Effect modification of the association between aerobic physical activity and diabetes-related mortality by race-ethnicity: a population-based prospective study using NHANES III and 1999-2006 NHANES

Purpose: To examine potential effect modification by race-ethnicity of the relationship between physical activity (PA) and diabetes-related mortality risk using a sample of U.S. adults from the Third National Health and Nutrition Examination Survey (NHANES III). Methods: The sample (n=10,717) included adults (≥20 years) who attended the Mobile Examination Center (MEC). An age-standardized PA score (PAS) was calculated from the self-reported frequency and intensity of 12 leisure-time aerobic activities. The PA scores were then grouped into three categories: inactive (PAS = 0), insufficiently active (PAS >0 - Results: There was no interaction between PA and race-ethnicity (p=0.83). Compared to inactive NHW, there was a significantly lower risk for diabetes-related mortality for insufficiently active (Hazard Ratio [HR] 0.51, 95% Confidence Interval [CI] 0.29-0.89) and active (HR 0.57, 95% CI 0.33-1.00) NHW adults. Statistical significance was not achieved for any level of activity for NHB and MA. A significant p-for-trend was revealed only for MA. Conclusions: The results of this study suggest that accumulation of any volume of PA is associated with a significantly lowered risk for diabetes-related mortality only in NHW. A significant dose-response relationship was only observed among MA. However, the interaction between race-ethnicity and PA did not attain statistical significance. Thus, the presence of effect modification could not be determined from the current study, but further investigation is warranted. Given longer follow-up time in the NHANES III cohort, the ability to estimate more precise risk reductions will become plausible.