Longitudinal associations between serum 25-hydroxyvitamin D, physical activity, knee pain and dysfunction and...

To describe the longitudinal associations between physiological falls risk, and between-person and within-person effects of 25-hydroxyvitamin D (25OHD), physical activity (PA), knee pain and dysfunction in community-dwelling older people.

Data for 1053 participants (51% women; mean age 63 ± 7.4 years) studied at baseline, 2.5, 5, and 10 years were analysed. Falls risk (Z-score) was measured using the Physiological Profile Assessment. Knee pain and dysfunction were assessed using the Western Ontario and McMaster Universities Osteoarthritis index (WOMAC). Moderate-to-vigorous PA (MVPA) was measured using accelerometer. Linear mixed-effect regression models, with adjustment for confounders, were used to estimate the association between physiological falls risk and between-person and within-person effects of PA, 25OHD and WOMAC score.

Between-person effects showed that 10-year average physiological falls risk was lower in participants who had a higher 10-year average 25OHD (β = -0.005 per nmol/l, 95% CI: -0.008, -0.002), log-MVPA (β = -0.16 per minute, 95% CI: -0.22, -0.10) and lower mean WOMAC score (β = 0.005 per-unit score, 95% CI: 0.003, 0.01). Within-person effects showed that a higher physiological falls risk at any time-point was associated with higher than average WOMAC score (β = 0.002 per-unit score, 95% CI: 0.0003, 0.004) and lower than average log-MVPA (β = -0.15 per minute, 95% CI: -0.24, -0.06), but not 25OHD, at the same time-point.

Having higher WOMAC global score above an individual's average increases the risk of falling, whereas, increasing one's own MVPA level further reduces their risk of falling. The presence of between-person but not within-person associations for 25OHD suggests the former may be confounded by other factors.