This article by Eoin O’Connell reports a study in the American Journal of Clinical Nutrition (the top nutrition journal) that found a correlation between meat consumption and weight gain: The more meat you ate, the more weight you gained over five years. Meat is fattening! reported several newspapers.
Mr. O’Connell did something unusual for a health journalist: He thought for himself. I don’t mean he applied a formulaic criticism (e.g., “correlation does not equal causation”). That’s not thinking, that’s knee-jerking. Mr. O’Connell read the paper. And he noticed an interaction: The correlation between meat consumption and weight gain depended on activity level. The study involved about 400,000 people. The researchers put each person in one of four activity levels: inactive, moderately inactive, moderately active, and active. There was a correlation between red-meat consumption and weight only for the two most active groups (moderately active and active). The original article reported that this interaction was significant:
The relation between red meat and weight gain was also stronger in physically active subjects compared with moderately inactive or inactive subjects (P values for interaction = 0.02)
The obvious implication of this interaction, as Mr. O’Connell says, is that meat caused muscle gain. Weight differences between more-meat and less-meat eaters were due to differences in muscle mass. This puts an entirely different spin on the results. The alternative explanation is quite plausible. I once had a grad student who was a vegetarian. When he was an undergrad, he told me, he and his roommate would go to the weight room and do similar sets. His roommate, who ate meat, rapidly gained muscle; he did not. Of course, meat = animal muscle.
Mr. O’Connell continued to the really interesting part of his article:
Perhaps not so surprisingly, the consideration that muscle is a form of weight gain does not appear in the newspaper articles but much more surprising is the fact that it does not appear in the original journal article either.
The AJCN article has fifty authors. Not one of them, apparently, noticed this all-important point! Nor did the reviewers for this prestigious journal. The article concludes: “Our results are therefore in favor of the public health recommendation to decrease meat consumption for health improvement.” No, they’re not, if the more meat, more muscle explanation is correct.
Most prestigious journal. Fifty authors. Huge expense. Total F-up in the sense that the final conclusion is probably wrong. (To be fair, the paper has plenty of value in other ways.) Congratulations, Mr. O’Connell, for noticing.