Vegetarians Less Healthy?

Horrible media coverage of science is back at it again. I feel less and less that I'm making an overarching generalization every time I tell my students to ignore every headline they hear.

This time, the headlines are reading "vegetarians are less health and have a poorer quality of life". Social media reports seem to have some level of rejoicing that finally meat isn't on the chopping block (pun ✓). So far I've seen comments like, "well obviously", "duh", "you need meat". I think to myself, oh, was there a longitudinal trial that closely tracked vegetarians, and all of their lifestyle factors, that I missed? Of course not.

This report (1) comes from a cohort of Austrians, that is actually a subset of a larger study (2). The analysis of the larger study concluded " Our results show that a vegetarian diet is associated with a better health-related behavior, a lower BMI, and a higher SES"." Of course, that study didn't have sexy enough conclusions. However, this newer report concludes "our results showed that a vegetarian diet is associated with poorer health (higher incidences of cancer, allergies, and mental health disorders), a higher need for health care, and poorer quality of life". 

What are we to make of this? First, I'd like to make a request: can we stop pitting vegetarians vs omnivores? Furthermore, can we stop pitting diets against diets - does feeling your diet is really that much more superior fulfill you? --No, you don't need to eat meat to survive from a nutrition standpoint (3), and no, eating meat will not automatically give you cancer. No diet, by name, is inherently better than any other. All diets must be well-planned, and fit within the context of a healthy lifestyle and environment overall before you can fairly pit two diets against each other. 

The latter study that found negative results for vegetarians is far from universally translatable. It was a cross sectional study, meaning none of the data will tell us about causality, it's just finding associations at one time point. In this study, the researchers divided 1320 individuals up into vegetarians, carnivores (I think they meant omnivores) eating a diet rich in fruits/vegs, carnivores eating a diet less rich in meat, and carnivores consuming a diet rich in meat. First, this was all self-reported/identified; second, the study already tells you more about the types of omnivorous diets, and less about vegetarians. Do these vegetarians eat potato chips all day, do they eat a variety of whole plant foods, do they supplement with b12/vitamin D, do they eat soy meats at every meal - we don't know. The reports on this actually should've focused on the differences found between types of omnivorous diets. For all we know, this report shows that people who feel at risk of developing cancer due to family history, or people who have mental health disorders, choose to go vegetarian because they think it might help. Actually, there is evidence to suggest that the adoption of a vegetarian diet tends to follow onset of mental disorders (4).

The participants were grouped based on SES, sex, and age, and matched to one carnivore from each group. Already, from a design standpoint, we have an issue with statistical power. You're looking at the vegetarians, who are 2.2% of the population, then matching them to individuals who are 97.8% of the population, that you've further stratified. The study, for me, would've garnered more power if, instead of matching 1 veg to 1 of each of the other diet groups, they were matched to 2-3 people from each omnivorous diet group. This would reduce the chance for random error/chance associations. Update: Upon looking at this further, I realized that they were using chi-squared but didn't correct for multiple comparisons either. Since they're running chi square tests on 17 different variables, they should've used p values of .05/17, not .05. Once you correct for this, the results really fall apart. 

Overall, the fact that we know nothing about these vegetarian diets makes us able to extrapolate very little from this study. We'd need longitudinal data, with accurate dietary intake, supplementation data, anthropometrics, and lifestyle variables/screenings to really conclude that vegetarians are worse off than their counterparts. The findings are not generalizable to the vegetarian population as a whole, and generally stand in opposition to a wealth of literature suggesting improved health, compared to overall population statistics (5) - if anything, these findings merely warrant further investigation into this population of Austrians, and the need for a potential public health intervention. 

For those who would manipulate these findings into some sort of holy grail that finally shows we need meat, I'm sorry this doesn't prove vegetarianism is wrong. For those vegetarians who think being so makes you impervious to disease, I'm sorry this doesn't prove your diet is inherently better. Well-planned, science-based, nutritionally adequate diets lead to good health. And must be eaten in the context of a healthy lifestyle. Hitting the genetic lottery wouldn't hurt either. I don't know any definitive data suggesting eating vs not eating meat is better than the other. If you choose not to eat meat, do it for reasons beyond nutrition. If you choose to eat meat, why do you feel the need to prove that others should be?

1. http://www.ncbi.nlm.nih.gov/pubmed/24516625

2. http://www.ncbi.nlm.nih.gov/pubmed/24343044

3. http://www.eatright.org/About/Content.aspx?id=8357
4. http://www.ijbnpa.org/content/9/1/67
5. http://www.medscape.com/viewarticle/763435_2