The Empiricist

A large scale study has identified an increased risk of eating disorders among young vegetarians as among meat-eaters of the same age.

The study, called Project EAT-II: Eating Among Teens, surveyed over 2,500 teens and young adults on their eating habits.

Each participant was questioned about binge eating, whether they felt a loss of control of their eating habits and whether they used any extreme weight-control behaviors.

About 21 percent of teens who had been vegetarians said they used unhealthy weight-control behaviors, compared with 10 percent of teens who had never been vegetarians. Among young adults, more former vegetarians (27 percent) had used such measures than current vegetarians (16 percent) or those who’d never been vegetarians (15 percent), the study found.

On the other hand, they found some health advantages to being vegetarian, such as lower rates of obesity. Interestingly , news sites such as Forbes and The Guardian worked hard to exonerate vegetarianism as the culprit for any of the problems that were found.
Forbes quoted a scientist who had nothing to do with the study, a Dr Katz of Yale University, who explained that the higher rates of eating disorders were “not due to vegetarianism but the other way around: Adolescents struggling to control their diets and weight might opt for vegetarianism among other, less-healthful efforts.”
That could be true, but he is speculating. For instance, it could also be true that a vegetarian diet short-changes the body in certain ways that in the long term can trigger a response such as an eating binge. Why speculate?
The Guardian responded with a vigorous critique of the research methods used. “These findings are noteworthy and should lead to more research, but the study has a few important shortcomings.”
You would almost think that this research touched a nerve of some kind. So what were the shortcomings? In sum (according to the Guardian), they were

• use of self-report
• too broad a definition of vegetarianism
• too narrow a definition of “healthy diet”
• respondents all from one state

None of these are actually shortcomings. For example, the fact that the respondents are all from one state means that there is a possibility of a “state-specific” effect, but this is simply something to keep in mind before leaping to conclusions. And self-report can lead to distorted results but there are well-known ways of minimising such a possibility, and indeed for checking to see if it is the case.

Still, it’s great to see journalists scrutinising scientific findings with such rigour and skepticism. It’s just a pity they don’t do it more often.