Male Sexual Dysfunction Associated with Selective Serotonin Reuptake Inhibitors (SSRIs): A Pharmacovigilance Disproportionality Analysis of FAERS Data

Escitalopram and citalopram showed the strongest ED signals (ROR ~8) across 6.6 million FAERS reports; fluvoxamine's signal didn't reach significance

Journal: Journal of Affective Disorders | Published: 2026-02-07 | Type: Journal Article | PMID: 41662926 Authors: Li Shouying, Feng Jiaxin, Zhang Shilin (Department of Urology, The Affiliated Foshan Women and Children Hospital, Guangdong Medical University) Funding/COI: Funding not disclosed; authors declare no conflicts of interest

Summary

Using 21 years of FDA adverse event reports, this study quantified how often male sexual dysfunction — particularly erectile dysfunction — shows up in the FAERS database for six common SSRIs. Escitalopram and citalopram carried the highest disproportionality signals for both ED and general sexual dysfunction; fluvoxamine was the outlier with an ED signal that failed to clear statistical significance. The authors are careful to say this is a reporting pattern analysis, not a head-to-head risk comparison.

Claims

• 6,631,746 SSRI-associated adverse drug event reports analyzed from FAERS (Q1 2004–Q1 2025)
• Erectile dysfunction was the most frequently reported male sexual adverse event across all six SSRIs
• Escitalopram had the highest "sexual dysfunction" ROR at 16.42 (95% CI: 14.11–19.12); citalopram was close at 15.09 (95% CI: 12.98–17.54)
• For ED specifically, escitalopram again led: ROR 7.93 (95% CI: 6.99–8.99); citalopram at 7.70 (95% CI: 6.81–8.69)
• Sertraline and fluoxetine fell in the middle range for ED: ROR 6.11 and 4.97, respectively
• Fluvoxamine's ED signal did not reach significance: ROR 1.08 (95% CI: 0.35–3.36) — confidence interval spans 1.0
• Fluvoxamine's "sexual dysfunction" signal was detectable (ROR 9.64) but with wide uncertainty (95% CI: 5.01–18.56) due to low report volume

Study Quality

This is a pharmacovigilance disproportionality analysis — a legitimate method for hypothesis generation from passive surveillance data, but not a clinical study. The core statistic, the reporting odds ratio, answers one question: is this adverse event over-represented in reports for this drug compared to the rest of the database? It does not answer whether the drug causes the event at any particular rate in the real-world population. The denominator is unknown; incidence cannot be calculated from FAERS.

The critical unaddressed confounder is indication bias: depression itself causes sexual dysfunction, and every patient in this dataset was (presumably) being treated for depression. Disproportionality methods do not control for this. Two decades of data also means wildly different prescribing volumes, labeling language, and public awareness campaigns — all of which drive spontaneous reporting rates independent of actual harm. The authors acknowledge most of these limitations, which is responsible; it just means the findings are more useful for prioritizing further research than for clinical inference.

Red Flags

• Fluvoxamine's ED signal is non-significant — yet the paper's framing implies all six drugs produced meaningful signals; this requires more prominent acknowledgment
• Indication bias: depression reduces libido and erectile function independently; FAERS cannot separate drug effect from disease effect
• Reporting bias: SSRIs with higher public profiles, more media coverage, or FDA label warnings generate more spontaneous reports regardless of actual risk differences
• No denominator: ROR of 7.93 for escitalopram tells you it's over-reported relative to the database baseline, not that 1-in-13 escitalopram users gets ED
• Funding not disclosed — unusual omission for a FAERS study with no obvious industry incentive, but it should be stated explicitly rather than left blank
• Author affiliation (Department of Urology, Women and Children's Hospital) is an unusual setting for this analysis; no methodological issue, but context matters

Strengths

• Massive dataset: 6.6 million adverse event reports across 21 years is about as large as FAERS analysis gets
• Two concordant methods (ROR and PRR) both applied, with appropriate confidence intervals reported
• All six major SSRIs covered in a single analysis, enabling within-database pattern comparisons
• Authors explicitly and correctly state that signal variation "reflects reporting differences, not comparative risk" — rare and commendable restraint in this literature
• Pre-specified methodology appropriate to the data source

Verdict

This is competent pharmacovigilance work that confirms what clinicians have long observed: SSRI-associated sexual dysfunction shows up prominently in adverse event data, with escitalopram and citalopram generating the loudest signals and fluvoxamine the quietest (though fluvoxamine's ED signal didn't actually clear significance). The paper's value is as a signal-detection exercise, not an efficacy or safety trial — and to the authors' credit, they mostly resist overclaiming. The unfiled funding disclosure and the uncritical treatment of fluvoxamine's non-significant ED finding are the two things an editor should have caught. Read this as background context on reporting patterns, not as evidence that one SSRI is safer for sexual function than another.