What Is BMI, and What Does It Tell You?

How BMI Is Calculated

BMI uses a simple equation: divide your weight in kilograms by the square of your height in meters. A person who weighs 80 kg and stands 1.75 m tall has a BMI of 80 ÷ (1.75 × 1.75), which equals 26.1 kg/m². The imperial formula multiplies weight in pounds by 703, then divides by height in inches squared.

Belgian mathematician Adolphe Quetelet developed this ratio in 1832 as a statistical tool for studying populations, not individuals 1. It was never designed to diagnose disease in a single patient. Ancel Keys repurposed the formula in a 1972 paper and coined the term "body mass index," arguing it correlated better with body fat than other height-weight indices available at the time 2. The National Institutes of Health formally adopted BMI-based obesity classifications in 1998, aligning U.S. thresholds with World Health Organization standards 3.

The calculation takes seconds. That speed is precisely why public health agencies rely on it. Every large epidemiological study from Framingham onward has used BMI to stratify cardiovascular risk, because height and weight are universally available data points 4.

What the BMI Categories Mean

The WHO classifies adult BMI into four primary brackets: underweight (below 18.5), normal weight (18.5 to 24.9), overweight (25.0 to 29.9), and obese (30.0 and above) 5. Obesity is subdivided further into Class I (30.0 to 34.9), Class II (35.0 to 39.9), and Class III (40.0 or higher), sometimes called severe obesity.

These cutoffs are not arbitrary. They come from mortality data. A 2016 meta-analysis published in The Lancet pooled 239 prospective studies across four continents (N = 10.6 million participants) and found that all-cause mortality was lowest in the BMI range of 20 to 25 kg/m² among never-smokers without pre-existing disease 6. Each 5 kg/m² increase above 25 was associated with roughly 31% higher all-cause mortality.

The underweight category carries its own risks. BMI values below 18.5 correlated with increased mortality in the same Lancet analysis, driven by respiratory disease, infections, and frailty-related causes. This U-shaped curve, higher risk at both extremes, is one of the most replicated findings in obesity epidemiology.

Pediatric BMI works differently. Because children's body composition shifts as they grow, the CDC uses age- and sex-specific percentile charts rather than fixed cutoffs 7. A child at the 85th percentile or above is considered overweight; at or above the 95th percentile, obese.

Where BMI Falls Short

BMI cannot tell you what your body is made of. It treats all mass the same: muscle, bone, water, and fat contribute identically to the numerator. A 5'10" running back at 220 pounds registers a BMI of 31.6 (obese Class I) despite having 12% body fat. This is not a rare edge case. A study of 13,601 adults using dual-energy X-ray absorptiometry (DXA) found that BMI misclassified 48% of women and 25% of men who had excess body fat as "non-obese" 8.

Fat distribution matters at least as much as total fat mass. Visceral adipose tissue, the fat packed around abdominal organs, is far more metabolically active and inflammatory than subcutaneous fat stored under the skin. Two people with identical BMIs of 27 can have vastly different cardiometabolic profiles depending on where their fat sits. BMI is blind to this distinction.

The formula also carries a built-in bias toward taller individuals. Because it divides by height squared rather than height cubed, taller people tend to receive higher BMI values for the same body fat percentage, while shorter people receive lower values 9. Nick Trefethen, an Oxford mathematician, published a corrected formula in 2013 that uses an exponent of 2.5, but it has not been widely adopted in clinical practice.

Ethnicity introduces another layer of imprecision. Asian populations develop type 2 diabetes and cardiovascular disease at lower BMI thresholds than white European populations. The WHO expert consultation recommended that Asian-specific cutoffs (overweight at 23, obese at 27.5) be considered for public health action, though these remain advisory rather than universally implemented 10.

Dr. Francisco Lopez-Jimenez, a cardiologist at Mayo Clinic, has stated: "BMI is a reasonable screening tool, but it should never be used as a standalone diagnostic. We need to look at waist circumference, metabolic bloodwork, and ideally some measure of body composition before we make clinical decisions."

Waist Circumference and the Metrics That Fill BMI's Gaps

Waist circumference measures abdominal fat distribution directly and is a stronger predictor of cardiovascular events than BMI alone. The INTERHEART study (N = 27,098 across 52 countries) found that waist-to-hip ratio was a more significant predictor of myocardial infarction than BMI in every ethnic group studied 11. Current NHLBI guidelines flag disease risk as elevated when waist circumference exceeds 40 inches (102 cm) in men or 35 inches (88 cm) in women 3.

Combining BMI with waist circumference improves risk prediction substantially. The table below reflects NHLBI guidance:

| BMI Category | WC ≤ 40 in (men) / ≤ 35 in (women) | WC > 40 in (men) / > 35 in (women) | |---|---|---| | 25.0 to 29.9 | Increased risk | High risk | | 30.0 to 34.9 | High risk | Very high risk | | 35.0 to 39.9 | Very high risk | Very high risk |

Body fat percentage, measured by DXA, bioelectrical impedance, or hydrostatic weighing, provides a direct reading of adiposity. DXA scans can also quantify visceral fat area and appendicular lean mass. The American Association of Clinical Endocrinology (AACE) 2023 guidelines encourage clinicians to assess body composition when BMI and clinical presentation are discordant 12.

The Edmonton Obesity Staging System takes a different approach entirely. Rather than relying on a number, it grades obesity from Stage 0 (no apparent risk factors) through Stage 4 (severe end-organ damage) based on clinical findings 13. A 2011 study of 8,143 participants from NHANES III showed that Edmonton stage predicted mortality independently of BMI: patients in Stage 2 or 3 had significantly higher all-cause mortality regardless of whether their BMI was 30 or 40.

BMI and Metabolic Health: The "Healthy Obese" Debate

Some individuals with BMIs above 30 show normal blood pressure, normal fasting glucose, and healthy lipid panels. Researchers have called this phenotype "metabolically healthy obesity" (MHO). It is real, but it may be temporary.

A 2018 study in the Journal of the American College of Cardiology followed 3.5 million UK adults (mean follow-up 5.4 years) and found that even metabolically healthy obese individuals had a 50% higher risk of coronary heart disease compared to metabolically healthy normal-weight individuals 14. Over longer follow-up periods, roughly half of MHO individuals transition to metabolically unhealthy obesity. The Endocrine Society's 2024 clinical practice guideline on obesity pharmacotherapy notes that BMI should be interpreted "in the context of a comprehensive metabolic and functional assessment, not as a sole criterion for treatment decisions" 15.

The opposite phenotype, metabolically unhealthy normal weight (sometimes called "thin outside, fat inside" or TOFI), also exists. These individuals have BMIs in the normal range but carry excess visceral fat, elevated inflammatory markers, and insulin resistance. A BMI of 23 tells you nothing about this risk. Waist circumference and fasting metabolic panels catch what BMI misses here.

How BMI Influences Medical Decision-Making Today

Despite its limitations, BMI remains embedded in clinical protocols. The FDA requires a BMI of 30 or higher (or 27 with at least one weight-related comorbidity) for prescribing GLP-1 receptor agonists like semaglutide for chronic weight management 16. Bariatric surgery guidelines set the threshold at BMI 40, or BMI 35 with obesity-related conditions 17.

Insurance coverage decisions often hinge on BMI documentation. Prior authorization for anti-obesity medications typically requires chart-documented BMIs on at least two separate visits. This gatekeeping function means that patients with significant metabolic disease but a BMI of 26 can face coverage denials, while patients with a BMI of 31 and no metabolic abnormalities may qualify.

The AACE framework proposed in 2023 attempts to move beyond this rigid number. It introduces a complications-centric model where the presence and severity of obesity-related diseases (type 2 diabetes, obstructive sleep apnea, NASH, osteoarthritis) drive treatment intensity rather than BMI alone 12. This shift is gaining traction. Dr. W. Timothy Garvey, chair of the AACE Obesity Scientific Committee, wrote: "We need to treat obesity as a disease defined by its clinical impact, not by a number on a BMI chart."

When to Check Your BMI and What to Do With It

The U.S. Preventive Services Task Force (USPSTF) recommends that clinicians screen all adults for obesity by calculating BMI at routine visits 18. Adults with a BMI of 30 or higher should be offered or referred to intensive behavioral interventions. For children and adolescents aged 6 and older, the USPSTF similarly recommends BMI screening and referral to comprehensive behavioral programs when indicated.

If your BMI falls in the overweight or obese range, the next step is not panic. Request a fasting metabolic panel (glucose, HbA1c, lipid panel, liver enzymes) and have your waist circumference measured. These data points, combined with blood pressure and family history, produce a risk profile that is far more informative than BMI in isolation.

A BMI in the normal range does not guarantee metabolic safety. If you carry weight primarily in your midsection, have a family history of type 2 diabetes or cardiovascular disease, or belong to an ethnic group with lower BMI thresholds for metabolic risk (South Asian, East Asian, Southeast Asian), a proactive metabolic workup is warranted even at a BMI of 23 or 24.

Track BMI over time rather than fixating on a single measurement. A BMI that has risen from 24 to 28 over three years signals a trajectory that merits intervention, even though 28 by itself is only "overweight." Trends reveal more than snapshots.

The single most useful thing BMI can do for an individual patient is start a conversation. That number on the chart should prompt questions about diet, physical activity, sleep, metabolic markers, and body composition. It should not end the conversation. Pair BMI with waist circumference (measured at the iliac crest), request a fasting glucose and lipid panel, and discuss results with your clinician at your next visit.