Physical activity, body mass index and waist circumference change, and normal-range glycated hemoglobin on incident...

The purpose of this study was to 1) evaluate whether physical activity has a protective effect on incident diabetes among African Americans across combinations of body mass index (BMI) and waist circumference (WC), 2) evaluate the effect of changes on BMI and WC on incident diabetes, and 3) evaluate the effect of 'normal range' glycated hemoglobin (A1C) on incident diabetes.

Data from the prospective Jackson Heart Study were evaluated, with baseline data assessed between 2001 and 2004 and follow-up data occurring between 2009 and 2013. Physical activity was assessed via a validated questionnaire, with measured BMI, WC and A1C assessed via standard procedures.

The sample included 2,450 adults who did not have evidence of diabetes at the baseline assessment, with 286 incident diabetes cases occurring at the follow-up assessment. Physical activity did not have a protective effect against incident diabetes across different BMI and WC combinations. Notably, BMI change from baseline to follow-up was associated with incident diabetes (HR = 1.08; 95% CI: 1.03-1.13). Further, higher levels of A1C within the 'normal-range' was associated with incident diabetes (HR = 7.51, 95% CI = 2.66-21.25).

Increases in BMI over time and higher A1C within the normal range were associated with incident diabetes. Serial monitoring of BMI, as well as A1C, even among those with a 'normal' A1C, may be warranted by clinicians. Future work evaluating this novel three-way model (physical activity, BMI and WC) should consider utilizing an objective measure of physical activity.