Body weight and healthy aging, what BMI can and cannot show

Body mass index shows up on nearly every intake form in American medicine, and it is treated, by default, as a single number that means roughly the same thing at every age. The evidence behind that assumption is thinner than the number’s ubiquity suggests, and the picture in older adults specifically runs in a direction many people do not expect. This page separates what BMI actually measures from what intentional weight loss trials have and have not shown.

What BMI measures, and what it cannot see

BMI is weight in kilograms divided by height in meters squared, nothing more. The CDC states this limitation directly: “BMI does not distinguish between fat, muscle, and bone mass,” and it “does not indicate what types of fat people have,” nor where in the body a person carries fat, despite fat distribution itself being independently relevant to cardiometabolic risk. Two people with an identical BMI can have very different body compositions, one carrying substantially more muscle and less fat than the other, and BMI alone cannot tell them apart. It is, per the CDC’s own framing, a screening tool “moderately to strongly associated” with more direct measures of body fat, not a direct measurement of body fat itself.

The “obesity paradox” in older adults, stated honestly

In adults under 65, higher BMI categories are consistently associated with higher all-cause mortality risk once BMI climbs well into the obese range. That relationship changes shape in older adults. Winter and colleagues (American Journal of Clinical Nutrition, 2014), pooling 32 studies and nearly 198,000 community-dwelling adults aged 65 and older, found that using a BMI of 23.0 to 23.9 as the reference point, adults with a BMI of 21.0 to 21.9 had a 12% higher mortality risk (hazard ratio 1.12) and adults with a BMI of 20.0 to 20.9 had a 19% higher mortality risk (hazard ratio 1.19), while mortality risk did not begin climbing again until BMI exceeded 33.0. The authors’ own conclusion was direct: “being overweight was not found to be associated with an increased risk of mortality” in this age group, while the increased risk sat at the lower end of the range usually described as healthy.

A separate, larger analysis reaches a similar shape across all adult ages, not only older adults. Flegal and colleagues (JAMA, 2013), systematically reviewing 97 studies with a combined sample of more than 2.88 million people and 270,000 deaths, found overweight (BMI 25 to under 30) associated with significantly lower all-cause mortality than normal weight (hazard ratio 0.94), and grade 1 obesity (BMI 30 to under 35) not significantly associated with higher mortality at all (hazard ratio 0.95). Only grade 2 and 3 obesity (BMI 35 and above) showed significantly higher mortality (hazard ratio 1.29).

Both findings are observational, and both are genuinely contested for the same reason: reverse causation is a serious competing explanation. Illness, including undiagnosed disease, frequently causes weight loss before it causes death, which can make low body weight look more dangerous than it independently is, and can make higher body weight in the reference group look protective by comparison. Smoking is a further confounder, since smokers tend to weigh less and die earlier from smoking-related causes rather than from low body weight itself. Winter and colleagues’ own recommendation reflects this complexity: rather than treating higher BMI as protective, they recommend monitoring older adults at the lower end of the range for “modifiable causes of weight loss.” The paradox describes a population-level statistical pattern, not a demonstrated protective mechanism of carrying more weight.

What randomized trials show about intentional weight loss

The trial evidence on losing weight on purpose is more mixed than either “weight loss is unambiguously protective” or “weight loss trials don’t show benefit” suggests, because the outcome tested matters enormously.

For preventing new disease, the evidence is strong. The Diabetes Prevention Program (Knowler and colleagues, NEJM, 2002) randomized 3,234 adults with elevated blood glucose to placebo, metformin, or a lifestyle program targeting at least 7% weight loss and 150 minutes of weekly activity. Over an average 2.8 years, the lifestyle intervention reduced diabetes incidence by 58% (95% CI 48% to 66%) compared with placebo, more effective than metformin’s 31% reduction.

For preventing cardiovascular events in people who already had type 2 diabetes, the picture is different. Look AHEAD (NEJM, 2013) randomized 5,145 overweight or obese adults with type 2 diabetes to an intensive weight-loss lifestyle intervention or to diabetes support and education. The intervention group lost significantly more weight throughout the trial and had greater improvements in fitness and most cardiovascular risk factors. The trial was stopped early for futility at a median 9.6 years: the composite outcome of cardiovascular death, nonfatal heart attack, nonfatal stroke, or hospitalization for angina occurred at nearly identical rates in both groups (hazard ratio 0.95, 95% CI 0.83 to 1.09, not statistically significant). A large, well-conducted randomized trial of intentional weight loss did not reduce cardiovascular events in this population, a genuine null result worth stating plainly rather than working around.

Muscle and bone during weight loss, and what exercise mode changes

A separate concern in older adults specifically is that weight loss, whatever its other effects, tends to reduce muscle and bone mass along with fat mass. Villareal and colleagues (NEJM, 2011) randomized 107 obese adults 65 and older to a control group, a diet-only group, an exercise-only group, or a combined diet-plus-exercise group. Body weight fell by about 10% in the diet group and 9% in the diet-exercise group, but lean body mass and hip bone mineral density fell less in the diet-exercise group (3% and 1% reductions) than in the diet-only group (5% and 3% reductions), while physical function improved most in the combined group. A follow-up trial by the same group, Villareal and colleagues (NEJM, 2017), randomized 160 obese older adults undergoing weight loss to aerobic training, resistance training, combined training, or no exercise, and found the combined-training group had the largest gains in physical function while lean mass and hip bone density declined less in the combined and resistance groups than in the aerobic-only group. Across both randomized trials, the exercise type paired with weight loss changed how much muscle and bone were lost along the way, not just how much weight came off.

Where claims outrun the evidence

The obesity-paradox findings are sometimes stretched into a claim that gaining weight, or staying at a higher BMI, actively protects older adults; the observational data, complicated by reverse causation and smoking, does not establish that carrying more weight is itself protective. In the other direction, treating any amount of intentional weight loss, by any method, as automatically improving cardiovascular outcomes overstates what Look AHEAD’s specific, well-powered null result actually found for that specific composite endpoint in that specific population.

Weighing this evidence against an individual situation

BMI categories, mortality meta-analyses, and these trial results all describe patterns across large groups with a wide range of body composition, muscle mass, chronic disease, and medication use, not a specific number appropriate for any one person. Deciding whether, how, and by what method to pursue weight change, particularly for someone managing a chronic condition, taking medications affected by weight or nutrition, or already experiencing muscle loss, is a question this population-level evidence was not built to answer and belongs with a clinician who knows the relevant history.