Joint Association of Diabetes and Physical Activity With Falls Among Midlife and Older Adults: 2018 Behavioral Risk Factor Surveillance System.

PURPOSE: We examined joint associations of diabetes and physical activity (PA) with falls. DESIGN: Complex survey (cross-sectional) design using the 2018 Behavioral Risk Factor Surveillance System (Response rate was 53% (landline phones) and 46% (cellular phones)). SETTING: National survey in the U.S. SUBJECTS: Adults ≥45 years who self-reported diabetes status, PA, and falls (n = 295,282; 98.5% of eligible samples; N = 130,103,093) were classified into 4 groups: no diabetes-PA, no diabetes-no PA, diabetes-PA, diabetes-no PA. MEASURES: Self-reported PA, diabetes, falls, and major health characteristics. ANALYSIS: Poisson regression models were used to estimate the association of groups with any (≥1 fall) falls. RESULTS: Compared to no diabetes-active groups, no diabetes-inactive (Relative Risk (RR) = 1.22; 95% CI: 1.18, 1.26), diabetes-active (RR = 1.25; 95% CI: 1.20, 1.30), and diabetes-inactive (RR = 1.46; 95% CI: 1.41, 1.51) groups were more likely to report falls, independent of tested covariates. CONCLUSIONS: Leisure-time PA may mitigate the likelihood of falls in adults with and without diabetes. Our findings could be useful for healthcare providers or clinicians to promote the importance of PA in midlife and older adults who are at risk of falls and/or diabetes. More detailed longitudinal information on objectively-estimated PA and a more frequent fall calendar are warranted to prevent recall bias and temporal ambiguity (causality between PA and falls).