Statistics and lies

‘There’s no evidence to suggest that diet and lifestyle make any difference to whether people get cancer or not,’ said the doctor on the radio. Which is true, if you only count top quality randomised double-blind trials on statistically significant groups of people, replicated and verified by independent researchers.
But those are few and far between. That’s partly because of the expense, and partly because it’s very difficult to control all the variables in order to get a meaningful result. ‘Diet and lifestyle’ covers a multitude of sins. But even if you could eliminate all variables but the ones you want to study, we know that there is a ‘half-life’ to medically established facts. In other words, what is demonstrably true now may not be true in twenty years’ time, no matter how carefully the study is set up.
In practice, most of us operate from a set of beliefs - some of which will be logically incompatible - borne of personal experience, anecdotal evidence, and wishful thinking. We quote statistics when it suits what we want to believe. And from that perspective, there is abundant evidence that diet and lifestyle do make a difference. Of course it’s better to eat vegetables rather than cake, to take moderate exercise, to manage stress sensibly, and so on. It’s better, not only because our physical health will improve, but also because it’s good for us to take an active part in maintaining our own wellbeing. It empowers us. We take responsibility, and perhaps that matters more than anything.