Re: Nutritional and socioeconomic factors in relation to prostate cancer mortality: a cross-national study.

In the discussion section of the recent ecologic mortality study of prostate cancer by Hebert et al. (1), several misleading statements are made concerning the cohort study of dietary and other lifestyle factors and prostate cancer among Seventh-day Adventist men that was published nearly 10 years ago (2). Specifically, Hebert et al. state that, in regard to the dietary findings presented in that study, “These results need to be interpreted cautiously because the authors do not provide details regarding their dietary assessment methodology.” However, in a companion paper published in the same issue as the prostate cancer report, details were provided for the methodology of the entire cohort study, including the details of the dietary assessment methodology (3). In addition, the development of the dietary assessment technique used in the study was described years earlier, at the inception of the cohort study(4). Hebert et al. add that “their results for intake of other foods (e.g. beef, poultry, and milk) do not agree with the majority of the prostate cancer literature.” This statement is untrue. Not only did our results suggest a role for moderately increased prostate cancer risk in association with increasing animal fat intake (multivariate adjusted relative risk for daily meat intake 4 1.41; 95% confidence interval 4 0.79–2.51), it was one of the first studies to identify a protective relationship between intake of several types of vegetables, notably beans, lentils, peas, raisins, and tomatoes and prostate cancer risk. The findings regarding tomato intake have been supported in several studies, particularly in other welldesigned and conducted prospective cohort studies of prostate cancer incidence (5). Finally, Hebert et al. also erroneously state that “The results of that study may be unique because their sample comprised a homogeneous group of predominantly white middle-class vegetarian men.” Indeed, the Adventist cohort was characterized by a wide degree of variation in socioeconomic status [ ee Table 7 in(3) for the distribution of education and occupation variables in male members of the cohort]. Additionally, dietary habits ranged from the pure vegetarian (i.e., the vegans) to the “omnivorous” diet in which several kinds of animal products are consumed to one degree or another. The Seventh-day Adventist church recommends, but does not require, church members to be vegetarian. An inspection of the personyears distributions of the meat-intake variables in the tables of our paper supports this wide range of dietary habits. I hope this clarifies the interpretation of the findings from the Adventist Health Study with regard to diet and prostate cancer risk. The results from any international ecologic mortality study of a disease, such as prostate cancer that relies on food disappearance data as the “exposure variable,” suffers from a multitude of methodologic limitations, which the authors have correctly noted. In addition, prostate cancer mortality as an outcome is influenced by several predictive factors, including screening practices, survival, and lifestyle as much as by biologic factors that