While the limitations of single alcohol measures have been discussed within the literature, much of the focus has been directed towards the need to disaggregate heterogeneous non‐drinkers 8, 9 and infrequent drinkers 9 reliably, owing to their disparate risks of assorted health conditions 4, 10. Such discussions overlook the risk of misclassification error among current drinkers whose alcohol consumption changes as a function of age 7. At least three studies have reported the longitudinal stability of intake within baseline‐defined categories of drinking, with each having pooled heterogeneous groups of infrequent and non‐drinkers and modelled changes as a function of follow‐up time 11, 12, 13. An understanding of how alcohol consumption varies within a broad spectrum of disparate drinking groups is therefore limited, especially within a life‐course context.
