Estimating the risk of irreversible post-SSRI sexual dysfunction (PSSD) due to serotonergic antidepressants
Joseph Ben-Sheetrit; Yehonathan Hermon; Shlomo Birkenfeld; Yehiel Gutman; Antonei Benjamin Csoka; Paz Toren
Annals of General Psychiatry2023
In plain terms
Researchers searched the health records of about 12,300 healthy, non-smoking men aged 21 to 49 at one Israeli health service. About 1 in 12 men who had taken an antidepressant that acts on serotonin (a brain chemical; most had taken an SSRI) had been prescribed an erection drug at some point, compared with 1 in 40 men who had not; for most, the records don't show the erection drug came after the antidepressant. Four of the 866 men on these antidepressants, about 1 in 216, started an erection drug within a year of the antidepressant and were still taking it at least a month after stopping the antidepressant. Their records showed about 4 to 15 months of erection-drug use by the end of the study.
What it doesn’t show: It can't prove the antidepressants caused the problems or show that they were permanent: it used existing records, counted only erection problems treated with a prescribed erection drug in a hand-picked group of healthy men, followed the four men for months rather than years, and worked out the 1 in 216 from every man who took the drug, not only those who stopped it.
Summary (paraphrased)
Using Israeli health-maintenance-organization (Clalit Health Services) records over a 19-year period, the authors compared 866 healthy, non-smoking men aged 21-49 who had used serotonergic antidepressants with 11,436 eligible men who had not (unmatched controls), using PDE-5 inhibitor prescriptions as the marker of erectile dysfunction (ED). Serotonergic antidepressant use was associated with roughly threefold higher odds of a PDE-5 inhibitor prescription at any time in the study period after adjustment, although only 24 of the 74 users with such a prescription had their first one after their first antidepressant. Four of 866 exposed men (0.46%, about 1 in 216) met full PSSD case criteria, which required ED treatment starting within 12 months of the antidepressant and continuing at least 1 month after stopping it. The authors reported a prevalence of 4.3 per 100,000 among male Clalit members aged 21-49 in its Tel-Aviv district (4 of 92,476), although cases were sought only among the 12,302 who met the inclusion criteria, and concluded the risk was small but real and consistent with prior PSSD literature.
Evidence
Extracted from the full text; page numbers refer to the paper. The library’s own notes are labelled as such.
- Design
- Cohort studyRetrospective cohort study of health-maintenance-organisation records, comparing men who used serotonergic antidepressants with men who did not (logistic regression), plus a stepwise count of PSSD cases among the users
- Setting
- Clalit Health Services (CHS) database, Tel-Aviv district, Israel; a 19-year retrospective period (calendar years not stated)
- Population
- Men aged 21–49 with a last recorded BMI under 25, who had never smoked, with no recorded medical or psychiatric condition or medication associated with ED (other than a history of depression or anxiety and the antidepressants themselves), no 5α-reductase inhibitor use (finasteride, dutasteride) and no alcohol or substance use disorder.
- Size
- 12,302 · 12,302 men (866 antidepressant users, 11,436 non-users), selected from 92,476 male CHS members aged 21–49 among 309,417 members in the district
- Exposure
- Any serotonergic antidepressant during the study period. Among users, SSRIs 79.9%, tricyclics 15.6%, SNRIs 8.1%, Hypericum (St John's wort) 6.1%, and smaller numbers of mirtazapine, mianserin, trazodone, vortioxetine, maprotiline and phenelzine; 76.4% used only one antidepressant.
- Compared with
- Men meeting the same criteria with no serotonergic antidepressant in the study period (not matched; the groups differed slightly in age, socioeconomic status and BMI)
- Outcome
- ED, defined as one or more prescriptions of a PDE-5 inhibitor (erection drug) at any time in the study period. PSSD, counted among users only: no ED treatment before the antidepressant, first PDE-5 inhibitor within 12 months of starting the antidepressant, antidepressant stopped (not through non-compliance), no active depression or anxiety diagnosis, and PDE-5 inhibitor use continuing at least 1 month after stopping.
- Follow-up
- Records over a 19-year period (dates not given); the 4 PSSD cases had 3.6–14.8 months of PDE-5 inhibitor treatment by the end of the study period
Key results
- PDE-5 inhibitor use in 74 of 866 users (8.5%) vs 286 of 11,436 non-users (2.5%); crude odds ratio 3.6 (95% CI 2.8–4.8). Recomputed from these counts, the odds ratio is 3.64 and the risk ratio about 3.4 · p. 4; Tables 3–4, p. 6
- Adjusted for age, socioeconomic status, BMI, depression and anxiety, odds ratio 3.2 (95% CI 2.3–4.4); depression and anxiety had no significant effect beyond antidepressant use, age, BMI and socioeconomic status · pp. 4–5; Table 4, p. 6
- Case-finding among the 866 users: 74 had a PDE-5 inhibitor; 24 had no earlier ED treatment and a first antidepressant prescription before their first PDE-5 inhibitor; 9 started the PDE-5 inhibitor within 12 months of the antidepressant; 7 had stopped the antidepressant; 5 had not stopped through non-compliance and had no active depression or anxiety diagnosis; 4 kept using PDE-5 inhibitors for at least 1 month after stopping · Fig. 1B, p. 5
- PSSD in 4 of 866 users (0.46%; 866 ÷ 4 = 216.5, reported as "1 in 216"). The four were aged 32–42, had taken escitalopram, paroxetine, imipramine or nortriptyline for 4–8 months before starting a PDE-5 inhibitor, and had 3.6–14.8 months of PDE-5 inhibitor treatment by the end of the study period. The denominator is all 866 users; the paper does not report how many of them stopped their antidepressant · pp. 6–7; Fig. 1B, p. 5
- Reported prevalence 4.3 per 100,000 (4 of 92,476; recomputed 4.33). The denominator is all male CHS members aged 21–49 in the district, but cases were sought only among the 12,302 (13.3%) who met the inclusion criteria · p. 7; Fig. 1A, p. 5
Limitations the authors note
- Retrospective design, men only, and reliance on physician-diagnosed conditions for medical history (which they argue is reduced by excluding on both diagnoses and related medications)
- Counting only ED, rather than other forms of sexual dysfunction, may underestimate the true prevalence of PSSD
- The estimate reflects otherwise healthy men and may underestimate prevalence in the general population
- Requiring PDE-5 inhibitor treatment sets a high threshold for detection; many men may not seek help because of shame or unawareness
Also worth weighing (library’s note)
- The odds ratios count PDE-5 inhibitor use at any time in the study period. In Fig. 1B, 50 of the 74 users with a PDE-5 inhibitor did not have a first antidepressant prescription before their first PDE-5 inhibitor (with no earlier ED treatment), so the threefold association does not show that ED treatment followed the antidepressant
- The Methods' exclusions for PDE-5 inhibitor use before the antidepressant, non-compliance, and active depression or anxiety appear as steps in the PSSD case count (Fig. 1B), not in the cohort flow used for the odds ratios (Fig. 1A)
- The 0.46% (1 in 216) divides 4 cases by all 866 users, including men who may not have stopped (only 7 of the 9 with early ED treatment had stopped). No comparable count of persisting PDE-5 inhibitor use is given for the 11,436 non-users, so it is not an excess over background; no confidence interval is given, and it rests on 4 cases
- The 12-month window from antidepressant to first PDE-5 inhibitor (chosen to limit misattribution; the authors cite a mean delay of about 30 months in seeking ED treatment) removed 15 of the 24 users whose ED treatment began after the antidepressant
- Persistence required at least 1 month of PDE-5 inhibitor use after stopping, and the cases had 3.6–14.8 months recorded; this follow-up does not show the "irreversible" course named in the title
- Serotonergic antidepressants here include tricyclics, mirtazapine, mianserin, trazodone and St John's wort; two of the four case drugs named are tricyclics (imipramine, nortriptyline), so the figure is not SSRI-specific
- Highly selected sample, 12,302 of 92,476 male members aged 21–49 (13.3%); one health-service district; results may not generalise
- Table 2 has internal inconsistencies. "Four or more" antidepressants is given as 26 (0.9%), but the counts then sum to 884, not 866, and 26 of 866 is 3.0%; 8 men would give both 0.9% and a total of 866. SSRIs are given as 892 (79.9%), more than the 866 users; 692 would give 79.9%. The main results do not depend on these figures
Funding: No specific grant from any funding agency in the public, commercial or not-for-profit sectors Interests: None declared
What it can support (library’s note): An association, in healthy Israeli men aged 21–49, between serotonergic antidepressant use and having a PDE-5 inhibitor prescription at some point, and a count of 4 users (of 866) whose erection-drug treatment began within a year of the antidepressant and continued at least a month after stopping it. It does not establish causation, that ED treatment followed the antidepressant in most cases, or a general risk of PSSD.
Why it’s in the corpus
One of the only quantitative PSSD estimates in the literature: 4 of 866 healthy, never-smoking men aged 21-49 with BMI under 25 (about 1 in 216), counting only erectile dysfunction treated with a PDE-5 inhibitor that began within 12 months of the antidepressant and continued at least 1 month after stopping it (3.6-14.8 months recorded). The denominator includes men who did not stop, there is no comparison with non-users and two of the four cases involved tricyclics, so it is a count under a narrow definition rather than a general PSSD risk; core epidemiology citation for the corpus.
Related records
- Peer-reviewed paperPSSD Discord pick
Post-SSRI Sexual Dysfunction (PSSD): A comprehensive review of epidemiology, pathophysiology, and clinical management
Xie, M.; Huang, P.; Wang, S.; Ma, J.; Zeng, Y.; Sun, J. · Journal (SAGE; 2026)
Systematic review (PubMed/Web of Science/Scopus to April 2024) finding persistent symptoms in up to 74% of affected cohorts beyond 6 months; pathophysiology multifactorial — 5-HT1A desensitization, neurosteroid (allopregnanolone) depletion via 5α-reductase…
HTR1APSSD-0032026PFSPSSDCore corpus - Peer-reviewed paper
Post-SSRI sexual dysfunction: barriers to quantifying incidence and prevalence
Healy, David; Mangin, Dee · Epidemiology and Psychiatric Sciences
Examines why PSSD incidence/prevalence remain unquantified — inconsistent coding scatters cases across health records under vague labels, blocking the large database studies needed for clear estimates.
PSSD-0132024PSSDCore corpus - Peer-reviewed paper
Post-SSRI sexual dysfunction & other enduring sexual dysfunctions
David Healy · Epidemiology and Psychiatric Sciences
Healy traces the history of enduring sexual dysfunction after antidepressants from 1991 regulator reports through the first formal PSSD descriptions in 2006, arguing the syndrome has major clinical, research, and regulatory implications that psychiatry has…
LIT-0222019PSSDLiterature sweep · Oct 2026