Glucose tolerance in rural diabetic Thais, first-degree relatives and non-diabetic controls

To determine whether non-insulin-dependent diabetes mellitus (NIDDM) in a rural Thai population is characterised by insulin resistance and hyperinsulinaemia, 17 unselected diabetic outpatients from a regional hospital, five first-degree relatives and 10 healthy controls were studied. Subjects in these groups were matched as closely as possible for age and sex, and mean body mass indices were similar (mean +/- S.D.; 21.8 +/- 5.5, 20.6 +/- 1.4 and 21.8 +/- 2.3 kg/m2, respectively, P > 0.5). Beta-cell function (%B) and insulin sensitivity (%S), expressed relative to values for non-diabetic Caucasians, were assessed mathematically using the 'CIGMA' model and plasma glucose and insulin achieved after a standard 1-h glucose infusion. The diabetic patients had higher fasting plasma glucose concentrations than the controls (8.6 +/- 4.0 vs. 4.6 +/- 0.4 mmol/l, P < 0.01) but plasma insulin levels were comparable (geometric mean [-S.D.-+S.D.]; 4.0 [1.7-9.4] vs. 4.0 [1.7-9.2] mU/l, P > 0.1). %B in the diabetic group (21% [10-41]) was lower than in the controls (128% [88-187], P < 0.001) while %S tended to be higher (185% [86-400] vs. 111% [49-251], 0.1 > P > 0.05). Relatives had intermediate values of both variables. %S and %B correlated poorly in the diabetic group (P > 0.1) but together accounted for 90% of the variation in basal plasma glucose (multiple r = 0.95, n = 17, P < 0.0001). Beta-cell dysfunction appears the primary defect in diabetic patients from a Thai subsistence farming population. Insulin resistance may not always characterise NIDDM in geographical areas where a 'thrifty genotype' would be expected; other factors associated with diabetes in developing countries (such increased susceptibility to serious infections) may also influence diabetes prevalence.