An exciting headline on the BBC News website recently:
Alzheimer’s is a devastating illness and anything that can help prevent or delay its onset, or lessen it’s severity, is to be welcomed. The headline, and in part the article that follows, suggest that diet can reduce brain shrinkage in later life and so act protectively against Alzheimer’s. There are a couple of problems here though, that should be quickly apparent if you read the article with a sceptical eye. The first is in the suggestion that the article has anything to do with Alzheimer’s: the research wasn’t conducted on Alzheimer’s sufferers, and participants weren’t followed longitudinally to see if there was a differential incidence of Alzheimer’s developing. So, the research can’t actually tell us anything about Alzheimer’s. To be fair to the article, this is mentioned. In the penultimate paragraph.
A bigger problem, potentially, is with the interpretation of the research itself. The article takes an unambiguous position that a diet high in vitamins and omega 3 fatty acids caused a reduction in brain shrinkage with age. However, the research didn’t find this. Rather, it found a correlation between blood nutrients and brain volume: as a quasi-experiment, even if the test used a difference statistic then the result is essentially correlational. And correlation doesn’t prove causation. There are a number of possible reasons for the results found, given the pre-existing evidence that education and intellectual effort increases brain complexity and volume. Off the top of my head, it may be that people who are well educated tend to have higher brain volumes, and also tend to eat healthier diets. Or people from higher socio-economic groups tend to be both more highly educated, and are more likely to follow (and be able to afford) a healthy diet.
To eliminate these possibilities, you’d hope that the original research controlled for factors including education and socio-economic status. It’s behind a paywall so I can’t check, but if they did then the BBC didn’t think to mention it. The other result reported, that there was a difference in performance on cognitive tasks in a sample of people without clinical deficits, suggests that there’s some relationship between diet and cognitive performance, but without knowing the educational history of the participants it’s impossible to decide whether diet causes differences in performance, as the article suggests; or whether it’s again a matter of better educated people tending to have better diets. The latter is certainly a strong possibility, and you’d hope that the health editor who wrote the article would discuss this.