CORRESPONDENCE: LETTER TO THE EDITOR
Problems in Measurement of Body "Fatness"
Gerson T. Lesser, MD*
* Jewish Home and Hospital, 120 West 106th Street, New York, New York 10025 (Email: glesser{at}jhha.org).
Gelber et al. (1) make an important contribution in comparing anthropometric measurements as predictors of cardiovascular disease (CVD). These authors and an editorialist (2), accept that "fatness" relates to CVD, and point out the controversy regarding total fat content versus some "central" fat accumulation in promoting CVD. Both entertain reservations as to whether body mass index (BMI) measures general obesity or whether waist circumference (WC) (or some modification, such as waist-to-hip ratio [WHR] or waist-to-height [WHtR]) measures "central" obesity.
BMI, calculated from body weight and height2, does not separate the components of its 2-compartment model: lean body mass (LBM) and total body fat (WF). Equating elevated BMI with "obesity" or "adiposity" incorrectly assumes that LBM is constant at any given height so that any BMI change represents a change of WF. Gelber et al. (1) and Litwin (2) are aware of this, but perhaps not of the resulting magnitude and frequency of potential error. Using a direct measurement of body fat, within a homogeneous cohort of similar height, LBM varies at least ±20% from the group mean, and an individual's deviation from average weight at this height can be all WF, all LBM, or any combination of these (3). For young women of identical height, weight, and BMI, a ±20% range of LBM is associated with a remarkable range of fat content (WF/body weight), from 10% to 40% (calculated from Lesser et al. [3]).
The interpretation of BMI (and, to a lesser extent, of WHR and WHtR) is further complicated by aging changes within the LBM. The Gelber et al. (1) subjects were in the fifth to ninth decades. Most older people lose muscle and bone mass; many also lose appreciable height (4). Such LBM changes vary widely among individuals and defy reasonable adjustment, and BMI use in this population becomes highly questionable.
The "central versus total fat" controversies do not seem firmly based, as BMI is not a trustworthy measure of total body fat, and there is only limited evidence that WC represents some vaguely defined central "fat compartment." WC is probably less perturbed by aging changes, although vertebral compressions can shorten the trunk and "falsely" increase waist measurement. Hip circumference is affected somewhat by fatness but probably best correlates with the size of the frame/LBM. Height also correlates with LBM, with the noted ±20% range of LBM at a given height. Thus, the ratios of WHR and WHtR probably serve as approximate adjustments of WC to individual size. WHtR, an admittedly coarse measurement, is probably the best available approximation of overall "fatness," WF/body weight, for large groups and over broad ranges of age, and was statistically the best model fit (1).
In large populations, almost any reasonable tool, even "eye-balling," should provide some approximation of fatness and correlate with fat-related conditions. Behnke (5), who pioneered human densitometry, later accommodated to a tape measure and simple procedures for the practical estimation of body fatness. Until improved methods are available, we can best understand relationships of body composition and future morbidity by appreciating the limitations of current anthropometric tools.
 |
References
|
|---|
1. Gelber RP, Gaziano JM, Orav EJ, Manson JE, Buring JE, Kurth T. Measures of obesity and cardiovascular risk among men and women J Am Coll Cardiol 2008;52:605-615.[Abstract/Free Full Text]2. Litwin SE. Which measures of obesity best predict cardiovascular risk? J Am Coll Cardiol 2008;52:616-619.[Free Full Text] 3. Lesser GT, Deutsch S, Markofsky J. Use of an independent measurement of body fat to evaluate overweight and underweight Metabolism 1971;20:792-804.[CrossRef][Web of Science][Medline] 4. Sorkin JD, Muller DC, Andres R. Longitudinal change in the height of men and women: Implications for interpretation of the body mass index Am J Epidemiol 1999;150:969-977.[Abstract/Free Full Text] 5. Behnke AR. Anthropometric evaluation of body composition throughout life Bull NY Acad Sci 1963;110:450-464.
|
Top
References