Does Ozempic Cause Erectile Dysfunction? What the Studies Behind the Headlines Show

Search for whether Ozempic causes erectile dysfunction and page one will contradict itself.

One result reports an increased risk of erectile dysfunction and low testosterone. Another says there is no good evidence for it. A third says some men report their erectile function improving.

All three are describing the same small body of research. The reason they disagree is that almost none of them say what the underlying studies actually measured — and the studies do not agree with each other either.

Here is what each of them found.

The Short Answer

No randomised clinical trial has shown that semaglutide causes erectile dysfunction.

Two observational studies have found more recorded erectile dysfunction among men taking GLP-1 medications. The larger relative risk comes from a database study with no active comparator; a 2026 study that added one found a much smaller association that disappeared when the researchers applied a statistical check for hidden bias.

Pointing the other way: a genetic causal-inference study found reduced erectile dysfunction risk, an FDA adverse-event analysis found no disproportionate reporting signal, and hormone research consistently finds testosterone rising during GLP-1 treatment.

Why testosterone rises is itself contested. One 2026 meta-analysis argues the effect is independent of weight loss; a 2025 meta-analysis and a separate 2026 review attribute it to weight loss and improved insulin sensitivity.

The honest summary is that this question is open.

First: Which Drug Is Actually Being Studied?

This distinction is skipped almost everywhere, and it matters for reading any of the research below.

Ozempic and Wegovy contain the same active ingredient — semaglutide — but they are different products with different FDA-approved indications and different dosing. Ozempic is approved for type 2 diabetes and certain cardiovascular and kidney indications. Wegovy is approved for chronic weight management.

The study that generated the erectile dysfunction headlines examined semaglutide prescribed for weight loss in non-diabetic men — not a brand. A separate 2026 study examined GLP-1 receptor agonists as a class in men with type 2 diabetes, which is a different population again.

So a headline naming one brand is already a step removed from what was measured.

The Study Behind the Headlines

Nearly every article claiming Ozempic causes erectile dysfunction rests on one paper: Able and colleagues, published in the International Journal of Impotence Research.

Its title is more careful than the coverage of it: “Prescribing semaglutide for weight loss in non-diabetic, obese patients is associated with an increased risk of erectile dysfunction.”

Associated with. That is the authors' own wording. The causal claim was added later, by other people.

What it did

The researchers searched TriNetX, a federated network of electronic health records, for non-diabetic men aged 18 to 50 with a BMI over 30 who were prescribed semaglutide after 1 June 2021. They excluded men with any prior erectile dysfunction diagnosis, any previous PDE5 inhibitor prescription, prior testosterone deficiency or therapy, and several surgical and cardiovascular histories.

That left 3,094 men, matched one-to-one against 3,094 obese non-diabetic men who had never received semaglutide. After matching, both groups had a mean age of about 38 and a mean BMI of about 39.

What it found

• Erectile dysfunction diagnosis or a new PDE5 inhibitor prescription: 1.47% in the semaglutide group versus 0.32% in the comparison group. Relative risk 4.5 (95% CI 2.3–9.0).
• Testosterone deficiency diagnosis: 1.53% versus 0.80%. Relative risk 1.9 (95% CI 1.2–3.1).

Read Those Numbers Again, in Absolute Terms

“4.5 times the risk” and “about one extra recorded case per 87 men” describe the same result. Only one of them gets published.

The absolute difference in erectile dysfunction was 1.15 percentage points. For testosterone deficiency it was 0.73 points. Applied to the study's own group sizes, those percentages correspond to roughly 45 men versus 10 — a small number of events, which is why the confidence interval on the headline figure runs from 2.3 all the way to 9.0.

A fourfold range is not a precise estimate. It is a signal that the underlying event count is small.

None of this makes the finding wrong. It makes it a much smaller finding than “Ozempic causes erectile dysfunction” conveys.

What a Database Study Can and Cannot Show

TriNetX is a network of health records. A study built on it is not a trial: nobody was randomised, nobody was blinded, and the researchers observed what clinicians happened to record.

The outcome is a diagnostic code

The study measured erectile dysfunction that was documented in a medical record or treated with a prescription — not erectile dysfunction assessed with a standardised instrument such as the IIEF questionnaire.

The comparison group's rate was 0.32%, roughly three men in a thousand. Population estimates of erectile dysfunction among obese men aged 18 to 50 are considerably higher than that. So the gap between what was counted and what men actually experience is likely to be large. (That is our reading of the numbers, not a statement the authors make. The paper's full text is behind a paywall, so its own limitations section could not be read for this article.)

Could detection bias explain part of the difference?

A man prescribed semaglutide is, by definition, in active medical care — attending appointments, having his history taken, being asked questions. A matched comparison patient who never received it may have had far less clinical contact.

More appointments produce more recorded diagnoses of many conditions. Matching two groups on age and BMI does not equalise how often they see a clinician.

This is methodological interpretation, not something the study demonstrates — it does not report encounter frequency. It is worth stating because a later study tested for exactly this kind of hidden bias, and the result is described below.

No active comparator

The study compared semaglutide users against people taking nothing. There was no arm on a different weight-loss medication, and no structured lifestyle-intervention arm.

That means it cannot separate an effect of the drug from an effect of rapid weight loss by any means, from the metabolic severity that prompted the prescription, or from the treatment-seeking behaviour of men who obtain one. The decision to prescribe semaglutide is not random.

The 2026 Study That Added an Active Comparator

In April 2026, researchers at the University of Pennsylvania published exactly the study the Able paper was missing.

Using a method called target trial emulation — designing an observational analysis to mimic a randomised trial as closely as the data allow — they compared men with type 2 diabetes starting a GLP-1 receptor agonist against men starting a DPP-4 inhibitor, a different diabetes medication. That is an active comparator: both groups are being treated, both are under care, both have the condition.

After weighting, the analysis covered 4,910 GLP-1 initiators and 5,524 DPP-4 inhibitor initiators, drawn from US electronic health records between 2019 and 2024, with a separate external validation cohort.

The result went in the same direction as Able, not the opposite one. Incident erectile dysfunction was higher among GLP-1 users: 35.2 versus 28.0 per 1,000 person-years, hazard ratio 1.26 (95% CI 1.08–1.46). In absolute terms that is about 7 additional recorded cases per 1,000 men per year — roughly one per 139.

Then comes the part that matters most.

The researchers applied negative control outcome calibration. The principle is simple: check the data for outcomes the medication has no plausible reason to affect. If those show associations too, the dataset carries hidden bias, and the main result is recalibrated to account for it.

After that calibration, the association was attenuated and no longer statistically significant.

The authors' own conclusion: GLP-1 use was "modestly associated with an increased rate of ED," and "these observational findings may reflect residual or selection bias and do not establish causation."

So the best-designed observational study on this question found a signal, formally tested whether that signal might be an artefact of who gets prescribed what and who sees a doctor more often, and could not rule it out.

What Causal-Inference Research Suggests

Observational data can only show association. One method is designed to strengthen causal inference.

Mendelian randomization uses naturally occurring genetic variants as stand-ins for an exposure. Because those variants are allocated at conception, they are largely free of the confounding that troubles observational research.

A 2024 drug-target Mendelian randomization study in Frontiers in Endocrinology examined genetically-proxied GLP-1 receptor agonist action against erectile dysfunction risk. It found reduced risk: odds ratio 0.493 (95% CI 0.430–0.565), with a mediation analysis attributing part of the effect to obesity (6.83%), hypertension (3.22%), cardiovascular disease (3.06%) and type 2 diabetes (2.89%).

The caveat is substantial. Mendelian randomization models lifelong genetic modulation of a drug target — not the effect of taking semaglutide for twelve months. The two are related but not interchangeable, the mediated proportions account for only about a sixth of the effect, and the authors call for randomised trials. This is not proof that GLP-1 medications protect erectile function.

What it is: the one study in this field designed for causal inference, pointing the opposite way to the two observational ones.

The Paper That Gets Cited Backwards

A 2025 analysis of the FDA Adverse Event Reporting System is frequently cited online as supporting a link between GLP-1 medications and sexual dysfunction.

It found no such signal.

Across twenty years of reports, the researchers identified 182 cases of male sexual dysfunction associated with GLP-1 receptor agonists, and calculated a reporting odds ratio of 0.41 (95% CI 0.36–0.48). A value below 1 indicates that these reports were not disproportionately elevated relative to the reporting background for other drugs in the database. The authors describe a "weak association" and state that "overall patient risk remains low."

What this does and does not mean. FAERS is a spontaneous-reporting system built for signal detection. It cannot estimate how often something actually happens, and reporting behaviour varies enormously between drugs and over time. The absence of a disproportionality signal is not evidence of safety — but it is certainly not evidence of harm, which is how the paper is usually cited.

Testosterone Goes Up. Why Is Contested.

If GLP-1 medications damaged male hormonal function, the hormone studies would show it. They consistently show the reverse.

A 2025 systematic review and meta-analysis in Andrology pooled seven studies covering 680 men and reported a standardised mean difference of 1.39 (95% CI 0.70–2.09) for total testosterone. Weight, BMI, waist circumference and HbA1c all fell.

A 2026 systematic review in the Journal of Sexual Medicine, covering ten studies and 639 men, found the same direction, and added a distinction that is easy to miss: total testosterone rose while LH and FSH were preserved or increased. That pattern differs from exogenous testosterone therapy, which commonly suppresses those gonadotropins through feedback on the hypothalamic-pituitary-gonadal axis. Free testosterone was less consistent in both reviews, often offset by a simultaneous rise in sex hormone-binding globulin.

The disagreement is about the mechanism.

The 2025 Andrology meta-analysis found that the size of the testosterone increase correlated with how much weight was lost, and stated that the available literature "does not allow to demonstrate a direct action of GLP-1RAs on the testicular function." A separate 2026 systematic review in Pharmaceuticals, covering 29 studies, concluded that the effect "is largely mediated by weight loss and improvement of insulin resistance rather than direct androgenic action."

But a 2026 meta-analysis in Andrology from a different research group reported the opposite inference: that the testosterone findings were "independent from the observed weight loss, suggesting a direct, rather than an indirect, mechanism of action on the hypothalamus-pituitary-testis axis."

Three syntheses, published within about a year of each other, two attributing the effect to weight and one to the drug. That is an unresolved question, not a settled one.

Where That Leaves the Weight Explanation

Weight loss is very likely an important contributor. Obesity is associated with functional suppression of the male reproductive axis, and testosterone rises with weight loss regardless of how the weight comes off: a 2013 meta-analysis of 24 studies found increases of roughly 2.87 nmol/L after a low-calorie diet and 8.73 nmol/L after bariatric surgery, with the degree of weight loss the single best predictor of the rise.

What current evidence does not establish is whether the hormonal effect of GLP-1 medications is entirely indirect, or whether GLP-1 receptor signalling also acts on the male reproductive axis independently. Those pathways cannot be separated confidently from the studies published so far. Related reading: Do GLP-1 Medications Affect Your Sex Drive? and Low Libido in Men: Causes, Testosterone and Treatment Options.

What Nobody Can Tell You From This Research

The literature describes populations. It cannot say why one man's erections changed after starting a medication.

Erectile difficulty has causes that have nothing to do with weight-loss drugs — cardiovascular disease, diabetes, medication effects, hormonal conditions, sleep disorders and depression among them — and several are conditions where erectile dysfunction can be an early warning sign. That is the reason a persistent change is worth raising with a provider rather than attributed to whatever started most recently. Related reading: Why Viagra or Cialis Sometimes Doesn't Work.

Decisions about starting, continuing or stopping a GLP-1 medication belong to the clinician who prescribed it. Nothing in this article is a reason to change how a prescribed medication is taken.

Frequently Asked Questions

Does Ozempic cause erectile dysfunction?

No randomised trial has shown that it does. Two observational studies have found more recorded erectile dysfunction among men taking GLP-1 medications: a database study of semaglutide for weight loss (1.47% versus 0.32%, relative risk 4.5, an absolute difference of about one case per 87 men) and a 2026 active-comparator study in men with type 2 diabetes (hazard ratio 1.26). In the second, the association was no longer statistically significant after the researchers corrected for hidden bias, and its authors state the findings "do not establish causation."

Does Wegovy cause erectile dysfunction?

Wegovy and Ozempic both contain semaglutide, but the database study behind the headlines examined semaglutide prescribed for weight loss in non-diabetic men rather than either brand specifically. It found an association with more documented erectile dysfunction and PDE5 inhibitor prescriptions. It did not establish causation, and no trial has tested either brand for this outcome.

How many men in that study actually developed ED?

The percentages correspond to roughly 45 men out of 3,094 taking semaglutide, against roughly 10 out of 3,094 in the comparison group. Small event counts are why the confidence interval on the relative risk spans 2.3 to 9.0 — a fourfold range.

Why is the control group's ED rate so low?

The study counted erectile dysfunction that was recorded in a medical record or treated with a prescription, not erectile dysfunction measured with a standardised questionnaire. A 0.32% rate among obese men aged 18 to 50 is well below population estimates, which suggests a large gap between what was counted and what men experience.

Does semaglutide lower testosterone?

The hormone research points the other way. A 2025 meta-analysis of seven studies (680 men) and a 2026 systematic review of ten studies (639 men) both found total testosterone increased during GLP-1 treatment, with LH and FSH preserved or increased. The database study did report more testosterone deficiency diagnoses, at 1.53% versus 0.80%.

Is the testosterone effect from the drug or from losing weight?

This is genuinely disputed. A 2025 meta-analysis found the testosterone rise correlated with weight lost and could not demonstrate direct testicular action; a 2026 review of 29 studies attributed the effect to weight loss and insulin sensitivity. A separate 2026 meta-analysis reported the effect was independent of weight loss and inferred a direct mechanism. Current evidence cannot separate those pathways confidently.

Do GLP-1 medications improve erectile function?

That has not been demonstrated by a randomised trial. A Mendelian randomization study found reduced erectile dysfunction risk (odds ratio 0.493), but it models lifelong genetic effects rather than a course of treatment. Meanwhile the two observational studies found the association running the other way. The evidence is mixed.

Should I stop taking my GLP-1 if I notice a change?

That is a question for the clinician who prescribed it, and it is not one this article can answer. Sexual function has many causes, and a change that coincides with starting a medication is not proof the medication caused it.

AmalfiDuo Sexual Wellness Programs

AmalfiDuo does not dispense GLP-1 medications.

If you're exploring ways to support your sexual wellness, AmalfiDuo offers private, provider-guided care programs for adults. A licensed clinician reviews your health information to determine whether treatment is appropriate.

Explore Sexual Wellness Programs

A prescription is never guaranteed.

References

1. Able C, Liao B, Saffati G, et al. Prescribing semaglutide for weight loss in non-diabetic, obese patients is associated with an increased risk of erectile dysfunction: a TriNetX database study. International Journal of Impotence Research. 2025;37(4):315–319. PMID 38778151.
2. Tang H, Lu Y, Zhang B, Zhang D, Asch DA, Chen Y. GLP-1 receptor agonist and risk of erectile dysfunction in men with type 2 diabetes: a target trial emulation. EClinicalMedicine. 2026;94:103857. PMID 42005929.
3. An H, Xie K, Gan H. GLP-1 receptor agonists and the risk of erectile dysfunction: a drug target Mendelian randomization study. Frontiers in Endocrinology. 2024;15:1448394. PMID 39605940.
4. Pourabhari Langroudi A, Chen AL, Basran S, et al. Male sexual dysfunction associated with GLP-1 receptor agonists: a cross-sectional analysis of FAERS data. International Journal of Impotence Research. 2025;37(8):661–667. PMID 40240532.
5. Salvio G, Ciarloni A, Ambo N, et al. Effects of glucagon-like peptide 1 receptor agonists on testicular dysfunction: A systematic review and meta-analysis. Andrology. 2025;13(8):2022–2034. PMID 40105090.
6. Deameh MG, Ramez M, Rowaiee R, et al. Effects of GLP-1 receptor agonists on male reproductive hormones, semen parameters, and metabolic outcomes: a systematic review. Journal of Sexual Medicine. 2026;23(2):qdaf381. PMID 41498523.
7. Corona G, Sparano C, Romeo M, et al. Emerging Effects of Glucagon-Like Peptide-1 Receptor Agonists and Sodium-Glucose Cotransporter 2 Inhibitors on Male Sexual Hormones and Behaviors: Systematic Review and Meta-Analysis. Andrology. 2026. PMID 42011503.
8. La Vignera S, Condorelli RA. Effects of Incretin-Based Therapies on Testosterone Levels and Incretin Response in Men with Hypogonadism: A Systematic Literature Review. Pharmaceuticals. 2026;19(8):1321. PMID 42653816.
9. Corona G, Rastrelli G, Monami M, et al. Body weight loss reverts obesity-associated hypogonadotropic hypogonadism: a systematic review and meta-analysis. European Journal of Endocrinology. 2013;168(6):829–843. PMID 23482592.

This article is provided for general educational purposes only and is not a substitute for individualized medical advice, diagnosis or treatment. It reports published research and does not recommend starting, stopping, continuing or changing any medication. The studies described are retrospective, observational or genetic analyses and cannot establish cause in an individual case. Questions about a prescribed medication should be discussed with the clinician who prescribed it. AmalfiDuo Journal articles are written by AmalfiDuo Editorial and are not medically reviewed — see our Editorial Policy.

Ozempic® and Wegovy® are trademarks of Novo Nordisk A/S. AmalfiDuo is not affiliated with or endorsed by the owners or manufacturers of these brands.