Results from the post‐hoc logit models indicate that downward trajectories observed among heavier baseline drinkers (men: > 150.0 g/week; women: > 100.0 g/week) were unlikely to have been a consequence of sudden transitions to complete abstention, but of a general reduction in drinking with increasing age. Reasons for this attenuation are likely to be complex, including a response to declining health or a proactive health precaution 24, 25. This longitudinal convergence of drinking trajectories with increasing age suggests that the categorization of drinkers using a single baseline measure may be especially problematic when applied to cohorts of older populations. Specifically, with higher volumes of alcohol consumption associated elsewhere with an increased risk of adverse health conditions 26, 27, the misclassification of former heavy drinkers as moderate consumers may lead to an overestimation of risk among older moderate drinkers. This convergence may explain why reductions in the risk of coronary heart disease 28 or all‐cause mortality 29 at moderate volumes of consumption appear less pronounced within adults who were older at baseline.
