Persistent Sexual and Psychological Symptoms After Finasteride Discontinuation: A Cross-Sectional Observational Study
Paweł Jędrzejczyk, Tomasz Ząbkowski, Jarosław Ratajski, Kamil Ciechan, Tomasz Waldemar Kaminski, Patryk Uciechowski, Tomasz Syryło
Journal of Clinical Medicine2026
In plain terms
Doctors in Poland assessed 129 men who had stopped finasteride, taken for hair loss or an enlarged prostate, and still had symptoms, using standard questionnaires about erections, low mood and anxiety. Erection scores barely changed over the six months after stopping. Low-mood scores improved somewhat and anxiety scores, which started low, improved slightly, but the authors say most men did not return to normal. Men who were older, or who had taken more of the drug in total, tended to have worse scores.
What it doesn’t show: It can't show how common these problems are or that the drug caused them: it included only men who already had symptoms, had no comparison group, and does not report the figures behind its main finding about age and total dose.
Summary (paraphrased)
This 2026 cross-sectional study enrolled 129 men with prior finasteride exposure for hair loss or prostate enlargement and measured sexual function, depression, and anxiety with validated instruments (IIEF, PHQ-9, GAD-7). Erectile function remained impaired at follow-up while depressive and anxiety scores improved only partially without normalizing in most participants. The authors reported that greater cumulative finasteride exposure and older age were independently associated with worse sexual and psychological symptom severity in multivariable modelling, but the paper did not present the regression estimates (no odds ratios, confidence intervals or p-values). The authors called for prospective studies to define risk factors and long-term trajectories.
Evidence
Extracted from the full text; page numbers refer to the paper. The library’s own notes are labelled as such.
- Design
- Cross-sectional studyCross-sectional study (the authors' label) of men reporting symptoms after stopping finasteride, with retrospectively collected exposure data, questionnaire scores reported at baseline and 1, 3 and 6 months, median-split comparisons and regression
- Setting
- Not described. The authors are from clinics and hospitals in Warsaw, Poland; the study was approved by the review board of the Military Institute of Medicine (National Research Institute) on 25 June 2025. Recruitment route and dates are not stated.
- Population
- Adult men who had taken finasteride at any dose for male pattern hair loss or benign prostatic hyperplasia, had stopped, and reported symptoms at evaluation (no minimum duration). Mean age 47.1 years (median 49, IQR 31.5–62); 59 took 1 mg/day and 70 took 5 mg/day. The number in each indication group is not reported.
- Size
- 129 · 129 men analysed; the number assessed at each follow-up time point is not reported
- Exposure
- Finasteride 1 mg or 5 mg/day; median treatment 24 months (IQR 12.5–36); cumulative exposure (daily dose × months) median 44 (IQR 18.5–153); median 8 months (IQR 1–17) from stopping to evaluation.
- Compared with
- None (no unexposed group); comparisons within the cohort by median splits of age, cumulative dose and testosterone, and across time points
- Outcome
- Erectile function (IIEF; version not stated), depressive symptoms (PHQ-9) and anxiety (GAD-7), given in clinician-guided interviews; serum testosterone, DHT, LH, FSH, SHBG, oestradiol and prolactin; self-reported somatic symptoms.
- Follow-up
- Scores reported at baseline and 1, 3 and 6 months after stopping; how and when these were collected is not described
Key results
- Erection scores (IIEF) did not change: mean 15.2 ± 0.46 (SEM) at baseline, 15.3 at 1 and 3 months, and 15.4 ± 0.47 at 6 months · p. 5, Fig. 1A
- PHQ-9 fell from 12.4 ± 0.41 to 10.7 (1 month), 10.1 (3 months) and 9.1 ± 0.41 (6 months), a drop of 3.3 points. The paper says mean scores stayed in the moderate range throughout; it does not state its cut-offs, and on the instrument's widely used bands a mean of 9.1 falls in the mild band · p. 5, Fig. 1B
- GAD-7 fell from 3.29 ± 0.23 to 2.54 ± 0.20 at 6 months. The conclusions describe anxiety as elevated (p. 11), but on the GAD-7's widely used bands both means fall in the minimal range · p. 6, Fig. 1C; p. 11
- Symptoms began during therapy or immediately after stopping in 83 of 129 men (64%); sleep disturbance in 75 (58%); fatigue in 86 (67%) · p. 4; Table 1, p. 5
- Treatment duration, cumulative dose and age are reported as "the strongest predictors of symptom severity". The methods say results are given as odds ratios with 95% confidence intervals, but no estimates, intervals, p-values or regression table appear in the paper · pp. 4, 8
- Median splits: men with higher cumulative exposure were older, with lower testosterone, higher LH, poorer baseline IIEF and worse mental-status scores. Men with higher testosterone were younger, with higher DHT and better baseline erectile function, yet the text also states that testosterone did not differentiate sexual or psychiatric symptom severity · pp. 6, 8, 10; Figs 3–4
Limitations the authors note
- Modest sample size, with limited power for subgroup and indication-specific analyses
- Possible selection and recall bias from the observational design
- Only men reporting persistent symptoms were included, so prevalence, incidence and relative risk cannot be estimated, and more severe cases are over-represented
- No standard definition of persistent symptoms; persistence meant symptoms present at the evaluation
- No data on sexual function before finasteride, or on comorbidities such as diabetes and hypertension or on other medicines; unmeasured confounding cannot be excluded
- Men under 30 were under-represented
- Reasons for stopping finasteride were not collected
Also worth weighing (library’s note)
- Described as cross-sectional, yet it reports scores at baseline and 1, 3 and 6 months after stopping; baseline is not defined, and a median of 8 months (IQR 1–17) from stopping to evaluation is hard to reconcile with that schedule
- Cumulative dose is mg/day × months, so the 70 men on 5 mg accrue five times the monthly exposure of those on 1 mg; age correlated strongly with cumulative dose, and the hair-loss vs prostate split is not reported, so dose, indication and age are hard to separate
- Persistence had no minimum duration, and the lower IQR bound of time since stopping is 1 month, so at least a quarter of men were evaluated about a month or less after stopping
- No setting, recruitment route or study dates are described, and the IIEF version and score range are not stated
- Some statements conflict, e.g. ethics approval is described as not required (p. 4) but also as granted under protocol 40/25 (p. 11), and Table 1 has a "Currently on medication" row coded 1 for 29 men, unexplained given that all had stopped (p. 3)
Funding: No external funding Interests: None declared
What it can support (library’s note): An association, within men who already report symptoms after stopping finasteride, between older age or greater total exposure and worse sexual and mood scores, as stated by the authors without reported estimates. It cannot estimate how often such symptoms occur, establish causation, or separate dose from age and indication.
Why it’s in the corpus
A recent clinical cohort of 129 men reporting symptoms after stopping finasteride (no minimum duration; 70 took the 5 mg prostate dose) in which the authors link greater cumulative exposure and older age to symptom severity, relevant to the corpus's clinical axis. It reports no dose-response estimates, and cumulative dose is strongly correlated with age.