DEXA vs InBody is not really a contest, because the two scans do different jobs: DEXA gives you an accurate absolute picture, and a standardised bioimpedance scan gives you a cheap, repeatable read on which way you are going. For a serious training block the sensible pattern is DEXA at the start and the end, with an InBody scan every few weeks in between. If you can only have one, pick by the question you are asking. “What am I made of right now?” is a DEXA question. “Is this working?” is an InBody question.

We run InBody scanning at our studio, so the awkward part belongs early: for one accurate number, DEXA is better.

For the wider picture of how measurement fits into keeping muscle while you lose weight, start with our pillar guide on losing fat without losing muscle.

Expertise reviewed by Mariyan Voykov, Manager and Senior Coach at Body Transformation London. MSc Sports and Exercise Science, BSc Biomedical Science, Level 3 Personal Trainer.

How different are the two numbers?

Further apart than most people expect, and consistently in one direction: bioimpedance underestimates fat and overestimates fat-free mass, which is why you cannot add a few points to your printout and call it a DEXA result.

The device we run is an InBody 270, and one study has compared that model directly with DXA. In 88 adults under a controlled protocol, percent body fat read 24.6% on the InBody against 27.7% by DXA, fat mass 17.8 kg against 19.7 kg, and fat-free mass 53.7 kg against 51.1 kg, all at p<0.0001 (Garcia 2020). So: body fat about 3.1 percentage points low, fat mass about 1.9 kg low, fat-free mass about 2.6 kg high.

One caveat travels with those numbers. That paper published paired t-tests and nothing else, so there is no limits-of-agreement figure and no correlation coefficient for this model against DXA anywhere in the literature. We can tell you the average gap. We cannot tell you how much it varies between individuals. The full accuracy picture is in is an InBody scan accurate.

Where does bioimpedance hold up well?

Three places, and they matter more than the accuracy gap suggests.

Whole-body muscle. Appendicular skeletal muscle index agreed closely with DXA, ICC 0.965, in 84 older adults (Buch 2022). If you care about total muscle rather than a precise fat percentage, the scanner is on firm ground.

Repeatability. Bioimpedance test-retest reliability under controlled conditions is close to perfect: ICC ≥0.998 for body fat percentage and 1.000 for fat-free mass, skeletal muscle mass and total body water (Looney 2024). No test-retest study of the 270 itself has been published, so the nearest evidence is the InBody 230, 720 and 770 across two test days, where test-retest ICC was ≥0.98 for body fat percentage, fat mass and fat-free mass (McLester 2020). The 230 shares the 270’s measurement engine, so we treat that as a reasoned inference about our device rather than a finding on it. A device can be consistently wrong and still be consistent, and consistency is what tracking needs.

Worth noting from that second study: “the research grade models, InBody720 and InBody770, added minimal benefit over the portable InBody230.” On the numbers both classes of machine actually print, the expensive one is not obviously better.

Subtraction. Because the bias against DEXA is largely systematic, most of it cancels when you subtract scan one from scan two. That is why a scan that reads four points off can still tell you something true about the last eight weeks.

Where does bioimpedance fall down?

Fat distribution, because it does not attempt it. Our scanner prints no visceral fat figure and no segmental fat analysis at all. It gives a five-segment breakdown of lean mass, which is genuinely useful for spotting a left-to-right or upper-to-lower imbalance, and nothing whatever about where your fat sits. If fat distribution is your question, DEXA is the only one of the two that answers it. Regional fat estimates are in any case the weak point of bioimpedance as a method: on devices that do produce them, arm fat mass managed only ICC 0.615 against DXA, with a mean difference of +2.32 kg (Buch 2022).

Fluid shifts, because it cannot see them. With only two measurement frequencies, 20 and 100 kHz, it cannot separate the water inside your cells from the water outside them, so it produces no extracellular-to-total-body-water ratio. On a higher-specification model that ratio is the standard check when lean mass appears to jump two kilogrammes in a week. Here there is no such check, so a hydration swing reads as a tissue change. It is the strongest argument there is for scanning under a fixed protocol.

Higher BMIs. The ESPEN technical review puts the method’s validity at up to about 34 kg/m², and notes that in obesity most predictive equations “are unable to predict static body composition and are not reproducible for individuals over time” (Kyle 2004). Reproducibility over time is exactly what tracking depends on, so this is not a small caveat.

Then the UK guidance point that cuts through all of it: NICE tells clinicians not to use bioimpedance as a substitute for BMI (what that does and does not mean). It is not a ban on scanning, but a scanner reading is not a superior replacement for BMI. It sits alongside BMI and waist-to-height ratio, not above them.

DEXA vs InBody: how small a change can either scan actually see?

This is the question that settles the practical answer, and almost nobody asks it. The least significant change is the smallest difference a device can attribute to real change rather than to noise. In resistance-trained men, DXA needed 1,204 g same-day and 1,615 g on consecutive days to call a change in fat mass real, and 1,461 g / 1,967 g for fat-free mass. Bioimpedance spectroscopy in the same study needed 2,331 g / 3,607 g for fat mass (Farley 2021). That was spectroscopy, a different device class from an eight-electrode multi-frequency scanner, so read those figures as the upper end of a range, not a verdict on any particular machine. ESPEN’s broader rule is that bioimpedance changes under 1.5 to 2 kg sit inside the noise.

For body fat percentage there is a more specific figure, and it is the one that catches people out. Minimum detectable change came out at 2.12 to 2.73 percentage points across two test days (McLester 2020). That study tested the InBody 230, 720 and 770, not the 270, but the 230 uses the same measurement engine as the 270, two frequencies at 20 and 100 kHz through eight-point tetrapolar electrodes, so we treat it as a reasoned inference about our machine rather than a direct finding on it. Either way: a body fat change under roughly 2 to 3 points is not something you can confidently separate from noise.

Read the DEXA column again, because it is the humbling one. Even the reference method cannot confidently call a fat-mass change much under 1.2 to 1.6 kg in one person. It is about twice as precise as bioimpedance here, and twice as precise still cannot resolve a good month.

The practical consequence: nobody can tell you from any scan that you gained 0.4 kg of muscle in four weeks. Not from an InBody, and not from a DEXA either. That is noise, and a coach reading it back as a win is selling you a rounding error.

What is actually involved in a DEXA scan?

You lie still on a table for a few minutes while a low-dose X-ray beam passes over you. Radiation dose for whole-body composition scanning is 0.15 to 4.7 microsieverts, against roughly 32 microsieverts for a plain chest radiograph (Maeda 2022). Natural background radiation runs at about 10 microsieverts a day (IAEA), so one scan is on the order of a few hours to half a day of ordinary background exposure. Small, but not zero, which is one reason nobody sensible does it monthly.

Access is the bigger constraint here. The NHS does not provide DEXA for body composition. It describes a DEXA scan solely as a bone density test used to assess osteoporosis risk, so body-composition DEXA in the UK is private, self-funded, and a trip to a centre that offers it. We will not quote a price, because the figures circulating online come from commercial providers and are undated.

DEXA vs InBody at a glance

So which one do you actually need?

If you want an accurate absolute baseline, a DEXA at the start and end of a serious block is the way to get one, twelve to twenty weeks apart. That gives you a genuine read on where your fat sits, a bone density figure the scanner cannot produce at all, and an endpoint measured the same way. If you are making decisions on absolute numbers, make them on the better numbers.

Use a standardised bioimpedance scan every three to four weeks in between. Its systematic bias largely cancels in a difference score, its repeatability is excellent, and it is quick enough that you will actually do it. Scanning at that interval is an accountability habit and a way of building the line, not a way of settling whether the last three weeks worked: a single three-week gap sits inside the noise, since a body fat change under roughly 2 to 3 percentage points or a mass change under 1.5 to 2 kg cannot be confidently separated from error. Read the line rather than the last point. The printout holds your last 10 tests, so ten three-weekly scans is roughly thirty weeks of trend on one sheet. That is what turns a set of readings into a trend.

Do not scan DEXA monthly. Three reasons, in order of weight. First, the precision floor: at 1,204 to 1,615 grammes for fat mass, most of what a monthly DEXA shows you is noise dressed as data. Second, access and cost, since every scan means a private booking and a trip. Third, dose, which is genuinely small but not a reason to be casual. Frequent DEXA buys you very little that a frequent standardised InBody scan does not, and it costs a great deal more.

Two readings are only worth comparing if both were taken under the same fixed protocol, which we set out step by step in is an InBody scan accurate.

Whatever your route to a deficit, the logic is the same. The scan is not the intervention. It earns its place only if something changes because of what it says.

If you want this measured properly

Bioimpedance scanning is part of the coaching at our Finchley Road studio: the same device, the same conditions, every time, and a coach who will say out loud which boxes on the printout we act on and which we disregard. What you get from it is a trend and a decision about your training, not a certificate. If that is the sort of measurement you want, book a call.

A body-composition DEXA is a private medical imaging service, arranged and reported by the provider. We do not order or interpret scans. If you have had one, we will factor the result into your programme.

Free InBody scan until 30 September

We are running complimentary body composition scans at our Finchley Road studio, Monday to Friday between 10am and 2pm. Thirty minutes: the scan itself, plus a coach taking you through what your numbers mean and which ones to ignore. No charge, no obligation, one per person.

Book your free scan or see what’s included.

This article is general information, not medical advice. If you have a medical condition, are pregnant, are taking any prescribed medication, or have a history of disordered eating, speak to your GP or a registered dietitian before changing how you train or eat.

Frequently asked questions

Is DEXA more accurate than InBody?
Yes, for a single absolute measurement. DEXA is the reference method bioimpedance is validated against. In 88 adults, the model we run read body fat 24.6% against DXA’s 27.7%, p<0.0001 (Garcia 2020), and in a separate study only 35% of InBody readings fell within 10% error of DEXA (Merrigan 2022). For tracking change under a fixed protocol the gap narrows, because most of that bias cancels between scans.

Can I get a DEXA body composition scan on the NHS?
No. The NHS describes DEXA solely as a bone density test used to assess osteoporosis risk, not as a body composition service. Body-composition DEXA in the UK is private and self-funded.

Is the radiation from a DEXA scan something to worry about?
The dose for whole-body composition scanning is 0.15 to 4.7 microsieverts, against about 32 microsieverts for a plain chest X-ray (Maeda 2022), with natural background radiation at about 10 microsieverts a day. One scan is on the order of a few hours to half a day of background exposure. Not a reason to avoid a scan, and one of several reasons not to have one monthly.

How often should I scan?
DEXA at the start and end of a training block of roughly twelve to twenty weeks, and a standardised bioimpedance scan every three to four weeks in between. Every three to four weeks is a good interval for accountability, but judge progress on the trend across two to three months rather than on any single gap between scans. Neither device resolves a short gap in one person: even DEXA cannot confidently call a fat-mass change much below 1.2 to 1.6 kg (Farley 2021).

Which scan should I use to track visceral fat?
DEXA, if that is the question you care about. Our scanner does not report visceral fat at all, so there is no number on it to track. Waist-to-height ratio, which NICE recommends alongside BMI in adults with a BMI under 35, is a free and better-supported proxy for central adiposity.

Sources

• Garcia JR, Manimaleth R, Czartoryski P et al. Journal of Exercise and Nutrition, 2020;3(3):10. https://journalofexerciseandnutrition.com/index.php/JEN/article/view/65
• McLester CN, Nickerson BS, Kliszczewicz BM, McLester JR. Journal of Clinical Densitometry, 2020;23(3):443–450.
• Merrigan JJ et al. Journal of Exercise and Nutrition, 2022;5(4). https://journalofexerciseandnutrition.com/index.php/JEN/article/view/133
• Buch A et al. Nutrition & Diabetes, 2022;12:45. https://doi.org/10.1038/s41387-022-00223-1
• Looney DP et al. Frontiers in Nutrition, 2024;11:1491931. https://doi.org/10.3389/fnut.2024.1491931
• Farley A, Slater GJ, Hind K. International Journal of Sport Nutrition and Exercise Metabolism, 2021;31(1):55–65. https://doi.org/10.1123/ijsnem.2020-0061
• Kyle UG et al. Clinical Nutrition, 2004. https://www.espen.org/documents/BIA1.pdf and https://www.espen.org/documents/BIA2.pdf
• Maeda SS et al. Advances in Rheumatology, 2022;62:7. https://doi.org/10.1186/s42358-022-00241-8
• IAEA, DXA bone mineral densitometry. https://www.iaea.org/resources/rpop/health-professionals/other-specialities-and-imaging-modalities/dxa-bone-mineral-densitometry/patients
• NHS, DEXA scan. https://www.nhs.uk/tests-and-treatments/dexa-scan/
• NICE NG246, Overweight and obesity management. Published 14 January 2025, last updated 8 January 2026. https://www.nice.org.uk/guidance/ng246/resources/overweight-and-obesity-management-pdf-66143959958725

{“@context”: “https://schema.org”, “@type”: “FAQPage”, “mainEntity”: [{“@type”: “Question”, “name”: “Is DEXA more accurate than InBody?”, “acceptedAnswer”: {“@type”: “Answer”, “text”: “Yes, for a single absolute measurement. DEXA is the reference method bioimpedance is validated against. In 88 adults, the model we run read body fat 24.6% against DXA’s 27.7%, p<0.0001 (Garcia 2020), and in a separate study only 35% of InBody readings fell within 10% error of DEXA (Merrigan 2022). For tracking change under a fixed protocol the gap narrows, because most of that bias cancels between scans."}}, {"@type": "Question", "name": "Can I get a DEXA body composition scan on the NHS?", "acceptedAnswer": {"@type": "Answer", "text": "No. The NHS describes DEXA solely as a bone density test used to assess osteoporosis risk, not as a body composition service. Body-composition DEXA in the UK is private and self-funded."}}, {"@type": "Question", "name": "Is the radiation from a DEXA scan something to worry about?", "acceptedAnswer": {"@type": "Answer", "text": "The dose for whole-body composition scanning is 0.15 to 4.7 microsieverts, against about 32 microsieverts for a plain chest X-ray (Maeda 2022), with natural background radiation at about 10 microsieverts a day. One scan is on the order of a few hours to half a day of background exposure. Not a reason to avoid a scan, and one of several reasons not to have one monthly."}}, {"@type": "Question", "name": "How often should I scan?", "acceptedAnswer": {"@type": "Answer", "text": "DEXA at the start and end of a training block of roughly twelve to twenty weeks, and a standardised bioimpedance scan every three to four weeks in between. Every three to four weeks is a good interval for accountability, but judge progress on the trend across two to three months rather than on any single gap between scans. Neither device resolves a short gap in one person: even DEXA cannot confidently call a fat-mass change much below 1.2 to 1.6 kg (Farley 2021)."}}, {"@type": "Question", "name": "Which scan should I use to track visceral fat?", "acceptedAnswer": {"@type": "Answer", "text": "DEXA, if that is the question you care about. Our scanner does not report visceral fat at all, so there is no number on it to track. Waist-to-height ratio, which NICE recommends alongside BMI in adults with a BMI under 35, is a free and better-supported proxy for central adiposity."}}]}