DEXA scan vs smart scale: how each works, how far they agree, and how to weigh in consistently

How does each method work?

Both methods split your body weight into fat and everything else. They get there in very different ways.

DXA (also written DEXA) scans the body with X-rays at two different energies. The tissues absorb the two beams differently, and software uses the difference to calculate bone mineral content and soft tissue composition, for the whole body and for regions such as arms, legs and trunk. The review that describes this also notes that the effective dose from a DXA scan is very small.¹ MedlinePlus, a US National Library of Medicine service, adds that you receive more radiation from a chest X-ray than from a DEXA scan, and describes the scan as painless, with the need to stay still.⁶

A smart scale uses bioelectrical impedance analysis (BIA). You stand on metal electrodes, a small current passes through the body, and the scale turns the resistance into estimates of fat, lean mass and water using built-in equations. A review of the method says those equations assume a constant hydration of fat-free mass and differ by the sample they came from.¹⁰

DEXA scan vs smart scale at a glance

The table lays out the main criteria. It is not a ranking. Each row states what the sources say, or says plainly that nothing was read.

Two cells deserve a note. The first is price: I did not find a price on a source I could read, so this page does not guess one. The second is the pacemaker row, which the next-to-last section covers in full.

How accurate is DEXA, and how accurate is a smart scale?

DEXA is often called the reference. The evidence is more careful than that. A review of the technique says precision is excellent but varies by region, and that the coefficient of variation is about 1% for spine and whole-body bone measurements in young healthy people, less good for osteoporotic and obese people. The same review finds the accuracy of DXA measurements can be problematic, with marked differences between the three makers it covers, caused by calibration and other factors. It concludes that DXA cannot be regarded as a gold standard for body composition, while noting that development was improving it.¹ That review dates from 1996, so ask what equipment a facility uses today.

A newer guideline from the American Society for Parenteral and Enteral Nutrition (ASPEN) retrieved 7,375 studies on adults with a disease or condition. On limited studies and expert opinion, it recommends DXA for fat mass in a variety of disease states. It says the validity of DXA for lean mass remains unknown in any clinical population. It made no recommendation for BIA, because validity data were limited and manufacturers keep their regression models proprietary.⁵

Those are statements about patients, not about healthy people at home. For healthy adults, studies compared consumer scales directly against DXA and MRI:

• In 19 obese adults, a consumer impedance device (Omron BF-500) measured fat lost over 12 months at 7.0 kg, the same as DXA. Correlations for fat mass were 0.91 to 0.99. It underestimated skeletal muscle by 1 to 2 kg.³
• In 200 healthy Chinese adults, two consumer single-frequency devices correlated with DXA for fat percentage at r = 0.89 and 0.90, with small but significant biases.⁴

Read together, these say that a scale can follow the direction of fat change reasonably well on average, and that an individual reading carries error. For the specific question of which scales to consider, see the guide to the best smart scales for body composition and the one on smart scale accuracy.

The hydration effect, with an example

Water is the largest source of day-to-day noise for impedance. In a study of 39 adults, participants had DXA and impedance scans before and after drinking 2 liters of water. Multi-frequency impedance read fat percentage higher by 2.1 points in men and 2.6 in women, and single-frequency by 1.3 and 2.1. The authors conclude that impedance wrongly classes the extra water as fat mass and that hydration status needs to be standardized.²

DXA was not immune. Its fat-free mass rose by 1.4 kg in men and 1.7 kg in women after the water, which is 3.1 and 3.7 pounds at 2.2046 pounds per kilogram.² A second study of 140 people drank water in four steps of 500 mL and measured with a foot-to-foot impedance device. Body fat overestimation grew with each step, up to 7.92% in men and 9.4% in women at 2,000 mL, while fat-free mass and total body water were underestimated.⁹

Here is the arithmetic for an example person, made up for illustration. Take someone at 180 pounds and 20% body fat, which is 36 pounds of fat. A reading 2.1 points higher shows 22.1%. If weight is held at 180 pounds for the calculation, that is 39.8 pounds of fat, an apparent gain of 3.8 pounds from drinking water alone.

The practical lesson is to compare like with like. If you weigh in after a glass of water one morning and after a liter the next, the difference may be water, not fat.

A weigh-in routine that makes readings comparable

No routine removes the error, but a fixed one stops it from changing from day to day. The steps below follow from the hydration and device findings above.²¹

1. Use the same device. Different makes give different values, for DXA and for impedance.¹⁵
2. Pick one time of day. First thing in the morning, before eating or drinking, keeps hydration closer to constant.
3. Hold what you drank. Do not drink a large amount just before. A liter or two shifted results in the studies.²⁹
4. Stand the same way. Same spot on the floor, bare feet, same posture.
5. Compare averages. Look at weekly averages, not single readings.
6. Repeat a DEXA the same way. Same facility and machine, similar time of day and hydration.

Steps two, four and five are good practice rather than findings from a specific study. Step three is the one the data support most directly. For more on timing, see the best time to weigh yourself and how to measure body fat at home.

Pacemakers and implanted devices

Smart scales pass a small current through the body, and that raises a safety question for people with a pacemaker or another cardiac implantable electronic device. A study published in Heart Rhythm used simulations and bench testing. It found interference voltages above the thresholds in the ISO 14117 standard, varying with the frequency and strength of the signal, and different effects across the three manufacturers' devices. Scales and rings produced lower interference than smart watches in the simulations. The authors do not recommend these consumer devices in people with cardiac implants.⁷

Another study measured something different: clinical BIA in 200 patients with implanted devices. It found no significant changes in battery voltage, lead impedance or pacing thresholds, and no over- or undersensing, and it concluded that BIA could be done safely in that setting.⁸ The two findings do not contradict each other. One tested clinical impedance measurements in a monitored visit, and the other tested consumer devices in simulation and on the bench. The cautious answer for home use is the one the first study gives: do not use an impedance scale unless your cardiologist says it is fine.

Chutex states on its Vita page that the scale is not for people with a pacemaker or other implanted electronic device. If you have one, a DEXA scan uses X-rays, not an electrical current, but ask your doctor and the facility before any test.

MedlinePlus also says to tell your provider before a DEXA scan if you are or could be pregnant.⁶

Which one fits which need?

Many people use both: a scan now and then for a reference point, and a scale daily for direction. If you do, treat the two as separate scales with separate numbers. A DEXA percentage and a scale percentage do not need to match for either to be useful.

Frequently asked questions