Are smart scales accurate? What validation studies show about body fat readings

What does a smart scale actually measure?

A scale that reports body fat uses bioelectrical impedance analysis, or BIA. A weak alternating current passes through the body and the scale records how much the body opposes it. A review in Contrast Media & Molecular Imaging defines bioimpedance as the opposition of a conductor to the flow of an alternating current, and notes that it varies with tissue composition and with the frequency used.⁷

Impedance is the only thing the scale senses. Everything else on the screen is calculated. The same review says BIA estimates parameters such as total body water and fat-free mass.⁷ A European consensus paper on muscle states that BIA equipment does not measure muscle mass directly. It derives an estimate from whole-body electrical conductivity, using a conversion equation calibrated against DXA-measured lean mass in a specific population.¹¹

A review in Aging Clinical and Experimental Research lists the assumptions behind those equations: the body behaves like a set of cylinders, fat-free mass holds virtually all the water and conducting electrolytes, and hydration stays constant.⁶ When one of those assumptions fails on a given morning, the estimate moves even though your body did not change.

Vita, the Chutex scale, is a Wi-Fi body composition scale with 8 electrodes. Its product page says the estimates are indicative and vary with hydration, meals, recent exercise and time of day.¹³ None of the studies in this guide tested Vita, so this page makes no accuracy claim about it.

Repeatability is not accuracy

Two different questions hide behind the word accurate. Repeatability asks whether the scale gives the same number twice. Accuracy asks whether that number matches a reference method. A scale can pass the first test and fail the second.

A team at Texas Tech tested 15 BIA devices on 73 healthy adults aged 19 to 50, 14 consumer models and one research model, against a four-compartment laboratory model built from air displacement plethysmography, DXA and bioimpedance spectroscopy.¹ The devices used hand-to-hand, foot-to-foot or eight-electrode hand-to-foot layouts.¹ Participants followed a strict protocol before each visit: no exercise for 24 hours and no food, fluid, caffeine or alcohol for 8 hours.¹

Repeat readings agreed closely: precision error ranged from 0.0 to 0.49 percent.¹ Accuracy against the lab model did not. The constant error ranged from minus 3.5 (SD 4.1) to plus 11.7 (SD 4.7) percentage points, the standard error of the estimate from 3.1 to 7.5 percent, and Lin’s concordance correlation coefficient from 0.48 to 0.94.¹

The authors add a warning about their own reliability numbers. A very low precision error may mean the device is not giving a true second estimate when tests are taken in immediate succession, and the test-retest figures capture technical precision, not day-to-day biological variability.¹ For change over 12 to 16 weeks, the constant error ranged from minus 0.4 to plus 1.3, the standard error from 1.7 to 2.6, and the concordance coefficient from 0.37 to 0.78. The authors conclude that performance varied widely and that select octapolar and foot-to-foot devices may be useful for tracking body composition over time.¹

Worked example: precise but offset

These numbers are made up to show the arithmetic. Say a scale shows 31.8, 32.0 and 31.9 percent body fat on three mornings. The spread is 32.0 minus 31.8, or 0.2 points, which looks reassuring. Now suppose a laboratory four-compartment test on the same person says 28.0 percent. The offset is 31.9 minus 28.0, or 3.9 points. The offset is about 20 times the spread (3.9 divided by 0.2 is 19.5), and it sits inside the minus 3.5 to plus 11.7 range reported in the study.¹

Worked example: is a small change real?

Again with made-up readings: a scale falls from 31.9 to 30.9 percent after a month. The drop is 1.0 point. The 15-device study found a least significant change of no more than 1.4 points for every device under standardized conditions, so 1.0 is below that line.¹ Over 12 to 16 weeks the standard error for tracking change was 1.7 to 2.6 points, which is larger still.¹ Reading these figures side by side is our arithmetic, not the authors’ conclusion, but it explains why a one-point wiggle on a home scale deserves caution.

What changes the reading?

Impedance depends on water, so anything that shifts water in the body can shift the estimate. Four studies put numbers on it.

The two effects differ in size. Five hours of quiet sitting moved the estimate by under 1 kg of fat,¹⁰ while a liter or two of water moved the fat percentage by several points.²³ The practical lesson is the same: do not compare a reading taken after a big drink with one taken fasted.

Timing matters because what you eat and drink changes body water. For the weight side of this, see the guide to the best time to weigh yourself.

How close do scales get to DXA and MRI?

Several studies compared home-style impedance devices with imaging. They agree on a pattern: decent on fat mass, less good on muscle, weak on visceral fat.

Pietiläinen’s group sums it up: BIA is at its best when assessing the amount or changes in fat mass, is useful for skeletal muscle, and is limited for visceral fat.⁴ For a side-by-side with a scan, see DEXA scan vs smart scale.

Pacemakers and other implanted devices

Impedance scales pass a small current through the body, so people with an implanted cardiac device are a special case. The evidence is mixed. A study in Heart Rhythm used simulations and benchtop tests with devices from three manufacturers. The simulations showed voltages above the limits of the ISO 14117 standard. The authors concluded that these consumer devices could interfere with cardiac implantable electronic devices and did not recommend their use in that group.⁸

A separate study in Clinical Nutrition tested 200 patients with implanted devices during routine checks. It found no change in battery voltage, lead impedance or pacing thresholds and no inappropriate sensing, and no changes at 6 and 12 months.⁹ The abstract does not say whether the devices tested were consumer scales.

Chutex states that Vita is not for people with a pacemaker or other implanted electronic device.¹³ If you have one, skip the electrodes and ask your cardiology team before using any impedance scale.

How to use a smart scale well

Given the numbers above, the sensible use is trend tracking under repeatable conditions. The steps below follow from the hydration findings and from the one-condition-at-a-time logic of the studies.

1. Pick one slot. Weigh at the same time of day, ideally before eating or drinking much, so water intake does not vary.
2. Standardize the setup. Use the same hard, flat floor, the same bare-foot contact and the same profile every time.
3. Skip the odd days. Do not compare a reading after a large drink or hard workout with an ordinary morning.
4. Read averages. Look at a weekly average and the direction over several weeks, not at single mornings.
5. Confirm what matters. For a decision about your health, ask a clinician about a reference test rather than relying on a home estimate.

To go deeper, read how a bioimpedance scale works, how to measure body fat at home and how two popular brands compare in Withings vs Renpho.

Frequently asked questions