A recent analysis of the EPIC-Norfolk cohort by Barnett et al found that retirement was associated with a decline in overall levels of physical activity as decreases in occupational and transport-related activity were not compensated for by increases in household and recreational activity.1 This study adds more accurate assessment of physical activity with a larger sample size than previous studies2 and allows the breakdown of physical activity across different domains, as well as investigation of socioeconomic differences. Although the main finding has been echoed in some previous studies,3 ,4 a 2012 systematic review by the same authors concluded that the impacts of retirement on overall physical activity were unclear.5 This lack of clarity was due to the paucity of accurate activity assessment in included studies, a gap that this study addresses and strengthens our confidence that retirement is indeed a time of declining physical activity. The study raises the possibility that intervening around the time of retirement may be one strategy to deal with low levels of physical activity, which is noted a serious concern worldwide.6

Barnett et al also found greater declines in physical activity for those previously in manual jobs compared with those in non-manual jobs.1 This is a worrying yet familiar trend, …