We are delighted with Dr. Baggerly’s strong interest in our report. Here, we try to clarify a few issues he has raised.

First, the widely accepted interpretation of the β coefficient is change in the dependent variable per unit change in the independent variable. We used this definition throughout our report, with 1 minor change: we presented results as change in the dependent variable per 10-unit change in independent variable. However, we did present β coefficients in Table 2 in a way in which merely changing 1 decimal position will convert the scale to per 1-unit change in the independent variable. When presenting the results for change in C-reactive protein, there is a choice to use either a logarithmic scale or to convert values back to a normal scale. Because the logarithm is not an additive transformation but a multiplicative one, we preferred to present results as geometric means. The graph presents univariate analysis and is consistent with the coefficients given in model 1 in Table 2.

Second, the association of obesity and inflammation is well known, so we preferred not to reiterate it in our report. Moreover, this was not the focus of our study and thus did not get the attention that some readers would have desired.