- © 2005 CMA Media Inc. or its licensors
James Douketis notes that waist circumference cut-offs have been validated, citing an excellent paper by Janssen and collaborators.1 Those authors demonstrated that the use of waist circumference cut-off points helps to identify subjects at increased health risk within normal-weight, overweight and class I obese BMI categories, but their study was not a validation of waist cut-offs. Accordingly, they acknowledged that their results do not imply that the specific cut-off values of 102 cm for men and 88 cm for women are the ideal threshold values denoting increased risk.1 They also pointed out that the waist circumference values that best predict health risk within different BMI categories are still unknown.1
We acknowledge that more research is needed on waist circumference cut-offs and on other indicators to better assess the health risks of people in the overweight category, especially given that decreasing the BMI cut-off for this category (from 27 to 25 kg/m2) has resulted in increased heterogeneity in terms of health risk.2
In addition, we wish to stress our concern about reducing the lower limit for the normal weight category (from 20 to 18.5 kg/m2). In our current sociocultural context, where thinness is highly valued, such changes to BMI categories could intensify people's (notably women's) excessive concern about body weight,3which unfortunately may lead them to seek rapid weight loss and to use unhealthy and even dangerous weight loss methods.4The impact of these changes in BMI classification is not trivial and must be recognized. Reducing the lower BMI cut-off for normal weight (to 18.5 kg/m2) may lead to risks associated with delayed identification of eating disorders, but such a change can also prevent early recognition of disordered attitudes and behaviours concerning eating and physical activity, a situation that precedes the onset of eating disorders and affects more adult women than do eating disorders.5
Although Douketis is correct in stating that the guidelines are not designed for intervention purposes in individuals, the BMI remains well known to the public. BMI calculators can be found easily on the Internet and in magazines, and it is impossible to control how people will interpret their BMI in the absence of advice from a health care provider.
Finally, we agree that more discussion is needed concerning the guidelines. It is essential for health care providers to be better informed on how to interpret and use the new weight classification system.