OR10-6 Sex Differences in Body Composition and Incident Diabetes Across BMI Categories in Older Versus Younger Adults

Abstract

Abstract Background Body composition often changes with age. While body mass index (BMI) is used to define obesity, the degree to which BMI accurately or consistently reflects fatness in older versus younger adults remains unclear. Importantly, it is uncertain whether diabetes risk is similar for older versus younger adults within BMI-defined categories. Objectives We evaluated the relationship of older (≥65 years) versus younger age (<65 years) to dual-energy x-ray absorptiometry (DXA)-derived body composition measures and incident diabetes, stratified by BMI-defined categories for non-overweight/non-obese (<25 kg/m2), overweight (25-29.9 kg/m2) and obesity (≥30 kg/m2). Design: Participants without diabetes at first DXA visit in the Baltimore Longitudinal Study of Aging were included. Linear regression models were used to investigate the association of age categories to body composition at baseline, and Cox proportional hazard models were used to study the relationship of age categories to incident diabetes, within BMI categories for men and women. Diabetes was defined based on self-reported history, medication use, and/or ADA diagnostic criteria (fasting or 2-hour OGTT glucose levels). Results A total of 993 men (n=463 older, n=530 younger) and 1109 women (n=385 older, n=724 younger) were examined. Comparing older versus younger participants, body weight was significantly lower across all BMI categories for both men and women (all p<0.05). Percent total fat mass (%TFM) and fat mass/lean mass ratio (FM/LM) were significantly higher in older versus younger men across all BMI categories and in non-overweight/non-obese women. Among men, in multiple regression analyses adjusting for race, education, height and weight, the relationship (β±SE) of older versus younger age (reference) to %TFM remained significant (non-overweight/non-obese: 1.84±0.59, overweight: 2.60±0.47, obese: 1.81± 0.87; all p<0.05), as did the relationship to FM/LM, across all BMI categories. Among women, the relationships of older versus younger age (reference) to %TFM and FM/LM remained significant for non-overweight/non-obese women. Over a median follow-up of 7.3 years (IQR=1.9-16.6), there were 165 incident cases of diabetes. Older obese men had an almost three-fold higher risk of developing diabetes than younger obese men (HR=2.72 [95%CI: 1.12, 6.63]), accounting for race, education, height, weight and %TFM. Among women, no differences by age group in incident diabetes were observed. Conclusions In our study, sex differences were observed in the relationship of older versus younger age to several measures of body composition, suggesting greater relative fat mass with aging, particularly in men. Further, our results suggest that BMI-defined categories for overweight and obesity do not fully account for changes in fat mass with aging and that the risk for diabetes is significantly greater for older compared to younger obese men. Further studies are needed to investigate the optimal BMI thresholds that predict diabetes risk in older adults. Presentation: Sunday, June 12, 2022 12:15 p.m. - 12:30 p.m.