Background
Migraine is a common and disabling disorder characterized by a sexual dimorphism: after puberty, women experience higher prevalence, attack frequency and duration than men, together with greater clinical burden and, in some studies, higher responsiveness to treatment [1–3]. Despite increasing attention to the biological determinants of sex differences, the mechanisms underlying migraine in men remain poorly understood, largely due to their historical underrepresentation in clinical research [4]. This knowledge gap limits the generalizability of current evidence and hinders the development of personalized therapeutic strategies.
Human provocation models represent a powerful approach to understand migraine pathophysiology: the experimental induction of a migraine attack in susceptible individuals allows the controlled exploration of causal mechanisms and the testing of mechanistic hypotheses [5]. Among pharmacological triggers, sildenafil, a selective phosphodiesterase type 5 (PDE5) inhibitor, increases intracellular cyclic guanosine monophosphate (cGMP) levels in smooth muscle cells by inhibiting cGMP degradation and can induce migraine without aura in susceptible individuals. Previous studies conducted almost exclusively in women with migraine without aura have reported a high rate of migraine (83–89%) following sildenafil administration, supporting the involvement of cGMP-dependent pathways in migraine [6, 7].
Direct pharmacological activation of the cGMP pathway can induce migraine despite CGRP receptor blockade, consistent with the downstream position of cGMP relative to CGRP signaling [8]. In contrast, the migraine-inducing potential of sildenafil in men with migraine without aura remains unknown. Addressing this question is clinically relevant for at least three reasons: (i) the phenotypic expression of migraine differs between men and women, implying possible differences in susceptibility to specific triggers; (ii) the sensitivity to cGMP-mediated mechanisms may be modulated by hormonal factors (e.g., estrogens, prolactin) more prominent in women, potentially influencing attack induction; and (iii) determining whether cGMP triggers attacks in men and to what extent compared with women could help delineate sex-related mechanistic targets and inform precision medicine approaches.
Based on these considerations, we designed a study to prospectively evaluate the migraine-inducing effects of sildenafil in men with migraine without aura, while including an open-label cohort of women for sex comparisons under similar experimental conditions. This design integrates a double-blind, randomized, placebo-controlled crossover trial in men with the contextual collection of data in women, avoiding reliance on historical cohorts. A placebo-controlled provocation study in women was not considered informative, as the migraine-inducing effect of sildenafil has already been demonstrated in previous randomized trials.
A female open-label cohort (n = 15), with a size comparable to earlier provocation studies, was included to allow contextual comparison with men under similar experimental procedures. The aim is to generate direct evidence on cGMP-dependent pathways in men and to assess potential sex-related gradients in their contribution to migraine pathophysiology.
Methods
This was a randomised, double-blind, placebo-controlled crossover trial in men with migraine without aura, complemented by an open-label parallel cohort in women. Men attended one screening visit and two experimental sessions separated by a wash-out of ≥ 7 days. Women attended one screening visit and a single experimental session.
All study visits were conducted at the Department of Clinical Pharmacology, Pharmacy and Environmental Medicine (University of Southern Denmark) between November 2024 and July 2025. Full methodological details were published in the protocol paper [9]. Here, we provide a concise overview together with any clarifications relevant to the present report.
Figure 1 illustrates the design of the randomized, double-blind, placebo-controlled crossover trial, whereas Fig. 2 depicts the open-label parallel study.
Fig. 1Flowchart of the randomised, double-blind, placebo-controlled, crossover trial
Fig. 2Flowchart of the open-label trial
Participants
Eligible participants were 18–50 years old and fulfilled the International Classification of Headache Disorders criteria for migraine without aura [10]. They had experienced 1–6 migraine attacks in the preceding 8 weeks. Exclusion criteria comprised other primary headaches (no more than 3 days/month of tension-type headache were allowed), migraine < 24 h before dosing, recent intake of caffeine, alcohol or analgesics, uncontrolled cardiovascular, hepatic, psychiatric or substance-use disorders, smoking, or use of α-blockers, ritonavir or other strong CYP3A4 inhibitors.
Preventive migraine treatments were allowed, provided that the dosage had been stable for at least 14 days prior to the screening visit and remained unchanged throughout the entire study period. Women of child-bearing potential provided negative urinary pregnancy tests at screening and on the dosing day and used reliable contraception throughout the study.
Randomization and blinding
In the male cohort, a computer-generated list prepared by an independent pharmacist allocated participants (1:1) to receive sildenafil first or placebo first. Allocation codes were concealed in sequentially numbered, opaque envelopes opened immediately before dosing. Participants, investigators and outcome assessors remained blinded until database lock.
Interventions
The investigational product was a single tablet (100 mg) of sildenafil. Placebo consisted of an identically encapsulated glucose monohydrate tablet (Supplementary Fig. 1 in the Additional File 1).
Women received open-label sildenafil (100 mg). Study medications were packaged, labelled and blinded by Glostrup Apotek. All treatments were encapsulated in identical opaque gelatin capsules to ensure indistinguishable appearance, weight and taste.
Each capsule was administered with a glass of water after an overnight fast, and no food or beverages other than water were allowed until 90 min post-dose.
Procedures and assessments
Participants rested supine for ≥ 15 min before baseline measurements. Heart rate and blood pressure were recorded at 10-min intervals for 90 min after dosing. Adverse events and headache intensity (0–10 numerical rating scale or NRS) were documented throughout the 90-min observation and for 12 h post-dose using a take-home diary, which was reviewed by a blinded investigator via telephone the next day.
A migraine-like attack was defined according to validated criteria [11]. Provoked migraine must fulfil at least one of two categories: category one describes headache with at least two of the following clinical features: unilateral location, pulsating quality, moderate to severe pain intensity, and aggravation or avoidance of routine physical activity. Additionally, headache must be accompanied by at least one of the following symptoms: nausea, vomiting, or photophobia and phonophobia.
Category two describes headache that mimics the patient’s usual migraine attacks and is treated with a rescue medication.
Sample size
The sample size in men was calculated based on data from previous sildenafil provocation studies in women, assuming a migraine incidence of 65% after sildenafil and 5% after placebo. With a two-sided α = 0.05 and 80% power, 12 participants were required according to McNemar’s test [9]. This calculation was performed for the placebo-controlled within-subject comparison in men, which represents the only endpoint that could be formally powered given the lack of prior placebo-controlled data in males.
The sex-based comparison between sildenafil-treated men and women was pre-specified as a co-primary endpoint in the study protocol, but no formal sample size calculation was feasible for this comparison. The female sample (n = 15) was chosen to provide a contextual biological reference, rather than to support formal hypothesis testing, under similar procedural conditions and in comparison with previously published placebo-controlled studies [6–8].
Statistical analysis
The primary outcome was the incidence of migraine-like attacks (i) after sildenafil versus placebo in men and (ii) after sildenafil in men versus women. Secondary outcomes included the incidence of any headache, comparisons of heart rate and mean arterial pressure area-under-the-curve (AUC₀-₉₀ min) between sildenafil and placebo in men, and the frequency of adverse events up to 12 h post-dose. Continuous variables are reported as mean ± standard deviation (SD) or as median and interquartile range (IQR), whereas categorical variables are expressed as numbers and percentages.
Results are presented together with 95% confidence intervals (CIs). Paired binary comparisons of migraine-like attacks, headache and adverse events in men (sildenafil vs. placebo) were performed using McNemar’s test. Differences in incidence between sexes were evaluated with Fisher’s exact test.
For the primary analyses and other pre-specified directional hypotheses (i.e., higher incidence with sildenafil, and women > men on the sildenafil day), one-sided p-values were used with α = 0.05. Hemodynamic AUCs were calculated from baseline until 90 min post-administration to obtain summary measures and to assess differences between sildenafil and placebo [12]. The within-subject AUC difference (sildenafil minus placebo) was computed and the mean difference was tested against zero using a paired Student’s t-test.
Analyses were performed using Python-based statistical routines (OpenAI, ChatGPT Data Analyst), employing functions equivalent to those implemented in GraphPad Prism 8.3.0. No data imputation was required owing to negligible missingness.
Ethics and monitoring
The study was conducted in accordance with the Declaration of Helsinki and the principles of Good Clinical Practice (GCP). Approval was obtained from the Danish National Committee on Health Research Ethics and the Danish Medicines Agency. The trial is registered in the Clinical Trials Information System (CTIS, number 2024-512014-17-02).
All participants provided written informed consent prior to any study procedures. The study was monitored by personnel from the GCP-enheden at Odense University Hospital (OPEN). All protocol deviations were documented and evaluated according to GCP procedures.
No substantive amendments were made before unblinding. One protocol deviation occurred, consisting of the inclusion of a male participant aged above 50 years (60 years old). The participant completed all study procedures without adverse events, and his data were included for analysis.
Results
A total of 27 participants (12 men and 15 women) were assessed for eligibility at screening. No screening failures occurred, and all enrolled participants completed the study. Their baseline characteristics are summarized in Table 1.
Overall, all participants experienced moderate to severe migraine attacks aggravated by routine physical activity. The mean age was 36.6 ± 11.2 years in men and 27.3 ± 5.3 years in women. Migraine features such as unilateral and pulsating headache, nausea, photophobia and phonophobia were descriptively less frequent in men than in women.
However, these differences were not formally tested for statistical significance. The mean number of migraine days over the previous three months and the use of preventive medication were comparable between sexes. The observed baseline migraine frequency in men and women corresponds to a population with episodic migraine and at least one attack per month, rather than very low-frequency migraine.
Table 1Baseline characteristics of study participantsMen with migraine (n = 12)Women with migraine (n = 15)Age (years)36.6 ± 11.227.3 ± 5.3Weight (kg)76.5 ± 10.265.9 ± 10.8Migraine featuresUnilateral headache42% (n = 5)80% (n = 12)Pulsating headache33% (n = 4)73% (n = 11)Moderate and/or severe intensity100% (n = 12)100% (n = 15)Headache worsened by physical activity100% (n = 12)100% (n = 15)Nausea58% (n = 7)93% (n = 14)Vomiting17% (n = 2)47% (n = 7)Photophobia92% (n = 11)93% (n = 14)Phonophobia75% (n = 9)100% (n = 15)Mean number of migraine days in the last three months3.8 ± 3.03.9 ± 2.7Use of preventive medications25% (n = 3)13% (n = 2)Data are expressed as mean ± standard deviation or percentages and absolute numbers
Crossover trial (men)
Headache incidence over the 12-hour observational period was higher after sildenafil (n = 10, 83%, 95% CI: 52–98%) compared with placebo (n = 3, 25%, 95% CI: 6–57%) (p = 0.033). In sildenafil-induced headache, localization was mainly bilateral (70%), with a predominantly pressing, non-throbbing quality (70%). None reported nausea or vomiting.
Photophobia (two cases) and phonophobia (one case) occurred as symptoms associated with the headache. Three men (30%) used analgesics and experienced relief. A total of three men (25%, 95% CI: 6–57%) developed migraine-like attacks after taking sildenafil, compared with two men (17%, 95% CI: 2–48%) after taking placebo (p = 0.500).
Headache intensities on the NRS after sildenafil and placebo are displayed in Fig. 3. Individual characteristics of headache and associated symptoms are reported in the Supplementary Table 1 (Additional File 1).
Fig. 3Headache intensity (0–10 on a numerical rating scale) in 12 male patients with migraine following the administration of sildenafil (A) and placebo (B). Each black dashed line represents an individual patient.
The red line indicates the median headache intensity at each time point. The first portion of each graph (0–90 min, in grey) corresponds to the on-site observation phase, with time expressed in minutes. The second portion represents the home monitoring phase (up to 12 h), with time expressed in hours
Open-label trial (women)
All women with migraine (100%) developed headache following sildenafil administration, and 10 out of 15 (67%) experienced migraine-like attacks. The majority of sildenafil-induced headaches in women were unilateral (80%, n = 12) and mostly described as pressing in quality (67%, n = 10). Associated symptoms included nausea (60%, n = 9), photophobia (80%, n = 12), and phonophobia (40%, n = 6).
About 60% of participants (n = 9) required the use of rescue medication within 12 h of sildenafil intake. The proportion of participants reporting migraine-like attacks after sildenafil was significantly higher among women (67%) than among men (25%; p = 0.038). Headache intensities on the NRS after sildenafil in women with migraine are displayed in the Supplementary Fig.
2 (Additional File 1). Individual characteristics of headache and associated symptoms are reported in the Supplementary Table 2 (Additional File 1).
Hemodynamic variables
In the crossover trial, mean AUC values of mean arterial pressure over the 90-minute observation period were 8991 ± 967 mmHg * min after placebo and 8588 ± 1149 mmHg * min after sildenafil. The mean within-subject difference (sildenafil minus placebo) was − 403 mmHg * min (95% CI: -796 to -11). A paired t-test showed that sildenafil significantly reduced the AUC compared with placebo (p = 0.022).
For heart rate, mean AUC values were 6595 ± 665 bpm * min after placebo and 6861 ± 988 bpm * min after sildenafil. The mean within-subject difference was + 266 bpm * min (95% CI: -145 to + 676), which did not reach statistical significance (paired t-test, p = 0.091). The individual data underlying these figures are available in the Supplementary Fig.
3 (Additional File 1).
Adverse events
During the 12-hour observation period, adverse events were more frequently reported after sildenafil compared with placebo in men (8/12, 67% vs. 2/12, 17%; p = 0.031) (Table 2). The most frequently reported events after sildenafil were flushing (n = 3), warm sensations (n = 3), and nasal congestion (n = 3).
Additional adverse events included neck stiffness (n = 2), fatigue (n = 2), concentration difficulties (n = 1), drowsiness (n = 1), red eyes (n = 1), and pressure over the head (n = 1). In the placebo condition, adverse events were rare, with only one participant each reporting fatigue or pressure over the head. Among women (n = 15), the most common events following sildenafil were flushing (n = 10), warm sensations (n = 7), neck stiffness (n = 4), fatigue (n = 4), concentration difficulties (n = 3), blurred vision (n = 2), dizziness (n = 2), palpitations (n = 2), thirst (n = 2), and drowsiness (n = 2).
Less common events included lacrimation, aura, mood swings and dry eyes (all n = 1).
Table 2Adverse events reported during crossover and open-label trials (12-hour observational period)Women with migraine (n = 15)Men with migraine (n = 12)Men with migraine (n = 12)Sildenafil (100 mg)Sildenafil (100 mg)PlaceboFlushing (n = 10)Flushing (n = 3)Fatigue (n = 1)Warm sensations (n = 7)Warm sensations (n = 3)Pressure over the head (n = 1)Neck stiffness (n = 4)Stuffy nose (n = 3)Fatigue (n = 4)Neck stiffness (n = 2)Concentration difficulties (n = 3)Fatigue (n = 2)Blurred vision (n = 2)Concentration difficulties (n = 1)Dizziness (n = 2)Drowsiness (n = 1)Heart palpitations (n = 2)Red eyes (n = 1)Thirst (n = 2)Pressure over the head (n = 1)Drowsiness (n = 2)Lacrimation (n = 1)Aura (n = 1)Mood swings (n = 1)Dry eyes (n = 1)
Discussion
In this randomized, placebo-controlled crossover study, sildenafil triggered headache in 10 of 12 men with migraine, whereas only 3 reported headache after placebo. Migraine-like attacks occurred in 3 of 12 participants following sildenafil, compared with 2 of 12 after placebo. Using the same sildenafil dose and outcome definitions, 10 of 15 women experienced a migraine-like attack after sildenafil administration.
This sex difference in the induction of migraine-like headache suggests a greater female susceptibility to cGMP-related signaling activation.
Biological sex differences
The sex difference in migraine prevalence (approximately 3:1, women to men) remains unexplained [1, 2]. Sex shapes therapeutic pathways, including those involving CGRP, but the precise differences between men and women with migraine are unclear [13, 14]. In this context, provocation studies provide a controlled framework to dissect these mechanisms and compare their relevance across sexes.
In our study, the cGMP-dependent pathway appeared less relevant in men, as sildenafil failed to elicit more migraine-like attacks than placebo. In contrast, 10 out of 15 (67%) women showed sildenafil-induced migraine, consistent with previous studies conducted in female cohorts [6–8]. Women used analgesics more frequently (60% vs.
25%), experienced a greater number of migraine-associated symptoms and reported adverse events more often. Pharmacologically, sildenafil increases intracellular cGMP by inhibiting PDE5, which is involved in the degradation of cGMP in both vascular cells and trigeminal ganglion. The exact mechanism by which sildenafil triggers headache is not fully understood, but it is thought to modulate pain-related processes within the trigeminovascular pathway.
Preclinical studies suggest the involvement of specific ion channels, although this has not yet been confirmed in humans [15]. One of these channels is the potassium/sodium hyperpolarization-activated cyclic nucleotide-gated ion channel 2 (HCN2), which is expressed in nociceptive neurons and vascular smooth muscle cells and can be directly activated by cGMP binding [16, 17]. Hormonal modulation may also play a role: estrogens and prolactin, which influence vascular reactivity and nociception, could enhance cGMP-driven mechanisms in women, whereas lower baseline levels of these hormones in men may attenuate the same cascades [18, 19].
In addition to potential pharmacodynamic differences, sex-related variations in sildenafil pharmacokinetics could further influence the observed responses. As in previous sildenafil provocation studies, a fixed 100 mg dose was used without weight adjustment to ensure comparability with the existing literature. Sex-related pharmacokinetic differences cannot be excluded, but the primary randomized within-subject comparison in men is not affected by interindividual differences in body weight.
Factors such as differences in volume of distribution, hepatic metabolism and CYP3A4 activity may affect systemic exposure to sildenafil, although no studies have investigated sex-specific pharmacokinetic profiles of sildenafil. At present, these hypotheses remain untested, and further human studies focusing on hormonal modulation and pharmacokinetic factors are warranted to clarify these mechanisms.
Rethinking diagnostic criteria
Although headache was common in men after sildenafil, the clinical features of these attacks less frequently met the criteria for a migraine-like attack [11]. In our study, headache occurred in most men (83%) and in all women (100%), but only 25% of men fulfilled the criteria for a migraine-like attack compared with 67% of women. The difference could be related to the presence of associated symptoms: nausea was reported by none of the men but by 60% of women, photophobia by 17% vs.
80%, and phonophobia by 8% vs. 40%. In addition, 25% of men and 60% of women used analgesics within 12 h after sildenafil.
When the diagnostic thresholds used in provocation studies are applied to experimentally induced migraine, they may not account for sex-related differences in symptom expression. Previous evidence suggested that men tend to report fewer trigger factors than women, even when menstruation is excluded from the analysis [20]. Moreover, men experience shorter migraine attacks, fewer accompanying symptoms and more often bilateral or non-pulsating pain compared with women [21–24].
The lower rate of migraine-like attacks in men may reflect a different way in which migraine symptoms manifest [25], rather than a lack of underlying trigeminovascular activation. Future provocation studies may benefit from integrating physiological or biomarker-based endpoints to more precisely characterize the comparability of migraine between men and women.
Study limitations
This study has some limitations. Sex-related differences at the baseline, which are well documented in the literature, may have contributed to the differential susceptibility to sildenafil-induced migraine observed between men and women. Although men were older than women at baseline, both cohorts fall within the adult age range typically investigated in migraine provocation studies, and there is currently no evidence that age within this range affects susceptibility to pharmacological migraine triggers.
Only participants with migraine without aura were included, which restricts generalizability to other migraine subtypes. In addition, hormonal status in women was not standardized and plasma concentrations of sildenafil were not measured, preventing pharmacodynamic-pharmacokinetic analyses that might have clarified sex-related differences in exposure. Moreover, the trial in women was conducted as an open-label study rather than a placebo-controlled comparison.
This may introduce expectation effects, but previous provocation studies with other triggers, such as CGRP, have shown that their migraine-inducing capacity is comparable between controlled and uncontrolled designs [26, 27], and our findings in women were consistent with those reported in earlier sildenafil studies [6–8]. The open-label female cohort was intended to serve as a contextual biological reference under similar experimental conditions, given that placebo-controlled sildenafil provocation data in women were already available from previous studies. At the same time, our study includes features that support the robustness of the findings.
This is the first randomized, placebo-controlled crossover trial specifically evaluating cGMP-mediated headache mechanisms in men with migraine, conducted under standardized and monitored conditions. The sample size in men was prospectively calculated to ensure adequate statistical power, and the parallel female cohort served as an internal comparator under matched experimental conditions, allowing direct assessment of sex-related differences. Despite the non-orthodox nature of the sex-based comparison, the observation of a statistically significant difference between men and women with a relatively small sample size argues in favor of a robust underlying signal rather than a spurious finding.
Conclusion
In this study, sildenafil provoked headache but not migraine-like attacks in men with migraine without aura, whereas women showed a markedly higher susceptibility to migraine induction. These findings suggest that cGMP-mediated mechanisms may contribute less prominently to migraine in men than in women. Biological sex modulates migraine and there is a need for future studies integrating hormonal, pharmacokinetic and mechanistic assessments [28].
Inconsistent results observed with other migraine triggers, such as vasoactive intestinal peptide (VIP) and substance P, may also point to sex-related differences in peptide signaling and further support the concept of sexually dimorphic mechanisms [29, 30]. New investigations may inform more sex-sensitive approaches to migraine therapy.
Supplementary Information
Below is the link to the electronic supplementary material.
Supplementary Material 1