An article written by Dr Edward Leatham, Consultant Cardiologist     © 2025 E.Leatham

For busy people, or to tune in when on the move, Google Notebook AI audio podcast and an explainer slide show are available for this story beneath.

Smart scales that promise “body fat”, “muscle mass”, “visceral fat” and “metabolic age” have moved from fitness enthusiasts into everyday cardiometabolic care. Patients bring screenshots to clinic. Some want reassurance. Others feel demoralised because the numbers don’t “make sense”—especially during GLP-1–assisted weight loss, where appetite, hydration, glycogen, and inflammation are changing at speed.

Used badly, smart scales can derail progress. Used well, they can be part of an armoury of behavioural tools that helps people stick with the unglamorous basics—protein prioritisation, strength training, low-glycaemic food choices, and consistent routines—long enough to lower their defended energy “set point”, with or without medication support.

This article offers a balanced, clinic-friendly way to think about smart body composition scales (bioelectrical impedance analysis; BIA): what they can do, what they can’t, and how a professional team can use them to support metabolic health.

What smart scales are actually measuring

Most consumer “body composition” scales use bioelectrical impedance analysis (BIA). A small electrical current passes through the body; the device measures resistance and reactance and uses proprietary equations to estimate fat mass and fat-free mass. In principle, BIA can be useful, but in practice the estimates are extremely sensitive to variables that matter a great deal during metabolic change—particularly total body water, intracellular/extracellular fluid shifts, and glycogen.¹,²

That sensitivity is not a minor technicality. It’s the reason BIA is best viewed as a trend tool, not a diagnostic instrument.

The accuracy problem: correlation is not agreement”

Some manufacturers advertise extremely high accuracy versus DEXA.

Two key points matter clinically:

1. Correlation is not agreement. A device can correlate well with DEXA across a population while still being meaningfully wrong for an individual. High correlation can coexist with wide limits of agreement and systematic bias.
2. Independent validation is the gold standard. The best evidence we have—independent of any single brand—is that consumer smart scales are not accurate for body composition and should not replace DEXA in patient care. A 2021 evaluation of smart scales concluded exactly that.¹

Why GLP-1 therapy makes smart-scale readings noisier

GLP-1 receptor agonists (and dual agonists) are highly effective for weight loss and metabolic improvement. But the rapid changes they induce can amplify BIA artefact.

Here’s why.

1) Glycogen and water shift

As appetite falls and carbohydrate intake often drops, glycogen stores tend to reduce. Glycogen binds water; when glycogen falls, intramuscular water falls too. BIA frequently mislabels these changes as “lean mass” or “muscle” loss.

2) Inflammation and extracellular water shift

As visceral fat and inflammatory tone reduce, extracellular fluid distribution can change. BIA is sensitive to these shifts.²

3) Lean mass is not the same as muscle

Even in high-quality clinical trials, reported “lean mass” changes can be heterogeneous, and interpretation requires care. Contemporary endocrine literature has emphasised that the discussion around GLP-1 therapy and fat-free mass needs context and basic body composition principles, not panic.⁵,⁶

This is why patients can be getting stronger, functionally better, and metabolically healthier—while a smart scale insists “muscle is falling”.

The real risk: scales can impede progress if expectations aren’t managed

There are three common failure modes in clinic:

1) “The scale says I’m losing muscle—so I’m going to eat more”

This can be catastrophically counterproductive if it pushes patients out of an energy deficit before visceral fat has meaningfully reduced.

2) “My numbers don’t match my effort—so this isn’t working”

If patients feel punished by their data, adherence collapses. This is a behavioural problem, not a physiology problem.

3) Clinicians overreact to the wrong signal

If a clinician treats BIA “muscle loss” as definitive and escalates calories or reduces GLP-1 prematurely, the patient may regain VAT and lose the metabolic momentum that was being built.

The solution is not to ban smart scales. The solution is to place them at the correct tier, with explicit rules for interpretation.

The case for smart scales: behaviour change needs tools

Now the other side of the ledger: why use smart scales at all?

Because lowering a defended energy set point is hard. Hunger and reward systems are powerful, ancient, and deeply wired. People don’t fail because they are lazy; they fail because biology pushes back. In this context, behavioural scaffolding matters.

A smart scale can be helpful because it:

• Creates a consistent measurement ritual (routine drives habit)
• Provides feedback loops (even if imperfect)
• Strengthens accountability (“someone is reviewing this with me”)
• Helps patients stay engaged through plateaus

This is the key clinical nuance: the value is often psychological and behavioural, not biochemical. The data don’t need to be perfect to be useful—provided the clinical team frames them correctly.

A pragmatic “metabolic clinic” way to use smart scales

Think in tiers. In a VAT-reduction clinic supported by GLP-1 therapy but anchored in lifestyle, your hierarchy might look like this:

Tier 1: VAT burden (primary target)

• CT or DEXA VAT where available
• Waist circumference / waist-to-height ratio as routine surrogate

Read more Medical imaging is the only accurate way to assess body composition

Tier 2: Muscle function (must be preserved)

• Strength progression (fixed-load reps)
• 30-second sit-to-stand
• Grip strength where practical

Grip strength and functional measures matter because they track outcomes far better than a “muscle mass” estimate on a consumer device; low grip strength is consistently associated with higher mortality and cardiovascular risk.⁷,⁸

Tier 3: Metabolic behaviour feedback (selected patients)

• CGM patterns for carbohydrate-sensitive phenotypes
  Used as feedback for food selection and meal sequencing, not as diagnosis.

Tier 4: Engagement and adherence tools

• Smart scale weight trend and fat % direction
• Food tracking to protect protein targets
• Coaching touchpoints and reinforcement

Smart scales belong here: useful, but never sovereign.

Which smart-scale metrics are worth watching?

If you’re going to use BIA devices, the most defensible approach is:

Metrics that can be useful (trend only)

• Body weight (weekly/monthly trend)
• Body fat % (direction over weeks, not days)

Metrics to treat with caution

• “Muscle mass”, “skeletal muscle”, “lean mass”
• Segmental muscle breakdown (arms/legs)
• “Visceral fat” scores generated from impedance (highly model-dependent)

The broader wearable-BIA literature supports this cautious stance: these devices can sometimes perform acceptably for broad measures like body fat percentage in certain populations, but variability—especially for skeletal muscle estimates and in people with higher body fat—limits their use for precise, individual-level assessment.⁹,¹⁰

How to prevent harm: reset expectations early

If your clinic uses smart scales, script the message before anxiety starts:

• “This scale is for engagement and trend, not truth.”
• “Ignore muscle numbers—trust strength and waist.”
• “During GLP-1 therapy, hydration and glycogen changes can make the scale look dramatic.”
• “If waist is falling and strength is stable or improving, you are preserving muscle function.”

This is not spin. It is disciplined interpretation.

Bottom line

Smart body composition scales are neither miracle devices nor useless toys. They are imperfect instruments that can either help or hinder metabolic care depending on how they are framed.

If you treat them as diagnostic—especially during GLP-1 therapy—you risk anxiety, inappropriate calorie increases, and stalled VAT reduction.

If you treat them as behavioural tools—embedded in a tiered system that prioritises VAT imaging/waist and functional strength—then they become part of a modern metabolic toolkit that helps patients do the hardest thing in medicine: sustain behaviour change long enough for biology to shift.

In the end, the goal isn’t to win an argument about “accuracy”. The goal is to help patients build a physiology that requires less pharmacological support over time: lower VAT, resilient muscle function, and a calmer appetite set point—with technology working for them, not against them.

References

1. Frija-Masson J, Beaudeux JL, et al. Accuracy of smart scales on weight and body composition. J Med Internet Res. 2021. Available from: https://pmc.ncbi.nlm.nih.gov/articles/PMC8122302/
2. Annunziata G, et al. Association between bioelectrical impedance analysis parameters and hydration/fluid distribution (discussion of resistance/reactance and TBW/ICW/ECW). [Article] 2025. Available from: https://pmc.ncbi.nlm.nih.gov/articles/PMC12392491/
3. Hume Health. Body Pod product page—“98% DEXA-like accuracy” claim. Accessed 10 Jan 2026. Available from: https://humehealth.com/pages/hume-body-pod
4. Hume Health. Science page—“98% high correlation” with DEXA for fat-free mass and body fat percentage. Accessed 10 Jan 2026. Available from: https://humehealth.com/pages/science
5. Tinsley GM, et al. Fundamental body composition principles provide context for interpreting GLP-1RA trial changes in fat-free mass. J Endocr Soc. 2024. Available from: https://academic.oup.com/jes/article/8/11/bvae164/7775409
6. Neeland IJ, et al. Changes in lean body mass with GLP-1–based therapies: heterogeneity and interpretation. [Review] 2024. Available from: https://pubmed.ncbi.nlm.nih.gov/38937282/
7. López-Bueno R, et al. Handgrip strength thresholds and risk of all-cause and cardiovascular mortality: systematic review and dose-response meta-analysis. [Journal] 2022. Available from: https://www.sciencedirect.com/science/article/pii/S1568163722002203
8. Zhang F, et al. Association of handgrip strength and risk of cardiovascular disease in middle-aged and older adults. [Journal] 2024. Available from: https://pmc.ncbi.nlm.nih.gov/articles/PMC11480190/
9. Mehra A, et al. The evolution of bioimpedance analysis: wearable BIA, accuracy discrepancies vs DXA, and validation needs. [Narrative review] 2025. Available from: https://www.sciencedirect.com/science/article/abs/pii/S0899900724002508
10. Carrier B, et al. Wearable-BIA device validity compared with DXA: acceptable BF% accuracy in some groups but limitations for skeletal muscle and higher BF%. Frontiers in Sports and Active Living. 2025. Available from: https://www.frontiersin.org/journals/sports-and-active-living/articles/10.3389/fspor.2025.1644082/full

Related posts

1. Sarcopenia: Are We Diagnosing the Wrong Muscle Problem?
2. Medical imaging is the only accurate way to assess body composition
3. A New Year Reset: Why Your Waist Matters More Than Your Scales
4. PCSK9, visceral fat, and the modern metabolic environment
5. Your Genes and Fat: Why Some People’s Cholesterol Rises More Than Others
6. Epigenetics and Metabolic Health: How Lifestyle Rewrites Your Genes
7. So what does determine your LDL (‘bad’) Cholesterol?
8. Examples of CT VAT scans and normal ranges for VATI
9. Why everyone is talking about VAT
10. LDL: the lower the better
11. How to Lose Visceral Adipose Tissue (VAT) and Improve Metabolic Health: A Guide to Sustainable Weight Loss
12. Visceral Fat, Mitochondria, and the Energy Trap: Why We Store Fat Where We Shouldn’t
13. Why everyone is talking about VAT