation. Diabetes Care 12:701-708, 1989

5. Fajans SS: Scope and heterogeneous natureof MODY. Diabetes Care 13:49-64, 1990

6. Pinhas-Hamiel O, Dolan LM, Daniels SR,Standiford D, Khoury PR, Zeitler P:Increased incidence of non-insulin-depen-dent diabetes mellitus among adolescents.JPediatr 128:608-615,1996

7. Scott CR, Smith JM, Craddock M, PihokerC: Characteristics of youth-onset non-insulin-dependent diabetes mellitus andinsulin-dependent diabetes mellitus atdiagnosis. Pediatrics 100:84-91, 1997

8. Gortmaker SL, Dietz WH, Sobol AM,Wehler CA: Increasing pediatric obesity inthe United States. Am J Dis Child141:535-540, 1987

9. National Diabetes Data Group: Classifica-tion and diagnosis of diabetes mellitus andother categories of glucose intolerance. Dia-betes 28:1039-1057, 1979

10. Kadish AH, Little RL, Sternberg JC: A newand rapid method for the determination ofglucose by measurement of rate of oxygenconsumption. Clin Chem 14:116-131,1968

11. Hales CH, Randle PJ: Immunoassay ofinsulin and insulin-antibody precipitate.Biochem] 88:137-146, 1963

12. Diabetes Epidemiology Research Interna-tional Group: Geographic patterns of child-hood insulin-dependent diabetes mellitus.Diabetes 37:1113-1119,1988

13. Hammer LD, Kraemer HC, Wilson DM,Ritter PL, Dornbusch SM: Standardizedpercentile curves of body mass index forchildren and adolescents. Am J Dis Child145:259-263, 1991

14. Bloch CA, demons PS, Sperling MA:Puberty decreases insulin sensitivity. J Pedi-atr 110:481-487, 1987

15. Guthrie RA, Guthrie DW, Murthy DY, Jack-son RL, Lang J: Standardization of the oralglucose tolerance test and the criteria fordiagnosis of chemical diabetes in children.Metabolism 22:275-282, 1973

16. Glaser N, Araya A, McFeely ME, Jones KL:Non-insulin-dependent diabetes mellitusin childhood (Abstract). J Invest Med43:134A, 1995

17. Todd JA: Transcribing diabetes. Nature384:407-408, 1996

18. United States Census Data, Los AngelesCounty Health Department, 1990

19. Sherman JB, Alexander MA, Dean AH, KimM: Obesity in Mexican-American andAnglo children. Prog in Cardiovasc Nurs10:27-34, 1995

20. Raffel LJ, Neufeld ND, Landon C, VadheimCM: Obesity in Hispanic and non-Hispanicadolescents (Abstract). Clin Res 41:75A,

199321. McCance DR, Pettit DJ, Hanson RL, Jacob-

sson LT, Bennett PH, Knowler WC: Glu-cose, insulin concentrations, and obesity inchildhood and adolescence as predictors ofNIDDM. Diabetologia 37:617-623, 1994

22. Uusitupa M, Siitonen O, Aro A, Pyorala K:Prevalence of coronary heart disease, leftventricular failure, and hypertension inmiddle-aged, newly diagnosed type 2 (non-insulin-dependent) diabetic subjects. Dia-betologia 28:22-27, 1985

23. Harris MI, Klein R, Welborn TA, KnuimanMW: Onset of NIDDM occurs at least 4-7yr before clinical diagnosis. Diabetes Care15:815-819,1992

24. Manson JE, Nathan DM, Krolewski S,Stampfer MJ, Willett WC, Hennekens CH:A prospective study of exercise and theincidence of diabetes among U.S. malephysicians. JAMA 268:63-67, 1992

2 5. Bogardus C, Ravussin E, Robbing DC, WolfeRR, Horton ES, Sims EAH: Effects of physi-cal training and diet therapy on carbohy-drate metabolism in patients with glucoseintolerance and non-insulin-dependent dia-betes mellitus. Diabetes 33:311-318,1984

26. Weissberg-BenchellJ, Glasgow AM, TynanWD, Wirtz P, TurekJ, WardJ: Adolescentdiabetes management and mismanage-ment. Diabetes Care 18:77-82, 1995

27. Helmrich SP, Ragland DR, Leuing RW, Paf-fenbarger RS: Physical activity and reducedoccurrence of NIDDM. N Engl J Med325:147-152,1991


Dow


early presentation of type 2 diabetes in mexican - diabetes care

Documents