BMI and Life Expectancy: What the Research Says
By Lifespan Compass · 2026-11-04
Body mass index — BMI — is the most widely used number in weight and health. It appears on medical charts, in fitness apps, and in our own life expectancy quiz. This article explains what the research actually shows about BMI and longevity, what the official cutoffs mean, and where BMI falls short.
What is BMI?
BMI is a simple ratio: your weight in kilograms divided by your height in metres squared. A person who weighs 70 kg and is 1.75 m tall has a BMI of about 22.9.
The World Health Organization classifies adult BMI into four bands:
- Underweight: below 18.5
- Normal weight: 18.5 to 24.9
- Overweight: 25.0 to 29.9
- Obese: 30.0 and above
Obesity is further divided into Class I (30.0–34.9), Class II (35.0–39.9), and Class III (40.0 and above). These cutoffs have been the global reference since the 1990s. They are screening lines for populations, not personal verdicts — a distinction that matters, as we will see.
The U-shaped curve
The largest study ever done on BMI and mortality is the Global BMI Mortality Collaboration, published in The Lancet in 2016. It pooled individual data from 239 prospective studies in 32 countries, covering 10.6 million adults. To avoid a classic trap — sick people losing weight, which can make thinness look dangerous — the main analysis focused on nearly 4 million people who had never smoked, had no chronic disease when the study started, and survived at least five years.
The result was a clear U-shaped curve. All-cause mortality was lowest at a BMI of 20.0 to 25.0 and rose on both sides. Below it, people with a BMI of 15.0 to 18.5 had about 50% higher mortality than those in the normal range. Above it, mortality was about 7% higher at BMI 25.0–27.5 and about 20% higher at 27.5–30.0 — and for every 5 additional BMI points above 25, the risk of death rose by roughly 30% (29% in North America, 39% in Europe and East Asia). At the extreme end, grade 3 obesity — a BMI over 40 — carried nearly triple the mortality of the normal range.
In other words, both substantial underweight and obesity are linked to shorter lives, with the lowest risk sitting in the upper half of the "normal" band and the lower end of overweight barely above it.
Why both ends of the curve are risky
The two ends of the U have different explanations.
At the high end, the mechanism is well understood: excess body fat, especially around the organs, drives heart disease, stroke, type 2 diabetes, and several cancers — the diseases behind most premature deaths in wealthy countries. The link is dose-dependent, and it held in every region studied.
At the low end, the story is more mixed. Some risk is real: very low weight can reflect malnutrition, muscle loss, or an eating disorder, and frailty in old age is itself dangerous. But some is reverse causation — people who are already ill lose weight, so low BMI can be a consequence of disease rather than a cause. That is why the Lancet collaboration restricted its main analysis to healthy never-smokers, and the U-shape survived even that filtering.
The cutoff debate: are 25 and 30 the right lines?
The WHO cutoffs are useful, but they are not universal truths. Three caveats matter.
First, risk rises continuously. There is no cliff at 25 or 30: a BMI of 25.1 is barely different from 24.9. The cutoffs are administrative conveniences, not biological boundaries.
Second, the lines shift by population. Asian populations tend to develop diabetes and cardiovascular disease at lower BMIs, and the WHO has identified lower action points for many Asian groups — 23 for overweight and 27.5 for obesity.
Third, age changes the picture: in adults 65 and older, the lowest-mortality range appears to sit a little higher, roughly 23 to 28, according to a 2019 meta-analysis of 44 studies. A little extra weight in old age may protect against frailty and fractures.
Why BMI is imperfect
BMI's biggest flaw is what it cannot see. It measures total weight against height, but it cannot tell muscle from fat. A muscular athlete can have a BMI over 30 and very little body fat. An older person can have a "normal" BMI while having lost most of their muscle and gained fat around the waist — a combination called sarcopenic obesity that BMI completely misses.
Fat distribution matters at least as much as total fat. Deep visceral fat around the organs is far more dangerous than fat on the hips and thighs, so two people with the same BMI can carry very different risks.
That is why experts increasingly recommend pairing BMI with a second measure. In January 2025, a Lancet commission of 58 international experts recommended combining BMI with waist circumference, waist-to-height ratio, or a direct measure of body fat, rather than relying on BMI alone. A tape measure around the waist adds real information BMI cannot provide.
What to do with your number
If your BMI is in the normal range, that is reassuring — but it is not a clean bill of health. Blood pressure, blood sugar, cholesterol, smoking, activity, and family history all matter independently.
If your BMI is above 25, the most useful question is not "how do I get below 25?" but "what is driving it, and what can I change?" The evidence points to habits, not crash diets: regular activity, mostly whole foods, less ultra-processed food and sugary drinks, enough sleep, and less alcohol. Even modest, sustained weight loss — 5 to 10% of body weight — measurably improves blood pressure, blood sugar, and joint pain.
If your BMI is below 18.5, especially after unintentional weight loss, discuss it with a doctor: unexplained weight loss can be an early sign of illness, and in older adults low weight is closely tied to frailty and falls.
And whatever your BMI, remember that it is one number among many. Our quiz uses it because it is the best simple, widely available proxy for weight-related risk — but a single number never captures a whole person.
Curious where you stand? Try the free life expectancy quiz — it combines your BMI with smoking, exercise, diet, sleep, and other factors for a fuller picture.
This article is for general health education only. It is not medical advice, and it cannot predict any individual's lifespan. BMI is a screening tool, not a diagnosis. If you have questions about your weight or your health, please talk to your doctor or another qualified health professional.
Sources
- Global BMI Mortality Collaboration. Body-mass index and all-cause mortality: individual-participant-data meta-analysis of 239 prospective studies in four continents. The Lancet 2016;388:776–786. https://www.thelancet.com/journals/lancet/article/PIIS0140-67361630175-1/fulltext
- World Health Organization adult BMI classification (underweight below 18.5, normal 18.5–24.9, overweight 25.0 and above, obese 30.0 and above). WHO fact sheet on obesity and overweight. https://www.who.int/news-room/fact-sheets/detail/obesity-and-overweight
- Lancet Diabetes & Endocrinology Commission on clinical obesity (2025). https://doi.org/10.1016/S2213-8587(24)00316-4
- Dose-response relationship between body mass index and risks of all-cause mortality and disability among the elderly: a systematic review and meta-analysis. Clinical Nutrition, 2019;38(4):1511–1523. https://doi.org/10.1016/j.clnu.2018.07.021
- Asian BMI action points (WHO expert consultation, 2004)