The problem is that how well anyone's particular life circumstances fit into a statistical model (or not) can be a very difficult thing to discern. There is a combinatorial explosion of factors that everyone is subject to, and most epidemiological statistics can address and account for a handful of them at a time. So what ends up happening is folks look at the statistics, shrug, and continue their own n=1. The best they can do is be honest with themselves about "how is that working out for you?"