Mortality is not lower among adults with diabetes who were overweight or obese at diagnosis than among those who were of normal weight, as has been suggested by proponents of the "obesity paradox," according to a report published online Jan. 15 in the New England Journal of Medicine.
An analysis of data from two large, long-term, prospective cohort studies showed a clear J-shaped association between body mass index (BMI) at diagnosis of type 2 diabetes and all-cause mortality. Excess adiposity did not confer any mortality benefit, said Deirdre K. Tobias, Sc.D., of the department of nutrition, Harvard School of Public Health, Boston, and her associates.
Previous studies that reported such a benefit had much smaller sample sizes and did not adequately control for smoking status. "Smoking is a concern in analyses of body weight and mortality because it is associated with decreased body weight but an increased risk of death," they noted.
istockphoto.com
Among diabetic patients, mortality rates were higher for overweight or obese adults with the disease than their peers.
Their findings, added to the known association between obesity and other critical public health concerns such as cardiovascular disease and cancer, clearly show that "maintenance of a healthy body weight should remain the cornerstone of diabetes management, irrespective of smoking status," Dr. Tobias and her colleagues said.
For their study, the investigators analyzed data on 11,427 participants in the Nurses’ Health Study and the Health Professionals Follow-Up Study who developed diabetes between baseline (1976 and 1986, respectively) and January 2010. The mean age at diagnosis was 62 years (range, 35-86 years) among the women in the NHS and 64 years (range, 41-91 years) among the men in the HPFS.
The primary outcome measure of the study was death from any cause through January 2012. There were 3,083 such deaths during a mean follow-up of 15.8 years.
For the study population as a whole, there was a J-shaped association between BMI and all-cause mortality. Compared with participants who had a normal BMI of 22.5-24.9 kg/m2 at the time of diabetes diagnosis, those with lower BMIs and those with higher BMIs both had significantly elevated mortality. The hazard ratio was 1.29 for subjects in the lowest BMI category (18.5-22.4 kg/m2) and 1.33 for those in the highest BMI category (35.0 kg/m2 or higher).
Smoking status exerted a significant effect on the association between obesity and mortality. Among nonsmokers, this association showed strong linearity, with a direct correlation between increasing BMI and increasing mortality. This association was not linear among smokers, confirming that adequate control for smoking status is essential in such analyses, Dr. Tobias and her associates said (N. Engl. J. Med. 2014 Jan. 15 [doi:10.1056/NEJMoa1304501]).
The results of sensitivity analyses supported those of the main study, including one analysis that used different, more traditional cutoff points for the BMI categories of normal weight, overweight, and obesity.
When the data were analyzed by cause of death, the association between BMI and cardiovascular mortality was strongly linear among nonsmokers and less strong but still linear among smokers, as was the association between BMI and cancer death. The association between BMI and death from respiratory disease, suicide, and accidents was J-shaped.
This study was limited in that the NHS and HPFS cohorts were relatively homogenous, comprising well-educated white participants almost entirely, so the findings may not apply to other racial/ethnic and economic groups.
This study was supported by the National Institutes of Health and the American Diabetes Association. Dr. Tobias reported no potential conflicts of interest; one of her associates reported ties to Merck and Novo Nordisk.