In his recent article, writing for Discover, Gary Taubes takes Harvard School of Public Health and UC San Diego researchers to task for impersonating actual scientists and making dietary recommendations.
…every time that these Harvard researchers had claimed that an association observed in their observational trials was a causal relationship—that food or drug X caused disease or health benefit Y—and that this supposed causal relationship had then been tested in experiment, the experiment had failed to confirm the causal interpretation—i.e., the folks from Harvard got it wrong. Not most times, but every time.
His blistering criticism is aimed at the messy job of running an observational study and then confusing correlation with causation. This appears to happen all the time in Nutrition, and which is partly why bad ideas persist (your diet should consist mostly of carbs) and why stupid ideas gain traction (eat more chocolate to lose weight).
The problem, as he sees it, is that those who provide advice like we have seen recently about the dangers of eating red meat, etc… they forgo the critical step of actually testing their hypothesis.
To confound matters further, Taubes asserts these selfsame science impersonators don’t even account for a more likely explanation for the correlation they see in their observational studies: The Compliance Effect.
All in all, it’s a great read and a worthy resource to have at the ready in case someone flings one of the recent recommendations research papers at you purporting to illustrate the dangers of eating meat.
- Gary Taubes (Discover): Chocolate & Red Meat Can Be Bad for Science: Why Many Nutrition Studies Are All Wrong
- Walter Willet, et al (Archives of Internal Medicine): Pseudo-Science: Red Meat Consumption and Mortality
- Beatrice Golomb, et al (Archives of Internal Medicine): Association Between More Frequent Chocolate Consumption and Lower Body Mass Index