Author + information
- Thais Coutinho, MD and
- Francisco Lopez-Jimenez, MD∗ ()
- ↵∗Division of Cardiovascular Diseases, Mayo Clinic, 200 First Street, Southwest, Rochester, Minnesota 55905
We thank Profs. Song, Wang, and Zhang for their interest in our study (1).
Because our study was a meta-analysis with individual patient data, information on statins was not uniformly available in all studies included, and for this reason the use of statins was not originally included in the models as a potential confounder. Multivariable Cox proportional hazard models in the 6,313 of the 15,547 participants with data on statin use and including statin use as covariate showed that normal-weight central obesity still had the highest mortality among all participants (analyses not shown). Furthermore, in the setting of secondary prevention, the relative risk reduction attributed to statins is approximately 20% (2), representing an association of lesser magnitude than the increased risk found with normal-weight central obesity. Thus, it is unlikely that the results of our study were driven by differences in statin use among subjects with different body adiposity patterns.
We recognize that there might be some residual confounding attributed to differences among groups in weight change over time, diet, psychosocial stress, socioeconomic status, and physical activity. Unfortunately, we had limited information on these factors to completely adjust in the multivariate models. However, we are not aware of any evidence suggesting that subjects with normal-weight central obesity would be more stressed, poorer, less likely to lose weight, or have worse nutrition than people with general obesity or other patterns of fat distribution, to suspect that these variables would explain all of the association between normal-weight central obesity and higher mortality.
- American College of Cardiology Foundation