Smart scales and body composition (BIA): what the numbers actually mean

Smart scales using bioelectrical impedance analysis (BIA) estimate body composition by passing a tiny electrical current through the body. The current crosses different tissue types at different rates. Fat, muscle, and water all conduct differently. Algorithms convert impedance values into body fat percentage, muscle mass, and various other metrics.

The technology is widespread, accessible, and meaningfully noisier than most users realise.

What BIA can and can't measure directly#

Directly measures:

• Whole-body impedance to small AC current

Estimates from impedance + body weight + entered demographics (age, sex, height):

• Total body water
• Fat mass
• Lean (fat-free) mass
• Skeletal muscle estimate
• Visceral fat estimate (consumer scales, questionable)
• Bone mass estimate
• Metabolic age (vendor construct)

The directly measured impedance is a real number. The estimates derived from it depend on population-based equations that vary in accuracy by sex, ethnicity, body composition, and hydration state.

How accurate is it really#

Validation studies comparing consumer BIA scales to DEXA (the clinical gold standard):

• Body fat percentage: typically ±3–5% absolute error vs DEXA in well-controlled studies. Individual scales vary.
• Lean mass: generally similar magnitude of error
• Visceral fat estimates: weakest area; consumer scales rarely correlate strongly with DEXA-measured visceral adiposity
• Hydration state: large source of day-to-day variability

A 4-electrode whole-body BIA (consumer scale, standing on metal pads) is less accurate than 8-electrode segmental BIA (also has hand grips) which is less accurate than research-grade BIA, which is less accurate than DEXA.

What changes the readings#

Within an individual, daily BIA readings can vary by:

• 0.5–2% body fat from hydration alone
• More after exercise (acute fluid shifts)
• More after high-carbohydrate meals (glycogen + water storage)
• More after alcohol (dehydration)
• More with menstrual cycle in women
• Less if you weigh consistently first thing in the morning, same hydration state

A single reading is far less informative than a 7-day rolling average measured under consistent conditions.

The consumer scale brands#

Withings Body+ / Body Pro 2 / Body Smart: well-established UK options; reasonable accuracy; clean app; integrates with Health Mate and Apple Health.

Garmin Index S2: integrates tightly with Garmin Connect ecosystem; reasonable accuracy.

Eufy / Renpho: cheaper consumer options; accuracy is comparable to more expensive scales in independent testing; basic apps.

Omron — more clinically oriented; sometimes 8-electrode versions available; better validation history.

Tanita: long history of BIA; some consumer and clinical product overlap.

What's worth tracking#

For most users:

Strongly trustworthy from BIA scales:

• Body weight trend (any scale)
• Resting heart rate if measured (some Withings models)
• Direction of body composition change over months (e.g. lean mass up, fat down with a training programme)

Less trustworthy:

• Absolute body fat percentage
• Specific visceral fat numbers
• Bone mass changes month-to-month
• Single-day "metabolic age" type metrics

Largely meaningless:

• "Metabolic age" composite scores
• Specific muscle mass numbers compared to vendor "average" populations

Comparison to alternative methods#

DEXA scan: clinical gold standard. £80–£150 in UK private settings. Useful for an annual or pre-/post-programme measurement; not for weekly tracking.

Calipers: require skilled measurement; accuracy depends on technique; minimal setup cost. Most users measure inconsistently.

3D body scanning: newer technology (Naked, Styku); volumetric measurement; less validated against DEXA than BIA in many cases.

Smart mirrors and full-body imaging: emerging; expensive; novel validation profile.

For a non-athlete trying to track training progress, the BIA scale trend is usually adequate. For pre-/post-comparison of a 12-week intervention, an opening and closing DEXA scan is more informative.

Hydration timing for consistent readings#

If you want consistent BIA readings over weeks:

• Weigh in first thing in the morning
• After voiding, before eating or drinking
• Without clothes or in consistent minimal clothing
• Same time of day
• Not the day after intense exercise (acute fluid shifts)
• Not after alcohol
• Same scale, same location

These conditions stabilise much of the day-to-day noise.

What the marketing overstates#

• "Tracks visceral fat accurately." Generally no. Consumer scales don't validate well against DEXA for visceral fat.
• "Measures metabolic age." Vendor construct; not a clinical metric.
• "Reveals hidden health risks." Marketing language; not a clinical screen.
• "Tells you your true body composition." Estimates, not measurements.

What we'd actually say#

A smart scale is a useful tool for tracking weight and rough body composition direction over time. Don't trust specific percentages too precisely.

For tracking a training programme, the same scale used consistently over months tells you the direction. The absolute body fat number matters less than whether it's trending up or down.

For an accurate baseline, periodic DEXA scans are the better tool. For day-to-day signal, BIA scales are convenient.

The marketing of specific composition metrics overstates what consumer BIA can actually deliver. Treat the numbers as ballpark, useful primarily for trend tracking.

When to talk to a clinician#

• Significant unintended weight loss or gain
• Concerns about visceral adiposity: BIA isn't the right tool; DEXA or imaging is
• Symptoms suggesting eating disorder
• Suspected sarcopenia in older adults: clinical assessment, not BIA, is the right route
• Significant disagreement between scale data and clinical body composition (e.g. apparent muscle loss with no explanation)