When suspicions turn to real evidence

Clinical studies sometimes surprise us, upturning conventional wisdom and forcing us to rethink our understanding of a condition. At other times they can disappoint, when promising treatments turn out not to have any real benefits for patients.

A study reported this week (doi:10.1136/bmj.l1255) fits another tradition: supporting and quantifying a link that has long been suspected to exist but for which evidence has proved hard to find. In this case, the …