C lark and colleagues raised an important point. Total body weight is comprised of fat mass, lean mass, and bone mass. Not only is fat mass correlated with lean mass and total body weight, but it also may exert both positive and negative effects on bone.(1) For example, adipose tissue may stimulate bone growth by its mechanical loading effect. It also can express or secrete a variety of biologically active molecules, including leptin and adiponectin, which may suppress bone growth, as reported by us (2) and others. (3) In evaluating such complex relationships between fat mass and bone phenotypes, some studies have adjusted for lean mass, (4,5) whereas others have adjusted for total body weight. (6,7) To date, the findings on the relationship between fat mass and bone phenotypes are conflicting in the literature. The aim of our study (8) was to examine the relationship between fat mass and bone phenotypes after accounting for a mechanical loading effect and other potential confounders.To this end, we believe that adjustment for total body weight [or body mass index (BMI)] best accounts for interindividual difference in overall mechanical loading. Furthermore, adjustment of total body weight also serves the purpose of correcting for the influence of body size on bone in children and adolescents. (9,10) Our analyses showed that percent fat mass (PFM) was inversely associated with bone phenotypes when body weight was adjusted. Such inverse associations remained significant when age-and gender-specific tertiles of lean mass were further added into the regression model. Our results indicate that the inverse associations between fat mass and bone mass cannot be fully explained by the positive association between lean mass and bone mass. If lean mass is adjusted instead of total body weight, then the association of fat mass with bone phenotypes would reflect the net result of its positive and negative effect on bone.In our cohort, with adjustment for lean mass rather than body weight, there was a positive trend between PFM and total-hip bone mineral content (BMC) in boys and girls. However, we found that this trend varies by age-and gender-specific BMI: In leaner subjects, the trends are more positive, whereas in heavier subjects, the trends are more negative. Our study population is relatively lean, with an average BMI of 18.8 AE 2.2 and 19.8 AE 2.3 kg/m 2 in boys and girls, respectively. We speculate that the net association of fat mass with bone may be more negative in overweight or obese Chinese children.In short, from a mechanistic, clinical, and public health perspective, it is important to characterize both the positive and negative effects of fat mass on bone phenotypes. The potential negative effect of fat mass on bone in children and adolescents can best be assessed with adjustment for total body weight in order to account for interindividual difference in overall mechanical loading and body size.