Body mass index (BMI) is a common measure of body fat, but new research shows that having a BMI in the "normal weight" range is not always the healthiest for every person. In fact, for many people, having a BMI in the overweight range may be linked with the lowest risk of dying over a 13-year period, the research suggests.

Researchers looked at data on about 400,000 people in the U.S. who were ages 50 to 71 at the start of the study, in 1995. The researchers followed up with them through 2009, and about 112,000 people in the study had died by then. The findings showed that the “best” BMI for the people in the study in terms of their lifespan was 26, on average.

A person's BMI is calculated based on their height and weight. Usually, a BMI below 18.5 is considered underweight, from 18.5 to 24.9 is considered normal weight, from 25 to 29.9 is considered overweight, and 30 and over is considered obese.

But these categories lump together people who actually have quite different BMIs, said study author Howard Karloff, a computer scientist formerly with Yahoo Labs and AT&T Labs, who worked on the new study because of his expertise in crunching large sets of data with numerous interacting variables. In the usual BMI categories, people with body weights that differ by 20 percent may be grouped together, Karloff said. [9 Healthy Habits You Can Do in 1 Minute (Or Less)]

Previous studies have suggested that BMI is not a perfect way to measure a person's health. For example, researchers reported in 2015 that colorectal cancer patients with a BMI higher than 25 lived longer than those with a BMI under 25. Another study, published in 2011, showed that people with a BMI of 26 to 29 were more likely to survive after having surgery than people with a BMI of 23 or less. 

In the new study, the researchers examined how BMI might relate to people's risk of dying. But instead of considering people's BMIs in the usual categories (of normal weight, overweight and so on), the researchers used the individuals' BMI numbers.

For each person in the study, the researchers looked at a dozen factors that could influence their risk of dying during a given period, including their age, race, education level, alcohol consumption, marital status and exercise level. The researchers calculated each person's "personalized optimal BMI," which is the BMI that, based on these variables, would be associated with the lowest relative risk of death for that person.

The findings showed that any individual's "optimal" BMI depends on his or her own features, and that a "one-size-fits-all" recommendation about what BMI people should strive for may not be best, Karloff said.

"We were able to generate personalized recommendations that, according to our model … are more accurate than the uniform recommendations given by the CDC [Centers for Disease Control and Prevention]," Karloff told Live Science.

Dr. Rexford Ahima, a professor of medicine at the University of Pennsylvania Perelman School of Medicine who was not involved in the study, said that"Body mass allows [researchers] to compare relative weights of people across populations, but was never intended to be used as a healthy tool."

A person's BMI is "just a measure of fatness," Ahima told Live Science. "It says nothing about your risk of developing any disease or dying."

The study was published online Jan. 8 in the journal Obesity.

Copyright 2016 LiveScience, a Purch company. All rights reserved. This material may not be published, broadcast, rewritten or redistributed.