Two methodological issues concerned us about the excellent study by Hu et al.1 The first is that the authors did not report the results of including more than 1 obesity measure in the same regression equation. When we used waist to hip ratio (WHR) as a risk factor, it was to supplement rather than to replace overall obesity measures such as body mass index (BMI).2 The combination of BMI and WHR was associated more strongly with diabetes and other diseases than was BMI alone or the combination of BMI and waist circumference.3 These associations suggested that both the total amount of adipose tissue (as reflected by BMI) and its location (as reflected by WHR) were supplemental risk factors for some diseases. It would be valuable to know whether they were supplemental risk factors for stroke.
Hartz AJ, Rimm AA. Waist to Hip Ratio as a Supplement to Body Mass Index. Arch Intern Med. 2008;168(2):237–238. doi:10.1001/archinternmed.2007.63