Body recomposition vs weight loss: what DXA reveals

Key takeaways

• The bathroom scale reports total mass. DXA separates fat mass, lean mass, and bone mineral content, and the two often disagree about whether something is working .
• Losing fat and gaining lean at the same time (body recomposition) is clearest in people new to resistance training or carrying higher body fat; a randomized trial showed it in four weeks on high protein .
• Consumer bioelectrical-impedance (BIA) scales disagree widely with DXA, with limits of agreement too large for tracking small short-term changes; watch the trend, not one reading .
• Trials report body weight, fat mass, lean mass, and waist size as separate endpoints. Reading total weight alone hides whether the loss was fat or muscle.
• In GLP-1 and dual-agonist body-composition substudies, lean mass was about a quarter of the weight lost on tirzepatide and roughly 40 percent on semaglutide .

What a scale reading actually includes

A bathroom scale reports total body mass, everything inside you at the moment you step on: fat, muscle, bone, organs, blood, cellular water, gut contents, and a full or empty bladder. In a healthy adult that number can swing a kilogram or two day to day, driven by hydration, sodium, glycogen, and gut transit. None of that swing is fat.

This matters because the scale is the easiest thing to measure, so it becomes the main feedback signal in self-run weight programs. Lose 0.5 kg of fat in a week but refill 1 kg of glycogen after a hard training block, and the scale calls it a failure while your body composition improved. Drop 2 kg of mostly water in the first week of a diet, and the scale calls it a win while fat has barely moved.

What DXA actually resolves

Dual-energy X-ray absorptiometry (DXA) started as a bone-density scan and is now the clinical reference for both bone mineral density and three-compartment body composition: fat mass, lean soft tissue, and bone mineral content. On the same scanner under the same conditions, repeat-scan precision runs about 0.5 to 1 percent for total lean and fat mass .

DXA measures non-fat, non-bone soft tissue and labels it lean. That compartment is mostly muscle, but it also includes connective tissue, organ mass, residual water, and the intramuscular fat the scan cannot separate from surrounding tissue. The clinical term is 'lean soft tissue' or 'fat-free mass', not 'muscle', for exactly that reason.

• Repeat-scan precision on one scanner: about 1 percent for total fat mass and about 0.5 percent for total lean mass .
• Cross-scanner comparability is worse. Different manufacturers' machines can disagree on the same body even after cross-calibration, so track on one scanner.
• Hydration shifts the lean reading, because soft-tissue water counts as lean. A dehydrated scan reads lower lean mass.
• Bone mineral content moves slowly. A large short-term BMC change usually means scanner artifact, not biology.

The trial measures total weight; your question is usually fat or muscle

Most weight-loss trials, GLP-1 programs included, use total body weight as the primary endpoint, because the scale is universal and FDA labeling has long been written around weight-percentage targets. STEP 1 and SURMOUNT-1 reported the headline percentages and ran DXA body-composition substudies on a smaller subset .

Your question is rarely how much total mass you lost. It is usually did I lose fat, did I keep muscle, or do I look different than last month. Those are DXA questions, and the scale cannot answer them. The number a trial headlines and the number you care about are different, even when they move in the same direction.

What body recomposition actually looks like in published trials

Body recomposition means losing fat and adding lean at the same time, often near a stable scale weight. In one randomized trial, young men doing resistance and interval training six days a week on a 40 percent energy deficit gained about 1.2 kg of lean mass and lost about 4.8 kg of fat on 2.4 g/kg/day of protein, versus roughly no lean change on 1.2 g/kg/day .

Three things have to line up. Resistance training is the signal to hold muscle. Protein is the material to rebuild it. And the calorie deficit is what forces the body to pull from fat stores. Take one away and the effect mostly disappears. The protein target has a ceiling: across 49 trials, added protein improved resistance-training gains in fat-free mass only up to about 1.6 g/kg/day .

• In people newer to resistance training, simultaneous gain-and-lose is well demonstrated. Longland's young men added lean and lost fat in four weeks .
• In already-trained athletes it is harder and the evidence is mixed. A 2022 review found a calorie deficit mostly costs lean mass, and higher training volume at best limits the loss rather than reliably building muscle .
• The fat-free-mass response to the same resistance training weakens with age, so older lifters work harder for the same lean, and protein and training volume matter even more with the years .
• Take a GLP-1 or have surgery without a training and protein plan, and fat and lean come off together. The GLP-1 section below has the numbers.

BIA, calipers, smart scales, and where they fit

Bioelectrical impedance analysis (BIA) and consumer smart-scales pass a small current through the body and estimate the fat-versus-lean split from the impedance. BIA is fast and cheap. It is also sensitive to hydration, recent meals, and hard exercise in the prior 12 hours. Against DXA in 3,655 paired measurements, BIA overestimated fat-free mass by roughly 3 to 8 kg and underestimated fat mass by 3 to 6 kg, with limits of agreement too wide to trust a single reading for small short-term changes .

Skinfold calipers measure subcutaneous fat at set sites and convert it to a body-fat estimate through population equations. In the hands of one trained operator they can catch a change of a point or two of body fat. With a different operator each time, non-standard sites, or sloppy technique, the noise swamps the signal.

• DXA: the clinical reference for tracking composition over time. Cost runs roughly $50 to $150 per scan in the US, depending on the clinic.
• BIA: fine for trends under tightly fixed conditions (same time of day, same hydration, same fed or fasted state). Not for absolute numbers.
• Calipers: useful with one trained operator and a fixed protocol; otherwise mostly noise.
• Smart scales: BIA inside, same caveats. Daily wobble usually exceeds real composition change.
• Air-displacement plethysmography (BodPod): near DXA precision, less widely available.
• Underwater weighing: the old reference, rarely used now.

What this means on a GLP-1 or dual agonist

On a GLP-1 or a dual agonist alone, without a resistance-training and protein plan, fat and lean come off together. The body-composition substudies show it plainly: in SURMOUNT-1 about a quarter of the weight lost was lean mass , and in the STEP 1 DXA subset the lean share was closer to 40 percent, per a 2025 summary of that subset .

The useful move is to add resistance training and hit a higher protein target while on the drug, so a smaller share of the loss is muscle. Newer combinations push total weight loss further: CagriSema (cagrilintide plus semaglutide) reached about 20 percent weight loss in REDEFINE 1 . Whatever the drug, the muscle-preserving levers are the same and independent of it: resistance training and enough protein.

The honest read

The scale and DXA answer different questions, and most of the confusion about whether a program is working comes from reading the wrong one. If you are changing body composition, DXA is the measurement that shows it. The scale is a cheap daily proxy that can mislead in both directions, so use it for the trend and let DXA settle whether the change was fat or muscle.

For research and educational purposes only. Not medical advice.

pepSmart has not commissioned independent clinical review of this article.

More on how we write and source these pieces: Editorial process and contributor disclosure and Sourcing posture.

Spot an error? Email corrections via /about.

Sources: 10 entries, primary canon (peer-reviewed trials, body-composition substudies, and meta-analyses on PubMed and PMC) plus two peer-reviewed reviews acknowledged inline, last reviewed 2026-07-08.

Related tools

• Tirzepatide dose calculator - Run tirzepatide-focused vial draw math.
• GLP-1 conversion calculator - Convert a GLP-1 mg dose to U-100 units and ml.
• GLP-1 ramp planner - Preview a linear educational dose-step table.
• Peptide half-life calculator - Estimate single-dose decay from cited half-life constants.
• PK simulator overview - Public overview of the Pro pharmacokinetic simulator.
• Body-recomp simulator overview - Public overview of the Pro body-recomposition simulator.