GLP-1s and Testosterone in Men: Where the Evidence Stops

A testosterone panel comes back low, and the result gets read out at a US men’s-health clinic. Somewhere in that visit a GLP-1 gets floated — a weight-management drug, offered as the answer to a hormone number.

Underneath that setup is a real finding. A 2025 meta-analysis pooled seven studies and 680 participants and found that GLP-1 receptor agonist treatment raised total serum testosterone.

The headline travels. The rest of that paper’s conclusion usually doesn’t: the authors wrote that the limitations of the current literature do not allow a direct action on testicular function to be demonstrated. They reported the rise and, in the same breath, said the evidence couldn’t show where it came from — which leaves a candidate they had no way to rule out, and it has nothing to do with the testicles.

Why a weight-management drug like Wegovy shows up in a testosterone workup

Two men can leave the same lab holding an identical number and not have the same problem.

For one of them, the hardware is damaged — the testicles themselves, or the pituitary and hypothalamus that signal them. For the other, nothing is broken. The system is intact, and the metabolic state around it is holding the signal down — the literature calls that second version functional hypogonadism, and older papers call it obesity-associated hypogonadotropic hypogonadism.

Everything hinges on functional. A signal held down is a signal that can let up — which is the entire reason a drug approved for weight management turns up inside a hormone workup.

Does weight loss on its own raise testosterone?

Weight loss by itself was associated with a rise in both bound and unbound testosterone. The finding landed in 2013, before any of this was a GLP-1 question. Corona and colleagues pulled 24 studies together — men on low-calorie diets, men who had bariatric surgery — into a systematic review and meta-analysis in European Journal of Endocrinology. In multiple regression, the degree of body weight loss came out as the best determinant of the rise in total testosterone.

A retrospective real-world cohort lands in the same place — 71 men with obesity and uncontrolled type 2 diabetes, reviewed after they were prescribed glucose-lowering therapy. Among those who lost more than 10% of body weight, 94% had total testosterone at or above 300 ng/dL, the threshold that paper worked with.

That 94% gets misread. It came out of a cohort prescribed glucose-lowering therapy broadly, with no GLP-1-only protocol behind it. So it’s a weight-loss number that happened to be collected in a diabetes clinic, and pinning it on semaglutide (Wegovy) or tirzepatide (Zepbound) credits those drugs with something the cohort never separated out.

If the degree of weight loss is the strongest determinant that analysis measured, then anything producing a large weight change will look like it’s doing something to testosterone. No trickery required — that’s just what a confound looks like when the biggest measured driver moves on the drug’s schedule.

What the pooled GLP-1 studies found in men

Salvio and colleagues pooled the GLP-1 receptor agonist studies that measured testicular function in men — a systematic review and meta-analysis in Andrology, 2025. Seven studies made it into the quantitative analysis, 680 participants between them. Treatment produced a significant increase in total serum testosterone, reported as a standardized mean difference of 1.39 ng/mL, with a 95% confidence interval of 0.70 to 2.09.

A standardized mean difference is an effect-size measure across pooled studies. It isn’t a line on anyone’s lab report saying the number climbed by 1.39. The interval sits clear of zero. Significant means that and nothing more.

So the rise is real. Pooled, peer-reviewed. That much of the pitch isn’t invented.

So is the drug acting on the testicles?

Salvio and colleagues take it up themselves, in the conclusion of the same paper. They wrote that GLP-1 receptor agonists may have a role in the therapy of functional hypogonadism related to overweight and obesity — and right beside it, that the limitations of the current literature do not allow a direct action on testicular function to be demonstrated. Two sentences. One paper.

There are no rival camps arguing across a journal here: one group ran the numbers, published what came out, and said the literature wasn’t strong enough to explain it.

The snag isn’t hard to name. GLP-1s produce weight loss. The variable the 2013 analysis ranked first is already moving while the drug is on board, and a pooled analysis has no way to pull the two apart.

“GLP-1s raise testosterone” is a supported claim. “GLP-1s act on the testicles” is a different claim, and the people best placed to make it said the evidence wasn’t there.

No regulator has licensed semaglutide, tirzepatide, or liraglutide for hypogonadism or male infertility, so any use aimed at a testosterone number sits outside the approved indication.

Testosterone therapy, GLP-1s, and what happens to LH and FSH

GLP-1 therapy and testosterone therapy left different marks here, and the difference matters most to men who want children. Deameh and colleagues published a systematic review in The Journal of Sexual Medicine in 2026, covering ten studies and 639 men on GLP-1 receptor agonists — liraglutide, semaglutide, dulaglutide, and exenatide. Luteinizing hormone and follicle-stimulating hormone were preserved or increased with GLP-1 receptor agonist use, in contrast to the suppression observed in the testosterone therapy comparator groups.

For anyone already on testosterone therapy, that contrast belongs in a conversation with whoever wrote the prescription — it isn’t grounds for changing a regimen on the strength of a search result.

One head-to-head against testosterone replacement does exist. It ran on liraglutide — the molecule in Saxenda. Jensterle and colleagues randomized 30 men with obesity-associated functional hypogonadism into two arms and ran the trial open-label for 16 weeks. The liraglutide arm lost 7.9 kg on average; the transdermal testosterone arm lost 0.9 kg. Each figure is that arm’s own mean change from its own baseline, not an increment on the other. And only the liraglutide arm showed a significant increase in luteinizing hormone and follicle-stimulating hormone.

The bottom two rows come from that single trial — 30 men split across both arms, not 30 per arm. Keep the scale in view. Sixteen weeks, open-label, one molecule; liraglutide is neither semaglutide nor tirzepatide, and findings don’t hop between molecules just because they share a receptor.

Semen parameters improved in some men and not in others

The 2026 review also reported improvements in semen parameters in men it described as obese or hypogonadal — the review’s own wording. Then, in the next clause: no significant changes were found in healthy individuals.

Keep both halves. Strip the second one and you’ve manufactured a fertility drug out of a weight-management drug — the same shape as the testosterone story, just measured somewhere else.

For a couple actively trying to conceive, the practical reading stays narrow. Nobody has licensed a GLP-1 as a fertility treatment, and the benefit the review reported was in obese or hypogonadal men.

What STEP 1 and SURMOUNT-1 told the registry they would measure

Registry entries are public, boring, and unusually honest. Each one records what a trial said it would measure, before it ran.

STEP 1 — semaglutide, registered as NCT03548935 — lists two primary outcome measures: the percent change in body weight from baseline at week 68, and the number of participants who achieve 5% or more body weight reduction. Both are weight measures. Its registered secondary measures include changes in waist circumference, systolic blood pressure, and body composition for lean body mass.

SURMOUNT-1, the tirzepatide trial filed as NCT04184622, registered the percent change from baseline in body weight and the percentage of participants who achieve at least 5% body weight reduction in the primary treatment period at week 72. On the same registry page, its secondary outcome measures include change from baseline in waist circumference and in hemoglobin A1c.

For STEP 1, that registered primary endpoint returned a mean change in body weight from baseline to week 68 of -14.9% in the semaglutide group, compared with -2.4% with placebo. Each figure is that group’s total change from its own baseline, and both groups received the same lifestyle intervention. The placebo figure isn’t a floor to stack the other one on top of.

The labels also record who was in the room. Section 14 of the Wegovy label puts Study 1 at a baseline mean age of 46 years, with 74% of participants female. For SURMOUNT-1, the Zepbound label gives a baseline mean age of 45 years and 68% female. Those percentages describe who enrolled. Not what the drug did.

Registry record, journal report, label clinical-studies section: three views of one trial, not three separate studies. STEP 1 and SURMOUNT-1 are the separate ones — different trials, different molecules, different durations — and nothing above sets one against the other.

The fertility line on the label comes from rats

The fertility statement people quote from the US labels sits in section 13.1, the nonclinical section. The exact wording matters.

For Wegovy: in a combined fertility and embryo-fetal development study in rats, no effects were observed on male fertility. For Zepbound: in fertility and early embryonic development studies, no effects of tirzepatide were observed on sperm morphology, mating, fertility, and conception.

Both of those come from rat studies, which is why they sit where they do, and animal data doesn’t convert into a human safety conclusion in either direction — neither reassurance nor alarm. Reading a label well mostly comes down to knowing which shelf a statement came off.

Read the indications section just as literally. The US FDA label says semaglutide (Wegovy) is indicated in combination with a reduced calorie diet and increased physical activity. One of the indicated groups is adults with overweight in the presence of at least one weight-related comorbid condition. That’s weight management. Lifestyle change is written into the indication itself.

The hard stop applies to men exactly as it applies to anyone else: a personal or family history of medullary thyroid carcinoma, or Multiple Endocrine Neoplasia syndrome type 2, rules both drugs out. Acute pancreatitis sits one tier lower. A warning and precaution. The Wegovy label says that if acute pancreatitis is suspected, the drug should promptly be discontinued and appropriate management started.

Under both of those is the everyday layer, the one that comes up week to week: the common gastrointestinal reactions. They aren’t hard stops. But symptoms that won’t settle, or abdominal pain that keeps getting worse, mean it’s time to call the prescriber.

One more piece of label literacy. “Boxed warning” is US FDA language. The EMA, MHRA, Health Canada, and TGA run their own formats, approvals, and wording, so a US label doesn’t tell you the status where you live. Brands don’t map cleanly across borders either. In the US, semaglutide is Wegovy for obesity and Ozempic for type 2 diabetes, while tirzepatide — a GIP and GLP-1 dual agonist, not a pure GLP-1 — is Zepbound for obesity and Mounjaro for diabetes. In each case it’s one molecule under two names, and the two labels word their indications differently.

The men’s-health clinic version of this conversation

The US list price for Wegovy runs about $1,349 a month. What anyone actually pays moves with the plan, the prior authorization, the clinic, and the dose — so the real figure lives with a prescriber and a formulary, not in an article like this one.

The pitch holds together without anyone having to prove a mechanism: a number that reads low, a drug that moves weight, and one visit where both are on the table. One question rarely makes it into the room: what is the drug actually treating here? If the testosterone number moves because the weight moved, the answer is the weight.

Five questions worth bringing to the appointment

Name the question first. A testosterone number and a fertility plan pull in different directions.

• “Is my reading the functional kind that travels with weight, or is something else going on?”
• “If my weight comes down, what would you expect the hormone numbers to do, and when would you recheck them?”
• “We’re trying to conceive — does that change what you’d suggest, and where does testosterone therapy fit into that?”
• “Here’s the thyroid-cancer history in my family. Does any of that rule these drugs out for me?”
• “If the testosterone rises as the weight comes down, what happens to both numbers if I stop?”

Take the marketing off and a small, durable distinction survives.

Testosterone rose in the men who were studied. Where the rise came from is the open question — testicles or weight, nobody has shown which — and the researchers who ran the numbers wrote that limit into their own conclusion.

So if a GLP-1 turned up in your search history because of a testosterone panel, that’s where the published work stops. The rest — your history, your other numbers, whether any of this belongs in your case — was never what a pooled analysis set out to answer. That part needs someone with your chart open.

Start managing your GLP-1 with Blueshot

AI coaching, injection scheduling, and weight tracking in one app