What the 24-week study actually found

The retrospective, observational, single-arm analysis followed 122 adults with overweight or obesity and no diabetes for 24 weeks. About 80% were women. Participants received tirzepatide, beginning at 2.5 mg weekly and increasing to 5 mg after four weeks, with individualized dose adjustments. They also received a standardized low-carbohydrate diet and tailored physical-activity guidance.

Mean body weight declined from 95 kg to 80 kg, an approximately 16% reduction. The body-composition estimates pointed toward fat-dominant loss:

• Fat mass fell by roughly 35%.
• Fat-free mass fell by roughly 5%.
• Estimated skeletal muscle mass did not change significantly.

Those findings should be reported as three separate results, not condensed into “16% weight loss with no muscle loss.” A small decline in fat-free mass was observed, while the estimated skeletal-muscle measure was statistically unchanged.

Body-composition terms are not interchangeable

• Fat-free mass is all non-fat tissue, including muscle, bone, organs, connective tissue, and water.
• Lean mass, as measured by DXA, generally means fat-free mass minus bone mineral content, also called lean soft tissue.
• Estimated skeletal muscle mass is a model-derived subset from a device. It is not a direct measure of muscle fibers, strength, or muscle quality.

These distinctions are central to interpreting tirzepatide body-composition data. A change in fat-free or lean mass cannot automatically be called a change in skeletal muscle.

Why stable estimated muscle is encouraging, not proof of preservation

The study had no randomized placebo or untreated comparison group. Because every participant received tirzepatide plus dietary and physical-activity support, the results cannot isolate the medication as the cause of the weight or body-composition changes.

It was also retrospective and observational, drawn from a specific clinical setting, with analysis apparently focused on people who had both baseline and 24-week follow-up measurements. The cohort was relatively small, and 24 weeks is a short period for judging longer-term treatment or weight-maintenance outcomes.

Most importantly, a statistically nonsignificant change in device-estimated skeletal muscle mass does not establish biological equivalence. It can reflect measurement variability or limited statistical power. It also does not demonstrate preserved muscle strength, power, physical performance, or muscle quality. The 122-person study did not establish preservation through direct strength or functional testing.

BIA can track trends, but it does not directly measure muscle

The real-world study used bioelectrical impedance analysis, or BIA. BIA estimates body compartments from electrical properties and prediction equations, with assumptions about total body water and fat-free-mass hydration. Its results can be affected by hydration, food and beverage intake, recent exercise, body position, temperature, and device-specific equations.

That does not make BIA useless. Under standardized conditions, it can help follow trends. But BIA is less definitive for muscle tissue than imaging-based approaches:

• DXA estimates whole-body and regional fat mass, lean soft tissue, and bone mineral.
• MRI can assess muscle volume and muscle-fat infiltration more directly.

DXA still has limitations, including hydration and scanner-related constraints. MRI is more informative for muscle volume and intramuscular fat, but is more costly and less scalable. Neither one alone substitutes for testing strength and function.

The SURMOUNT-1 DXA results show why lean mass matters

A randomized, double-blind, placebo-controlled SURMOUNT-1 DXA substudy offers a useful, but not directly interchangeable, comparison. It included 160 participants with baseline and week-72 DXA data. Participants also received diet and activity counseling, including a 500-kcal daily deficit goal and advice to increase activity to at least 150 minutes per week.

With pooled tirzepatide, body weight decreased by 21.3%, fat mass by 33.9%, and lean mass by 10.9%. In absolute terms, this was approximately 15.9 kg of fat mass and 5.6 kg of lean mass. About 74% of weight lost was attributed to fat mass and 26% to lean mass.

This is still a favorable fat-dominant pattern, but it shows why claims of complete muscle preservation go beyond the evidence. DXA lean mass is not identical to skeletal muscle, just as BIA-estimated skeletal muscle is not direct evidence of muscle fiber mass or function. The DXA substudy also did not directly test muscle strength or physical performance.

Muscle volume, muscle fat, and function can move differently

A 2025 MRI substudy in adults with type 2 diabetes found reduced thigh muscle volume with tirzepatide-associated weight loss, alongside reduced muscle-fat infiltration. Reduced infiltration may be favorable, and the muscle-volume changes were broadly consistent with changes expected during weight loss.

However, that was a post-hoc exploratory analysis in a population with type 2 diabetes, not the same as the 122 adults without diabetes in the real-world study. It reinforces a broader point: muscle health includes quantity, quality, strength, power, and physical performance. No single body-composition number answers all of those questions.

Practical priorities during GLP-1 and GIP/GLP-1 treatment

Clinical guidance supports using nutrition and training to manage the risk of lean-tissue loss during pharmacologically assisted weight reduction. The American Diabetes Association recommends optimizing protein intake through dietary counseling, resistance training to preserve lean mass, and regular assessment of muscle strength and function.

Protein should not be framed as a stand-alone solution. A joint professional-society advisory emphasizes that protein intake alone is unlikely to preserve muscle without structured resistance training. Protein goals need individualization for body size, age, kidney function, dietary pattern, and clinical circumstances.

These are risk-management strategies, not proof that they caused the 24-week study results. The observational study did include dietary and activity guidance, but it was not designed to test a particular protein plan or resistance-training program.

How to interpret the 16% result

The most accurate reading is that this 24-week real-world cohort experienced substantial, fat-dominant weight loss while BIA-estimated skeletal muscle mass was statistically stable. It is a promising finding, not proof that tirzepatide preserves skeletal muscle.

The observed outcome should not be generalized to every dose, duration, body size, or level of adherence. Better answers require longer studies with appropriate comparison groups, robust body-composition methods, and direct measures of strength and physical function.