In Reply: Dr Cheng raises a statistical concern regarding the comparison of the magnitude of the ORs for BMI and physical inactivity in relation to cardiovascular biomarkers. We agree that when comparing the association of 2 different measures, such as BMI and physical inactivity, with a dependent variable, the 2 measures should be standardized for comparison. Some investigators argue for using standardized regression coefficients when comparing the relative contributions of each measure. However, standardized coefficients have limited applicability to nonnormally distributed predictors.1 In our study, BMI and physical activity were not normally distributed. For such analyses, it is preferable to calculate the coefficients for quantiles of BMI and physical inactivity1 as we did in Table 4, providing an accurate and readily interpretable comparison of the 2 measures with respect to their associations with the cardiovascular biomarkers. From a clinical and public health perspective, we also chose to display the results using clinical cutpoints for BMI and physical inactivity, as demonstrated in the Figure.
Mora S, Lee I, Ridker PM. Physical Activity, Body Mass Index, and Cardiovascular Biomarkers in Women—Reply. JAMA. 2006;296(8):931–932. doi:10.1001/jama.296.8.932-a
Customize your JAMA Network experience by selecting one or more topics from the list below.
Create a personal account or sign in to: