BMI and Body Fat: What Each Measure Is For

A BMI of 27 is not a statement about a person’s body fat. It is a statement about a ratio of mass to height squared, compared against a reference population. The number arrives without an error bar, but it has one. Quetelet’s index, now called BMI, was designed to describe populations, not individuals, and it cannot distinguish muscle from fat. Body-fat estimates from tape measurements carry errors of 3 to 5 percentage points; consumer scales can be worse. This page gives both figures, the equations that produce them, and the conditions under which each changes.

What BMI is

BMI is mass in kilograms divided by height in meters squared. The equation is:

BMI = weight_kg / (height_m * height_m)

A person weighing 70 kg at 1.75 m:

BMI = 70 / (1.75 * 1.75)
BMI = 70 / 3.0625
BMI = 22.9

That figure is a screening cut-point. The World Health Organization defines overweight as BMI 25.0 to 29.9 and obesity as BMI 30.0 and above for adults. These thresholds come from associations between BMI and mortality in large populations, not from a physiological switch at 25. The error around an individual BMI is not a single number; it depends on body composition, age, and ethnicity. A 2013 review in Science noted that BMI misclassifies muscular individuals as overweight and elderly individuals as normal when they have lost muscle and gained fat.

What body fat percentage is

Body fat percentage is the mass of adipose tissue divided by total body mass. It cannot be measured directly without dissection, so every method estimates it. The two most common field methods are skinfold calipers and bioelectrical impedance.

Skinfold calipers use the Jackson-Pollock equations, which predict body density from the sum of skinfolds at three or seven sites, then convert density to fat percentage using the Siri equation:

Body fat % = (495 / density) - 450

Bioelectrical impedance sends a small current through the body and estimates fat-free mass from the resistance. Consumer scales use this method. The error for skinfolds is typically 3 to 5 percentage points when performed by a trained tester; for consumer scales, the error can exceed 8 percentage points and varies with hydration, recent exercise, and skin temperature.

The table: what changes the figure

Worked example: from BMI to body fat

Suppose a 45-year-old woman weighs 68 kg, is 1.65 m tall, and has a waist circumference of 82 cm. Her BMI is:

BMI = 68 / (1.65 * 1.65)
BMI = 68 / 2.7225
BMI = 25.0

That places her at the lower boundary of the overweight category. But BMI does not say how much of that 68 kg is fat. A tape-based body fat estimate uses the US Navy method, which for women is:

Body fat % = 495 / (1.29579 - 0.35004 * log10(waist + hip - neck) + 0.22100 * log10(height)) - 450

Using waist 82 cm, hip 98 cm, neck 32 cm, height 165 cm:

waist + hip - neck = 82 + 98 - 32 = 148
log10(148) = 2.170
log10(165) = 2.217
Body fat % = 495 / (1.29579 - 0.35004*2.170 + 0.22100*2.217) - 450
Body fat % = 495 / (1.29579 - 0.7596 + 0.4900) - 450
Body fat % = 495 / 1.0262 - 450
Body fat % = 482.4 - 450
Body fat % = 32.4%

The estimate is 32.4%, with a typical error of 3 to 5 percentage points. So the plausible range is roughly 27% to 37%. For a woman in her 40s, 32% is within the average range; 27% is lean, 37% is above average. The BMI of 25.0 does not tell you which of those is true.

What each measure is for

BMI is useful for tracking weight status across populations and for flagging individuals who may benefit from further assessment. It is not a diagnostic tool. A BMI above 30 in a population predicts higher rates of type 2 diabetes and cardiovascular disease, but an individual with BMI 31 who is muscular may have a lower risk than an individual with BMI 24 who has low muscle mass and high visceral fat.

Body fat percentage is more directly relevant to health because it distinguishes fat mass from fat-free mass. But it is harder to measure accurately outside a laboratory. The error bars matter more than the number. A body fat estimate of 25% from a consumer scale could mean 20% or 30%. That range is too wide to guide daily decisions.

Waist circumference is a third measure. It correlates with visceral fat, which is more strongly associated with metabolic risk than total fat. WHO cut-points are 94 cm for men and 80 cm for women of European descent, with lower thresholds for South Asian and Chinese populations. The measurement error is smaller than for body fat, but it still depends on where the tape is placed.

What people get wrong

The most common mistake is treating BMI as a personal verdict. It is natural because the number is easy to calculate and comes with a category label. But the label was assigned by comparing a person to a population distribution, not by measuring their body. A muscular athlete with BMI 28 is not overweight in the sense that predicts disease; an elderly person with BMI 22 may have lost muscle and gained fat, a condition called sarcopenic obesity, which BMI cannot detect.

The second mistake is assuming body fat scales are accurate. They are not. They measure resistance, which depends on total body water. Drinking a liter of water before stepping on the scale can change the reading by several percentage points. The number is a estimate, not a measurement.

The third mistake is using BMI to set a weight-loss goal. A BMI of 25 is not a physiological target. The relationship between BMI and mortality is J-shaped: risk is elevated at both low and high BMI, and the lowest risk is often in the 22-25 range for some populations, but the curve is shallow. For an individual, the more useful target is a waist circumference below the cut-point and a body fat percentage within a range that a clinician can help interpret.

Energy balance and why the line is not straight

BMI and body fat change when energy intake and expenditure differ. The energy balance equation is:

Change in body energy stores = Energy intake - Energy expenditure

Energy expenditure has three main components: resting metabolic rate (60-70%), physical activity (20-30%), and thermic effect of food (10%). Resting metabolic rate is predicted by the Mifflin-St Jeor equation:

Men: RMR = 10 * weight_kg + 6.25 * height_cm - 5 * age + 5
Women: RMR = 10 * weight_kg + 6.25 * height_cm - 5 * age - 161

For the 45-year-old woman weighing 68 kg at 165 cm:

RMR = 10*68 + 6.25*165 - 5*45 - 161
RMR = 680 + 1031.25 - 225 - 161
RMR = 1325 kcal/day

Total daily energy expenditure depends on activity. If she is lightly active, multiply by 1.375:

TDEE = 1325 * 1.375 = 1822 kcal/day

A deficit of 500 kcal/day would predict about 0.45 kg of weight loss per week, using the approximation that 1 kg of body weight corresponds to 7,700 kcal. But that approximation is wrong in two directions. Early in a diet, water loss makes the scale drop faster than fat loss. Later, as weight falls, RMR falls because the equation depends on weight. A 5 kg loss would reduce RMR by about 50 kcal/day, shrinking the deficit unless intake is adjusted. The line is not straight.

This is why a BMI target is a poor goal. The energy balance equation changes as the body changes. A more useful approach is to track waist circumference and body fat percentage alongside weight, and to accept that the error bars on all three are wide.

When to seek help

This page is not medical advice. BMI and body fat estimates are screening tools, not diagnoses. Anyone with a history of disordered eating should not use calorie counting or weight tracking without clinical support. In the United States, the National Eating Disorders Association helpline is available at 1-800-931-2237. In the UK, Beat’s helpline is 0808 801 0677. These figures are for adults and do not apply to children, pregnant people, or athletes in training.

FAQ

Q: Is BMI accurate for individuals? A: Not very. BMI was designed to describe populations, not individuals. It cannot distinguish muscle from fat, so it misclassifies muscular people as overweight and elderly people as normal when they have lost muscle. The error for an individual can be 1-2 BMI units in either direction.

Q: What is a healthy BMI? A: The World Health Organization defines a healthy BMI as 18.5 to 24.9 for adults. But this range comes from population mortality data, not from a physiological threshold. An individual with BMI 26 who is muscular may be healthier than an individual with BMI 23 who has high visceral fat.

Q: How accurate are body fat scales? A: Consumer bioelectrical impedance scales typically carry errors of 5 to 8 percentage points, and the error varies with hydration, recent exercise, and skin temperature. A reading of 25% could mean 20% or 30%. They are useful for tracking trends over time under consistent conditions, not for diagnosis.

Q: Can I lower my BMI without losing muscle? A: Yes, but the rate matters. A deficit of 500 kcal/day from a mixed diet with adequate protein (1.2-1.6 g/kg body weight) and resistance training tends to preserve muscle better than a larger deficit without exercise. The Mifflin-St Jeor equation predicts that resting metabolic rate falls as weight falls, so the deficit shrinks unless intake is adjusted.

Q: What is the difference between BMI and body fat percentage? A: BMI is a ratio of weight to height. Body fat percentage is the proportion of total mass that is adipose tissue. BMI cannot tell you how much of your weight is fat; body fat percentage can, but it is harder to measure accurately. Both have error bars, and neither is a diagnosis.