Abnormalities of islet B-cell function, insulin action, and fat distribution in women with histories of gestational diabetes: relationship to obesity.

Although obese women with histories of gestational diabetes mellitus (former GDM) are highly predisposed to develop noninsulin-dependent diabetes mellitus (NIDDM), lean former GDM women are less predisposed. To explore reasons for this difference, we performed measures of islet B-cell function and insulin action in eight lean former GDM women [ideal body weight (IBW), 107 +/- 2% (mean +/- SEM)], 11 obese former GDM (IBW, 161 +/- 11%), and 19 normal women subjects who were individually pair-matched to former GDM for % IBW and age. The first phase (0-10 min) insulin secretory response to iv glucose was significantly lower in both lean and obese former GDM compared to that in normal women (3,480 +/- 548% vs. 8,234 +/- 1,337% basal . min and 3,444 +/- 682 vs. 10,251 +/- 2,465). The second phase (10-60 min) insulin response to glucose was also significantly lower in lean former GDM women and tended to be lower in obese former GDM women compared to that in their respective controls. Insulin action was assessed by the insulin sensitivity index (SI) using Bergman's minimal modeling technique. SI values in lean former GDM women were similar to those in their controls (4.42 +/- 1.3 X 10(-4) ml min-1 microU-1 vs. 5.19 +/- 1.2 X 10(-4). In contrast, SI values in obese former GDM women were significantly lower than those in their controls (0.77 +/- 0.28 X 10(-4) vs. 2.04 +/- 0.43 X 10(-4). To assess whether differences in fat distribution and fat cell size were associated with these differences in insulin sensitivity, the waist to thigh circumference ratio, the waist to hip ratio, and abdominal fat cell diameter were measured. All three were significantly greater in the obese former GDM women than in controls. Thus, an abnormal central distribution of adiposity appears to be associated with the insulin action defect in obese former GDM women. We conclude that both lean and obese former GDM women have insulin secretion defects. Although a modest insulin action defect in lean former GDM women may have been missed by this technique, only in the obese former GDM women, who have a higher risk for future NIDDM, was an insulin action defect demonstrable. Thus, impairments of both insulin secretion and insulin action may be necessary to cause a marked predisposition toward NIDDM.