WARNING: the interactive features of this website use CSS3, which your browser does not support. To use the full features of this website, please update your browser.
To assess the relationship between physical activity levels and the risk of disease or death, researchers have to follow study participants over a period of time. Prior to this study, it was not known whether the length of that follow-up time makes a difference to the relationships seen. Using UK Biobank data, we investigated the relationships between physical activity and mortality and cardiovascular disease using four different lengths of follow-up. Keeping everything else equal, shorter follow-up times had stronger associations i.e. we might conclude that physical activity was more protective against mortality and cardiovascular disease than if the study was conducted over a shorter follow-up time. One reason for this might be that as the follow-up time increases, people change their behaviours and so their physical activity they reported no longer reflects their actual behaviour. This has the statistical effect of adding in noise, and so the relationship weakens. Another reason that we found some evidence for was that those who had a diagnosis of cardiovascular disease or cancer were more likely to report lower levels of activity. As these individuals are also at a higher risk of a future event, it overestimates the risks of low activity levels, strengthening the overall relationship. In conclusion, we advised future researchers to carefully consider the length of their study follow-up time and their methods to try and isolate the relationship of interest.
We also published a summary blog on the International Journal of Epidemiology's website https://ije-blog.com/2019/10/29/with-great-statistical-power-comes-great-responsibility/#more-2119