Early Pregnancy Insulin Resistance and Subsequent Gestational Diabetes Mellitus

We conducted two institutional review board–approved, prospective, nested, case-control studies within Massachusetts General Hospital’s Obstetrical Maternal Study (MOMS) comparing subjects with GDM and subjects with abnormal glucose loading test (GLT) results but no GDM with control subjects. Consecutive MOMS participants with singleton gestations between February 2002 and February 2004 who provided fasting blood samples between 16–18 weeks’ gestation, underwent GLT testing at 24–28 weeks’ gestation, and delivered after 34 weeks were eligible. To limit confounding by other etiologies of insulin resistance, women with pregestational diabetes, preeclampsia, or gestational hypertension were excluded. All eligible women who met the criteria for GDM (1) were included. Women with an abnormal 1-h post-GLT glucose level (≥7.8 mmol/l) but no GDM were randomly selected from 115 eligible women. Control subjects (∼2:1) were randomly selected from 1,016 eligible women who had normal 1-h post-GLT glucose levels. The mean age at GLT testing was comparable between case and control subjects (Table 1).

Plasma glucose was measured using standard techniques. Serum insulin was measured using a radioimmunoassay from Linco Research (St. Louis, MO). Clinical data were obtained from a previously validated and prospectively maintained electronic clinical record (2). Analyses were performed with Stata (Stata, College Park, TX). The homeostasis model assessment (HOMA) was used as an index of insulin resistance (3). The population that did not have GDM or an abnormal GLT served as the reference group in all analyses. P < 0.05 was considered statistically significant.

Baseline characteristics (Table 1) demonstrate that fasting glucose, fasting insulin, and HOMA were significantly higher in women who subsequently developed GDM compared with control subjects. Fasting glucose and HOMA were also significantly higher in the group with abnormal GLT and no GDM than in the control subjects.

The risk of GDM was significantly increased for each log-unit increase in HOMA in both univariate analysis (OR 4.9 [95% CI 2.1–11.2], P < 0. 01) and after adjustment for maternal age, race, BMI, gestational age at serum sampling, systolic and diastolic blood pressure, and parity (adjusted OR 4.2 [1.3–13.5], P = 0.02). Similarly, for each millimole increase in fasting glucose, the risk of GDM increased significantly in univariate- (OR 5.9 [2.2–15.6], P < 0.01) and multivariable-adjusted (adjusted OR 4.7 [1.5–14.9], P = 0.01) models. Neither HOMA (adjusted OR 2.5 [0.73–9.1], P = 0.14) nor fasting glucose (adjusted OR 3.4 [0.81–14.0], P = 0.09) was significantly associated with subsequent abnormal GLT with no GDM.

In this prospective study, we found that women in whom GDM was diagnosed at 24–28 weeks of gestation demonstrated higher levels of fasting glucose, fasting insulin, and HOMA at ∼17 weeks in pregnancy compared with women who had normoglycemic pregnancies. The risk of GDM increased significantly with increasing HOMA and fasting glucose levels, independent of other variables that are known to be associated with GDM. While there are prior reports with similar findings (4–7), such studies either examined small numbers of high-risk women (4,5) or used surrogate markers of insulin resistance (6). Similar results have been noted in nonpregnant populations, where HOMA has been shown to be associated with the later development of type 2 diabetes (8,9). Our findings lend support to the hypothesis that increased insulin resistance and increased fasting glucose at 16–18 weeks’ gestation are associated with the subsequent development of overt glucose intolerance later in pregnancy.

This work was supported by grants from the American Kidney Fund (to K.V.S.), the American Diabetes Association (to J.L.E. and R.T.), and the National Institutes of Health (HD39223 to R.T. and RR17376 to M.W.).