24-Hour Activity Composition is Associated with Lower Fall and Fracture Risk in Older Men.

Physical activity (PA), sedentary behavior (SB), and sleep are each individually associated with falls and fractures, but often are not examined simultaneously. Compositional data analysis examined the combined prospective associations between the proportion of time in PA, SB, and sleep relative to the remaining behaviors with recurrent falls (2+ falls in any year), any fractures, and major osteoporotic fracture (MOF) from tri-annual questionnaires, with adjudication for fractures, in 2918 older men aged 78.9 ± 5.1 years in the Osteoporotic Fractures in Men (MrOS) Study. Accelerometers were worn on the right tricep for seven consecutive 24-hour periods and measured PA (>1.5 METs), SB (≤1.5 METs), and sleep. Generalized Estimating Equations evaluated associations with recurrent falls. Cox proportional hazards regression estimated any incident fracture and MOF risk separately. Over four years of follow-up 1025 (35.2%) experienced recurrent falls; over 10 ± 4 years of follow-up, 669 (22.9%) experienced incident fractures and 370 (12.7%) experienced a MOF. Higher proportions of PA relative to SB and sleep were associated with a lower odds of recurrent falls [Odds Ratio (OR): 0.87, 95% CI: 0.76-0.99]. Higher proportions of SB relative to PA and sleep were associated with a higher odds of recurrent falls (OR: 1.38, 95% CI: 1.06-1.81) and higher risk of any fracture [Hazard Ratio (HR): 1.42, 95% CI: 1.05-1.92]. Higher proportions of sleep relative to PA and SB were associated with a lower risk of fracture (HR: 0.74, 95% CI: 0.54-0.99). No associations of activity composition with MOF were observed. When accounting for the co-dependence of daily activities, higher proportions of SB relative to the proportion of PA and sleep were associated with higher odds of recurrent falls and fracture risk. Results suggest reducing SB (and increasing PA) may lower fall and fracture risk in older men, which could inform future interventions.