Biohacking Kompakt

Tip · Movement

Preserving muscle on GLP-1

On weight-loss injections, lean mass is lost as well as fat: in STEP 1 just under 40 percent of the weight loss (DXA substudy, 140 participants), across more than 40 DXA reports 29 percent on average. A good part of this is water and glycogen, not muscle. The best-supported countermeasure remains strength training with adequate protein.

In every discussion about weight-loss injections the same number comes up: one third of the weight lost is said to be muscle. The number does appear like that in the studies, and yet it does not measure what its name claims.

In short

Anyone who loses weight on a GLP-1 injection loses lean mass as well as fat: in STEP 1 around 38 percent of the weight lost, on tirzepatide about 25 and on semaglutide plus cagrilintide about 28 percent. For an ordinary diet, the old rule of thumb is about 25 percent, so the gap is smaller than the headlines. Lean mass is also not the same as muscle: the usual measurement method counts water, glycogen and organ tissue as well. The best-supported countermeasure is strength training with adequate protein; two drugs are candidates, not a finished remedy.

Where the number comes from

The best-known source is STEP 1, the large study on semaglutide: 1,961 adults, 68 weeks. Among the participants whose body composition was measured, around 38 percent of the weight lost was lean mass, that is, everything except fat. This one number turned into the catchphrase of muscle loss.

In the other large studies the share is smaller. On tirzepatide, measured in SURMOUNT, it is about 25 percent. On the combination of semaglutide and cagrilintide, about 28 percent. STEP 1 is thus the outlier on the high side and not the norm.

The comparison figure that is almost never mentioned amid the excitement is dieting without medication. For that, a rough rule of thumb has applied for decades: about 75 percent fat, about 25 percent lean mass. Anyone who loses weight therefore always loses lean mass as well. That is not an injection phenomenon but a weight-loss phenomenon: a lighter body needs less supporting tissue.

Why lean mass does not mean muscle

Measurement is almost always done with DXA, an X-ray method that divides the body into fat, bone and the rest. This rest is called lean mass, and it contains far more than muscle: water. Glycogen, the sugar store in muscle, with each gram of glycogen binding about 3 grams of water. Liver tissue, which shrinks with weight loss. And fat that lies between the muscle fibers.

This leads to a systematic confusion. Anyone who empties their glycogen stores loses several kilograms of lean mass on the DXA report without having lost a single muscle fiber. The honest wording would not be “one third muscle loss” but “one third non-fat”: less dramatic and more accurate.

The gap between 38 and 25 percent does not disappear as a result, and it is a legitimate point. An obvious explanation: on the injection, weight goes down faster and further, and rapid weight loss tends to cost more lean mass.

For whom the problem is serious nonetheless

Muscle is lost, that is beyond question. What is open is for whom the loss matters. For older people it does, because many already have too little muscle beforehand and because muscle mass in old age helps determine independence. Anyone who loses 20 kilograms at the age of 70 and loses strength in the process pays a high price for the result.

Precisely for this reason, the enobosarm study QUALITY included only people aged 60 and over and measured stair climbing as performance instead of merely imaging tissue.

What is really supported against it

Against the loss there is one intervention with more studies behind it than both drugs combined: strength training and adequate protein. That is not a consolation prize for lack of anything better, but the best-supported muscle preservation there is, regardless of what causes the weight to drop.

Added to this is the choice of the measure. A DXA report is an image of tissue. Whether someone gets up the stairs and out of the chair is what decides everyday life. If body composition and function move in the same direction, that is a better signal than a single measurement alone.

Two drugs in development

Two candidates are far enough along to be worth discussing. Bimagrumab is an antibody that blocks the myostatin pathway. Myostatin is the body’s own brake on muscle growth; if it is removed, muscle grows. Enobosarm is a selective androgen receptor modulator, i.e. a SARM, from the class that has been traded on the gray market in the scene for years and is being properly studied for this purpose here for the first time.

Both shift body composition in the desired direction without slowing weight loss. Neither is approved for this, and their data differ greatly.

What is well supported

The phenomenon itself is robust: weight loss costs lean mass, on GLP-1 as well as on a diet, measured in studies with four-digit participant numbers over more than a year. It is also robust that the share varies between the active substances. The criticism of the measurement is well supported: that DXA does not distinguish muscle, water and organ tissue is not an objection from outside but is stated in the studies themselves. And strength training with adequate protein intake is the longest-studied measure for keeping muscle while losing weight.

What the studies show

STEP 1 and the comparison studies

STEP 1 studied semaglutide in 1,961 adults over 68 weeks. In the subgroup with measurement of body composition, around 38 percent of the weight lost was lean mass. SURMOUNT with tirzepatide came to about 25 percent, the combination of semaglutide and cagrilintide to about 28 percent. The yardstick for comparison is the rule of thumb used for decades for a diet with a calorie deficit: about 75 percent fat and 25 percent lean mass.

BELIEVE: bimagrumab plus semaglutide

Randomized study with 507 adults with obesity over 48 weeks in 9 groups, published in Nature Medicine and funded by Eli Lilly. On semaglutide alone, lean mass fell by 6.9 percent, on the combination of bimagrumab and semaglutide by only 2.3 percent, and on bimagrumab alone it rose by 1.0 to 1.1 percent. Weight fell more on the combination, namely by 17.8 kilograms versus 14.2 kilograms on semaglutide alone. The gap was largest for fat mass: minus 33.7 percent on the combination versus minus 21.1 percent.

QUALITY: enobosarm in people aged 60 and over

Study by the company Veru with 168 people aged 60 and over with overweight over 16 weeks. All received semaglutide, plus enobosarm or placebo. The loss of lean mass was 71 percent smaller than on placebo, clearly statistically significant. Fat loss was 27 percent higher, but this difference was not statistically significant. Total weight was practically the same, 4.4 kilograms on enobosarm versus 4.7 kilograms on placebo. In the stair-climbing function test, 19.4 percent on enobosarm had a decline in performance of at least 10 percent, versus 42.6 percent on placebo, that is, around 54 percent fewer. As of September 2026, there is no peer-reviewed publication on this. The follow-up study PLATEAU, with 239 enrolled participants aged 65 and over over 68 weeks, has weight change as its primary objective; muscle and stair performance are secondary objectives.

Two meta-analyses from 2026 and an MRI analysis

Across more than 40 DXA reports on GLP-1-based agents, the share of lean mass in weight loss averaged 29 percent, with wide scatter. Two meta-analyses from 2026 disagree on whether incretins cost more lean mass than a diet: one finds no significant difference (lifestyle 26 percent), the other 33 versus 22 percent for diet without exercise. They agree on training: 17.5 percent with strength training and 7.7 percent for diet with exercise, respectively. In a post hoc MRI analysis on tirzepatide in type 2 diabetes, funded by Eli Lilly, thigh muscle volume fell to the extent expected for the weight loss, while fat infiltration in the muscle decreased more than expected.

Where the data stop

The criticism of the measurement also applies to the studies that set out to counter the loss. The authors of BELIEVE state explicitly that they measured with DXA and not with MRI, that is, with exactly the method that cannot separate muscle and water. That is candid for a study whose selling point is muscle preservation, but it does not change the fact that the percentages are images of tissue and not measured function.

How to do it

The most effective part of this topic is not on a prescription: throughout the entire weight loss, build up two things in parallel, strength training for the large muscle groups and a protein intake that does not drop along with the calorie intake. This combination has more studies behind it than both drug candidates combined and works regardless of what causes the weight to drop. However, the studies evaluated here do not provide a protein amount per kilogram or a training frequency proven for exactly this situation.

Measure function rather than just tissue, i.e. stair climbing and getting up from a chair. A DXA report images tissue and counts water, glycogen and organ tissue as lean mass. QUALITY, with 168 people aged 60 and over over 16 weeks, found around 54 percent fewer participants with a decline in stair performance of at least 10 percent on enobosarm, according to a company announcement and not published in peer-reviewed form as of September 2026. For the two investigational substances themselves, no usage information is given here, because neither is approved: in BELIEVE, bimagrumab reduced the loss of lean mass from 6.9 to 2.3 percent, but measurement was also by DXA, and on bimagrumab alone 14 to 21 percent of participants discontinued treatment, in the combination groups 5.5 to 12.5 percent. That strength or independence is preserved in the end has not been shown for either substance.

Safety

The side effects of the candidates rarely make the headlines. In BELIEVE, 14 to 21 percent of participants discontinued treatment on bimagrumab alone, 5.5 to 12.5 percent in the combination groups, and 4 to 9 percent on semaglutide alone. Common in the combination groups were muscle spasms at 57 to 64 percent and acne at 43 to 55 percent. The few cases of pancreatitis and skin cancer were also distributed across the placebo and semaglutide groups and cannot be attributed to bimagrumab. Muscle spasms, diarrhea and acne are cited as common side effects of bimagrumab. For enobosarm, the results are available only as an announcement by the manufacturer; a peer-reviewed publication with complete safety data is lacking. Enobosarm is ostarine (MK-2866), a SARM that is on the doping list and not approved; there are case reports of liver damage, and gray-market products are not the study drug. Strength training and adequate protein are the part of this strategy without a notable risk profile.

BK-Score Thin human evidence

Human evidence5
Mechanism7
Safety data6
Hype gap4
Track record of use7

Evidence 5: there are several randomized studies with four-digit (STEP 1, 1,961 participants) to three-digit participant numbers (BELIEVE, 507; QUALITY, 168), but almost all endpoints are DXA surrogates, and the function data are thin: in BELIEVE, grip strength improved in only one of four combinations, and the stair measurement from QUALITY comes from a company announcement rather than a peer-reviewed publication. Mechanism 7: the myostatin pathway is confirmed in humans; bimagrumab alone increased lean mass by 1.0 to 1.1 percent over 48 weeks, and the dilution effect via glycogen, which binds about 3 grams of water per gram, is also measurable in humans. Safety 6: for bimagrumab there are controlled data over 48 weeks, with discontinuation rates of 14 to 21 percent on bimagrumab alone and frequent muscle spasms (57 to 64 percent) and acne (43 to 55 percent) in the combination groups; for enobosarm only 16 weeks without a full peer-reviewed publication. Hype 4: the headline figure of one third muscle loss measures non-fat rather than muscle and is nevertheless sold as muscle loss, while the primary literature itself names the DXA limitation. Use 7: strength training and protein have been widely used for decades, whereas the drug countermeasures have so far been used only in studies.

What is rated is the state of knowledge, not the effect. “Safety data 9” means well studied – not harmless.
Subjective assessment by Biohacking Kompakt based on published scoring rules – not a scientific rating and not a medical recommendation. Rules and all ratings (German)

Frequently asked questions about preserving muscle on GLP-1

Do you lose more muscle on a weight-loss injection than on a diet?

The measured share of lean mass in weight loss is around 38 percent on semaglutide in STEP 1, about 25 percent on tirzepatide, and about 25 percent for an ordinary diet according to the old rule of thumb. The gap is therefore smaller than is often claimed and depends heavily on which study you consult. That lean mass is lost at all applies to every form of weight loss.

Is lean mass the same as muscle?

No. In the usual measurement method, lean mass is everything except fat and bone, i.e. also water, glycogen and organ tissue. Each gram of glycogen binds about 3 grams of water, which is why the measured value can drop by several kilograms without any muscle fibers being lost.

What works best against the loss?

Strength training together with adequate protein. This combination has more studies behind it than both drug candidates combined and works regardless of what causes the weight to drop. It is therefore not a stopgap but the best-supported part of the topic.

How well studied is bimagrumab?

In the BELIEVE study with 507 adults over 48 weeks, lean mass fell by only 2.3 percent instead of 6.9 percent on the combination with semaglutide, and weight decreased more. However, measurement was by DXA, i.e. without separating muscle and water. In addition, 14 to 21 percent of participants discontinued on bimagrumab alone, 5.5 to 12.5 percent in the combination groups, and 4 to 9 percent on semaglutide alone.

What is special about the enobosarm study?

It measured stair performance and did not merely image tissue: on enobosarm, around 54 percent fewer participants had a marked decline in stair-climbing performance than on placebo. However, the figures come from an announcement by the manufacturer, not from a peer-reviewed publication, and 168 people over 16 weeks is few and short.

Should older people be especially careful on GLP-1?

The loss weighs more heavily for them, because many already have little muscle beforehand and because muscle mass in old age helps determine independence. That is why the enobosarm study QUALITY included only people aged 60 and over. Everything else belongs in the hands of the treating physician.

Related

Sources

Open in the database – with all tips, filters and BK-Score (German app)

Information only, not medical advice and not a usage recommendation. If you have pre-existing conditions, and before major changes, consult a physician. Last updated: 2026-09-19.