NSAIDs and Erectile Dysfunction: Do Common Painkillers Affect Erections?

The question of whether NSAIDs and erectile dysfunction are causally linked has circulated in men's health coverage since a 2011 study of more than eighty thousand men reported that regular users of non-steroidal anti-inflammatory drugs were significantly more likely to report erectile difficulty. Headlines followed. What did not follow, in most of that coverage, was the next fifteen years of data — which include a large prospective trial cohort that found no independent association once the reasons men take these drugs were accounted for. This article reviews what the evidence actually supports, because the answer turns out to be more clinically useful than a simple yes or no.

The Early Signal: Two Cohorts That Raised the Question

The finding that started the conversation came from the California Men's Health Study. Gleason and colleagues analysed 80,966 men aged 45 to 69 within a large integrated health system. Roughly 47% were regular NSAID users, and 29% reported moderate-to-severe erectile dysfunction. The unadjusted odds ratio was 2.40 (95% CI 2.27–2.53) — a striking number in isolation. After adjustment for age, race and ethnicity, smoking, diabetes, hypertension, hyperlipidaemia, vascular disease, coronary artery disease and body mass index, the association fell to an odds ratio of 1.38 but remained statistically significant ^[1]. The authors were explicit that the design was cross-sectional and that confounding by indication could not be excluded.

A smaller Finnish population cohort had reported something similar five years earlier. Shiri and colleagues followed 1,126 men aged 50, 60 and 70 who were free of erectile dysfunction at baseline. Over five years, the incidence of erectile dysfunction was 93 per 1,000 person-years among NSAID users versus 35 per 1,000 person-years among non-users. Adjusted incidence density ratios were 2.0 (95% CI 1.2–3.5) in NSAID users without arthritis and 1.9 (95% CI 1.2–3.1) in those with arthritis, and the authors argued the effect appeared independent of the indication for treatment ^[2].

Two studies, two continents, consistent direction. That is how an association gets into the public consciousness.

Confounding by Indication: Why Men Who Take NSAIDs Are Different

The central methodological problem is that regular NSAID use is not randomly distributed across a population. Men who take ibuprofen, naproxen or diclofenac several times a week are, on average, men with osteoarthritis, rheumatoid arthritis, chronic back pain, gout or established cardiovascular disease requiring an anti-platelet agent.

Every one of those conditions is itself associated with erectile dysfunction, through mechanisms that have nothing to do with the medication. Chronic pain suppresses sexual desire and interferes with sleep. Inflammatory arthritis impairs mobility and joint function in ways that affect sexual activity directly. Systemic inflammatory disease is associated with endothelial dysfunction. Atherosclerosis reduces cavernosal arterial inflow. The statistical term for this problem is confounding by indication: the drug is a marker for the disease, and the disease is what drives the outcome.

Adjusting for measured comorbidities helps but rarely eliminates the problem, because severity is poorly captured by a diagnosis code. Two men can both carry a diagnosis of osteoarthritis and have entirely different disease burdens — and the one taking NSAIDs daily is almost certainly the one with worse disease.

The PCPT Null Result and the 2025 Mendelian Randomization Data

The most informative test of the hypothesis came from Patel and colleagues, who analysed 4,726 men in the placebo arm of the Prostate Cancer Prevention Trial who had no erectile dysfunction at baseline and were followed prospectively with structured assessment ^[3].

In unadjusted terms, the association appeared again: non-aspirin NSAID use carried a hazard ratio of 1.16 (P = 0.02) for incident mild-to-moderate erectile dysfunction, and aspirin use a hazard ratio of 1.16 (P = 0.03) for severe erectile dysfunction. But when the investigators adjusted for the indications for which the drugs were being taken, the associations attenuated. Arthritis itself carried a hazard ratio of 1.56 and atherosclerotic disease a hazard ratio of 1.60 — both larger than the drug signal. The authors concluded that NSAID use is not independently associated with erectile dysfunction risk.

More recent work using genetic instruments points the same way, with one interesting exception. A 2025 analysis combining NHANES cross-sectional data from 3,989 men with Mendelian randomization found that non-aspirin NSAIDs were not associated with erectile dysfunction, while salicylic acid derivatives — aspirin — were, with an odds ratio of 3.28 (95% CI 1.27–8.50), and the genetic analysis supported a causal relationship for salicylates specifically ^[4]. That finding deserves replication before it changes practice, and it is worth noting that aspirin users are, again, a cardiovascular population. But it does suggest that lumping all anti-inflammatory drugs together obscures more than it reveals.

There is, at the time of writing, no systematic review or meta-analysis specifically addressing NSAIDs and erectile dysfunction. Claims of settled consensus in either direction are not supported.

The Mechanisms That Were Proposed — and Their Limits

Several plausible mechanisms were advanced to explain the early observational signal, and it is worth understanding why plausibility alone was never sufficient.

The prostaglandin argument is the most cited. Prostaglandin E1 is a potent cavernosal vasodilator acting through cyclic AMP, and it remains an established intracavernosal therapy for erectile dysfunction in the form of alprostadil ^[5]. Since NSAIDs inhibit cyclooxygenase and therefore reduce prostaglandin synthesis, the inference that they might blunt prostaglandin-mediated smooth muscle relaxation is reasonable on its face. The difficulty is that the principal physiological pathway for erection is nitric oxide and cyclic GMP, not prostaglandin and cyclic AMP, and no human study has demonstrated that therapeutic-dose oral NSAIDs measurably impair cavernosal haemodynamics.

The second argument concerns blood pressure and vascular risk. Non-aspirin NSAIDs raise blood pressure modestly and carry cardiovascular risk, a question examined directly in the PRECISION trial comparing celecoxib, naproxen and ibuprofen in patients with arthritis at cardiovascular risk ^[6]. Since hypertension and atherosclerosis are well-established contributors to vasculogenic erectile dysfunction, sustained NSAID use could contribute indirectly over years. This remains a reasonable long-term concern, but it is a slow, indirect pathway — not an explanation for the large effect sizes reported in cross-sectional data.

What This Means Clinically for NSAIDs and Erectile Dysfunction

Three practical conclusions follow.

First, a man should not stop taking a medically indicated anti-inflammatory because of concern about erectile function. The best-controlled prospective evidence does not support an independent effect, and untreated inflammatory disease has its own well-documented consequences for sexual function and quality of life. Medication changes belong in a conversation with the prescribing physician.

Second, the association is still clinically informative, just not in the way the headlines suggested. A man on regular NSAIDs who develops erectile dysfunction is flagging the presence of chronic inflammatory or atherosclerotic disease — and in the PCPT data those conditions carried the larger hazard ratios. New-onset erectile dysfunction in this group warrants cardiovascular risk assessment. Montorsi and colleagues documented that among 300 men with angiographically confirmed coronary artery disease, 49% had erectile dysfunction, and two thirds of those developed erectile symptoms before any cardiac symptom, at a mean interval of 38.8 months ^[7]. The smaller calibre of the cavernosal arteries makes them an earlier indicator of endothelial disease than the coronaries.

Third, the reference prevalence matters for interpretation. The Massachusetts Male Aging Study established that some degree of erectile difficulty affects roughly 52% of men aged 40 to 70 ^[8]. Against that baseline, small adjusted odds ratios are easy to generate and easy to over-interpret.

Conclusion

The relationship between NSAIDs and erectile dysfunction is best described as a real statistical association with a weak causal case. Two observational datasets found a signal; the most rigorous prospective analysis found it disappeared once the reasons for taking the drugs were accounted for; and the most recent genetic work implicates aspirin rather than NSAIDs as a class. The clinically useful reading is not that painkillers cause erectile dysfunction, but that men who need them regularly often carry the inflammatory and vascular burden that does. Further reading on vascular and inflammatory contributors to sexual function is available on the OnyxMD blog.

If you're exploring clinically-formulated options, OnyxMD offers physician-supervised treatment plans starting with a free online assessment at questionnaire.getonyxmd.com. Its on-demand formulation, Red Pill, pairs tadalafil with Pycnogenol, a standardised pine bark extract studied for its antioxidant and endothelial effects, and is prescribed only after review by a licensed physician.

---

These statements have not been evaluated by the FDA. This content is for informational purposes only and does not constitute medical advice.

References

1. Gleason JM, Slezak JM, Jung H, et al. Regular nonsteroidal anti-inflammatory drug use and erectile dysfunction. The Journal of Urology. 2011;185(4):1388-1393. doi:10.1016/j.juro.2010.11.092
2. Shiri R, Koskimäki J, Häkkinen J, Tammela TL, Auvinen A, Hakama M. Effect of nonsteroidal anti-inflammatory drug use on the incidence of erectile dysfunction. The Journal of Urology. 2006;175(5):1812-1816. doi:10.1016/S0022-5347(05)01000-1
3. Patel DP, Schenk JM, Darke A, Myers JB, Brant WO, Hotaling JM. Non-steroidal anti-inflammatory drug (NSAID) use is not associated with erectile dysfunction risk: results from the Prostate Cancer Prevention Trial. BJU International. 2016;117(3):500-506. doi:10.1111/bju.13264
4. Zhao C, Qu L, Liu F. Association between 20 medications and erectile dysfunction: results from a cross-sectional study and Mendelian randomization analysis. Medicine (Baltimore). 2025;104(32):e43157. doi:10.1097/MD.0000000000043157
5. Burnett AL, Nehra A, Breau RH, et al. Erectile dysfunction: AUA guideline. The Journal of Urology. 2018;200(3):633-641. doi:10.1016/j.juro.2018.05.004
6. Nissen SE, Yeomans ND, Solomon DH, et al. Cardiovascular safety of celecoxib, naproxen, or ibuprofen for arthritis. New England Journal of Medicine. 2016;375(26):2519-2529. doi:10.1056/NEJMoa1611593
7. Montorsi F, Briganti A, Salonia A, et al. Erectile dysfunction prevalence, time of onset and association with risk factors in 300 consecutive patients with acute chest pain and angiographically documented coronary artery disease. European Urology. 2003;44(3):360-365. doi:10.1016/s0302-2838(03)00305-1
8. Feldman HA, Goldstein I, Hatzichristou DG, Krane RJ, McKinlay JB. Impotence and its medical and psychosocial correlates: results of the Massachusetts Male Aging Study. The Journal of Urology. 1994;151(1):54-61. doi:10.1016/s0022-5347(17)34871-1