Body Roundness Index (BRI) Calculator

Calculator

What the Body Roundness Index measures

Thomas and colleagues started from a geometric observation: a human trunk is closer to an ellipse than to the cylinder or the flat plane most anthropometric indices implicitly assume. Take the height as the ellipse's major axis and the waist circumference as the perimeter of its minor cross-section, and the shape is fully described by one dimensionless number: eccentricity, written ε, which runs from 0 for a circle to just under 1 for a needle-thin ellipse. BRI is that eccentricity turned upside down and rescaled: 364.2 − 365.5 × ε.

The consequence is that BRI is height-independent by construction. Two people with the same waist-to-height ratio get the same index whether they are 155 cm or 195 cm tall, which is precisely what you want from a shape measure and precisely what BMI fails to deliver: BMI drifts upward with height because mass scales closer to height cubed than squared.

Where BMI collapses everything into one mass-per-area figure, BRI is blind to weight altogether. That is a strength and a weakness in one. It means a heavily muscled athlete with a flat stomach is not penalised for being dense, because BRI never sees the scale reading. It also means BRI cannot tell you anything about total body fat, lean mass or energy needs; it describes shape and only shape.

How to measure yourself for it

Everything hangs on the waist reading, which is by far the easiest of the two numbers to get wrong. Stand relaxed with your feet together, find the midpoint between the bottom of your lowest rib and the top of your hip bone (for most adults that is at or a little above the navel) and pass a flexible, non-stretch tape level all the way around. Check in a mirror that the tape has not ridden up at the back, which is the most common error and reads several centimetres low.

Breathe out normally and take the reading at the end of that exhale, with the tape snug against the skin but not compressing it. Measure first thing in the morning, after the bathroom and before eating or drinking, and take two or three readings and average them. Waist circumference moves two to three centimetres across an ordinary day on food and fluid alone, which is more than a month of genuine change usually amounts to.

Height should be measured without shoes, heels together, standing as tall as you can with your eyes level. Weight is needed for A Body Shape Index and the conicity index, not for BRI itself; a hip measurement, taken around the widest part of the buttocks with your feet together, unlocks the Body Adiposity Index.

• A centimetre of waist error shifts BRI by roughly 0.1 to 0.2 of a point, enough to cross a band boundary if you are near one.
• Sucking in can shave three to five centimetres off a waist reading; it is the single most common way people mislead themselves.
• Use the same tape, the same time of day and the same landmark every time. The trend is far more reliable than any one reading.

Reading your band and the U-shaped mortality curve

The bands on this page are quintiles of the US adult population, taken from an analysis of 32,995 participants in the National Health and Nutrition Examination Survey between 1999 and 2018, followed for a median of ten years. They are descriptive, not diagnostic: they tell you roughly where you sit among American adults, not whether a doctor would act on the number.

What makes them worth showing is the shape of the association with death from any cause. Mortality was lowest in the middle quintile, running about 4.45 to 5.46. Above it, risk climbed steadily: a quarter higher in the fourth quintile and about half as high again in the top one. Below it, risk climbed as well: the leanest fifth of the distribution, under 3.41, carried a raised hazard of a similar size to the fourth quintile. The curve is U-shaped, not a straight line.

That matters for how you read your own result. A very low BRI is not a prize. Some of the raised risk at the bottom of the range is almost certainly reverse causation (illness that causes weight loss before it causes death), but the honest statement from the data is that people at both ends fared worse than people in the middle, and this page marks the lowest band as unusual rather than as best.

Where BRI came from and what the 2024 analysis found

Thomas and colleagues published the index in Obesity in 2013. They pooled three data sets containing demographics, anthropometry, dual-energy X-ray absorptiometry fat mass and magnetic-resonance-imaging visceral adipose tissue volume, built two elliptical models of the body, and derived BRI from the eccentricity of the better-fitting one. Against DXA fat percentage and MRI visceral fat, BRI performed slightly better than BMI, waist circumference or hip circumference (a modest improvement), and the paper says so.

BRI then sat quietly in the literature for a decade. What revived it was a 2024 cohort analysis in JAMA Network Open, which tracked BRI in the US survey population from 1999 to 2018 and linked it to the national death index. Two findings drove the attention it received: average BRI rose steadily across those two decades, with the sharpest increases among women and older adults, and the relationship with all-cause mortality was U-shaped rather than monotonic.

It is worth being precise about what that study does and does not establish. It is observational, so it shows association rather than cause. It reports population quintiles, not clinical thresholds, and no professional body has adopted a BRI cut-off. What it does establish is that a purely geometric index computed from two tape measurements carries real mortality signal in a nationally representative sample, which is a reasonable argument for measuring your waist, and a poor argument for treating any particular BRI number as a target.

ABSI and conicity: the other shape indices

A Body Shape Index, published by Krakauer and Krakauer in 2012, asks a different question: is your waist bigger than your weight and height predict? It divides waist circumference by BMI raised to the two-thirds power times the square root of height, all in metres, which strips out the part of your waist that simply follows from being large. The result is a number near 0.08 for most adults, and the useful part is the deviation, not the value.

ABSI deliberately has no universal cut-off. The original paper entered it into survival models as an age- and sex-specific z-score against population norms, and that is still how research uses it: an ABSI half a standard deviation above the mean for your age and sex means something, while a raw ABSI of 0.081 on its own means very little. In the 1999–2004 survey population, death rates rose roughly exponentially with above-average ABSI, and 16 per cent of population mortality hazard was attributable to high ABSI, against 5 per cent for BMI.

The conicity index, proposed by Valdez in 1991, is older and simpler in spirit. It compares your waist with the waist a perfect cylinder of your height and weight would have; the constant 0.109 is the density term that makes that cylinder read exactly 1.00. Values above 1.00 mean the trunk bulges at the waist and tapers away from it, the double-cone shape the index is named for. Like ABSI, it has no agreed clinical threshold: it is a continuous descriptor, best read against your own earlier readings.

The Body Adiposity Index belongs in a different category. Bergman and colleagues proposed it in 2011 as hip circumference divided by height in metres to the power 1.5, minus 18, fitted in Mexican-American adults and checked in African-American adults, and it estimates body fat percentage rather than describing shape. Independent validations have not been kind: BAI tends to over-read at low body fat and under-read at high, and multiple studies have found it no more accurate than BMI, waist circumference or hip circumference alone. This page computes it because people ask for it, and labels it the weakest number on the page.

How shape indices compare with plain waist-to-height

Here is the fact that reframes the whole family. BRI's ellipse has a semi-minor axis of waist ÷ 2π and a semi-major axis of height ÷ 2, so the ratio that goes into the eccentricity term is waist ÷ (π × height). BRI therefore depends on the waist-to-height ratio and nothing else. It is a strictly increasing transform of a number you can work out on a phone calculator in three seconds, which means the two can never rank two people differently or disagree about which direction you are moving.

Run the arithmetic through and the correspondences are fixed. A waist of exactly half your height gives a BRI of 3.36. The bottom of the second quintile, BRI 3.41, is a ratio of 0.503. The middle quintile that carried the lowest mortality, 4.45 to 5.46, spans ratios of about 0.555 to 0.601, comfortably above the 0.5 line that waist-to-height guidance treats as the point where central adiposity starts to matter. The top quintile begins at a ratio of about 0.662.

That is not a contradiction; it is two different kinds of number sitting side by side. The 0.5 threshold is a risk boundary drawn from cardiometabolic outcome data and endorsed in public-health guidance as a simple rule anyone can remember. The BRI quintiles are a description of where American adults actually are, and American adults are on average well past 0.5. Read the BRI band as 'where I sit in the population' and the waist-to-height ratio as 'where the evidence draws the line', and they stop appearing to argue.

The practical upshot: if you want one number to act on, use waist-to-height, because it has a threshold and you can compute it in your head. If you want to understand shape as a continuous quantity, or you are reading a paper that uses BRI, ABSI or conicity, this page gives you all four from the same measurements so you can see how they move together.

Limitations and who BRI misreads

The quintile bands come from one national population, and there is no ethnicity-specific band set for BRI. That is a real gap, because waist thresholds demonstrably are population-specific: the International Diabetes Federation lists a 90 cm male waist cut-off for South Asian, Chinese and Japanese populations against 94 cm for Europid ones. Applying US quintiles to a South Asian adult will under-call central adiposity, in the same way that a single BMI cut-off does.

BRI has not been validated in children or adolescents, whose body proportions change continuously through growth, and BMI-for-age percentile charts remain the right tool there. It is invalid in pregnancy, where the waist measurement no longer describes fat at all, and it is unreliable in anyone with significant abdominal distension from ascites, bloating or a hernia. Very short and very tall adults sit at the edges of the derivation sample, and the index has not been characterised well at those extremes.

Finally, the whole index rests on one tape measurement taken by one person, usually without training. Measurement error at the waist is the dominant source of uncertainty in every number on this page, and unlike the formula it does not average out across a single reading. Take the measurement carefully, take it the same way each time, and judge the direction of travel across three or four monthly readings rather than reacting to a single value.

• No universal clinical cut-off exists for BRI, ABSI or the conicity index; these are continuous descriptors, not diagnoses.
• Bands are US population quintiles; they are not adjusted for ethnicity, and they were not derived in children, adolescents or pregnancy.
• BRI ignores weight entirely, so it says nothing about total body fat, lean mass or how much muscle you carry.
• The evidence linking BRI to mortality is observational, a strong association, not a demonstrated cause.

What to do with the number

If your BRI sits in the upper bands and your waist-to-height ratio is over 0.5, the two agree and the response is the ordinary one: a moderate calorie deficit worked out from your maintenance needs, enough protein to protect muscle while you lose fat, and resistance training to keep what you have. Waist tends to respond earlier and more visibly than the scale, so it is a better weekly signal than body weight.

If your BRI is in the lowest band, the useful move is to look at what else is going on rather than to celebrate. A very low reading in an otherwise healthy, athletic adult is unremarkable; a very low reading alongside unintentional weight loss, fatigue or appetite change is worth a conversation with a GP.

Either way, re-measure monthly under the same conditions, log the raw waist figure rather than only the index, and compare across three or four readings. A single BRI carries all the measurement error of a single tape pull; a trend across a quarter does not.

How it's calculated

Worked example: 175 cm tall, 82 kg, 94 cm waist, 102 cm hip

1. Semi-minor axis: 94 ÷ (2 × 3.14159) = 14.96 cm, half the width of the ellipse across the waist.
2. Semi-major axis: 175 ÷ 2 = 87.5 cm, half the height.
3. Divide and square: 14.96 ÷ 87.5 = 0.1710, and 0.1710² = 0.02923.
4. Eccentricity: √(1 − 0.02923) = √0.97077 = 0.98527.
5. Body Roundness Index: 364.2 − 365.5 × 0.98527 = 364.2 − 360.12 = 4.08.
6. Read the band: 4.08 falls in the 3.41–4.44 quintile, below the middle of the US adult distribution but above the leanest fifth.
7. A Body Shape Index: BMI = 82 ÷ 1.75² = 26.78, so 26.78^(2/3) = 8.95 and √1.75 = 1.3229; ABSI = 0.94 ÷ (8.95 × 1.3229) = 0.0794, almost exactly the US adult average, so the waist is about what this weight and height predict.
8. Conicity index: √(82 ÷ 1.75) = 6.845, and 6.845 × 0.109 = 0.7461; C = 0.94 ÷ 0.7461 = 1.26.
9. Body Adiposity Index: 1.75^1.5 = 2.315, so 102 ÷ 2.315 = 44.06, and 44.06 − 18 = 26.1% estimated body fat.
10. Cross-check against waist-to-height: 94 ÷ 175 = 0.54, which is past the 0.5 line even though the BRI band sits below the population middle. Both readings are correct: 0.5 is an evidence-based risk threshold, while the BRI band is a population percentile.

Where this number is used in the real world

• Population epidemiology, the setting BRI was revived for, where height and waist were recorded on tens of thousands of people and imaging was not.
• Primary care and community health checks, as a second look when a BMI reading does not match the person in front of the clinician.
• Self-tracking through a fat-loss phase, because waist usually moves before scale weight does and BRI rescales that change onto a stable index.
• Research into visceral adipose tissue where MRI or DXA is impractical, since BRI was derived against MRI-measured visceral fat volume.
• Comparing shape fairly across very different heights, which BMI cannot do because it drifts upward with stature.
• Teaching and health communication, as a concrete demonstration of why fat distribution and total mass are different questions.
• Reading the literature: papers increasingly report BRI, ABSI or conicity, and this page reproduces all three from the same measurements so a published figure can be placed against your own.

Frequently asked questions

Keep going

A single number rarely tells the whole story. Alongside the body roundness result, the waist-to-height calculator, the waist-to-hip calculator, the RFM calculator, the BMI calculator and the body fat calculator each add a different angle on the same measurements. For the reasoning behind the numbers, read Measure your waist, BMI vs body fat vs waist and BMI for athletes.

Sources

1. Thomas DM, Bredlau C, Bosy-Westphal A, et al. Relationships between body roundness with body fat and visceral adipose tissue emerging from a new geometrical model. Obesity 2013;21:2264–71. doi.org/10.1002/oby.20408
2. Zhang X, Ma N, Lin Q, et al. Body Roundness Index and All-Cause Mortality Among US Adults. JAMA Netw Open 2024;7:e2415051. doi.org/10.1001/jamanetworkopen.2024.15051
3. Krakauer NY, Krakauer JC. A new body shape index predicts mortality hazard independently of body mass index. PLoS One 2012;7:e39504. doi.org/10.1371/journal.pone.0039504
4. Bergman RN, Stefanovski D, Buchanan TA, et al. A better index of body adiposity. Obesity 2011;19:1083–9. doi.org/10.1038/oby.2011.38
5. Valdez R. A simple model-based index of abdominal adiposity. J Clin Epidemiol 1991;44:955–6. doi.org/10.1016/0895-4356(91)90059-I