Estimating Body Mass Index After Amputation or with Limb Difference

Standard BMI calculation for amputees and wheelchair users is invalid. The formula was designed for a whole body with all four limbs present. The weight of the missing limb or the altered body geometry means the quotient produced by dividing weight in kilograms by height in metres squared tells you nothing reliable about body fat or health risk. You cannot take a BMI number from a standard scale and act on it if you have an amputation, a limb difference, or use a wheelchair. The answer is not to ignore body composition screening. Use correction factors or alternative anthropometric methods.

The core problem is arithmetic. BMI divides total body weight by the square of height. A person with a lower-limb amputation carries roughly 16 percent less body weight than their pre-amputation weight. The leg and foot are gone. But the height denominator assumes the original stature. The resulting BMI is artificially low. A person who would have been in the overweight range may land in the normal range, and a person who would have been obese may land in the overweight range. That misclassification matters when clinicians or screening programmes use BMI to flag risk of diabetes, hypertension, or cardiovascular disease.

The most widely cited approach to estimating body weight after amputation comes from a 1995 paper by Osterkamp published in the Journal of the American Dietetic Association. Osterkamp provided percentages of total body weight assigned to each body segment, based on cadaver studies. To estimate corrected body weight, divide the measured weight on the scale by one minus the decimal proportion of the missing segment.

Applying the Correction Figures

For a below-knee amputation, the missing segment is the foot and lower leg, which accounts for approximately 5.9 percent of total body weight. The formula is: estimated total body weight = measured weight ÷ (1 − 0.059). For a below-knee amputation of one leg, that is measured weight ÷ 0.941. For a full leg including the thigh, the proportion is approximately 15.9 percent, so the divisor is 0.841. For a full arm including the hand, about 4.9 percent, divisor 0.951. For a forearm and hand, about 2.3 percent, divisor 0.977. For a hand only, about 0.7 percent, divisor 0.993.

Limitations of the Correction

You then use the corrected weight with the person's pre-amputation stature, or with a stature estimated from arm span or demispan if the original figure is unknown. The corrected weight divided by the square of that stature gives a BMI that is more comparable to population norms than the uncorrected number. This is the best available method, but it has limitations. The Osterkamp percentages come from a small number of cadavers, not from living amputees, and they assume average body proportions that may not match a given individual. No WHO, CDC, or NICE guideline endorses a specific BMI threshold adjustment for amputees. The correction is a clinical workaround, not a validated standard.

If you use a wheelchair or have a limb difference that prevents standing stature assessment, you cannot get the height component of BMI from a standard stadiometer. Two surrogate measures exist.

Demispan

Demispan is the distance from the centre of the sternal notch to the tip of the middle finger, measured with a tape along the fully extended arm. It correlates with stature in ambulatory populations and can be used to estimate height via published regression equations.

Seated Height

Seated height is measured from the seat surface to the top of the head with the person sitting upright, with hips and knees at 90 degrees. Seated height alone does not give total height, but if the person has a known pre-amputation stature, seated height can track changes over time.

Neither method has a universal conversion formula validated for wheelchair users or people with limb difference. The error in height estimation from these surrogates adds to the error from the weight correction. For a clinical setting, the most practical path is to use the person's best known pre-amputation stature if it is documented. If it is not, use demispan with the most population-appropriate regression equation available. Do not rely on self-reported height. The systematic review by Connor Gorber et al. (2007) found that self-report overestimates height by 0.6 to 2.0 cm and underestimates weight by 0.4 to 1.5 kg, yielding a BMI that is 0.5 to 1.5 kg/m² too low on average.

When BMI cannot be reliably calculated, the most useful alternative methods are waist circumference, waist-to-height ratio, and skinfold assessment. These do not require total body weight and do not depend on the assumptions about missing limbs or altered geometry that break BMI.

Waist Circumference and Waist-to-Height Ratio

Waist circumference is measured with a tape at the midpoint between the last palpable rib and the top of the iliac crest, at the end of normal expiration, per the WHO protocol. The NIH/NHLBI protocol instead uses the top of the iliac crest. Choose one protocol and stick with it for consistency. For a wheelchair user, the seated position alters abdominal geometry, and no validated threshold exists for this population. The best you can do is to apply the general risk cut-offs from the issuing body and note the uncertainty. The NIH/NHLBI thresholds for increased cardiometabolic risk are above 102 cm for men and above 88 cm for women, from their 1998 Clinical Guidelines. The International Diabetes Federation (IDF) uses lower thresholds for non-European populations: for South Asian and Chinese men, ≥94 cm, and for women, ≥80 cm. For Japanese men, ≥85 cm, women ≥90 cm.

Waist-to-height ratio divides the waist circumference by height in the same units. NICE guideline CG189, published in 2022, advises that a ratio of 0.5 or above is a population-level health signal of central adiposity. This ratio removes the need for sex-specific cut-offs and adjusts for stature. For a wheelchair user, height must still be estimated from demispan or a known pre-amputation value, but the error in the ratio is lower than in BMI because the height term is linear, not squared.

Skinfold Measurement

Skinfold assessment using callipers at three or seven sites estimates subcutaneous fat and from it body fat percentage. The accuracy depends on technician skill and the prediction equation chosen. For wheelchair users, standard skinfold equations developed on ambulatory populations may overestimate or underestimate body fat because of muscle atrophy in the lower body and altered fat distribution. The error is ±3-5 percentage points compared to DXA, and inter-rater variability is ±2-3 percentage points. If you have access to a trained measurer, skinfolds are more informative than a corrected BMI that carries compounding errors.

Bioelectrical Impedance Analysis and DXA

Bioelectrical impedance analysis (BIA) passes a small current through the body and estimates fat and lean mass. For wheelchair users, standard BIA equations assume ambulatory body geometry and hydration. The leg segment is excluded or misrepresented, yielding fat-free mass errors of 3-8 percent. BIA is also highly sensitive to hydration status, recent exercise, and device algorithm. A single reading under uncontrolled conditions is not reliable. Dual-energy X-ray absorptiometry (DXA) is considered a reference method for body composition, but it may misclassify tissue in regions of severe disuse atrophy or heterotopic ossification. DXA is not a routine screening tool. It is a clinical or research method.

For someone with an amputation or limb difference, body composition measurement disability means that every method has a known failure mode. The question is not which method is perfect. None is. The question is which method has the smallest error for your specific situation.

• If you have a lower-limb amputation: Use the Osterkamp correction for weight. Measure height from demispan or known pre-amputation stature. Calculate corrected BMI. Then use waist-to-height ratio as a cross-check. If the waist-to-height ratio is above 0.5, central adiposity risk is elevated regardless of what the corrected BMI says.
• If you use a wheelchair and have no amputation: Do not use BMI at all. Use waist circumference measured in the seated position, with the same protocol each time. Apply the NIH/NHLBI or IDF thresholds. Add waist-to-height ratio using a demispan-derived height estimate. If you can get a skinfold assessment from a trained technician, use it as a body fat estimate. Do not trust a consumer BIA scale.
• If you have a limb difference that does not involve amputation: Height measurement is the primary barrier. Use demispan if the arm is unaffected. If both arms are affected, you need a clinical consultation to find a valid height surrogate. Do not use self-reported height from memory. The error is too large.

What Goes Wrong Most Often

The single most common mistake is treating a corrected or surrogate BMI number as if it carries the same meaning as a standard BMI number. It does not. The thresholds from WHO, NIH/NHLBI, and CDC were developed on general ambulatory populations. No large-scale prospective study of BMI and all-cause mortality specific to wheelchair users or amputees exists. The Global BMI Mortality Collaboration study (10.6 million participants, 2016) found a hazard ratio nadir at 20.0-25.0 kg/m² in the general population, but that does not apply to these groups. You cannot use a corrected BMI to decide whether you are at risk. You can only use it as a screening signal that something may need further investigation.

If you need a body composition screening and have an amputation, limb difference, or use a wheelchair, your single most practical step is to get a waist-to-height ratio measurement done by a trained clinician who uses a consistent protocol. That measurement, combined with a corrected BMI from the Osterkamp method if applicable, gives you two independent signals. If either signal crosses the population risk threshold, waist-to-height ratio above 0.5, or waist circumference above the NIH/NHLBI cut-off, the appropriate response is not a self-diagnosis. It is a medical consultation that can include DXA, dietary assessment, and metabolic blood work. Do not treat a number from a corrected BMI as a verdict. Treat it as a question that requires a more complete answer.