Commentary: Body mass index persists as a sensible beginning to comprehensive risk assessment

Abstract

The axiom that body weight increases with height stimulated Ancel Keys and colleagues to address the public health quandary of the appropriate body weight for health and longevity. Awareness of the limitations of the standard height-weight tables from the life insurance industry established the need for a simple, physical indicator of body size that differentiates body fatness levels with increasing weight of adults in the assessment of risk of mortality. Keys et al. cleverly integrated the resources of physical anthropology, nutrition and physiology and demonstrated that body weight divided by standing height squared (Wt/Ht; kg/m) or the body mass index (BMI), compared with other weight and height ratios, was a valid indicator of adiposity because it increased with greater levels of body fatness among groups of men. This critical finding evolved to the development of specific ranges of BMI to categorize underweight ( 30 kg/m) with gradations (class I, 30_34.9; class II, 35–39.9; and class III >40 kg/m, respectively). Thus BMI was designated a measure of weight status, and serves as a prominent variable in population research to elucidate relationships between adiposity and metabolic variables associated with increased risk of cardiovascular disease, stroke, type 2 diabetes, cancer and mortality. Widespread use of BMI as a measure of adiposity and a predictor of health risk requires an understanding of its limitations for an individual. Despite consensus that BMI accounts for appreciable variance (60–70%) in measured fatness in groups of adults, it is an unreliable indicator of the body composition of an individual. Correlation coefficients relating BMI to fatness and fat-free mass are similar (r1⁄4 0.7 to 0.8) and indicate its non-specificity in prediction of body composition for an individual. BMI does not adequately distinguish fat from muscle and bone, so individuals with the same BMI can have vastly different amounts of muscle and body fat. Also BMI does not account for age, gender, ethnicity and physical fitness in differences in fat mass at the same body weight. Further, it does not differentiate inter-individual disparities in adipose tissue distribution (abdominal vs subcutaneous) so individuals with the same BMI can have the same fat mass but different amounts of abdominal fat that affects their risk of insulin resistance and some chronic diseases. The focus on increased BMI as an independent predictor of health risk is under scrutiny because accumulating evidence challenges the standard relationships between weight status and metabolic health, cardiovascular disease and mortality for certain individuals. One group is the metabolically healthy obese (MHO) who have a high level of insulin sensitivity, normal lipids, low inflammation and no hypertension, and have BMI classified as obese. Longitudinal data indicate a protective effect of this metabolic profile for type 2 diabetes and cardiovascular disease in adults with MHO. The second group is characterized as the metabolically obese with normal weight (MONW) and includes people with normal BMI but altered metabolic profiles that are indicative of the metabolic syndrome. Individuals with MONW, compared with others with a healthy metabolic profile and normal weight,