What the evidence really says about GLP 1 sexual function in men

Reports that weight loss drugs change libido have spread faster than the research behind them, and the published work on GLP 1 sexual function is thinner than the discussion suggests. A magazine feature this week gathered the available studies and the expert caveats. Looking at the same papers, the picture is of small samples, opposite directions of effect and no trial built to answer the question. Here is what each study actually measured and in whom.

Where the survey numbers come from

The figures driving the conversation come from a survey, not a trial. In work run by the Kinsey Institute with a dating website in 2025, more than half of people taking these drugs said the medication had affected their sex lives. Within that group the direction split almost evenly. About 18 percent said desire went up and 16 percent said it went down, while 16 percent reported better sexual function and 12 percent reported worse. Men were roughly twice as likely as women to report any shift, in either direction. Those are self reports from people who knew they were taking a drug and knew what they had read about it.

Surveys of this kind are useful for showing that something is worth studying and useless for showing what causes it. There is no comparison group, no blinding and no measurement beyond what respondents chose to say. People who notice a change are also more likely to answer a survey about change. The balanced split is itself informative, though, because a strong pharmacological effect in one direction would be expected to show up as an imbalance. What the survey does establish is that a noticeable minority of users perceive a difference, which is reason enough to look at the clinical literature rather than dismiss the question.

What the clinical studies on GLP 1 sexual function found

The largest human dataset comes from an exploratory analysis of REWIND, a randomised cardiovascular trial of dulaglutide. Among 3,725 men with type 2 diabetes, new erectile dysfunction occurred at 21.3 cases per 100 person years on the drug against 22.0 on placebo, a hazard ratio of 0.92. That is a small reduction in risk, not an increase. Pulling the other way, a matched records study of 3,094 men prescribed semaglutide and 3,094 controls, all under 50 with obesity and without diabetes, found new erectile dysfunction diagnoses or prescriptions in 1.47 percent against 0.32 percent. The relative risk looks dramatic at 4.5, but the absolute difference was about one case in a hundred men.

The relative risk looks dramatic at 4.5, but the absolute difference was about one case in a hundred men.

Records studies like that one count diagnoses, not symptoms. Men starting a weight loss drug see a doctor more often, and a consultation is where a diagnosis gets written down, so detection alone could produce part of the gap. The only randomised trial that measured desire directly was small and short. Twenty six healthy men of normal weight were given dulaglutide or placebo for four weeks each in a crossover design, and 24 completed it. There was no significant change in sexual desire scores and no significant difference in testosterone, follicle stimulating hormone, luteinising hormone or sperm measures. Four weeks in healthy volunteers tells us little about months of treatment in men who are losing weight.

Reading the evidence side by side

Before comparing the numbers it helps to sort the studies by what kind of evidence each one is. A randomised trial can support a causal claim about the outcome it was built to measure. An observational study of medical records can show associations and little more, because the people taking a drug differ from those who are not in ways no adjustment fully removes. A spontaneous reporting system records what doctors and patients choose to report, which reflects attention as much as biology. The table below keeps those categories visible, because the disagreement between these results is partly a disagreement between methods rather than between findings.

Set out together, the studies are not really measuring the same thing. One counts diagnoses in records, one counts events in a trial of a different drug, one measures a questionnaire score in healthy volunteers, and one scans spontaneous reports to a regulator, where 182 cases produced a reporting odds ratio of 0.41, meaning these drugs were reported less often than other medicines rather than more. A 2004 randomised trial in 110 obese men found that losing weight through diet and exercise improved erectile function scores, and a 2025 review of 28 studies found sexual desire improved after weight loss by diet or surgery. Weight loss itself therefore moves the outcome.

Why the question is hard to settle

That last point is the core difficulty. Everything the drugs do arrives together: appetite falls, weight falls, blood sugar improves, nausea and fatigue are common early, mood may shift and sleep often changes. Any of these can move sexual function on its own, and obesity and diabetes cause sexual problems before treatment starts. Separating a direct drug effect from all of that needs prospective studies designed for the purpose, with validated questionnaires at fixed time points and a comparison group losing weight another way. Clinicians quoted in the feature make the same point, and one notes that the drugs in this class differ from one another, so a result for one should not be assumed for the next.

Animal work hints at a possible biological route but cannot carry much weight here. Infusing a GLP 1 receptor agonist into specific reward related regions of the mouse brain reduced sexual interaction behaviours, which shows the receptors sit in circuits that matter rather than showing what happens in people at prescribed amounts. For now the honest summary of GLP 1 sexual function is that no reliable estimate exists in either direction. Anyone noticing a change should raise it with their prescriber, since low testosterone, depression and vascular disease all have their own management. Our map of 175 health outcomes on these drugs shows how often such questions remain open. The same caution applies to what happens when treatment stops.

Conclusion

The research on GLP 1 sexual function does not support a confident claim in either direction. The biggest trial analysis points to slightly fewer cases of erectile dysfunction, a records study points to more diagnoses in a small absolute number of men, and the only randomised test of desire lasted four weeks in 24 healthy volunteers. Weight loss by any route improves sexual function in men with obesity, which is the one finding the evidence does support.

Source Evidence reviewed from PubMed records including PMID 38778151 (semaglutide records study), PMID 34153269 (REWIND analysis), PMID 39232425 (dulaglutide crossover trial), PMID 40240532 (adverse event reports) and PMID 15213209 (lifestyle trial in obese men)

Amino Weekly reports research. It is not medical advice and never gives doses.