BMI is one of the most widely used clinical tools, yet also one of the most misunderstood. Used correctly, it is a useful population-level screening tool. Used incorrectly, it leads to diagnostic errors, unfair treatment of patients, and missed clinical risk. This guide covers everything doctors actually need to know.
What BMI Actually Measures
Body Mass Index = weight (kg) ÷ height² (m²). It was developed by the Belgian mathematician Adolphe Quetelet in 1832 — not as a clinical tool, but as a population-level statistical measure of human body proportions. The WHO adopted it as a clinical screening tool in the 1990s, principally because it is cheap, universal, non-invasive, and moderately correlated with body fat percentage at the population level.
What BMI does not measure:
- Body fat percentage
- Fat distribution (visceral vs. subcutaneous)
- Muscle mass
- Bone density
- Metabolic health
- Cardiovascular risk (directly)
WHO BMI Classification
| BMI Range | Classification |
|---|---|
| < 18.5 | Underweight |
| 18.5 – 24.9 | Normal weight |
| 25.0 – 29.9 | Overweight |
| 30.0 – 34.9 | Obese Class I |
| 35.0 – 39.9 | Obese Class II |
| ≥ 40.0 | Obese Class III (Morbid obesity) |
These cut-offs were derived from predominantly White European populations. As discussed below, they require modification for other ethnic groups.
Limitations Every Clinician Must Know
Muscle vs. Fat
BMI does not distinguish between lean mass and fat mass. A professional rugby player with a BMI of 31 and 12% body fat is not obese in any clinically meaningful sense. Conversely, a sedentary adult with a BMI of 24 may have a very high body fat percentage — a phenomenon called “normal weight obesity” — with significant associated metabolic risk.
Fat Distribution
Visceral adiposity — fat deposited around the abdominal organs — carries substantially greater metabolic and cardiovascular risk than subcutaneous fat deposited on the hips and thighs. BMI cannot distinguish between these two patterns. Waist circumference and waist-to-hip ratio provide additional, clinically important information about fat distribution that BMI cannot.
Ethnicity
This is the most clinically important limitation for a diverse patient population. Large-scale epidemiological studies have consistently shown that South Asian populations (Indian, Pakistani, Bangladeshi, Sri Lankan) carry equivalent metabolic and cardiovascular risk at lower BMI values than White European populations. The same is true, to a lesser extent, for East and South-East Asian populations.
WHO/NICE recommendations for South Asian populations:
- BMI ≥ 23.0 kg/m² = increased risk (equivalent to 25.0 in White European populations)
- BMI ≥ 27.5 kg/m² = high risk (equivalent to 30.0 in White European populations)
Failure to apply ethnic-specific cut-offs leads to systematic under-diagnosis of metabolic risk in South Asian patients — a patient population already disproportionately affected by type 2 diabetes and cardiovascular disease.
Sex Differences
Women naturally carry a higher proportion of body fat than men at equivalent BMI values. This is physiologically normal and does not represent equivalent metabolic risk. However, post-menopausal redistribution of fat to central/visceral patterns does increase metabolic risk in women — context that BMI alone cannot capture.
Age
In older adults (particularly over 70), the relationship between BMI and mortality is paradoxical — there is a “BMI paradox” where slightly overweight individuals have better outcomes than those in the “normal” range. This likely reflects the protective effect of lean muscle mass and nutritional reserve in the context of acute illness. Sarcopenic obesity — high fat, low muscle mass — is the particularly high-risk pattern in older adults and cannot be identified by BMI alone.
Waist Circumference: The Missing Measurement
NICE guidelines recommend measuring waist circumference in addition to BMI for all patients where weight management is being assessed. Waist circumference directly reflects visceral adiposity and is a better predictor of cardiometabolic risk than BMI alone.
Risk thresholds (measured at the level of the umbilicus):
| Risk Level | White European Men | White European Women | South Asian Men | South Asian Women |
|---|---|---|---|---|
| Increased risk | ≥ 94 cm | ≥ 80 cm | ≥ 90 cm | ≥ 80 cm |
| High risk | ≥ 102 cm | ≥ 88 cm | ≥ 90 cm | ≥ 80 cm |
BMI in Clinical Decision-Making
Despite its limitations, BMI remains useful as a screening tool for initiating further assessment. The correct clinical approach:
- Calculate BMI using height and weight
- Apply ethnic-specific cut-offs where relevant
- Measure waist circumference
- Assess body composition context (athletic, sarcopenic, or neither)
- Combine with metabolic markers: fasting glucose, HbA1c, lipid profile, blood pressure
- Use the composite picture — not BMI alone — to guide clinical decisions
BMI-Based Drug Dosing
In pharmacology, body weight and BMI are used to guide dosing in several contexts:
- Renal dosing: eGFR calculations use actual body weight, which is affected by obesity
- Chemotherapy: Capped dosing at certain BSA thresholds is increasingly abandoned — guidelines now support full BSA-based dosing in obese patients for most chemotherapy regimens
- VTE prophylaxis: Patients with BMI ≥ 40 typically require higher molecular weight heparin doses — most trusts have obese patient LMWH dosing tables
- Anaesthesia: Obese patients have altered pharmacokinetics — some drugs should be dosed on Ideal Body Weight (IBW) (e.g. opioids, benzodiazepines), others on Total Body Weight (TBW) or Adjusted Body Weight (ABW). Always check your anaesthetic formulary.
MedDraftPro’s Ideal Body Weight Calculator and BMI Calculator are free tools for rapid clinical use at the bedside.
Communicating BMI to Patients
How you discuss weight and BMI with patients matters. The evidence shows that:
- Non-judgemental, person-centred language produces better outcomes than stigmatising or directive approaches
- Always ask permission before discussing weight — “Would it be helpful to talk about weight and how it might relate to your health today?”
- Focus on health behaviours (activity, sleep, stress, diet quality) rather than the BMI number itself
- Acknowledge the complexity — genetics, socioeconomic factors, medications, and mental health all influence weight
Summary
BMI is a screening tool, not a diagnostic criterion. Used thoughtfully — with appropriate ethnic adjustments, combined with waist circumference, and interpreted in clinical context — it is a useful first step in metabolic risk assessment. Used naively — as a single number that defines a patient’s health — it is inadequate and potentially harmful. Every clinician should understand both its utility and its limitations.