Vegetarianismo

Editorial
See corresponding article on page 943.

Is ‘‘vegetarianism’’ a serious risk factor for osteoporotic fracture?1,2
Susan A Lanham-New
In the West, there are now appreciable numbers of individuals who are classified as ‘‘vegetarian’’ (those who exclude meat, fish, and poultry) or ‘‘vegan’’ (those who exclude all foods of animal origin). A recent 2009 survey suggests that ’3.2% of USadults adhere to a vegetarian-based diet and ’0.5% of US adults consume no animal products at all (1). Similar 2009 figures have also been published for the United Kingdom (’3% are completely vegetarian) (2). Concomitantly, there has been considerable interest as to the health benefits and health-adverse effects of following such a dietary pattern. We have many more questions than answers, andcertainly the debate as to whether ‘‘vegetarianism’’ increases an individual’s risk of osteoporosis over the long term has been raging for well over a quarter of a century (3). From a public health nutrition perspective, it is critical to address whether adhering to particular dietary habits puts an individual at an increased or decreased risk of disease outcome. Given that we are now in an epidemic ofosteoporosis, with .10 million Americans affected and with estimated costs in the United States and Europe rising above $17.9 billion and V13.9 billion annually, respectively, we need conclusive evidence on how exogenous (modifiable) factors can significantly improve (or harm) bone health at the population level (4). In this issue of the Journal, Ho-Pham et al (5) report the findings of a Bayesianmeta-analysis that examines the effect of vegetarian diets on bone mineral density. This is a most timely and important piece of work. Results included 2749 individuals (ratio of females to males: 2:1) and showed that, overall, bone density was lower in those subjects who adhered to a vegetarian/vegan diet than in those who consumed an omnivorous one but at a level that is unlikely to be clinicallyrelevant. The particular strengths of this study are the careful selection of studies for inclusion in the analysis and the rigorous methodology of Bayesian-type meta-analysis. In particular, Bayesian analysis considers the probability of the hypothesis of treatment effect and is not reliant on P values but instead allows the reporting of direct probability statements that are of interest and ofimportance. That said, this study does not provide the ‘‘conclusive’’ evidence that pubic health specialists require. The numbers of subjects are relatively small given the number of vegetarians worldwide; the study design of all but one of the studies is cross-sectional rather than longitudinal/prospective; and although the quality of the studies selected is in one way a strength, this meta-analysisis not fully representative of the many studies published in this area. The results point to a significant (albeit very small) difference in bone density in those who adhere to a vegetarian/vegan lifestyle compared with those who adhere to a mixed, omnivorous one, but it is important to note that the results do not fully adjust for key confounding factors, such as for differences in 1) bodyweight, 2) physical activity levels, and 3) smoking, as well as for differences in the considerable genetic-ethnic backgrounds in the population studied (Asian compared with white). Indeed, several of the studies on vegetarianism and bone health published before 1984 (not included in this meta-analysis) were based on Seventh Day Adventists who had a significantly different lifestyle compared with thosewho follow an omnivorous diet (6). In this Bayesian meta-analysis, in .50% of the articles included, body weight was significantly lower in the vegetarian group compared with the omnivorous group, and it is well established that body weight is a key determinant of bone mineral density. It is also important to point out that, in the article by Ellis et al (1972), which is quoted in the study but...