We thank Hartwig and Davey Smith for their engagement1 with our article examining the effect of cycle commuting on mental ill-health, using an instrumental variable (IV) approach.2 Hartwig and Davey Smith raise two important points in their letter: the necessity of providing a theoretical justification for the candidate IV and the importance of exploring empirical support for its independence. We fully acknowledge the importance of both points and will use this opportunity to outline our approach to empirically testing the IV before addressing the criticisms of our paper directly. In our study, we employed three distinct analytical approaches, each susceptible to different sources of potential bias. Hartwig and Davey Smith highlight two of these approaches, but it is crucial to note the third: a standard observational study with adjustment of known measured confounders. Our multi-approach strategy enabled us to triangulate findings and therefore bolster confidence in our conclusions, as all three approaches point to a consistent result. ...
Language: en
LA - en SN - 0300-5771 UR - http://dx.doi.org/10.1093/ije/dyae095 ID - ref1 ER -