Author Affiliations: Department of Preventive Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois (Dr Carnethon and Mr De Chavez) (email@example.com); and Division of Endocrinology and Metabolism, Johns Hopkins University School of Medicine, Baltimore, Maryland (Dr Golden).
In Reply: Drs Zeng and Dong and Dr Yano and colleagues suggest that we should have explored the association of weight status at the time of diabetes with cardiovascular morbidity in addition to mortality. We agree that studying morbidity would have been informative; however, the pooled cohort approach that we used required harmonization of outcome definitions across studies.
Because protocols for determining nonfatal events varied across studies (eg, active community-wide surveillance for cohort participants vs contacts with cohort participants and their proxies only), we were concerned about the influence of heterogeneity on our ability to determine the outcome. While we speculate in our article that the absence of an association with cardiovascular mortality is due to the relatively smaller number of events, it is possible that elevated total mortality in normal-weight persons with diabetes is being driven by other causes. To test this hypothesis, future studies performed in a single cohort should include nonfatal outcomes.
Carnethon MR, Golden SH, De Chavez PJD. Weight and Mortality in Adults With Diabetes—Reply. JAMA. 2012;308(20):2080-2081. doi:10.1001/jama.2012.14570