Libido, erectile function and orgasm are among the most searched questions about these drugs and among the least studied. The weight-loss trials that produced the figures everyone quotes did not measure sexual function at all. What exists instead is a pharmacovigilance signal, a pair of database analyses that disagree with each other, and a handful of case reports.
What the weight-loss trials measured
STEP 1 randomized 1,961 adults with overweight or obesity to semaglutide 2.4 mg weekly or placebo for 68 weeks, and reported a mean weight change of −14.9% against −2.4%[1]. SURMOUNT-1 randomized 2,539 adults to tirzepatide or placebo for 72 weeks and reported −20.9% at 15 mg against −3.1% [2]. Neither trial listed a sexual function endpoint among its primary, secondary or prespecified exploratory outcomes. Those figures are the reason the drugs are prescribed, and they say nothing about this question. The same distinction runs through the withdrawal evidence, which measured weight and cardiometabolic variables and nothing else.
The reporting signal, and what a reporting signal is
A 2026 disproportionality analysis of the FDA Adverse Event Reporting System identified 49 semaglutide-associated reports of erectile dysfunction and calculated a reporting odds ratio of 1.53 (95% CI 1.16–2.03) [3]. Sitagliptin and dapagliflozin returned almost identical figures, 1.52 and 1.51, while dulaglutide, linagliptin and metformin returned no significant signal. The authors describe their own finding as preliminary and hypothesis-generating, and that description is the important part: a reporting odds ratio counts what was reported, not what happened. Nobody knows the denominator.
A time-to-onset analysis in the same paper found a random failure-type distribution rather than a cluster shortly after starting the drug. That is worth reading carefully. It means the reports do not line up with the moment a dose is introduced, which is what a drug-caused effect usually looks like.
The cohort data points the other way
A retrospective cohort study using the TriNetX network compared men aged 18 to 70 with type 2 diabetes and no prior erectile dysfunction, in three propensity-matched comparisons. Tirzepatide was associated with a lower risk of an erectile dysfunction diagnosis or a PDE-5 inhibitor prescription than sitagliptin (RR 0.70, 95% CI 0.64–0.76), than injectable semaglutide (RR 0.67, 95% CI 0.62–0.72) and than dulaglutide (RR 0.55, 95% CI 0.51–0.59)[4]. Its authors state that randomized trials are needed to confirm the finding.
Two honest readings sit side by side here. Weight loss and better glycemic control improve erectile function through mechanisms that are well described, so a drug that produces both might reasonably reduce risk. And the same drug might cause a distinct problem in a small number of people, which a population-level risk ratio would not show. A narrative review of the mechanistic literature published in 2025 reaches the same unresolved position[5].
Case reports are real, and they are not rates
A 2025 case report in Sexual Medicine describes anorgasmia beginning after a GLP-1 agonist was started [6]. One case is not a frequency. It is a documented event, which is more than an anecdote and much less than an incidence, and the correct response to it is to recognize the symptom rather than to estimate how common it is.
What was searched, and when
The searches behind this page were run against PubMed in September 2026, covering semaglutide and tirzepatide against libido, sexual function, erectile function and anorgasmia. No randomized controlled trial reporting a sexual function endpoint for either molecule was found. That is a statement about what a search returned on a date, not a promise that no such trial exists or will exist. This page records its method here because an absence claim is only checkable if the search behind it is stated — the same standard the editorial methodology applies to a price.
What to do with this
A reader who notices a change should treat it as reportable rather than as expected, and should raise it with whoever wrote the prescription. Which seller that is matters here: a service that bills a consultation separately, or that never says how often a patient is reviewed, is a harder place to raise a symptom than one where the visit sits inside the price. The per-visit fee schedules and the seller reviews record what each one publishes about that.