Having your DEXA scan results explained starts with one distinction: the report estimates fat, lean tissue and bone separately. Lean tissue includes muscle, water and organs. A lower lean-mass number doesn't tell you how much muscle changed. The useful comparison includes the scanner's reference and your earlier results.
DXA, often spelled DEXA, means dual-energy X-ray absorptiometry: a scan using two X-ray energies to estimate different tissues.
DEXA scan results explained: what each line measures
A body-composition report describes how body weight is divided. Some scanners add estimates beyond the basic fat, lean and bone measurements.
| Report line | Plain meaning |
|---|---|
| Total fat mass | Estimated fat across the whole body, usually in grams or kilograms. |
| Regional fat mass | Estimated fat within an area, such as the trunk, arms or legs. |
| Lean mass | Tissue other than fat and bone mineral, including muscle, organs and water. |
| Body-fat percentage | Fat mass divided by total scanned mass, multiplied by 100. |
| Visceral fat estimate | A software estimate of fat around abdominal organs, within a defined region. |
| Appendicular lean mass index, or ALMI | Arm-and-leg lean mass divided by height in meters squared, reported in kg/m². |
These definitions follow the SURMOUNT-1 DXA report's measurement table and ISCD's reporting positions, guidance from the International Society for Clinical Densitometry. Visceral fat and ALMI are optional outputs. Their absence doesn't mean the scan failed.
Reference ranges vary by scanner and population. Kelly's 2009 NHANES dataset in PLOS ONE built age-, sex- and ethnicity-specific curves from a U.S. health survey using Hologic scanners. Its authors explain that calibration, how the scanner's measurements are standardized, affects results. A clinician reads your result against the scanner's reference; a population average isn't a treatment target.
In SURMOUNT-1: tirzepatide's fat and lean changes
In that trial, fat mass fell more than lean mass. The detailed DXA findings appear in Look and colleagues' 2025 Diabetes, Obesity and Metabolism paper, a separate report from the main 2022 trial publication.
The substudy enrolled 255 adults; 160 had paired scans for analysis: 124 receiving tirzepatide and 36 receiving placebo, an inactive comparison treatment. Participants had obesity or overweight without type 2 diabetes. The three tirzepatide dose groups were combined.
| Change from the starting amount at 72 weeks | Tirzepatide | Placebo |
|---|---|---|
| Body weight | −21.3% | −5.3% |
| Total fat mass | −33.9% | −8.2% |
| Total lean mass | −10.9% | −2.6% |
| Estimated visceral fat mass | −40.1% | −7.3% |
The full substudy report also gives fat losses of 15.9 kg versus 3.6 kg and lean losses of 5.6 kg versus 1.2 kg. Roughly three-quarters of the combined fat-and-lean loss was fat in both groups. These are trial averages, not expected changes on someone's next scan.
In STEP 1: semaglutide's fat and lean changes
In that trial, lean tissue fell in kilograms even as it became a larger share of body weight. Both statements can be true because fat fell faster.
STEP 1's main results appeared in the New England Journal of Medicine in 2021. Its exploratory DXA analysis covered 140 adults without diabetes: 95 assigned semaglutide and 45 placebo, over 68 weeks.
The 2021 Journal of the Endocrine Society abstract reports semaglutide-group declines of 19.3% in total fat, 9.7% in lean tissue and 27.4% in regional visceral fat. Those percentages describe changes from each tissue's starting amount.
For an exact comparison in kilograms, the posted STEP 1 registry results give these observed means, averages from participants with available scans:
| Change at 68 weeks | Semaglutide, 83 participants | Placebo, 39 participants |
|---|---|---|
| Total fat mass | −9.3 kg | −1.5 kg |
| Total lean mass | −5.8 kg | −1.8 kg |
These observed means use fewer participants than the abstract's statistically calculated percentages. They aren't interchangeable estimates. The semaglutide muscle-loss explainer covers the clinical question behind a falling lean-mass result.
Neither substudy directly compared semaglutide with tirzepatide. Putting their tables side by side cannot establish which preserves more muscle.
Percent change, percentage points and muscle are different
A percent change compares a tissue amount with its own starting amount. Percentage points describe a change in its share of the body.
The STEP 1 investigators' poster reports a 3.0-percentage-point increase in lean tissue's share with semaglutide, versus 0.1 with placebo. That increase didn't mean muscle was gained: total lean mass still fell.
The distinction matters in newer trials too. As of October 1, 2026, the day this article was written, Regeneron's October 1 COURAGE update reports separate DXA lean-tissue and MRI thigh-muscle findings with experimental trevogrumab added to semaglutide. MRI uses magnets to measure muscle more directly. The trevogrumab page follows that trial. A scan of tissue amount still isn't a test of strength.
Tesamorelin's visceral-fat number came from CT
In Falutz's 2007 New England Journal of Medicine trial, 412 adults with HIV and excess abdominal fat were studied for 26 weeks. Visceral fat fell 15.2% with tesamorelin and rose 5.0% with placebo, according to the paper's indexed abstract.
The measurement came from CT, computed tomography, which uses X-rays to create cross-sectional images. The study measured abdominal fat area in square centimeters. A DXA estimate uses software to infer deep fat within a selected region, often reporting mass or volume. Those numbers aren't interchangeable, even when both reports use the word “visceral.”
The tesamorelin clinical trials page owns the fuller trial story. This result concerns an HIV population with excess abdominal fat. It doesn't establish a DXA target or a general weight-loss result for someone without HIV.
Bone scores and the comparison a clinician can use
Bone-density T- and Z-scores usually come from a separate hip-and-spine assessment, rather than the body-composition report. A T-score compares density with a young-adult reference. A Z-score compares it with an age-matched reference, accounting for sex.
ISCD's guidance generally prefers T-scores for women after menopause and men 50 or older, and Z-scores for younger adults. These bone scores don't grade muscle or body fat.
For repeat body-composition scans, the useful record includes the scanner, software, units and body region measured. Similar hydration and positioning matter too. SURMOUNT-1 used the same scanner and software for each participant under similar conditions. Small differences can reflect measurement variation alongside real tissue change.
Through Promise, tirzepatide, semaglutide and tesamorelin are dispensed as compounded medications, prepared by a pharmacy for a prescription. That is different from an FDA-approved product: the formulations offered here are not FDA-approved. The FDA explains that compounded drugs aren't reviewed for safety, effectiveness or quality before marketing. The trial figures above don't establish the same results for these preparations.
A licensed provider may still prescribe — that decision is between you and your doctor. At Promise, a licensed provider reviews every request, and not everyone qualifies.
