乌鸦传媒

Adults over 65 at the high end of the healthy body mass index (BMI) range were at lower risk of mortality, and those at the low end were at highest risk, according to a recent meta-analysis.

The association between all-cause mortality and BMI created a U-shaped curve with a broad base (P-nonlinearity <0.001). The "nadir of the curve for BMI and mortality was between 24.0 and 30.9, with the lowest risk being between 27.0 and 27.9 (HR 0.90, 95% CI 0.88-0.92)," wrote , of Deakin University in Melbourne, Australia, and her co-authors, in the American Journal of Clinical Nutrition.

They said that mortality risk did not increase with excess weight in this population until BMI was ≥33 (HR 1.08 for BMI of 33.0-33.9, 95% CI 1.00-1.15).

Risk of mortality was highest at a BMI lower than 23, the authors said. Using a BMI of 23.0 to 23.9 as the reference, there was a 12% greater risk of mortality for those with a BMI in the range of 21.0-21.9 (HR 1.12, 95% CI 1.10-1.13) and a 19% greater risk for those with a BMI in the range of 20.0-20.9 (HR 1.19, 95% CI 1.17-1.22), the authors said.

The analysis included 32 studies identified through Medline, CINAHL and the Cochrane Library, and hand searches between 1990 and September of 2013. All of the studies included population-based cohorts, and they followed participants for an average of 12 years. The final analyses included 197,940 community-living adults age 65 or older. There were 72,469 deaths, in total, recorded in the studies.

The World Health Organization (WHO) defines a healthy body weight range for adults as a BMI between 18.5 and 24.9. But "this range has been based primarily on studies in younger adults, for whom the risks of diabetes, cardiovascular disease, certain cancers, and mortality associated with increased body weight are well documented," the authors wrote. It may be "overly restrictive" for older adults, they said.

Given the current findings, "it behooves us to reconsider current weight-for-height guidelines," wrote , of the Baltimore VA Medical Center, in an editorial that accompanied the article.

Sorkin wrote that the findings of the current study are consistent with conclusions drawn by clinical director of the National Institute on Aging from 1977 to 1998, after reviewing "the handful of articles that examined weight-for-height," and converting data on weight and height from the U.S. and Canadian life insurance companies to BMIs.

Andres' conclusions were controversial then, because "they questioned the beautiful hypotheses that increasing weight is associated with increasing mortality and weight should remain unchanged through adult life," Sorkin wrote.

Subsequent studies suggested that older adults with a BMI in the overweight range (25 to 29.9) were at similar or lower risk of all-cause mortality than those in the normal-weight range, the authors said. Many of those studies, however, were concerned with the risk associated with a high BMI, and not with those at the lower end of the scale.

Given the current findings, "we must be open to the possibility that the hypothesis that best weight-for-height in older adults is the same as that seen in younger adults may be wrong," said Sorkin.

One explanation is that older adults are susceptible to undernutrition due to physiologic changes, chronic disease, polypharmacy, and psychosocial changes, the authors said. It often goes unrecognized, they noted.

"Therefore it is important to understand the association between BMI and mortality in the older population," they wrote. Monitoring weight status in individuals with a BMI of less than 23, they said, "would seem appropriate to detect weight loss promptly and address modifiable causes."

There were limitations in this analysis. It assessed only mortality risk associated with BMI, rather than weight change or body composition. "Weight change may be more important for older adults in terms of health risks," the authors said.

Few of the studies included in it included standardized assessments of physical activity, and physical activity at different BMI levels may have influenced results. Furthermore, the relationship between morbidity and BMI was beyond the scope of this study.

The study also focused only on older individuals living in the community. "The relationship between BMI and mortality may be different for those in institutionalized or residential care who are sicker and frailer," the authors said.

"We were interested in understanding the mortality risks associated with BMI among the 'independent' living older population, because these are the likely recipients of dietary advice based on weight status," they said.

Comment