Decreased Insulin Secretion but Not Insulin Sensitivity in Normal Glucose Tolerant Thai Subjects

Reduced insulin secretion and insulin sensitivity has been demonstrated in normal glucose tolerant (NGT) subjects whose 2-h plasma glucose levels after an oral glucose tolerance test were 5.6–7.7 mmol/l compared with those with 2-h plasma glucose levels <5.6 mmol/l (1,2,3). These data are from high-risk ethnic subjects. Whether these data are true in Asians is uncertain.

We studied insulin secretion and insulin sensitivity in 51 NGT and 15 impaired glucose tolerant Thai subjects. Subjects were grouped according to 2-h plasma glucose levels after an oral glucose tolerance test into four groups (Table 1). Insulin sensitivity was determined by euglycemic-hyperinsulinemic clamp and expressed as glucose infusion rate. Insulin secretion was determined by homeostasis model assessment (HOMA) of steady-state β-cell function (%B) from a HOMA2 model (available at http://www.dtu.ox.ac.uk/homa) and adjusted with glucose infusion rate (HOMA%B_adjusted) to obtain the accurate result of insulin secretion. For statistical analysis, ANOVA was used for group comparison, and between-group differences were compared using Bonferroni post hoc analysis.

As shown in Table 1, age, BMI, and waist circumference were significantly different between groups. Glucose infusion rate was also different between groups, but the difference disappeared after adjustment with age, BMI, and waist circumference. HOMA%B_adjusted of gr.IV was significantly lower than those of gr.I (P = 0.003) and gr.II (P = 0.039) but was not different from that of gr.III. The difference of HOMA%B_adjusted between groups could still be demonstrated after adjustment with age (P = 0.005).

This study demonstrated that insulin secretion adjusted for insulin sensitivity in NGT subjects started to decline progressively from 2-h plasma glucose >5.6 mmol/l. This study agrees with others (1,2,3). This is the first study in Asians where the declined β-cell function adjusted for insulin sensitivity is demonstrated in NGT subjects. These findings are in accordance with those from studies of other ethnic populations, including Mexican Americans, African Americans, Hispanics, and Caucasians, indicating that the results are not ethnic specific. It can be hypothesized that these high-normal oral glucose tolerance test subjects may have an increased risk of developing diabetes; therefore, lifestyle modification should be implemented early in this group as in impaired glucose tolerant subjects.