Despite these benefits, the data set lacks prospective alcohol consumption data during early adulthood and advanced old age. Based on existing research 7, it is hypothesized that the latter is probably marked by a continued convergence towards lower volumes of consumption. This is intimated by the non‐linear trajectories presented in Fig. 2. As such, longitudinal transitions between baseline categories may be even more pronounced during periods of the life‐course not captured by the Whitehall II study. However, while correlations between intercepts and rates of change were negative, there is a possibility that the convergence of drinking trajectories was a result of regression to the mean and not a consequence of age‐related factors.
