Waist-to-Hip Ratio Calculator (No Borrowed Thresholds)
Waist-to-Hip Ratio Calculator: the Ratio, the Measurement Error That Dominates It, and Why No Threshold Table Appears on This Page
Waist divided by hip, computed from the figures as typed so no unit conversion enters the ratio, with the arithmetic of the measurement error printed beside it — because where the tape goes matters more here than anything else on the page, and published comparisons of waist sites in the same people differ by up to 6.9 cm in men and 10.1 cm in women. There is no risk threshold on this page and that is deliberate. The sex-specific cut-offs in universal circulation come from a WHO expert consultation, WHO publications are licensed NonCommercial, and this site carries advertising; no other authoritative body publishes a waist-to-hip cut-off, and the one US national guideline that addresses the question recommends against using the ratio at all and gives none. What the page prints instead is the published evidence on what the ratio predicts, and the two waist thresholds that are cleanly licensed.
These are measurements of body size. They are not a judgement about you and they are not a diagnosis. They cannot see muscle, bone, where fat sits, or how healthy you are — two people with the same figure can be in very different health. Treat anything here as one rough signal among many, and speak to a doctor or dietitian before making a substantial change to how you eat.
Waist divided by hip, with the measurement error that governs it
Waist 86 cm, hip 98 cm, men’s threshold rows
One division, and the arithmetic of everything that happens before it
- waist
- a circumference whose level is not uniquely defined. The US national survey protocol takes it just above the uppermost lateral border of the right ilium; other protocols use the midpoint between the lowest rib and the iliac crest, the narrowest point of the torso, or the umbilicus. Differences between pairs of sites in the same people have been measured at 0.2 to 6.9 cm in men and 0.1 to 10.1 cm in women
- hip
- the circumference at the widest part of the buttocks. A maximum rather than a bony landmark, so it depends on the measurer’s eye. NHANES does not collect it at all in its current body-measures protocol, which is why no public-domain reference distribution for this ratio exists
- as entered
- the ratio is unitless, so the page divides the typed figures and never converts. Converting both to centimetres first changes the floating-point quotient in about a third of inch pairs and the second decimal place in five of 100,651 swept for this page. The conversion at exactly 2.54 is used only for the rows quoted in centimetres
- 1/hip
- the sensitivity to the waist. At a hip of 98 cm, one centimetre of waist is 0.0102 of the ratio, so the last digit of a two-decimal figure is worth about one centimetre of tape placement and half of it is worth about half a centimetre
- −waist/hip²
- the sensitivity to the hip, which is slightly smaller in magnitude whenever the waist is smaller than the hip, and of the opposite sign. Because the two errors are independent and work in opposite directions, the ratio is noisier than either measurement
- no threshold
- the page applies none, prints none and paraphrases none. The reasoning is in the note under the result and in the table of sources below. This is the one page in this set where “no cleanly licensed boundary could be sourced” is the finding rather than a gap
Worked example
Waist 86 cm, hip 98 cm, men's threshold rows
The division. 86 / 98 = 0.8775510…, printed as 0.88. That is the entire arithmetic of this page and it is exact: a quotient of two doubles is correctly rounded, so unlike BMI there is no squaring to lose precision in, and a sweep of 771,801 waist-and-height pairs and a comparable sweep of waist-and-hip pairs found no case where the ratio was misplaced relative to exact arithmetic.
Now the part that actually determines the answer. One centimetre of waist is 1/98 = 0.0102 of the ratio, so the second decimal place is worth almost exactly one centimetre of tape placement. One centimetre of hip is 86/98² = 0.0090, of the opposite sign. The errors are independent, so they compound: four centimetres of combined uncertainty, which is unremarkable between two careful measurers using two different published protocols, is worth about 0.08 of the ratio — eight times the last printed digit.
How large that is against a real protocol difference. A published comparison of waist measurement sites in the same people found differences between pairs of sites ranging up to 6.9 cm in men. Applying 6.9 cm to this waist gives 92.9 / 98 = 0.948 against 0.878 — 0.07 of the ratio from nothing but where the tape went. In women the same study found up to 10.1 cm between extreme sites, which on a hip of 105 cm is 0.096 of the ratio. The commonly quoted sex-specific cut-offs differ from each other by 0.05. The measurement error is larger than the distance between the two thresholds people compare themselves against.
What the page does not do next, and why. It does not print a category. The sex-specific cut-offs almost every other waist-to-hip calculator shows are from a World Health Organization expert consultation; WHO publications are CC BY-NC-SA 3.0 IGO, the NonCommercial term applies to a site carrying advertising, and reproducing them — or paraphrasing them, or taking them from a third party that republished them — is the same breach three ways. Copyright does not launder.
And no substitute could be found, which turned out to be the interesting result. The 1998 US national guideline on the identification, evaluation and treatment of overweight and obesity in adults is a public-domain US Government work and would have been the obvious source. It gives no waist-to-hip cut-off. It states that waist circumference has been found to be a better marker of abdominal fat content than the ratio, that whether the ratio adds anything beyond the waist is uncertain, and that between the two the waist carries greater prognostic significance — and concludes that in clinical practice abdominal fat should be assessed by measuring the waist. Indian national guidance, in its 2025 revision, reaches a similar conclusion and prefers waist-to-height ratio. So the absence of a threshold here is not only a licensing outcome; it is also the direction the guidance itself points.
What is printed instead. The ratio, with its sensitivity. The evidence on what the ratio predicts, which is strong and is about position in a distribution rather than about a line: in 27,098 people across 52 countries the top quintile of waist-to-hip ratio carried an odds ratio of 2.52 for a first heart attack against the bottom quintile, while BMI's association vanished once the ratio was adjusted for. And the two waist thresholds that are cleanly licensed: 102 cm for men and 88 cm for women from the 1998 US guideline, and 90 cm for men and 80 cm for women from Indian national guidance. For this reader at 86 cm, that is 16 cm below the US figure and 4 cm below the Indian one — and the 4 cm is comfortably inside the protocol error computed two steps above, which is exactly the point.
What this page will not tell you. Whether this ratio is a problem, what it should be, how to change it, or anything about your health. It divides one tape measurement by another, says how much the answer moves when the tape moves, and names the published evidence and the published thresholds with their sources and their licences. Everything past that needs a clinician with your blood pressure, your lipids, your glucose and your history.
Every source of a waist-to-hip threshold this page could find, what it says, and whether it can be used here
| Source | Does it give a waist-to-hip cut-off? | Licence position for a commercial, advertising-supported site | Used here? |
|---|---|---|---|
| World Health Organization expert consultation on waist circumference and waist-hip ratio (2008 report, published 2011) | Yes — the sex-specific pair in universal circulation | WHO publications are licensed CC BY-NC-SA 3.0 IGO. The NonCommercial term applies to this site. Reaching the same figures through a third party that has republished them is the same breach: copyright does not launder | NO. Not reproduced, not paraphrased, and the figures are not printed anywhere on this page |
| NHLBI Obesity Education Initiative, Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults (NIH, 1998) | No. It gives none, and it recommends against using the ratio: it states that waist circumference is a better marker of abdominal fat content, that whether the ratio adds anything beyond the waist is uncertain, that the waist carries greater prognostic significance, and that abdominal fat should be assessed by measuring the waist | US Government work, public domain | YES, for what it does give: waist circumference figures of 102 cm in men and 88 cm in women, and its stated position on the ratio, which this page reports |
| Misra and colleagues, Indian consensus statement (J Assoc Physicians India 2009) and its revision (Diabetes Metab Syndr 2025) | No. The 2025 revision explicitly prefers waist-to-height ratio and does not set a waist-to-hip threshold | A journal article. Figures are cited with attribution; recommendation text and criteria tables are not reproduced, which is this project’s standing position | YES, for the waist figures it gives: 90 cm in men and 80 cm in women |
| National Cholesterol Education Program ATP III and the International Diabetes Federation metabolic syndrome definitions | No. Both use waist circumference, not the ratio | varies; not needed | not needed |
| NHANES body-measures protocol (NCHS) | No threshold, and no data: NHANES measures waist circumference and does not collect hip circumference in its current protocol | US Government work, public domain | YES, for the waist measurement protocol described in the field hints above, and for the fact that no public-domain reference distribution for this ratio exists |
| NICE (UK) | not examined for a figure | The NICE UK Open Content Licence is UK-only and forbids displaying the licensed information next to advertising. Ruled out for this site in an earlier batch | NO |
What the ratio predicts, from the largest study of it, as positions in a distribution rather than as a line
| Measure | Odds ratio for a first myocardial infarction, top quintile against bottom | After adjustment for BMI | Population attributable risk, top two quintiles |
|---|---|---|---|
| Waist-to-hip ratio | 2.52 (95% CI 2.31–2.74) | — | 24.3% (22.5–26.2) |
| Waist circumference | — | 1.77 (1.59–1.97) | — |
| Hip circumference | — | 0.73 (0.66–0.80) — a LARGER hip went with LOWER risk | — |
| BMI, unadjusted | 1.44 (1.32–1.57) | — | 7.7% (6.0–10.0) |
| BMI, adjusted for waist-to-hip ratio | 1.12 (1.03–1.22) | — | — |
| BMI, fully adjusted | 0.98 (0.88–1.09) — no association left at all | — | — |
Where the tape goes, and how much it is worth
| Protocol | Landmark | Published difference from other sites, same people | Effect on the ratio at a hip of 98 cm |
|---|---|---|---|
| US national survey (NHANES) | just above the uppermost lateral border of the right ilium, crossed at the midaxillary line, at the end of a normal expiration, to 0.1 cm | reference | — |
| Midpoint protocol | midway between the lowest rib and the iliac crest | within the 0.2 to 6.9 cm range reported between pairs of sites in men and 0.1 to 10.1 cm in women | up to 0.070 in men, up to 0.103 in women |
| Minimal waist | the narrowest point of the torso. In the published comparison this site correlated best with visceral fat in men (r = 0.70) | as above | as above |
| Umbilicus | at the navel. This is the site the men’s US Navy body fat equation was fitted on | as above | as above |
| Hip | the widest part of the buttocks. A maximum found by eye, not a bony landmark; the original US Navy women’s report instead specifies just below the gluteal fold | no comparable published range; the less reproducible of the two | 0.0090 per centimetre at this waist, opposite in sign |
A ratio with strong evidence behind it, no usable threshold in front of it, and a measurement error larger than the distance between the thresholds other pages print
This page does something unusual for a calculator, which is to refuse to categorise. It is worth explaining at length, because the refusal is the most useful thing on it. Waist-to-hip ratio is a well-studied measurement with real evidence behind it: in the largest case-control study of heart attack risk ever run, across 27,098 people in 52 countries, the odds of a first myocardial infarction rose across every successive quintile of the ratio, reaching 2.52 in the top quintile against the bottom, and BMI’s own association with heart attack disappeared completely once the ratio was adjusted for. A reader is entitled to ask what their own figure means against that. The problem is that the answer everyone gives — a pair of sex-specific cut-offs — comes from a World Health Organization expert consultation, and WHO publications are licensed CC BY-NC-SA 3.0 IGO. The NonCommercial term applies to this site, because it carries advertising. Those figures are therefore not on this page, and they are not paraphrased either: paraphrasing a table to get around the licence on the table is the same breach with the numbers retyped, and taking them from one of the hundreds of sites that have republished them is the same breach again, because copyright does not launder.
So the work was to find a substitute, and the search produced a more interesting answer than a substitute would have been. The obvious place to look was the 1998 US national guideline on the identification, evaluation and treatment of overweight and obesity in adults, produced by an NHLBI expert panel and in the public domain as a US Government work. It does not give a waist-to-hip cut-off. What it gives instead is a position: that waist circumference has been found to be a better marker of abdominal fat content than the ratio, that whether the ratio adds anything about disease risk beyond the waist is uncertain, that between the two the waist carries greater prognostic significance, and that in clinical practice abdominal fat should be assessed by measuring the waist. In other words, the body most likely to have published a threshold considered the measurement and recommended against using it in this form. Indian national guidance arrives at a similar place from a different direction: the 2025 revision of the Misra consensus prefers waist-to-height ratio with a cut-off above 0.5 and does not set a waist-to-hip threshold. The metabolic syndrome definitions from the National Cholesterol Education Program and the International Diabetes Federation both use waist circumference rather than the ratio. And NHANES, the largest continuous health survey in the world, measures the waist and the arm and does not collect hip circumference at all — which is why there is no public-domain reference distribution for this ratio to place a reader in.
There is also an arithmetic argument against a threshold, independent of the licensing, and it is strong enough that this page would be reluctant to print one even if a usable source turned up. A published comparison of waist measurement sites in the same individuals found differences between pairs of sites ranging from 0.2 to 6.9 cm in men and from 0.1 to 10.1 cm in women. At a hip of 98 cm, one centimetre of waist is 0.0102 of the ratio, so 6.9 cm is 0.070 and 10.1 cm is 0.103. The two sex-specific cut-offs in common circulation differ from each other by 0.05. The error introduced by choosing a different, equally published place to put the tape is therefore larger than the distance between the two thresholds themselves, and comfortably larger than the distance most readers sit from whichever one applies to them. A categorisation whose output is determined by the measurer’s choice of landmark is not a categorisation of the reader.
What the page prints instead is the sensitivity, and it prints it in the units of the answer. The change in the ratio per centimetre of waist is one over the hip. The change per centimetre of hip is the waist over the hip squared, with the opposite sign. Both are in the rows above for the figures you entered, along with the number of centimetres of waist that the last printed digit is worth, the ratio recomputed with the waist two centimetres larger and two centimetres smaller, and the ratio recomputed with the waist shifted by the largest between-site difference published for your sex. A reader who looks at those rows will not read the second decimal place as a fact about their body, which is the correct conclusion and one that no amount of prose achieves as efficiently as the arithmetic does. The errors in the two measurements are independent and work in opposite directions, so the ratio is noisier than either of its parts — a point in favour of the simpler alternative on the waist-to-height page, where the denominator is a height, which does not change between Tuesday and Thursday.
Two waist thresholds are on the page, and they are thresholds for the waist, not for the ratio. They are here because they are the only cleanly licensed numbers in this territory and because a reader who has come looking for a line is better served by a well-sourced line about a related measurement than by no number at all. The 1998 US guideline gives 102 cm for men and 88 cm for women, applied to adults with a BMI between 25 and 34.9, and notes that above a BMI of 35 the waist loses its additional predictive value; those figures were derived in a largely white North American population. Indian national guidance gives 90 cm for men and 80 cm for women, on the evidence that abdominal fat and its metabolic consequences appear at a smaller waist in South Asian populations. Twelve centimetres separate them for men and eight for women. This site’s readership is largely Indian, so the lower pair is likely to be the more relevant one, and the page prints both and switches neither — the same discipline it applies to the two sets of BMI cut-offs on the BMI page.
How to take the two measurements, since this is the part that determines the answer. For the waist: at the end of a normal breath out, not held in and not pushed out, with the tape horizontal and parallel to the floor, snug against the skin without compressing it, over bare skin or the thinnest possible layer, read to the nearest tenth of a centimetre. For the level, pick one published protocol and never change it: the US survey protocol marks a line just above the uppermost lateral border of the right ilium and takes the tape there; others use the midpoint between the lowest rib and the iliac crest, the narrowest point of the torso, or the umbilicus. Write down which one you used. For the hip: the widest part of the buttocks, tape horizontal, feet together, which means finding a maximum by eye rather than a bony landmark and is the reason the hip is the less reproducible of the two. Take each measurement three times and use the median; if the three differ by more than a centimetre, the problem is the technique rather than the body. A mirror helps more than a second pair of hands for keeping the tape level.
Where this measurement is actually load-bearing is in the medical set, and this page is the right place to take it from. Waist circumference is an input to several calculators that do carry published, cited scoring with coefficients that can be checked: the fatty liver index and the hepatic steatosis index combine it with blood results to estimate hepatic steatosis, and the estimated glucose disposal rate uses it in a measure of insulin sensitivity. A waist measured carefully once, with the protocol noted, serves all of them. That is a better use of the ten minutes than refining a ratio whose threshold nobody can lawfully give you.
And the thing this page will not do under any circumstances. It will not tell you that your shape is wrong. The ratio is a quotient of two tape measurements; the evidence behind it is a graded association across populations, not a verdict on an individual; the error in it is larger than most of the distances anybody cares about; and the word that would be attached to a number here would have been decided by a committee rather than measured. If the figure above concerns you, the useful next step is a doctor who can measure blood pressure, lipids and glucose, which are the things the ratio is a rough proxy for and which can be measured directly.
Frequently asked questions
Why does this page not tell me whether my ratio is high?
Because it cannot do so honestly. The sex-specific cut-offs you have seen elsewhere come from a World Health Organization expert consultation, and WHO publications carry a NonCommercial licence term that this site, which carries advertising, cannot satisfy — so they are not reproduced here, not paraphrased, and not taken from a site that has republished them. The search for a usable substitute found none: the 1998 US national guideline gives no waist-to-hip cut-off and actively recommends measuring the waist instead, Indian national guidance prefers waist-to-height ratio, and the main metabolic syndrome definitions use the waist alone. There is also an arithmetic objection that would apply even with a clean source in hand: the difference between two published places to put the tape is worth more of the ratio than the gap between the two thresholds people compare themselves against. What the page gives you instead is the ratio, how much it moves when the tape moves, the published evidence on what it predicts, and the two waist thresholds that are cleanly licensed.
What is a normal waist-to-hip ratio, then?
This page will not give you one, and the reason is the same: a figure describing the middle of a distribution would have to come from a reference distribution, and there is no public-domain one for this ratio because NHANES, the survey that would hold it, does not collect hip circumference. What can be said from the evidence is about ranking rather than normality. In the 27,098-participant heart attack study the risk rose smoothly across every fifth of the ratio’s distribution, with no step and no natural break — which is both why the measurement is worth taking and why a threshold has always had to be chosen rather than found. If you want a figure to compare yourself against, the cleanly licensed ones on this page are for the waist alone: 102 cm for men and 88 cm for women from the 1998 US guideline, and 90 cm and 80 cm from Indian national guidance.
Should I use waist-to-hip ratio or waist-to-height ratio?
On the evidence, waist-to-height, and the pages say so to each other rather than competing. Three reasons. It needs one fewer awkward measurement, and the one it drops — the hip — is the less reproducible of the two, so the ratio is less noisy. Its denominator is a height, which does not change between measurements, so repeat readings reflect changes in the waist rather than in the measurer’s eye for the widest part of the buttocks. And it has a boundary value that could be sourced cleanly, from a CC BY paper and from Indian national guidance, which is more than can be said for this one. Waist-to-hip ratio remains the measure that beat BMI in the largest heart attack case-control study ever run, so it is not a bad measurement; it is a measurement whose threshold is not available to this site.
Where exactly do I put the tape?
Pick one published protocol and use the same one every time, and write down which. For the waist the US national survey protocol marks a line just above the uppermost lateral border of the right ilium and takes the tape there, at the end of a normal breath out, horizontal and parallel to the floor, snug without compressing the skin, read to the nearest 0.1 cm. Other widely used protocols take the midpoint between the lowest rib and the iliac crest, the narrowest point of the torso, or the level of the umbilicus. A study comparing sites in the same people found differences of up to 6.9 cm in men and 10.1 cm in women between pairs of sites, which is worth up to 0.07 and 0.10 of the ratio. For the hip, the widest part of the buttocks, tape horizontal, feet together. Take three readings and use the median.
My ratio is exactly 1.00 or more. What does that mean?
Arithmetically, that your waist measured at least as large as your hip. Before reading anything into it, check two things: that the two figures went into the right fields, and that the hip was taken at the widest part of the buttocks rather than somewhere higher. If both measurements are right, a ratio at or above 1 sits well above the middle of the distribution in every published population, and the association in the heart attack study was graded across the whole range rather than switching on at a point — so a high figure is a reason to have blood pressure, lipids and glucose measured, which are the things this is a rough proxy for, rather than a finding in itself. The page still prints no category, because there is still no cleanly licensed one to print.
Does it matter whether I measure in centimetres or inches?
Not to the ratio, which is dimensionless — and this page divides the figures exactly as typed rather than converting, which is slightly more careful than it sounds. Converting both to centimetres first and then dividing gives a different floating-point answer in about a third of inch pairs, and on a sweep of 100,651 of them it changed the second decimal place in five. The conversion is used only for the rows quoted in centimetres, at exactly 2.54 cm per inch. What matters enormously more is that both measurements use the same unit and that the waist uses the same protocol as whatever figure you are comparing against.
Can the ratio tell me whether I am an “apple” or a “pear”?
It is the measurement that language came from, and the language is worth being careful with. The physiological finding underneath it is real and is visible in the data: in the heart attack study, once BMI was controlled for, waist circumference carried an odds ratio of 1.77 and hip circumference 0.73 — a larger hip went with lower risk, which is why the ratio outperforms either measurement alone. But a ratio is a single number summarising two, so it cannot tell you which of the two is unusual, and a shape label attached to a body is the kind of statement this set of pages does not make. The useful version of the finding is that where fat sits is more informative than how much there is, which is why the waist pages exist beside the BMI page rather than instead of it.
Should I measure first thing in the morning?
Measure at the same time of day, whenever that is, rather than at a particular time. Abdominal girth varies through the day with the contents of the gut and bladder and with posture and breathing, by a centimetre or two in most people, which at a hip of 98 cm is worth one or two in the second decimal place of the ratio. Since both the waist and the hip are affected and in the same direction, the ratio is a little more stable through the day than either measurement alone. Consistency is the whole technique: same time, same protocol, same measurer, three readings, take the median.
Related calculators
References
- National Heart, Lung, and Blood Institute, Obesity Education Initiative Expert Panel. Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults. NIH Publication No. 98-4083, 1998 (available as NCBI Bookshelf NBK2004). A US Government work in the public domain. THE KEY SOURCE FOR THIS PAGE’S POSITION, and the reason the absence of a threshold here is a finding rather than a gap: the guideline gives NO waist-to-hip ratio cut-off and recommends against the measure, stating that waist circumference has been found to be a better marker of abdominal fat content than waist-hip ratio, that whether the ratio imparts any independent information about disease risk beyond the waist is uncertain, that between the two the waist appears to carry greater prognostic significance, and that in clinical practice abdominal fat should therefore be assessed by measuring waist circumference. It is also the source of the waist figures used in the comparison rows: greater than 102 cm (40 inches) in men and greater than 88 cm (35 inches) in women, applied to adults with a BMI of 25 to 34.9, with the note that the measurement loses added predictive value above a BMI of 35.
- LICENSING POSITION ON THE WAIST-TO-HIP THRESHOLDS, recorded deliberately and at length because it is the central editorial decision on this page. The sex-specific waist-to-hip cut-offs in universal circulation are from a World Health Organization expert consultation on waist circumference and waist-hip ratio (Geneva, 2008; report published 2011). WHO publications are licensed CC BY-NC-SA 3.0 IGO. The NonCommercial term applies to this site because it carries advertising. Those figures are therefore NOT printed on this page, NOT paraphrased, and NOT taken from any of the many third-party sites that have republished them, because copyright does not launder. The alternatives considered and rejected: inventing a threshold (this page’s opinion dressed as guidance); deriving one from a population distribution (no public-domain distribution exists, because NHANES does not collect hip circumference); and reproducing a quintile table from the INTERHEART paper (a copyrighted Lancet article, and quintile boundaries of one case-control sample are not a threshold). The outcome recorded here is that no cleanly licensed waist-to-hip threshold could be sourced, which this page states on its face.
- Yusuf S, Hawken S, Ôunpuu S, Bautista L, Franzosi MG, Commerford P, Lang CC, Rumboldt Z, Onen CL, Lisheng L, Tanomsup S, Wangai P, Razak F, Sharma AM, Anand SS, on behalf of the INTERHEART Study investigators. Obesity and the risk of myocardial infarction in 27,000 participants from 52 countries: a case-control study. Lancet 2005;366(9497):1640–9. doi:10.1016/S0140-6736(05)67663-5. The strongest evidence for this measurement and the source of every figure in the evidence table above: odds ratio 2.52 (2.31–2.74) for the top quintile of waist-to-hip ratio against the bottom, adjusted for age, sex, region and smoking, with significantly greater odds at every successive quintile; waist circumference adjusted for BMI 1.77 (1.59–1.97) and hip circumference adjusted for BMI 0.73 (0.66–0.80); BMI 1.44 (1.32–1.57) unadjusted, 1.12 (1.03–1.22) after adjustment for waist-to-hip ratio and 0.98 (0.88–1.09) fully adjusted; population attributable risk for the top two quintiles 24.3% (22.5–26.2) for the ratio against 7.7% (6.0–10.0) for BMI. Results are cited; no table is reproduced.
- Misra A, Chowbey P, Makkar BM and colleagues. Consensus statement for diagnosis of obesity, abdominal obesity and the metabolic syndrome for Asian Indians. J Assoc Physicians India 2009;57:163–70. Source of the second waist threshold used in the comparison rows: abdominal obesity at a waist circumference of 90 cm or more in men and 80 cm or more in women, on the evidence that abdominal adiposity and its metabolic consequences appear at a smaller waist in South Asian populations. Cited as two figures with their provenance; no criteria table is reproduced.
- Misra A, Vikram NK, Ghosh A, Ranjan P, Gulati S and the India Obesity Commission. Revised definition of obesity in Asian Indians living in India. Diabetes Metab Syndr 2025;19:102989. doi:10.1016/j.dsx.2024.102989. Carries the 2009 waist figures of 90 cm and 80 cm forward. Cited additionally for its stated position, which supports this page’s: it notes the limitations of waist-to-hip ratio, does not set a waist-to-hip threshold, and prefers waist-to-height ratio with a cut-off above 0.5 as the abdominal adiposity measure.
- National Center for Health Statistics. National Health and Nutrition Examination Survey (NHANES) Anthropometry Procedures Manual. Centers for Disease Control and Prevention. A US Government work in the public domain. Source of the waist measurement protocol described in the field hints and the table above: a horizontal line drawn just above the uppermost lateral border of the right ilium and crossed at the midaxillary line, the tape placed at that mark parallel to the floor, snug without compressing the skin, measured at the end of a normal expiration and recorded to the nearest 0.1 cm. Cited also for a fact this page makes use of: the current body-measures protocol collects arm and waist circumference and does not collect hip circumference, which is why no public-domain reference distribution for waist-to-hip ratio exists.
- Waist circumference measurement sites and their association with visceral and subcutaneous fat and cardiometabolic abnormalities (2023; PubMed Central PMC10118742). The source of the protocol-difference figures that dominate this page’s error analysis: differences between pairs of waist measurement sites in the same individuals ranged from 0.2 ± 2.7 cm to 6.9 ± 6.7 cm in men and from 0.1 ± 3.7 cm to 10.1 ± 4.3 cm in women. The study also found the minimal-waist site correlated best with visceral adipose tissue in men (r = 0.70) and that no site showed clear superiority in women.
- Hodgdon JA, Beckett MB. Prediction of percent body fat for U.S. Navy men / women from body circumferences and height. Naval Health Research Center Reports 84-11 and 84-29, 1984. Cited here for two site definitions used in the table above, which are relevant because the same tape measurements feed the body fat page in this category: the men’s equation uses the abdominal circumference at the level of the umbilicus, the women’s uses it at the level of minimal abdominal width, and the women’s hip is specified as just inferior to the gluteal fold rather than at the widest part of the buttocks. US Government works.
- Floating-point behaviour of the ratio, established for this page by sweeping rather than by reasoning from the specification. The two-decimal rounding of a quotient of two doubles was compared against exact rational rounding over 1,320,000 waist-and-hip pairs on a one-millimetre grid; the float and exact roundings differ in 238 of them, all at values whose exact ratio ends in a half of the last displayed digit, and since this page applies no threshold none of them has any consequence. Separately, dividing two inch figures directly was compared against converting both to centimetres and then dividing, over 100,651 pairs: the raw quotient differs in 36,689 and the two-decimal value in five. The page divides as typed, which is both simpler and the one of the two that is exactly right.
CalcEngines health calculators are for education and for checking arithmetic that has already been decided elsewhere. They are not medical advice, they do not decide what to give, and they do not replace the judgement of a doctor, nurse, midwife or dietitian who knows the person in front of them. Every figure depends on the values you enter and on the assumptions stated on the page — check it against the prescription, the product label and your local policy before acting on it.