Relative weight classifications in the assessment of underweight and overweight in the United States

Abstract

We compared five recent relative weight classifications based on body mass index for their estimates of prevalence of underweight and overweight in the adult population of the United States and for their ability to predict subsequent morbidity and mortality. The sources of the classifications were: the 1990 Dietary Guidelines for Americans of the US Departments of Agriculture and Health and Human Services, the National Academy of Sciences, the National Center for Health Statistics, the World Health Organization, and the Canadian Minister of National Health and Welfare. These classifications were applied to the body mass index distributions of the second National Health and Nutrition Examination Survey (1976-1980) and to the Hispanic Health and Nutrition Examination Survey (1982-1984). Depending on classification, a wide range of prevalence for the total population was found: 9-17% of the US population were categorized as underweight, and 25-45% were categorized as overweight. White women had the highest prevalence of underweight in all but the National Center for Health Statistics classification. Black and Mexican American women had the highest prevalence of overweight under all classifications (range: 38.4-58.6%). Associations with health outcomes were determined using all cause hospitalization and mortality in the 1971-1987 follow-up of the first National Health and Nutrition Examination Survey. Underweight and overweight as defined by the National Academy of Sciences classification had the highest population attributable risk for hospitalization and death: 5.0% of hospitalizations and 11.0% of deaths among men and 4.2% of hospitalizations and 11.4% of deaths among women were associated with weights outside the healthy range. Under this classification a greater proportion of both hospitalizations and mortality were associated with overweight than underweight. For all classifications, a higher proportion of hospitalizations were associated with overweight than underweight. All classifications performed better at predicting death than hospitalization.