乌鸦传媒

A body mass index (BMI) in the overweight -- but not obese -- range was not independently associated with an increased mortality risk, a retrospective study relying on two decades' worth of National Health Interview Survey (NHIS) data found.

Of roughly half a million U.S. adults captured in the 1999-2018 surveys, a slightly lower risk for death over 9 years of median follow-up was observed among people with a BMI in the overweight range compared with a reference group within the normal range (BMI of 22.5-24.9):

• BMI 25-27.4: adjusted HR 0.95 (95% CI 0.92-0.98)
• BMI 27.5-29.9: adjusted HR 0.93 (95% CI 0.90-0.96)

Meanwhile, BMIs of 30 and above or below 18.5 were each associated with an increased risk of mortality compared with that reference group, Aayush Visaria, MD, MPH, and Soko Setoguchi, MD, DrPH, both of Rutgers Institute of Health in New Brunswick, New Jersey, reported in .

Compared with the reference group, BMIs in the obesity range (30 and above) were associated with a 21% to 108% increased mortality risk in further adjusted analyses that only included healthy never-smokers and omitted participants who died within 2 years of follow-up.

"The real message of this study is that overweight as defined by BMI is a poor indicator of mortality risk, and that BMI in general is a poor indicator of health risk and should be supplemented with information such as waist circumference, other measures of adiposity, and weight trajectory," Visaria told 乌鸦传媒.

The current study to suggest a lower mortality risk for individuals with an overweight BMI, but the results come at a time when BMI is being questioned more and more as a risk tool. Last month, the American Medical Association joined the debate, adopting a new policy that cited BMI's limitations -- including that it was primarily based on data collected from non-Hispanic white populations -- and recommending that BMI be used in conjunction with other valid measures of risk.

"Physicians should interpret adiposity measures in the context of their cardiometabolic health parameters, such as blood pressure, blood sugar, and cholesterol levels," said Visaria.

Most of the data on the links between BMI and mortality stem from 20th-century cohorts, the researchers noted, and the current study sought to determine the associations in a 21st-century population. The contemporary U.S. population has a higher mean BMI; is more diverse; life expectancy has increased by more than 10 years, regardless of BMI; and treatments for various obesity-related conditions have improved.

During the two-decade study period, mean BMI rose from 26.7 to 28.0, and obesity prevalence increased from 22% to 31%.

Visaria and Setoguchi's analysis included 554,332 non-pregnant adults ages 20 and older (mean age 46) who participated in NHIS surveys from 1999 to 2018 and had a recorded BMI. The cohort was evenly split between men and women, and racial and ethnic groups roughly matched that of the U.S. population (69% white, 14% Hispanic, 12% Black, 4.6% Asian).

Researchers divided participants into nine BMI groups ranging from below 18.5 to 40 and above. Overall, 35% had a BMI in the overweight range (25-29.9) while 27% had a BMI in the obesity range (30 or above).

Individuals with overweight had slightly higher rates of comorbidities compared with the reference group, including higher rates of diabetes (6-9% vs 4%, respectively), hypertension (26-32% vs 20%), and myocardial infarction (MI; 3.4-3.9% vs 2.6%). Adults with class I obesity (BMI of 30-34.9) had approximately triple the rates of diabetes (13%) and nearly double the rates of MI (4.2%) and hypertension (39%) compared with the reference group.

A total of 75,807 deaths occurred over a median follow-up of 9 years (maximum 20 years). The associations between BMI and mortality were largely similar between men and women and across racial and ethnic groups, according to the researchers.

The slightly reduced mortality risk observed in people with a BMI of 25-29.9 was more pronounced in adults 65 and up, while younger adults in that BMI range had an increased mortality risk. Older adults had no significant increase in mortality up to a BMI of 34.9.

The overall findings remained consistent in analyses that adjusted for confounders, restricted the population to healthy never-smokers, and excluded survey participants who died within 2 years of follow-up; and in a sensitivity analysis using National Health and Nutritional Examination Survey data from the same timeframe.

Study limitations, said Visaria and Setoguchi, included the use of self-reported BMI and disease-burden data, the relatively short follow-up time that may have led to underestimates of mortality in higher BMI groups, and that change in weight over time was not captured in the surveys.

Comment