Are smart scales worth it? Yes for weight trends, not for body-fat or metabolic-age readings

Does regular weighing help? Yes, inside a program

Randomized trials have tested the first question. The extra numbers come from a weak electrical current, usually passed through your feet, and the second section covers how accurate they are^{9, 10}.

The most useful summary is a 2015 systematic review and meta-analysis of 24 randomized trials in adults with overweight or obesity¹. Of those, 21 reported enough data to be pooled¹. The trials enrolled 23 to 415 people each (median 110)¹. Most participants were women, and 15 of the trials were run in the US¹.

The result depended on what came with the weighing (see table)¹. The one trial of advice to self-weigh on its own found no significant difference¹. Adding self-weighing to a behavioral weight-loss program improved loss by 1.7 kg across four trials¹.

A subgroup analysis compared trials by whether people were accountable to a therapist or researcher for their weights^{1, 12}. The 14 trials with that accountability showed 3.5 kg more loss than the comparison groups^{1, 12}. The two trials without it showed 2.3 kg more^{1, 12}. The figures include a 2016 correction for a data error in one trial¹². After the correction, the authors described the difference between the two groups as approaching significance (p = 0.05), not significant¹².

The evidence has other limits. Results varied widely between the trials of full programs^{1, 12}. For the add-on comparison, the prediction interval ran from 7.5 kg in favor of self-weighing to 4.1 kg against it^{1, 12}. So a future trial could plausibly find no benefit. Only seven of the 24 trials were judged at low risk of bias for concealing group allocation, and four were at high risk¹.

One 6-month study of 91 adults with overweight tested a smart scale directly². The intervention group got a cellular-connected scale for daily weighing, a web-based weight graph and weekly emails with tailored feedback and lessons². They weighed themselves on 6.1 days a week on average, versus 1.1 days for controls².

At 6 months the intervention group had lost 6.55% of their body weight, compared with 0.35% in a delayed control group². In the intervention group, 42.6% reached a 5% loss, versus 6.8% of controls². The catch is that the scale came bundled with coaching emails, so the trial doesn’t separate the device from the program around it².

Weighing as part of a brief program also limited short-term weight gain in one UK randomized trial³. It enrolled 272 adults with a BMI of 20 or more before the Christmas holidays³. The intervention group weighed themselves regularly and recorded and reflected on their weight³. They also got weight-management strategies for the holidays, plus information on how much activity it takes to burn off common festive foods and drinks³. The comparison group got a healthy-living leaflet³.

Over 4 to 8 weeks, the intervention group’s weight changed by -0.13 kg and the comparison group’s by +0.37 kg³. The adjusted difference was 0.49 kg less gain (95% CI 0.13 to 0.85 kg)³. The share of people who kept their gain to 0.5 kg or less didn’t differ significantly between groups³.

How accurate are smart-scale body-fat readings?

Smart scales estimate body composition with bioelectrical impedance analysis (BIA)^{9, 10}. The scale sends a very small, imperceptible electrical current through the body and uses the result to estimate fat mass and related numbers^{9, 10}. The European Society for Clinical Nutrition and Metabolism (ESPEN) guideline says BIA performs well in healthy people with stable fluid balance⁶. That holds when the prediction equation was validated for the person’s age, sex and race⁶. ESPEN advises against routine use at the extremes of BMI or with abnormal hydration until the methods are validated for those groups⁶.

Weight itself is the strong point. A Paris hospital study used DEXA (dual-energy X-ray absorptiometry) as the reference⁴. It measured 153 patients, each on one of three consumer smart scales⁴. They had been referred for a DEXA scan because of obesity or a chronic condition that can affect body composition⁴. For all three scales, the median difference from DEXA-measured weight was 0.3 kg or less⁴.

Body fat was the weak point. All three scales underestimated it, by a median of 2.2 kg, 4.4 kg and 3.7 kg of fat mass⁴. The spread was wide: for one scale, the middle half of errors ranged from 8.0 kg too low to 0.3 kg too high⁴. The authors concluded that smart scales aren’t accurate for body composition and shouldn’t replace DEXA in patient care⁴.

A study published in 2023 tested 15 BIA devices, 14 of them consumer-grade, against a laboratory four-compartment model⁵. The 73 participants were healthy adults aged 19 to 50⁵. Repeat readings were very consistent, with precision errors of 0.0 to 0.49 percentage points of body fat⁵.

Accuracy was another matter. Depending on the device, the average error ranged from about 3.5 percentage points too low to 11.7 points too high⁵. The standard error of the estimate, a gauge of individual error, was 3.1 to 7.5 points⁵. A device can be consistent and still consistently wrong.

Tracking change is somewhat better but still noisy. The study retested 37 participants 12 to 16 weeks later⁵. Average errors in the change in body fat were small (-0.4 to 1.3 points)⁵. But the standard error of the estimate for an individual’s change was 1.7 to 2.6 points⁵. That means a shift of a point or two over a few months can fall within the error.

Concordance with the reference ranged from 0.37 to 0.78, where 1.0 would be perfect⁵. The authors suggested some octapolar (hand-and-foot) and some foot-to-foot devices may be useful for tracking change when a research-grade device isn’t practical⁵. ESPEN likewise says following individuals over time is possible at a BMI of 16 to 34 without abnormal hydration⁶. It adds that results should be interpreted with caution⁶.

Metabolic age and the other extra numbers

Metabolic age is a manufacturer-defined score, not a clinical measurement. Tanita, a body-composition scale maker, says its monitors compare the rate at which you burn calories with the average for people your age⁷. It presents the result as an idea of “how old you are on the inside”⁷. Tanita’s page gives no formula, reference population, validation data or margin of error⁷.

The same page says building muscle and being more active raise basal metabolic rate (the calories you burn at rest), which lowers metabolic age⁷. On a smart scale, that means the score moves with weight and the impedance-based body-composition estimates. The studies above found those estimates imprecise^{4, 5}. None of the studies cited in this review assessed metabolic age, visceral fat ratings, body water or bone mass readings against a reference method. So there’s no basis here for treating those numbers as precise.

What the evidence does not show

The weighing trials support a habit inside a program, not the device or its extra features.

• Advice to weigh yourself with no other support hasn’t been shown to work: the single trial found a 0.5 kg difference that wasn’t statistically significant¹
• The trials that found benefits paired weighing with feedback, lessons, goals or weight-management advice, so they don’t show what a scale and app achieve on their own^1–3
• Long-term evidence is thin: only three trials reported follow-up after contact with participants ended, six to 18 months later¹
• Those three follow-ups were inconsistent, with differences of -8.0 kg, -0.3 kg (not statistically significant) and -7.5 kg¹
• The weight-loss trials enrolled mostly women with overweight or obesity, and four enrolled only women¹
• Most of the weight-loss trials were run in the US¹
• The holiday trial recruited adults in Birmingham, UK, and followed each person for only 4 to 8 weeks over the holiday period³
• Given who these trials enrolled, the results may not apply equally to men or other groups
• No study cited here shows that tracking smart-scale body-fat, muscle or metabolic-age numbers changes weight or health outcomes
• The accuracy studies tested particular models in particular groups, so results may not carry over to other models or people^{4, 5}
• Those groups were patients referred for DEXA at one Paris hospital and healthy adults aged 19 to 50^{4, 5}

Safety: mood, implanted devices and pregnancy

A common worry is that frequent weighing harms mood or body image. A 2016 meta-analysis pooled 29 independent tests⁸. It found no association between self-weighing and mood measures such as anxiety or depression, body-related attitudes or disordered eating⁸. It did find a small negative association with psychological functioning, such as self-esteem⁸.

At least some results varied with age, obesity status, the amount of weight lost, how long people weighed themselves and study design (randomized or correlational)⁸. The authors said the association may be more negative in some cases than in others⁸. On average the evidence is reassuring, but it doesn’t rule out harm for some people.

The body-composition current is a separate issue if you have an implanted electronic device. A 2023 University of Utah study, published in Heart Rhythm, tested the bioimpedance currents of a smart scale, a smartwatch and a smart ring⁹. The currents were run against three cardiac resynchronization devices in simulations and bench tests, not in patients⁹. The researchers reported that the currents interfered with the implants and could sometimes confuse them into operating incorrectly⁹. The press release says the research doesn’t convey an immediate or clear risk to patients, and that clinical studies are needed⁹.

Manufacturer safety pages take a cautious line. One smart-scale maker advises people with a pacemaker or other internal medical device not to use its scale unless body-fat measurement is turned off¹⁰. For metal implants, which it says it hasn’t tested, it recommends talking to a doctor first¹⁰.

The same maker recommends turning off fat-mass measurement during pregnancy, because the readings won’t be accurate¹⁰. It says there are no known risks for pregnant women using its scales¹⁰. Turning off body-fat measurement still leaves the weight reading, which is the part of the scale the trials support^{1, 10}.

What to look for in a smart scale, and what to ignore

No study compares smart-scale models for weight-management results. These points follow from the trials and accuracy studies above, not from head-to-head testing.

• Pick a scale you’ll actually use regularly: the programs that worked involved regular weighing, recording and reflecting on the trend^1–3
• If you want accountability, look for a clear trend graph and an easy way to share data with a coach, clinician or program^{1, 12}
• Accountability was linked to somewhat larger losses, though after the correction the difference was of borderline statistical significance^{1, 12}
• Check whether the scale can turn off body-fat measurement^{9, 10}
• At least one maker says people with a pacemaker or other internal medical device should use its scale only with that measurement turned off¹⁰
• The same maker recommends turning it off during pregnancy, because the readings won’t be accurate¹⁰
• Weigh under similar conditions each time, since ESPEN says BIA needs stable fluid balance to work well⁶
• Body-fat readings taken under different hydration conditions are harder to compare⁶
• Read the privacy terms, because consumer health apps may fall outside HIPAA¹¹
• The FTC’s Health Breach Notification Rule covers non-HIPAA businesses that offer personal health records, including health apps that pull data from sources such as fitness trackers¹¹
• Under that rule, disclosing health data without the person’s authorization can count as a breach that must be reported¹¹
• Ignore metabolic age: it’s a manufacturer-defined comparison, and the page that defines it publishes no method or error margin⁷
• Treat body-fat percentage, muscle mass and similar readings as rough estimates, since they can be several percentage points off for an individual^{4, 5}
• Changes of a point or two over a few months may be noise⁵

Sources

1. Is self-weighing an effective tool for weight loss: a systematic literature review and meta-analysis International Journal of Behavioral Nutrition and Physical Activity (Madigan CD et al.), 2015 · meta-analysis
2. The efficacy of a daily self-weighing weight loss intervention using smart scales and e-mail Obesity (Steinberg DM et al.), 2013
3. Effectiveness of a brief behavioural intervention to prevent weight gain over the Christmas holiday period: randomised controlled trial BMJ (Mason F et al.), 2018 · randomized trial
4. Accuracy of smart scales on weight and body composition: observational study JMIR mHealth and uHealth (Frija-Masson J et al.), 2021
5. Assessing the reliability and cross-sectional and longitudinal validity of fifteen bioelectrical impedance analysis devices British Journal of Nutrition (Siedler MR et al.), 2023
6. Bioelectrical impedance analysis, part II: utilization in clinical practice (ESPEN guideline) Clinical Nutrition (Kyle UG et al., ESPEN), 2004 · guideline
7. What is my metabolic age and what does it mean? Tanita UK · manufacturer
8. What is the psychological impact of self-weighing? A meta-analysis Health Psychology Review (Benn Y et al.), 2016 · meta-analysis
9. Shock to the System: Certain Wearable Gadgets Could Interfere With Implantable Cardiac Electronic Devices (press release on a Heart Rhythm study) University of Utah Health, 2023 · news
10. Body Cardio: frequently asked questions about safety Withings · manufacturer
11. Complying with FTC's Health Breach Notification Rule Federal Trade Commission, 2024 · regulator
12. Erratum to: Is self-weighing an effective tool for weight loss: a systematic literature review and meta-analysis International Journal of Behavioral Nutrition and Physical Activity (Madigan CD et al.), 2016