Physical activity, physical function and falls from mid-life to older age in Australian women: patterns and bidirectional associations

To examine the bi-directional prospective relationship between PA and falls, and PA and physical function. 

Data of the 1946-51 born women (n=13,714; mean age: 48 years) from the Australian Longitudinal Study on Women's Health. Women self-reported PA, physical function (Short-Form 36 physical functioning subscale) and falls every 3-year period from 1996 to 2019. Lagged generalised estimating equations using 3-year intervals examined bi-directional associations between PA and falls, and PA and physical function. All associations were adjusted for directed-acyclic graph-informed potential confounders (i.e., Accessibility Remoteness Index of Australia, body mass index, number of health conditions, education and ability to manage income).

Nearly two-thirds of women had at least a fall. Half (50%) of the women reported physical function limitation in at least one survey, and 14% of the women reported physical function limitation in more than three surveys. Only 16% of women maintained the recommended level of PA or above from mid-life to older age. Regarding the bidirectional association between PA and falls, participation in the recommended level of PA (OR_150-300min:0.87, 95%CI 0.79-0.96) or above (OR_≥300min:0.85, 95%CI 0.78-0.93) was associated with a reduced risk of subsequent injurious falls, compared to women with no PA. Compared to women without falls, women who reported non-injurious falls had lower odds of undertaking ≥300 minutes/week of PA three years later (OR:0.84, 95%CI 0.77-0.92). Women who reported injurious falls had lower odds of undertaking subsequent PA.

Considering the bidirectional association between PA and physical function, women who reported no PA or 150 minutes/week had, on average, higher odds of developing physical function limitation in the next three years (OR_0min: 3.3, 95% CI 3.08 to 3.53; OR1_-150min: 1.44, 95% CI 1.36-1.54), compared with women who reached the recommended level. Women who reported ≥300 minutes/week PA had lower odds of developing physical function limitation. In addition, women who reported physical function limitation had higher odds of undertaking PA less than the recommended level (OR_0min:3.32, 95% CI 3.1-3.5; OR1-_150min:1.42, 95% CI, 1.33-1.51), but lower odds of undertaking ≥300 minutes/week of PA.

PA was prospectively associated with a lower risk of non-injurious and injurious falls or developing physical function limitation. Physical function limitation and non-injurious and injurious falls decreased the likelihood of subsequently reporting PA. 

While participation in PA can improve physical function and reduce falls, physical function limitation, falls and fall-related injuries can be a potential barrier to PA. The bidirectional prospective associations found between PA and falls, and PA and physical function highlight the importance of programmes to ameliorate the cycle of physical inactivity and physical function limitation and falls.