Baby boomers and booze: we should be worried about how older Australians are drinking

In reply: We acknowledge the importance of statistical precision and appropriateness, especially when dealing with large samples, as highlighted by Roselli in his comment on our report on alcohol use by older people.1 In our study, we used weighted Ns to correct for under‐represented groups in the sample. The use of relative weighting procedures for this purpose is well established. However, we appreciate that use of absolute weightings can result in an artificially large sample size.

We note that in conducting work similar to our research, using an independent data source from a single jurisdiction (Queensland), Roselli identified findings broadly consistent with our own. As with other similar studies overseas, these findings provide further support for our central observations regarding alcohol use trends among older people. We have subsequently re‐examined our data using non‐weighted Ns (Box) and found significant decreases between 2004 and 2016 for low risk drinkers, and significant increases in the same period for high risk drinkers among Australians aged 50 years or more. When risky and high risk drinkers were combined, this group also increased significantly over time (Box). These findings and other similar reported studies offer confidence that the conclusions of our original study remain valid, correct and important.