Human dose-ranging pharmacokinetic study. Drake et al. enrolled 249 men with androgenetic alopecia who underwent scalp biopsies before and after receiving 0.01, 0.05, 0.2, 1, or 5 mg daily finasteride or placebo for 42 days. Scalp skin DHT, scalp skin testosterone, and serum DHT were measured via radioimmunoassay from punch biopsy specimens and blood draws.

Finasteride 1 mg/day reduced scalp skin DHT by 64.1% and serum DHT by 71.4%. Doses as low as 0.2 mg/day achieved near-maximal suppression of both scalp DHT (56.5%) and serum DHT (68.6%). The 5 mg BPH dose reduced scalp DHT by only 69.4% — marginally more than 1 mg. Scalp testosterone increased as expected (less conversion to DHT), but testosterone is not the primary driver of follicle miniaturization.

This is the foundational pharmacology study that justified the 1 mg dose. It directly demonstrates that finasteride reaches the target tissue (scalp), inhibits the target enzyme (type II 5-alpha reductase), and reduces the target androgen (DHT) — all measured in human scalp biopsies from AGA patients. The dose-response data shows a plateau effect: 1 mg captures ~95% of the maximal scalp DHT reduction achievable at any dose, which is why the hair loss dose is one-fifth of the BPH dose. This is exactly the kind of mechanistic evidence you want — not extrapolated from cell lines or animal models, but measured in the actual human tissue where the drug needs to work.

• Drake L et al. (1999). The effects of finasteride on scalp skin and serum androgen levels in men with androgenetic alopecia. J Am Acad Dermatol PMID 10495374

Two randomized, double-blind, placebo-controlled Phase III trials (Studies 087 and 089). 1,553 men aged 18-41 with male pattern hair loss (Hamilton-Norwood types III vertex, IV, and V) were randomized to finasteride 1 mg/day or placebo. 1,215 men continued in blinded extension for a second year. Primary endpoint: hair count in a 1-inch diameter circle (5.1 cm2) of balding vertex scalp, counted from macrophotographs. Secondary endpoints: investigator assessment, patient self-assessment, and standardized global photography with expert panel review.

Mean baseline hair count was 876 hairs in the target area. Finasteride-treated men gained hair while placebo men continued to lose, producing a net treatment difference of +107 hairs at 1 year and +138 hairs at 2 years (p<0.001 for both). By investigator assessment, 48% of finasteride patients showed increased hair growth at 1 year versus 7% on placebo. Patient self-assessment confirmed that finasteride slowed hair loss, increased hair growth, and improved appearance of hair at all timepoints (p<0.001 for all comparisons).

This is the gold standard of hair loss evidence and one of the most robustly positive pivotal trials in dermatology. Two independent multi-center trials with identical designs produced consistent results. The 138-hair treatment difference at 2 years over a 5.1 cm2 area — roughly a 16% increase from baseline — was visible to patients, investigators, and blinded expert panels reviewing photographs. Every endpoint moved in the same direction with the same statistical significance. The trial design is essentially above reproach: randomized, double-blind, placebo-controlled, multi-center, adequate sample size, relevant endpoints including both objective counts and subjective assessments, and a 2-year duration that captures the full treatment arc. This is why finasteride carries an FDA approval and why it remains the reference standard against which all other hair loss treatments are compared.

• Kaufman KD et al. (1998). Finasteride in the treatment of men with androgenetic alopecia. J Am Acad Dermatol PMID 9777765

Multinational blinded extension of the two Phase III pivotal trials. Men from the original Kaufman 1998 cohort continued in up to four additional 1-year placebo-controlled extension studies for a total of 5 years of blinded treatment. Efficacy was assessed by hair counts in the vertex target area, investigator and patient assessments, and panel review of standardized clinical photographs.

Finasteride led to durable improvements in scalp hair over 5 years (p<0.001 versus placebo at all endpoints). Hair counts in finasteride patients peaked at approximately year 2 and were maintained above baseline through year 5, while placebo patients fell progressively below baseline. Finasteride treatment produced a 93% decrease relative to placebo in the 5-year likelihood of developing further visible hair loss (95% CI: 89-97%; p<0.001). No new safety concerns emerged during long-term use.

Five-year blinded data from the same pivotal cohort is extraordinary in hair loss research. Most trials run 6-12 months; having 5 years of placebo-controlled data addresses the critical question for a drug taken indefinitely: does it keep working? The answer is clearly yes. The 93% reduction in visible progression is a powerful clinical result. The main caveat is survivorship bias — 5-year completers are inherently enriched for patients who tolerated the drug and perceived benefit. Non-responders and those with side effects likely dropped out. This is standard for extension studies and doesn't invalidate the finding, but the 93% figure should be understood as the outcome for the completer population, not a guarantee for every patient who starts treatment.

• Kaufman KD et al. (2002). Long-term (5-year) multinational experience with finasteride 1 mg in the treatment of men with androgenetic alopecia. Eur J Dermatol PMID 11809594

Randomized, double-blind, placebo-controlled trial specifically designed to assess frontal/anterior mid-scalp response. Men with androgenetic alopecia involving the frontal/anterior mid-scalp region received finasteride 1 mg/day or placebo for up to 2 years. Hair counts were measured in a defined frontal scalp area. Investigator assessment and global photographic evaluation were secondary endpoints.

Finasteride produced statistically significant increases in frontal hair count compared to placebo (p<0.05), with significant improvements in patient, investigator, and global photographic assessments. However, the magnitude of response was less than at the vertex: in long-term follow-up, 89.7% of patients improved at the vertex versus 61.2% in the frontal area and 44.4% at the hairline over 5 years of treatment.

This study directly answers the common patient question: 'Will finasteride help my receding hairline?' The answer is yes — finasteride produces statistically significant improvement at the frontal scalp — but with an important reality check. The vertex responds best (nearly 90% improvement rate at 5 years), the frontal scalp responds moderately well (61%), and the temples/hairline respond least (44%). This gradient likely reflects regional differences in androgen receptor density, follicle stem cell biology, and the severity of miniaturization at the hairline. Patients expecting complete hairline restoration from finasteride alone will likely be disappointed. But slowing frontal recession and improving mid-scalp density are realistic and evidence-based outcomes. The combination with minoxidil (which works well at the frontal scalp) may help bridge this gap.

• Leyden J et al. (1999). Finasteride in the treatment of men with frontal male pattern hair loss. J Am Acad Dermatol PMID 10365924

Pooled analysis of multicenter, double-blind, placebo-controlled studies. Men aged 18-41 and 41-60 with vertex hair loss were randomized to finasteride 1 mg/day or placebo for up to 24 months. Standardized global photography was evaluated by a blinded expert panel — a more clinically meaningful endpoint than hair counts alone because it captures overall visual improvement rather than localized counts.

At 24 months, finasteride produced statistically significant hair growth versus placebo in all scalp regions (p<0.05). Among younger men (18-41), 66% showed improved vertex hair growth versus 7% on placebo. Among older men (41-60), 39% improved versus 4% on placebo. Slightly higher drug-related sexual adverse experiences were reported in the finasteride group regardless of age, consistent with prior trials.

Expert panel photography assessment is the clinical endpoint that matters most — it answers 'can you actually see the difference?' The answer is clearly yes for most patients. This study is particularly valuable because it stratifies by age, showing that while younger patients respond better (66% vs. 39% improvement rate), even men over 40 show meaningful visible benefit. The age-stratified data supports the clinical principle of starting treatment early: the earlier you intervene, the more follicles are still viable and capable of responding. But even for older patients, nearly 4 in 10 showing visible improvement on blinded expert review is a meaningful response rate for a daily pill with a well-characterized safety profile.

• Olsen EA et al. (2012). Global photographic assessment of men aged 18 to 60 years with male pattern hair loss receiving finasteride 1 mg or placebo. J Am Acad Dermatol PMID 22325459

Open-label observational study of 523 Japanese men with androgenetic alopecia treated with finasteride 1 mg/day for 10 years. Efficacy was assessed using the Norwood-Hamilton classification and modified global photographic assessment with a standardized 7-point rating score. Scalp photographs were evaluated at regular intervals throughout the study period.

At 10 years, 91.5% of patients showed improvement (photographic assessment score of 5 or higher) and 99.1% showed at least prevention of disease progression (score of 4 or higher). Mean Norwood-Hamilton classification improved by approximately 1 full grade, from 3.35 at baseline to 2.55 after 10 years of treatment. No serious adverse reactions were identified during the entire study period.

The 10-year duration is extraordinary and provides reassurance about long-term durability that no controlled trial can ethically offer (you cannot keep men on placebo for a decade in a progressive disease). The finding that 99% of men maintained at least their baseline after 10 years of a progressive disease is clinically meaningful — without treatment, virtually all would have continued losing hair. The caveats are standard for long-term open-label data: no placebo control, inherent survivorship bias (10-year completers are enriched for responders), and the Japanese study population may not fully generalize to other ethnicities. But these limitations are inherent to the study type, not design flaws. When you combine this 10-year data with the 5-year controlled data, the overall durability picture for finasteride is remarkably strong.

• Yanagisawa M et al. (2019). Long-term (10-year) efficacy of finasteride in 523 Japanese men with androgenetic alopecia. Clin Res Trials

Two evidence sources. (1) Phase III pivotal trial safety data from Kaufman 1998 (n=1,553): prospective, blinded adverse event collection with standardized reporting. (2) Mondaini 2007: 120 men with BPH randomized to finasteride 5 mg with or without counseling about sexual side effects — designed specifically to quantify the nocebo effect on reported sexual adverse events.

In the pivotal trials, drug-related sexual adverse events (decreased libido, erectile dysfunction, ejaculatory disorder) occurred in 3.8% of finasteride patients versus 2.1% on placebo — a treatment-attributable excess of ~1.7 percentage points. Most resolved with continued treatment or upon discontinuation. In the Mondaini nocebo study, men informed about sexual side effects reported erectile dysfunction at 30.9% versus 9.6% in uninformed men (p=0.03). Similar patterns for decreased libido (23.6% vs. 7.7%) and ejaculation disorders (16.3% vs. 5.7%).

The Phase III data from 1,553 blinded patients provides the most reliable estimate of true sexual side effect incidence: approximately 2-4% above placebo. This is real — finasteride does cause sexual side effects in a small minority of men — but it is also small, and most resolve with continued use or upon stopping the drug. The Mondaini nocebo data is critical for interpreting the gap between clinical trial rates (~4%) and the rates reported in online forums, post-marketing databases, and clinical practice (often 10-30%+). When patients are told to expect sexual side effects, they report them at 3x the rate of uninformed patients receiving the same drug. This does not mean side effects are imaginary — the 9.6% baseline rate in uninformed patients is still real — but it means that fear, expectation, and anxiety substantially amplify the experience. The clinical implication: informed consent is essential, but the framing matters. Catastrophizing the side effect profile may itself increase adverse outcomes through the nocebo effect.

• Kaufman KD et al. (1998). Finasteride in the treatment of men with androgenetic alopecia. J Am Acad Dermatol PMID 9777765
• Mondaini N et al. (2007). Finasteride 5 mg and sexual side effects: how many of these are related to a nocebo phenomenon?. J Sex Med PMID 17655657

Heterogeneous evidence base including: (1) Retrospective cohort studies — a 2017 study of 4,284 men found 0.8% persistent sexual symptoms after discontinuation; another found >1% persistent erectile dysfunction (lasting average 4.2 years) in men taking finasteride for 206+ days. (2) Case series and patient self-report registries. (3) FDA regulatory review (2022) examining whether existing evidence supports a causal link. (4) Phase III trial safety data showing resolution of side effects upon discontinuation in the majority of affected patients.

The evidence is genuinely contradictory. Multiple double-blind, placebo-controlled trials show no statistically significant excess of persistent sexual symptoms after drug discontinuation. However, retrospective cohort studies report persistent symptoms in 0.8-1.4% of users. The FDA reviewed the evidence in 2022 and concluded it 'does not provide reasonable evidence' of a causal link between finasteride and persistent sexual, depressive, or suicidal symptoms — though it has required label updates acknowledging reports. A confirmed biological mechanism for PFS has not been established, though neurosteroid depletion has been proposed.

This is the most contentious question in the finasteride literature, and intellectual honesty requires acknowledging the genuine uncertainty. The controlled trial data does not support PFS as a common or well-characterized phenomenon — in the pivotal trials, sexual side effects resolved in the majority of affected patients upon discontinuation. But the controlled trials were not designed to detect rare, persistent effects in susceptible individuals, and their follow-up after drug cessation was limited. Retrospective studies and patient registries describe real symptoms in a small minority, but retrospective data cannot establish causation, and self-selected patient populations (those who believe they are affected) introduce severe ascertainment bias. The FDA's refusal to confirm a causal link is not evidence of safety — it reflects the genuine inadequacy of existing data to answer the question either way. The honest position is: the risk of persistent effects after stopping finasteride, if it exists, is likely small (<1-2% based on the best available retrospective data), but it has not been definitively characterized or ruled out by prospective, adequately powered research. Patients deserve to know this uncertainty exists.

• Irwig MS (2012). Persistent sexual side effects of finasteride: could they be permanent?. J Sex Med PMID 22789024
• Kiguradze T et al. (2017). Persistent erectile dysfunction in men exposed to the 5α-reductase inhibitors, finasteride, or dutasteride. PeerJ PMID 28289563

Two key evidence sources. (1) Hu et al. 2015: Randomized comparative trial of 450 Chinese men assigned to finasteride alone, 5% minoxidil alone, or the combination for 12 months, with efficacy assessed by global photographic evaluation. (2) Zhu et al. 2020: Systematic review and meta-analysis of 5 RCTs evaluating combined finasteride plus topical minoxidil versus monotherapy for androgenetic alopecia.

The Zhu meta-analysis found that combination therapy produced significantly higher global photographic evaluation scores (p<0.00001) and significantly more patients with marked improvement (p<0.001) compared to either monotherapy. The combination achieved a mean increase of approximately 29.7 hairs/cm2 after 24 weeks — exceeding established clinically meaningful thresholds versus monotherapy alone. Adverse event rates were similar between combination and monotherapy groups, indicating no additive safety concern.

The synergy between finasteride and minoxidil is both biologically logical and clinically proven. Finasteride addresses the hormonal cause (blocking DHT-driven miniaturization) while minoxidil addresses the growth stimulus side (prolonging anagen phase, enhancing follicular blood flow via potassium channel opening). They work through entirely independent mechanisms, so combination benefits are expected and now confirmed by meta-analytic evidence. Multiple independent meta-analyses published between 2020 and 2025 converge on the same conclusion: the combination is superior to either drug alone, with no increase in adverse events. This has made the finasteride-minoxidil combination the de facto first-line recommendation in clinical practice for moderate-to-severe AGA. The effect is additive — the combination achieves hair density improvements that neither drug can match alone.

• Zhu Y et al. (2020). The Efficacy and Safety of Finasteride Combined with Topical Minoxidil for Androgenetic Alopecia: A Systematic Review and Meta-analysis. Skin Appendage Disord PMID 32166351
• Hu R et al. (2015). Combined treatment with oral finasteride and topical minoxidil in male androgenetic alopecia: a randomized and comparative study in Chinese patients. Dermatol Ther PMID 26031764

Multiple study types. (1) Animal teratogenicity studies required for FDA approval: pregnant animals exposed to finasteride during fetal development. (2) Two controlled clinical trials of finasteride 1 mg in premenopausal women with female pattern hair loss. (3) Observational studies of finasteride 2.5-5 mg in post-menopausal women. (4) Case reports and clinical reviews on use in post-menopausal women without hyperandrogenism.

Animal studies demonstrated that finasteride causes abnormalities of the external genitalia in male offspring when administered during critical stages of sexual differentiation (analogous to weeks 8-12 of human gestation). Two controlled clinical trials of finasteride 1 mg in premenopausal women showed no benefit over placebo. In post-menopausal women, higher doses (2.5-5 mg) showed benefit in observational studies — one study of 40 post-menopausal women found that 22 (55%) had significant improvement after 6 months — but controlled data remains limited.

The teratogenicity concern is unambiguous and well-established: finasteride blocks the DHT-mediated differentiation of male external genitalia during fetal development. Women of childbearing potential must not take finasteride and should not even handle crushed or broken tablets (the drug can be absorbed through skin). This is not a theoretical risk — it is a known, mechanistically understood consequence of 5-alpha reductase inhibition during a critical developmental window. For post-menopausal women, the picture is more nuanced. Finasteride 1 mg (the male AGA dose) does not appear to work in women, likely because female pattern hair loss has a different hormonal profile with less dependence on type II 5-alpha reductase. At higher doses (2.5-5 mg), some post-menopausal women respond, particularly those without clinical or laboratory signs of hyperandrogenism. But the evidence base is far weaker than for men — mostly observational studies with small samples and no long-term controlled data.

• Stout SM, Stumpf JL (2010). Finasteride treatment of hair loss in women. Ann Pharmacother PMID 20442354
• Suchonwanit P et al. (2020). Finasteride and Its Potential for the Treatment of Female Pattern Hair Loss: Evidence to Date. Drug Des Devel Ther PMID 32184564