Body weight classification

I agree with Simone Lemieux and associates1 that further discussion is warranted regarding the strengths and limitations of the 2003 Canadian Guidelines for Body Weight Classification in Adults.2 However, several points in the commentary might mislead health care providers about the interpretation of these guidelines and their use in clinical practice.

First, Lemieux and associates state that the waist circumference cut-off values used in the guidelines “have not yet been validated.” However, these cut-offs for assessing health risk have been validated in several studies, including 2 population-based studies involving 23 000 adults in Canada and the United States.3^,4

Second, Lemieux and associates claim that “people who are in the overweight range while showing low levels of abdominal adipose tissue generally display a risk profile similar to that of nonobese subjects,” thereby suggesting that body mass index (BMI) is not a useful predictor of health risk in the one-third of Canadian adults who are overweight.5 Although it is plausible that a multitiered classification of waist circumference might identify overweight people with a low waist circumference as having the same health risk as non-overweight people,6 there is a lack of consensus on how to stratify waist circumference according to health risk.

Third, Lemieux and associates are concerned that the lowering of the BMI cut-off for “underweight” in the guidelines from 20.0 kg/m² to 18.5 kg/m² might lead to false identification of elderly people or young women in the 18.5–20 kg/m² range as having a normal weight when they may be malnourished or, with young women, have an eating disorder. The potential health risks in adults older than 65 years with a BMI in the 18.5-to-low 20s range was acknowledged in the guidelines,2 and the need for further health assessment was suggested for such individuals. I concur as to the potential of under-recognizing eating disorders in young women with a BMI in the 18.5–20 kg/m² range. However, according to prespecified criteria,7 the diagnosis of anorexia nervosa is considered in people with a BMI less than 17.5 kg/m², which is within the underweight range in the guidelines.

Fourth, Lemieux and associates suggest that, on the basis of the weight classification guidelines, people with a BMI of 25.0–29.9 kg/m² will be labelled as overweight and may be under greater pressure to lose weight. This is not the aim of the guidelines, which clearly state that “the [BMI] cut-off points are not intended as targets for intervention purposes in individuals.” The possibility that misinformed individuals will use the guidelines to initiate weight loss cannot be excluded, particularly since 6% of men and 24% of women with normal weight are attempting weight loss.8 The intent of the guidelines is to increase awareness of body weight classification and to act as a catalyst for people to review their health status with their health care providers.

Finally, Lemieux and associates question the clinical applicability of the weight classification, indicating that “[m]any factors beyond BMI influence health risk” and that “a number of important clinical factors are absent from the report: special considerations for elderly people, ethnic differences, physical activity and diet.” However, the guidelines state that “at the individual level, the [BMI and waist circumference] system should be used as one part of a more comprehensive assessment of health risk,” and there are separate sections on how to interpret BMI and waist circumference in adults older than 65 years, ethnic and racial groups, and physically fit individuals. Table 1 outlines how BMI and waist circumference, which are easy-to-perform bedside measurements, can be used as part of a health assessment in adults.