Major Finding: Among white, nonsmoking, healthy men and women, all-cause mortality rates showed a linear increase in association with increasing BMI.
Data Source: A pooled analysis of 19 prospective cohort studies involving 1.46 million white adults followed for a mean of 10 years, to provide mortality risk estimates associated with BMI.
Disclosures: The study was supported by the National Institutes of Health. One of Dr. de Gonzalez's associates said he received consulting fees from Iovate Health Sciences USA.
Among white adults, both overweight and obesity are associated with increased all-cause mortality, a study has shown.
Mortality is lowest within a body mass index range of 20.0-24.9, said Amy Berrington de Gonzalez, D.Phil., of the division of cancer epidemiology and genetics at the National Institutes of Health, and her associates.
Underweight adults also appear to have a higher rate of all-cause mortality, although that association is not as strong, they said.
Some earlier studies suggested that being overweight (BMI, 25.0-29.9) either was beneficial or had little effect on all-cause mortality, while others found slightly increased risks. These inconsistencies may have been due to confounding by tobacco use or disease-related changes in weight.
“We examined the relationship between BMI and all-cause mortality in a pooled analysis of 19 prospective studies” with at least 5 years of follow-up in which smoking status and prevalent disease could be accounted for, the investigators said. The analysis was restricted to non-Hispanic white adults (mean age, 58 years at baseline) “because the relationship between BMI and mortality may differ across racial and ethnic groups,” Dr. de Gonzalez and her colleagues noted.
The 19 cohorts included 1.46 million people and 160,087 deaths during a median follow-up of 10 years.
A total of 58% of the study subjects were women.
When the analysis excluded smokers and people with prevalent disease, all-cause mortality was lowest at a BMI of 20.0–24.9 among both men and women and rose in a nearly linear fashion as BMI increased.
For example, in women who were overweight (BMI, 25.0–29.9), the estimated hazard ratio was 1.13; this rose to 1.44 for obesity class I (BMI, 30.0–34.9); 1.88 for obesity class II (BMI, 35.0–39.9); and 2.51 for obesity class III (BMI, 40.0–49.9).
Hazard ratios were similar for men, except that they were even higher for obesity classes II and III.
“In our study, there were more than five times as many deaths among participants in the highest obesity categories (BMI of 35.0–39.9 and 40.0–49.9) than in previous studies because severe obesity has become more common,” the investigators wrote. In the United States, the rates of these levels of BMI were estimated to be 11% among men and 17% among women in 2008, they added.
Statistical adjustments for other potential confounders such as alcohol consumption, physical activity level, education level, and marital status only attenuated these estimates slightly (N. Engl. J Med. 2010;363:2211-9).
Hazard ratios also increased for a BMI below 20.0, but the results varied widely according to length of follow-up and level of physical activity. In underweight people, increased mortality “is probably, at least in part, an artifact of preexisting disease,” Dr. de Gonzalez and her associates said.