Body shape and pants size as surrogate measures of obesity among males in epidemiologic studies.

This study aimed at characterizing anthropometric indicators that can be used as alternatives to measurements for assessing overall obesity over adulthood and abdominal obesity among men. We used data from a population-based case-control study of prostate cancer conducted in Montreal, Canada in 2005-2012. It included men aged ≤ 75 years, 1872 of which were newly diagnosed with prostate cancer, and 1918 others randomly selected from the electoral list. In-person interviews elicited reports of height as well as of weight, pants size and Stunkard's silhouette at 5 time points over adulthood, i.e., for the ages of 20, 40, 50 and 60 years, if applicable, and at the time of interview. Waist and hip circumferences were measured by interviewers following a validated protocol. Analyses were conducted on the overall sample of 3790 subjects, after having confirmed that results did not differ according to disease status. Stunkard's silhouette scale proved to be an easy-to-administer tool that reflects well reported body mass index, either recently or decades in the past among adult males. It was discriminatory enough to classify individuals according to commonly-used obesity categories. We observed that a model including age, reported pants size, silhouette and weight can reasonably predict current abdominal obesity. In conclusion, alternative anthropometric indicators can serve as valuable means to assess overall and abdominal obesity when measurements cannot be envisaged in the context of epidemiological studies.