The modelled impact of increases in physical activity: the effect of both increased survival and reduced incidence of disease

Physical activity can affect ‘need’ for healthcare both by reducing the incidence rate of some diseases and by increasing longevity (increasing the time lived at older ages when disease incidence is higher). However, it is common to consider only the first effect, which may overestimate any reduction in need for healthcare. We developed a hybrid micro-simulation lifetable model, which made allowance for both changes in longevity and risk of disease incidence, to estimate the effects of increases in physical activity (all adults meeting guidelines) on measures of healthcare need for diseases for which physical activity is protective. These were compared with estimates made using comparative risk assessment (CRA) methods, which assumed that longevity was fixed. Using the lifetable model, life expectancy increased by 95 days (95% uncertainty intervals: 68–126 days). Estimates of the healthcare need tended to decrease, but the magnitude of the decreases were noticeably smaller than those estimated using CRA methods (e.g. dementia: change in person-years, −0.6%, 95% uncertainty interval −3.7% to +1.6%; change in incident cases, −0.4%, −3.6% to +1.9%; change in person-years (CRA methods), −4.0%, −7.4% to −1.6%). The pattern of results persisted under different scenarios and sensitivity analyses. For most diseases for which physical activity is protective, increases in physical activity are associated with decreases in indices of healthcare need. However, disease onset may be delayed or time lived with disease may increase, such that the decreases in need may be relatively small and less than is sometimes expected. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1007/s10654-017-0235-1) contains supplementary material, which is available to authorized users.