Prevalence of Postprandial Hyperglycemia in Adolescents

Abstract

ment for PI, there was a difference in sex (P ϭ 0.002) in newborns of diabetic mothers with higher leptin values in fe-males (median 11.2 ng/ml, range 8.1– 20.2) than males (median 7.7, range 4 –12). No difference was observed be-tween the absence and presence of mater-nal pre-eclampsia. Reflecting maternal-hyperglycemia influence on fetal growth, offspring of mothers with preexisting di-abetes had higher PI and insulin levels than those of mothers with GDM or con-trol subjects (2.89 Ϯ 0.25 vs. 2.69 Ϯ 0.26 vs. 2.68 Ϯ 0.18, P ϭ 0.01, median 17.3 ␮IU/ml, range 7.1–25 vs. median 4.8, range 2.5– 8.5 vs. median 2.5, range 2.5– 6.2, P ϭ 0.001). Insulin levels correlate strongly to leptin levels independently of PI, but only in the GDM group (P ϭ 0.014). Cord-blood leptin reflects the fetal growth in newborns of diabetic mothers. It appears as a valuable marker of fat mass at birth and allows to quantify even the mild " maternal diabetes effect " on the progeny. Nevertheless, sexual dimor-phism already exists in utero and sex could affect leptin level independently of fat mass. The maternal diabetes effect on fetal leptin is likely to arise from fetal in-sulin overproduction, which subse-quently contributes to fat deposition. References 1. Gross GA, Solenberger T, Philpott T, Hol-comb WL Jr, Landt M: Plasma leptin con-centrations in newborns of diabetic and nondiabetic mothers. Am J Perinatol 15: 243–247, 1998 2. Hytinantti TK, Koistinen HA, Teramo K, Karonen SL, Koivisto VA, Andersson S: Increased fetal leptin in type I diabetes mellitus pregnancies complicated by chronic hypoxia. Diabetologia 43:709 – 713, 2000 3. Wolf HJ, Ebenbichler C-F, Huter O, Bod-ner J, Fö ger B, Patsch JR, Desoye G: Fetal leptin and insulin levels only correlate in large-for-gestational age infants. A population-based study P opulation-based studies determin-ing the prevalence of type 2 diabetes in adolescents are sparse. Therefore, we designed a feasibility population-based study to detect postprandial hyper-glycemia, an early manifestation of diabetes, in adolescents. The study population consisted of all students taking a mandatory sophomore health class at a suburban Chicago high school during the 1998 –1999 school year. Those who returned parental con-sent, student assent, and a parental ques-tionnaire entered the study. Within 90 – 120 min of completing a standardized lunch (ϳ100 g carbohydrate), students had a capillary blood glucose (CBG) exam for acanthosis nigricans (AN), as well as height and weight measurements. Study was approved by the institutional review board and the school board. Of 553 students, 284 (51%) enrolled and 255 (90%) completed the study. The ethnic profile of our sample (46% Cauca-sian, 38% African American, and 6% His-panic) closely paralleled that of the school. The mean age was 15.9 Ϯ 0.5 years; 53% were female; 38% had a first-or second-degree relative with diabetes; 3% had maternal gestational diabetes; and 27% had BMI Ն85th percentile. Mean CBG was 5.1 Ϯ 0.7 mmol/l. Postprandial CBG was significantly associated with BMI Ն85th percentile (P Ͻ 0.01) and a first-degree relative with diabetes (P ϭ 0.001). However, these two factors only accounted for 6% of the vari-ation in CBG (R 2 ϭ 0.06). Prevalence of AN in African Ameri-can, Hispanic, and Caucasian students was 51, 25, and 1%, respectively. AN was significantly associated with a first-or second-degree relative with diabetes (P ϭ 0.004), ethnicity (P Ͻ 0.001), and BMI Ն85th percentile (P Ͻ 0.001). No postprandial CBG was Ն7.8 mmol/l. The absence of undiagnosed hy-perglycemia is consistent with findings of the Third National Health and Nutrition Examination Survey (2) and speaks against population-based studies. Limit-ing screening to high-risk individuals may better improve efficacy and feasibility. Recently Sinha et al. (3) reported the prevalence of impaired glucose tolerance in obese Caucasian and African American adolescents to be 16 and 27%, respec-tively, and undiagnosed diabetes in Afri-can American teens to be 8%. The absence of hyperglycemia in our obese subjects (including 6 Caucasian and 17 African American) may be due to ascertainment bias, given small sample size, or because teens with a postprandial CBG Ͻ7.8 mmol/l may still have abnormal glucose tolerance. Interestingly, our results suggest that a normal but higher CBG is associated with adult risk factors for type 2 diabetes. We propose that subtle abnormalities in glucose homeostasis may present in ado-lescence and then track with the develop-ment of overt abnormalities in adulthood.