The Desire Paradox: Why GLP-1 Drugs Restore Your Testosterone But May Not Restore Your Want

If you are an obese man with low testosterone and your doctor prescribes a GLP-1 receptor agonist, three independent systematic reviews say your testosterone will rise. Your LH will stabilize or increase. Your metabolic trap will begin to break. On paper, the hormonal story ends here — restoration achieved.

But testosterone is not desire. It is a necessary condition, not a sufficient one. And a growing body of evidence — from animal models, pharmacovigilance data, a single extraordinary case report, and a theoretical framework published by one of sexual medicine's most prominent researchers — suggests that GLP-1 drugs may suppress desire through the same serotonergic pathway that makes SSRIs notorious for sexual side effects. The hormone goes up. The drive may not follow. And the reason no one has measured this properly is that the suppression is invisible — camouflaged by the very benefits these drugs produce.

The Restoration Evidence Is Now Settled

The hormonal side of this story has been confirmed three times over:

Study	Design	Key T Finding
Salvio et al., 2025	Meta-analysis, 7 studies, n=680	TT ↑1.39 ng/mL (p<0.0001)
Orra et al., BMC Urol 2025	Meta-analysis, 4 studies, n=219	Bioavailable T ↑ (p<0.001)
Deameh et al., J Sex Med 2026	Systematic review, 10 studies, n=639	Consistent ↑ in obese/T2DM/FHH; no effect in healthy men

The Deameh review — the most comprehensive to date — adds an important nuance: LH and FSH remain stable or increase with GLP-1 RA treatment, in direct contrast to testosterone replacement therapy, which suppresses both. This makes GLP-1 RAs fertility-sparing. The EAU 2025 guidelines now mention GLP-1 RAs for functional hypogonadism related to obesity — the first major urology guideline to do so.

The hormone is restored. The question no one is asking loudly enough: does desire follow?

The Suppression Evidence Is Building

In April 2026, Shevchouk et al. published the first peer-reviewed animal study directly demonstrating that GLP-1 receptor agonists suppress sexual behavior. In female rats, exendin-4 reduced mounting and lordosis intensity. In female mice, liraglutide and dulaglutide diminished time spent with males and altered hypothalamic neurotransmitter levels. The effects were drug-specific and species-specific — but the direction was consistent. GLP-1R agonism suppresses sexual behavior independent of metabolic effects.

This was the mechanistic piece that had been missing. Before Shevchouk, the suppression hypothesis rested on pharmacological reasoning and pharmacovigilance signals. After Shevchouk, it rests on observed behavior in controlled conditions.

But an animal study alone doesn't explain why clinicians aren't seeing the signal. For that, you need the camouflage model.

The Camouflage Effect

In November 2025, Gelfand, Tveit, and Simon published a clinical review in Obesity Pillars that names the problem. James A. Simon — one of sexual medicine's most published researchers — is a co-author. Their thesis: GLP-1 agonists decrease sexual desire via serotonergic mechanisms, but competing positive effects mask the decrease in clinical settings.

The mechanism they propose: 50-80% of nucleus tractus solitarius (NTS) preproglucagon (PPG) neurons — the neurons that mediate central GLP-1 effects — are innervated by serotonergic projections. GLP-1 receptor agonism increases serotonergic activity at the 5-HT2C receptor. This is the same receptor subtype responsible for SSRI-induced sexual dysfunction — a connection I traced in article #14 of this series.

The camouflage works because the suppression and restoration forces operate on the same patient simultaneously. An obese man on semaglutide is losing weight (improving body image), restoring testosterone (improving hormonal drive), enhancing vascular function (improving erectile capacity), and reducing inflammation (improving mood) — all while the same drug's serotonergic activation is quietly dampening the central desire circuit. The net clinical effect? The patient reports feeling "about the same" sexually. The clinician documents no sexual complaint. The signal vanishes into the noise of competing improvements.

The Tirzepatide Exception

One case report cracks the camouflage open. Visvabharathy et al. (Sexual Medicine, July 2025) documented a 71-year-old woman who developed complete anorgasmia within two weeks of starting liraglutide. Switching to semaglutide — a different GLP-1 RA — changed nothing. Switching to tirzepatide — a GLP-1/GIP dual agonist — restored orgasms within two weeks.

This drug-class differentiation matters. Tirzepatide is not just a GLP-1 RA. Its GIP receptor co-agonism may protect against the serotonergic desire suppression that pure GLP-1 agonists produce. This is consistent with FAERS pharmacovigilance data showing tirzepatide accounts for only 14.3% of erectile dysfunction reports — a lower proportion than liraglutide, dulaglutide, or semaglutide. And it aligns with my earlier analysis of why tirzepatide outperforms pure GLP-1 agonists on hormonal restoration: Cannarella's pilot showed 100% testosterone normalization with tirzepatide versus partial restoration with semaglutide alone.

One case report is not proof. But the Visvabharathy result generates a specific, testable hypothesis: GIP receptor co-agonism protects against GLP-1-mediated serotonergic desire suppression. This would explain why tirzepatide consistently outperforms pure GLP-1 RAs across multiple outcomes — not just desire, but testosterone normalization, lean mass preservation, and metabolic improvement. The GIP pathway may be a pharmacological shield against the 5-HT2C cost of GLP-1R activation.

The Serotonergic Triangle

This is not an isolated finding. It connects three threads in this series through a single receptor:

The convergence is pharmacologically precise. Testosterone powers the dopaminergic motivation circuit — the mesolimbic pathway from VTA to nucleus accumbens that drives wanting. Serotonin, acting through 5-HT2C, opposes this circuit. SSRIs flood serotonin directly. GLP-1 RAs reach it indirectly, through NTS PPG neuron projections. The endpoint is the same: a shift in the DA/5-HT balance that suppresses wanting while leaving hormonal levels intact or improved.

This matters clinically because it means an obese hypogonadal man who starts a GLP-1 RA and an SSRI simultaneously — a common scenario given the depression-obesity-hypogonadism overlap — faces converging serotonergic suppression of desire from two independent pathways. No study has examined this combination.

What Needs to Exist

The specific void is narrow and well-defined: a prospective, controlled trial enrolling obese men with documented functional hypogonadism, randomized to GLP-1 RA versus placebo (or versus TRT), measuring — simultaneously and longitudinally — total testosterone, free testosterone, LH, FSH, and a validated desire-specific instrument (not just the IIEF, which conflates desire with erectile function). Ideally stratified by GLP-1 RA type (semaglutide vs tirzepatide) to test the GIP-protection hypothesis.

This trial does not exist. No registry entry describes it. The Lengsfeld dulaglutide RCT in healthy lean men measured sexual function but enrolled the wrong population — the hormonal restoration effect occurs specifically in obese men with metabolic dysfunction, not in healthy eugonadal subjects. Measuring desire in men whose testosterone doesn't change tells you nothing about whether desire follows restoration.

The camouflage effect explains the absence. If clinicians don't see a clinical signal — because the competing improvements hide it — there is no clinical impetus to fund a trial measuring it. The void perpetuates itself. Meanwhile, tens of millions of patients are prescribed GLP-1 RAs, and the subset who are obese hypogonadal men receive a drug that restores their testosterone while potentially suppressing the very desire that testosterone is supposed to drive. The suppression may be partial. It may be temporary. It may be offset by benefits. We do not know, because no one has looked.