We need new terms for a new era. The concept of Body Mass Index — BMI — was created by a Belgian astronomer and statistician in 1835 as a way to describe the "average man." It is a number derived from a single, blunt calculation: a person's weight in kilograms divided by height in meters squared. That is the entire formula. No blood work. No scan. No clinical nuance.
And yet, BMI has become the de facto standard by which health organizations, insurers, electronic health records, and now major telehealth platforms screen patients, set policy, and determine who qualifies for life-changing GLP-1 medications. The WHO, the CDC, the FDA — they all lean on BMI. Ro, WW, Medvi, and other large-scale telehealth obesity programs rely on it as both an eligibility screen and a measure of treatment success. Some rely on self-reported BMI.
"BMI is unquestionably useful as an epidemiologic tool. But it is of doubtful accuracy and limited usefulness in treating specific individuals for obesity."
— Dr. Lael E. ForbesWhy BMI Falls Short
It cannot distinguish fat from muscle
Because BMI uses only height and weight, it treats all mass as equivalent. A highly muscular athlete may carry a BMI above 30 and be flagged as "obese," despite having low body fat and excellent metabolic health. Conversely, someone who is "skinny fat" — low muscle, high fat percentage — may show a perfectly normal BMI while harboring elevated metabolic risk. BMI both over-diagnoses and under-diagnoses in the same patient population.
It ignores where fat lives in the body
Visceral fat — the fat packed around organs in the abdomen — is strongly linked to cardiovascular disease, type 2 diabetes, and insulin resistance. Subcutaneous fat, stored under the skin around the hips and thighs, carries far less risk. Two patients can share an identical BMI yet have completely different fat-distribution profiles and therefore completely different health trajectories.
It doesn't account for age, sex, or ethnicity
Women naturally carry higher body fat percentages than men at the same BMI. Older adults lose muscle mass (sarcopenia) and accumulate fat, meaning a "normal" BMI can mask significant frailty risk. Certain Asian populations develop metabolic disease at far lower BMI thresholds than standard cutoffs reflect. Applying universal BMI thresholds across all demographics doesn't just limit personalized care — it actively misdirects it.
Adult BMI categories are not applicable to children and adolescents, whose bodies are still developing and require age- and sex-specific percentile interpretation. BMI is similarly invalid during pregnancy, where weight includes fetus, placenta, and amniotic fluid.
It tells you nothing about metabolic health
Some people with high BMIs are metabolically healthy — normal blood sugar, cholesterol, and blood pressure. Others with normal BMIs have insulin resistance, high triglycerides, and fatty liver. Using BMI to draw clinical conclusions is like using a pair of dollar-store reading glasses when what you actually need is a microscope.
There Are Better Tools
Waist circumference and waist-to-hip ratio capture abdominal fat distribution, making them stronger predictors of heart disease and type 2 diabetes than BMI alone. DEXA — Dual-Energy X-ray Absorptiometry — remains the gold standard for full body composition analysis, accurately measuring fat mass, lean muscle, and bone density. It is, however, expensive and not widely available in private clinical settings.
Bioelectrical Impedance Analysis (BIA) via systems like InBody passes small, harmless electrical currents through the body to generate detailed estimates of body fat percentage, muscle mass, and segmental water distribution. It is quick, non-invasive, and most useful for longitudinal tracking within an individual patient over the course of treatment.
"It isn't about reaching a target weight or BMI. It's about becoming healthier, more vital, and maintaining that health and vitality."
— Dr. Lael E. ForbesEnter FML: Percent Fat Mass Loss
The metric that matters is FML — percent fat mass loss. It is calculated simply:
FML = (Fat Mass Lost ÷ Total Weight Lost) × 100
FML measures what the body is actually losing — not just how much. Data from the STEP 1 trial show that of the average 13.6 kg lost with semaglutide, approximately 62% was fat mass and 38% was lean body mass. A subsequent meta-analysis found lean mass loss accounted for roughly 25% of total GLP-1 weight loss. This is not a rounding error — it is a clinically significant finding that BMI-based assessment misses entirely.
FML focuses the entire clinical conversation on what genuinely matters: the preservation of lean muscle while driving down fat. It correlates more meaningfully with improvements in metabolic health, physical function, and long-term treatment sustainability than any number a scale produces on its own.
The Bottom Line
BMI was invented nearly two centuries ago to describe statistical populations, not to diagnose and treat individual patients. In the GLP-1 era — where millions of people are losing weight rapidly, where muscle preservation is a genuine clinical concern, and where the stakes of accurate metabolic assessment have never been higher — we owe patients better than a 19th-century astronomer's approximation. We have the technology and the metrics to do this right. The question is whether we choose to use them.