Authors’ reply to Colquhoun and Buchinsky

Colquhoun makes an important comment about reverse causality in our study of active commuting.12 We excluded people with reported prevalent disease at baseline but cannot exclude the possibility that including those with subclinical or undiagnosed disease at baseline could have influenced the findings. In the manuscript we originally submitted we simply considered two groups of commuters—active and non-active. We tried to minimise the potential contribution of reverse causality by undertaking a sensitivity …