1696-P: Early Predictors of Maternal Metabolic Health—Utility of Regional and Overall Body Composition Measures

Pregnancy is characterized by reduced insulin sensitivity, which can advance to gestational diabetes in at-risk individuals. Identifying early predictors of GDM would help pregnancy outcomes by improving the time to intervention in this group.

Objective: To test body composition measures as potential early predictors of gestational insulin resistance.

Methods: At 19-20 weeks gestation, 32 nulliparous women (aged 27.2± 4.5 yrs) received body composition measures including percent body fat (PBF), fat and lean mass by bioimpedance; abdominal adipose via ultrasound (intra-abdominal, IAAT; subcutaneous, SAT); and skinfolds. Homeostasis assessment model of insulin resistance (HOMA-IR) was calculated from fasting glucose and insulin at 24-26 weeks and log-transformed for normality. Missing data was imputed by linear trend. One individual was excluded due to extreme BMI (>50). RESULTS: In analyses controlled for age and race, pre-pregnancy BMI and most measures of overall and regional body composition were associated with HOMA-IR (BMI: r=.523; PBF: r=.481; IAAT: r=.421; SAT: r=.611; fat mass: r=.569; lean mass: r=.434; p<.05 for all). In a regression model with age, race, and lean mass, fat mass was the only significant predictor of HOMA-IR (Model: R²= .345, p=.022; Fat Mass: Std. β= .501, p=.025; Lean Mass: β= .131, p=.535). Replacing lean mass with IAAT and SAT improved the association with HOMA-IR (Model R²=.559, p=.004), with SAT as the only significant predictor (β =.479, p<.01). Skinfold measures did not improve the model or achieve significance.

Conclusion: Body composition measures, especially SAT, may serve as potential early predictors of gestational insulin resistance. Adopting such early predictors may have a significant impact on maternal health by facilitating the implementation of earlier interventions for metabolic health.