At a university hospital in Basel, Switzerland, two dozen healthy young men spent two four-week periods answering questions about their sex lives and receiving weekly injections. During one period, they received the GLP-1 drug dulaglutide. During the other, they received placebo. The original question was whether a medicine known for reducing appetite might also reduce sexual desire.
It did not produce a statistically significant change in desire. But a secondary analysis published November 5 in the Journal of Neuroendocrinology found something unexpected: prolactin levels fell during dulaglutide treatment and rose during placebo treatment. The adjusted comparison narrowly met the threshold for statistical significance.
Prolactin is a pituitary hormone best known for its role in breastfeeding, but it also participates in reproductive and neuroendocrine regulation. The researchers suggest their finding could point toward effects of GLP-1 receptor activation beyond appetite and blood sugar. That is possible. The evidence, however, comes from just 24 men, one drug and four weeks of exposure.
The result is a scientific clue, not a new clinical use for GLP-1 medicines.
Why Researchers Measured Prolactin Levels in the First Place
GLP-1 receptor agonists are established treatments for type 2 diabetes, and some medicines in the class are also approved for obesity. Their effects extend beyond the pancreas and gut. Researchers have investigated how GLP-1 signaling interacts with brain circuits associated with appetite, motivation and reward, including possible effects on alcohol consumption and other reward-driven behavior.
That interest raised an uncomfortable question. If a drug can change how strongly someone wants food or alcohol, could it also change sexual desire?
The DESIRE trial, registered as NCT04687514, was designed to investigate that possibility. Its primary findings, published in eBioMedicine in September 2024, found no statistically significant negative effect of dulaglutide on sexual desire, measured with a standardized questionnaire. Testosterone, other reproductive hormones and semen parameters also showed no significant treatment-related differences.
Prolactin was a logical hormone to investigate separately. Dopamine normally restrains prolactin secretion from the pituitary gland. Persistently elevated prolactin, known as hyperprolactinemia, can suppress reproductive hormone signaling and contribute to reduced libido and sexual dysfunction.
The Basel researchers wanted to know whether GLP-1 receptor activation might influence this system. Prolactin was a prespecified secondary measurement in the original trial, and its results are the focus of the new paper.
The original question was about desire. The unexpected finding concerned a hormone.
Inside the Four-Week Dulaglutide Experiment
DESIRE was a single-center, randomized, double-blind, placebo-controlled crossover study at University Hospital Basel. Twenty-six men were randomized, and 24 completed the study. They were healthy, had normal reproductive hormone function and reported satisfactory sexual lives.
The participants had a median age of 24.5 years and a median body mass index of 23.9. They were not representative of the older or metabolically complicated populations commonly prescribed GLP-1 drugs.
Each man received both treatments in a randomized order, allowing researchers to compare responses within the same individual. The dulaglutide phase began with 0.75 mg in the first week, followed by 1.5 mg weekly for the next three weeks. During the placebo phase, participants received saline injections. A four-week washout separated the treatment periods.
Researchers collected morning blood samples before and after each treatment phase under standardized fasting conditions. Prolactin was measured alongside body weight, appetite and sexual-function assessments.
The crossover design was a strength because each participant served as his own comparison. But the small sample remained a major limitation. The study’s sample size was originally determined around sexual desire, not the ability to establish a precise prolactin effect.
Prolactin Fell. Sexual Desire Did Not Significantly Change.
The observed hormone changes moved in opposite directions.
During dulaglutide treatment, prolactin fell by an average of 31.1 milli-international units per liter (mU/L). During placebo treatment, it rose by an average of 44.6 mU/L. The simple difference between those mean changes was approximately 75.7 mU/L in favor of dulaglutide.
After adjusting for baseline prolactin, weight change and an interaction between treatment and baseline prolactin, the researchers estimated a treatment difference of minus 149.3 mU/L. The 95% confidence interval extended from minus 298.5 to minus 0.1 mU/L.
The men also lost an average of 2.8 kilograms during dulaglutide treatment, compared with a 0.4-kilogram gain during placebo. The association between prolactin reductions and weight loss was weak, with a correlation coefficient of 0.25.
Thirteen of the 24 participants had reductions in prolactin alongside reduced appetite during dulaglutide treatment. That observation was exploratory and did not establish that either change caused the other.
What did not change significantly was sexual desire. In the original analysis, the difference in the sexual-function questionnaire between dulaglutide and placebo was not statistically significant (p = 0.402). The new analysis also found no meaningful association between prolactin levels and reported sexual desire or sexual activity.
The hormone measurement changed. The reported sexual experience did not show a significant treatment difference.
The Statistical Detail That Makes This a Preliminary Finding
There is an important distinction between the observed hormone changes and the larger adjusted estimate presented in the November paper.
The raw difference between the two average changes was approximately minus 75.7 mU/L. In a presentation of the same study at the European Congress of Endocrinology in 2024, the researchers reported a paired estimate of minus 75.68 mU/L, with a 95% confidence interval stretching from minus 158.68 to plus 7.33 mU/L.
That earlier interval included zero, meaning the unadjusted analysis did not establish a statistically significant treatment difference at the conventional 95% confidence level.
The November 2026 publication used a mixed statistical model that adjusted for baseline prolactin concentrations, weight changes and their specified relationships with treatment. Its estimated difference was larger, and its confidence interval narrowly excluded zero.
These analyses do not represent two separate experiments or contradictory blood measurements. They are different ways of estimating the treatment effect from the same small dataset. But the distinction matters because the strength of the statistical conclusion depends on the analytical approach.
The published authors acknowledge several limitations. Prolactin fluctuates naturally, and the study relied on one morning measurement at each relevant visit rather than repeated sampling. The unexpected rise during placebo treatment may also have inflated the apparent treatment difference.
The study was not powered specifically for prolactin, and its participants were young men with normal baseline hormonal function. The results cannot establish what would happen after months of treatment or in people with elevated prolactin, diabetes or obesity.
A confidence interval ending at minus 0.1 is a reason to investigate further, not a reason to declare the question settled.
Could GLP-1 Drugs Influence the Dopamine-Prolactin System?
The proposed biological explanation is interesting because it reaches beyond the usual metabolic story of GLP-1 medicines.
Prolactin secretion is regulated by dopamine, and GLP-1 receptors are present in brain and pituitary regions involved in neuroendocrine signaling. That creates a plausible route through which GLP-1 receptor activation might influence prolactin levels, directly or indirectly.
The authors also note earlier experimental findings linking prolactin signaling to feeding behavior. They raise the possibility that changes in prolactin could participate in the broader appetite-related effects of GLP-1 medicines.
But this study did not measure dopamine activity, manipulate prolactin independently or establish that prolactin changes contributed to weight loss. The weak correlation with weight and the exploratory appetite observations are insufficient to show a causal pathway.
There are several alternative explanations. The hormone change might be downstream of reduced food intake or weight loss. It might reflect a more direct neuroendocrine response. Or part of the observed difference could be explained by normal biological variation in a small sample.
The wider history of peptide research is full of examples in which measuring a hormone changed how scientists understood an established pathway. Our profile of Roger Unger describes how glucagon measurement helped challenge the idea that diabetes was solely an insulin problem.
Today’s developers are also combining hormone targets into increasingly complex molecules, including the GLP-1/GIP/glucagon strategy described in our retatrutide science explainer. Those combinations make it increasingly useful to understand what individual receptor pathways do outside their most obvious metabolic functions.
That does not mean the Basel prolactin finding applies to retatrutide, semaglutide or tirzepatide. It means the endocrine effects of GLP-1 signaling deserve further study.
What the Next Trial Needs to Answer
The next step is replication with more participants, repeated prolactin measurements and a statistical plan capable of distinguishing small hormone effects from natural variation.
Future studies should also include women, older adults, people with obesity or diabetes and participants with different baseline prolactin concentrations. Longer exposure would help determine whether the effect persists beyond four weeks.
If researchers want to establish a dopamine-mediated mechanism, they will need measurements or experimental approaches capable of testing that pathway rather than inferring it from prolactin alone.
Clinically, the new paper gives no reason to prescribe dulaglutide to lower prolactin and no evidence that doing so would improve sexual function. It also does not establish the absence of every possible sexual effect during longer-term treatment. The participants were healthy and had normal hormone function, and the study was small.
The result is useful precisely because it is limited. In 24 men, four weeks of dulaglutide produced a modest observed prolactin decline relative to a rise on placebo, while an adjusted analysis suggested a larger treatment difference. Sexual desire showed no statistically significant change.
The experiment began with a question about sex drive and ended with a question about pituitary biology. That is a worthwhile finding, provided nobody mistakes the question for an answer.
A hormone moved. A behavior did not significantly change. The mechanism remains open.
Sources
- Reber V et al. Effects of glucagon-like peptide-1 analogs on prolactin levels: A secondary analysis of a randomized, double-blind, placebo-controlled crossover trial. Journal of Neuroendocrinology. November 5, 2026;38(11):e70308.
- PubMed Central. Full open-access text of Reber et al., 2026. Includes methods, adjusted analysis, hormone measurements and limitations.
- Lengsfeld S et al. Effects of the glucagon-like peptide-1 receptor agonist dulaglutide on sexuality in healthy men: a randomised, double-blind, placebo-controlled crossover study. eBioMedicine. September 2024;107:105284.
- Mueller C et al. Effects of glucagon-like peptide-1 analogs on prolactin: a randomized, double-blind, placebo-controlled crossover trial. European Congress of Endocrinology abstract P549, May 2024. Earlier unadjusted analysis.
- ClinicalTrials.gov. Effects of Glucagon-Like Peptide-1 Analogs on Sexuality (DESIRE). NCT04687514, completed trial registration.
