点击显示 收起
1 From the Pennington Biomedical Research Center, Baton Rouge, LA.
See corresponding article on page 379.
The article by Janssen et al (1) in this issue of the Journal examined whether body mass index (BMI; in kg/m2) adds to the risk associated with waist circumference (WC). The authors made several important points. On the basis of data from the National Health and Nutrition Examination Survey conducted between 1988 and 1994, they showed that the BMI is related to cardiovascular disease (CVD) risk factors in a graded fashion—ie, as the BMI increases, the CVD risk increases. However, other conditions, such as diabetes, sleep apnea, cancer, and osteoarthritis, were not included in the risks evaluated in their analysis, which limits their overall conclusions about the relation of BMI and WC to CVD risk. Moreover, they showed that most of the information about CVD risks that is obtained from BMI can also be obtained from WC. Finally, Janssen et al showed that, when WC was used as a continuous variable, it accounted for the risk of CVD better than it did when it was dichotomized with the use of the risk algorithm published by the National Institutes of Health (2). These are important findings, and they led me to propose a new way of classifying WC (Table 1).
View this table:
TABLE 1. Proposed new cutoffs for waist circumference in adults
However, the analysis by Janssen et al has some limitations that lead me to caution against replacing BMI with WC, for fear of throwing the baby out with the bath water. The concept of BMI [weight (kg)/ stature (m2)] was originally proposed by Quetelet (3) in 1835 on the basis of his analysis of the way in which the body grows in 2 and 3 dimensions. Wide acceptance of this concept, however, took > 100 y (4). At the beginning of the 20th century, the life insurance industry had identified central adiposity as a risk factor for death (5). However, height and weight tables provided by that industry, rather than measures of central adiposity, were widely used to establish overweight, and the Framingham Study used those tables as the initial criterion for overweight (Metropolitan relative weight, as it was called). With the publication of the Diet and Health report by the National Research Council in 1989 (6) and the subsequent Dietary Guidelines, the use of BMI became more widely accepted; the cutoffs were defined as 18.5-24.9 for normal weight, 25-29.9 for overweight, and =" BORDER="0"> 30 for obesity, which was further divided into grades 1-3, with BMIs of 30-34.9, 35-39.9, and =" BORDER="0"> 40, respectively (4).
Janssen et al stated, "It is possible that WC alone could be used as an indicator of health risk and that measures of BMI would not be required." In my judgment, this would be a bad idea. The BMI is composed of weight and height. To propose that clinicians should not obtain these measures is irresponsible, because height and weight are important indicators of health status and are easy for clinical personnel to measure. Janssen et al also stated that "most members of the population cannot readily calculate their BMI." This difficulty is easily remedied, because there are both tables and nomograms from which BMI can be determined. A table that allows anyone to determine his or her BMI easily by using either pounds and inches or kilograms and centimeters is shown in Table 2. Those interested can be instructed thus: Simply locate your height in inches along the left side or in centimeters along the right side, and then move toward the center along the same line until you come to the cell with the weight closest to your own in either pounds or kilograms. Your BMI is given above and below this cell, in the rows of boldface numbers at the top and bottom of the page.
View this table:
TABLE 2. BMI obtained by using either pounds and inches or kilograms and centimeters1
The National Heart, Lung, and Blood Institute clearly recognized that obtaining the BMI was only the first step for the clinician who was making a risk assessment (2). An assessment of central adiposity was also needed. Several criteria are available for assessing central adiposity, but the most practical are WC and WC divided by hip circumference (also called the waist-to-hip ratio, or WHR). The WHR served as the key index that led to the recognition of the importance of central adiposity as a major risk factor for the diseases associated with obesity (7). In studies that compared WC and WHR with respect to validity as a standard for measurement of central fat, WC was as good as or better than WHR (8). Because WC is a single measure, whereas WHR is composed of 2 separate measures, WC has largely replaced WHR in estimations of central adiposity. Han et al (9) proposed cutoffs for high risk of 102 cm for men and 88 cm for women. These cutoffs have been widely used (2), but the report by Janssen et al suggests that a newer approach is needed.
The proposal in Table 1 is an attempt to open a new dialogue on this issue. The mean (± SD) value for WC obtained by using the data from the National Center for Health Statistics (Table 1 in the article by Janssen et al) provides the basis for this proposal. Although the data probably are somewhat skewed, 2 SDs around the mean for the WC of 40 cm (80-120 cm in men and 70-110 cm in women) would include 95% of the population. Very high and very low risks were defined as those above or below the 5th or 95th percentile, respectively. The low and high categories were the lower and higher 2 SDs, respectively. The advantage for the clinician and the public alike of a simple criterion for WC categories is obvious. This criterion is similar to the basis for selecting 25 kg/m2 as the upper limit of normal for BMI and 200 mg/dL as the beginning of the high-risk category of high cholesterol.
Central adiposity is a key criterion of the metabolic syndrome (10). Including the newly proposed categories in the algorithm for defining the metabolic syndrome may improve the usefulness of criteria for evaluating the risk of central obesity.
In summary, nothing in the article by Janssen et al dissuades me from making measurement of the BMI the first evaluation as proposed in the algorithm from the National Heart, Lung, and Blood Institute's evidence report (2), but, at any given BMI, the use of a measure of central adiposity may improve the criteria for assessing risk. For assessing the risk of diabetes, sleep apnea, osteoarthritis, and cancer, there is as yet insufficient evidence to throw out BMI in favor of WC as the first step in the line of assessment.
REFERENCES
Related articles in AJCN: