Objectives: The aim was to examine the associations between level of physical activity (PA) and non-hospital medical costs, and between physical activity and hospitalisations in older women from 1999 to 2013.
Design: Longitudinal observational study.
Methods: Data were collected from participants in the Australian Longitudinal Study on Women's Health, who completed surveys in 1999 (aged 73–78 years), 2002, 2005, 2008 and 2011. Annual cost data (from the Medicare Benefits Schedule) were available for 1999–2013 and hospital admissions data were available for 2002–2010. Costs were expressed in 2013 Australian dollars (AUD). Prospective associations between self-reported physical activity (categorised as inactive, low, moderate or high) and costs/admissions were examined using quantile regression (for costs) and logistic regression fitted with generalised estimating equations (for hospitalisation).
Results: Median annual costs were AUD122 (95% confidence interval [CI] = 199, 45), AUD284 (CI = 363, 204) and AUD316 (CI = 385, 247) lower in low, moderate and highly active women, respectively, than in those who were inactive [AUD1890 (interquartile range = 1107–3296)]. Odds of hospitalisation were also lower in the low (odds ratio [OR] = 0.88, CI = 0.80–0.96), moderate (OR = 0.77, CI = 0.70–0.85) and highly active (OR = 0.78, CI = 0.71–0.85) women, than in the inactive group.
Conclusions: In inactive older Australian women, a small increase in physical activity may be sufficient to obtain substantial cost savings for the health system and to reduce hospital admissions.