Overweight based on BMI alone ‘not an appropriate measurement’ for all-cause mortality

Key takeaways:

• All-cause mortality risks were similar among adults with BMIs in the overweight range and those with lower BMIs.
• The risk was significantly higher for those with a BMI of 30 k/mg² and greater.

BMI in the overweight range was not associated with an increased risk for all-cause mortality, highlighting the limitations of using BMI to predict mortality risk independent of other risk factors, researchers found.

“Overweight as a medical condition is still important but will likely need to be diagnosed taking into account body composition and body fat distribution.” Aayush Visaria, MD, MPH, a first-year internal medicine resident at Rutgers Robert Wood Johnson Medical School, told Healio. “This was not all that surprising to me because clinically we do intuitively already tend to ignore BMI if it doesn't match the phenotype or clinical picture of someone's health. However, now we have more rigorous data demonstrating BMI’s limitations.”

Visaria and Soko Setoguchi, MD, DrPH, a professor of medicine and epidemiology at the same institution, noted in PLoS One that previously suggested links between all-cause mortality and BMI have been inconsistent, “especially with regards to overweight and class I obese individuals.” Data from past studies primarily came from the 1960s to 1990s and focused on predominantly non-Hispanic white populations, according to the researchers.

BMI itself has also drawn questions on its validity when used alone. At its House of Delegates Meeting, the AMA adopted a policy that advised physicians to use BMI in conjunction with other measurements when assessing for overweight or obesity.

To fill gaps in the literature, Visaria and Setoguchi analyzed the BMI and mortality data of 554,332 adults, which were retrieved from the 1999-2018 National Health Interview Survey and 2019 U.S. National Death Index.

Of the participants, the mean age was 46 years and 50% were women. During the study period, 75,807 deaths occurred.

The researchers found that mortality rates were consistent across various BMI categories. Compared with a BMI of 22.5 to 24.9 kg/m², the adjusted HR for mortality was:

• 0.95 (95% CI, 0.92-0.98) for BMI of 25 to 27.4 kg/m²; and
• 0.93 (95% CI, 0.9-0.96) for BMI of 27.5 to 29.9 kg/m².

The findings remained after the analysis was restricted to those who were healthy and never smoked and excluded those who died within the first 2 years of follow-up.

However, Visaria and Setoguchi noted that the risk for all-cause mortality was significantly elevated in those with BMI of:

• 30 to 34.9 kg/m² (HR = 1.08; 95% CI, 1.04-1.11);
• 35 to 39.9 kg/m² (HR = 1.12; 95% CI, 1.07-1.16);
• 40 kg/m² and over (HR = 1.31; 95% CI, 1.24-1.37); and
• 18.5 kg/m² and less (HR = 1.9; 95% CI, 1.79-2.01).

Notably, adults aged 65 years and older did not show any increase in all-cause mortality across BMI of 22.5 to 34.9 kg/m².

There are several potential reasons why adults with a higher BMI may have similar all-cause mortality rates as those with normal BMI, according to the researchers. For example:

• adults with overweight may have greater survival skills in adverse health events, like chronic conditions or severe infections;
• adults with overweight without a disease might be metabolically healthy “and have a more favorable body composition consisting of higher lean mass”;
• lean adults with diseases such as hypertension or diabetes may develop more treatment-resistance diseases, whereas adults with overweight may manage or reverse these diseases with weight loss strategies; and
• the U.S. population has become more racially and ethnically diverse, adding different body compositions than those of previous cohorts.

The take-home message of the findings “is that overweight as defined by BMI is not an appropriate measurement tool, at least for all-cause mortality,” Visaria said.

“I think it is important to distinguish overweight BMI from overweight as a disease process,” he said. “To truly measure body composition and adiposity, physicians should consider supplementing BMI with other measures such as waist circumference, waist-to-height ratio, waist-to-hip ratio.”

Visaria added that bioimpedance scales “are also an option to estimate total body fat [percentage], but they still need to be validated and are known to have some margin of error.”

“Additionally, physicians should interpret adiposity measures in the context of their cardiometabolic health parameters such as blood pressure, blood sugar and cholesterol levels,” he said.