Hair-loss drugs work. The trials measured hairs per square centimetre, not a full head

Male-pattern hair loss has two drugs with decades of randomised evidence behind them, and both produce a measurable increase in hair density. The unit of that measurement is worth sitting with: in the pooled trial evidence, the best-performing treatments added somewhere between 12 and 20 hairs per square centimetre over 24 weeks against a control. That is a real effect on a progressive condition. It is not the restoration the category advertises.

What the network meta-analysis found

A Bayesian network meta-analysis searched PubMed and Scopus on 30 April 2025 for randomised trials reporting change in total hair density at 24 weeks in men with androgenetic alopecia [s1]. It identified eight conventional monotherapies and seven non-conventional over-the-counter products — ketoconazole 2% topical, a marine complex, procyanidin 0.7% topical, rosemary oil, melatonin, saw palmetto and watercress 2% [s1].

A discrepancy in the source: the abstract reports 24 eligible trials, the results section states that 25 studies were used, and the study-characteristics table lists 24. The paper does not reconcile the counts.

The estimates against control, drawn from the paper's node-splitting analysis, were 20.0 hairs/cm² for topical minoxidil 5% (95% credible interval 3.7 to 41.0), 19.0 for oral dutasteride 0.5 mg (9.5 to 28.0), 15.0 for topical finasteride 0.25% (0.85 to 30.0) and 12.0 for oral finasteride 1 mg (4.20 to 19.0) [s1]. On the overall ranking metric, oral dutasteride 0.5 mg came first, with a surface-under-the-cumulative-ranking value of 95.8% in the base analysis and 94% in a severity-adjusted sensitivity analysis [s1].

Read those credible intervals rather than the point estimates. Topical minoxidil's runs from 3.7 to 41.0 — a range spanning "barely detectable" to "substantial" — and topical finasteride's lower bound sits at 0.85, effectively at no effect.

The intervals also overlap almost completely, which is the finding most likely to be lost in translation. This analysis does not establish that minoxidil beats finasteride, or that dutasteride beats either. It establishes that all four separate from control and that ranking them against each other is beyond what the data supports.

The trials are old, and the drugs are older

The regulatory history explains why the evidence looks the way it does. Topical minoxidil 2% solution was approved as a prescription product for male androgenetic alopecia in 1988, moved to over-the-counter status in 1996, with the 5% solution following over the counter in 1997 and the 5% foam approved in 2006 [s1]. Oral finasteride 1 mg was approved in 1997 [s1]. The pivotal minoxidil trials in the network date from 1986 to 1988; the largest finasteride trial, enrolling 779 men on treatment against 774 controls, is from 1998 [s1].

Roughly 50 million men in the United States have androgenetic alopecia, and the market for hair loss therapies was valued at $7.6 billion with a projected value of $13 billion by 2028 [s1]. Almost none of that spending is on the compounds with the strongest evidence.

One genuinely useful mechanistic finding sits in the same paper. Minoxidil is a pro-drug that must be converted to minoxidil sulfate by the enzyme sulfotransferase 1A1, and people with low enzyme activity are more likely to be poor responders. In a study cited by the authors, 75% of those using a SULT1A1 adjuvant alongside daily 5% topical minoxidil for 60 days experienced hair regrowth against 33% on minoxidil plus a placebo adjuvant (p = 0.023) [s1]. Trials of topical minoxidil almost never measure the enzyme, which means the pooled estimate is an average across responders and non-responders rather than an estimate of the effect in anyone.

The adverse-effect literature, reported carefully

Finasteride's safety record is the most contested question in this field, and it deserves to be reported without either dismissal or amplification.

The regulators have moved. The FDA, the European Medicines Agency, the UK Medicines and Healthcare products Regulatory Agency and Health Canada have all issued alerts regarding possible associations between oral and topical finasteride and psychiatric adverse events and sexual dysfunction, in some cases persisting after the product is discontinued [s1].

The pooled trial evidence finds a real but modest signal. A meta-analysis of 15 randomised, double-blind, placebo-controlled trials covering 4,495 men found that 5-alpha reductase inhibitors carried a 1.57-fold risk of sexual dysfunction (95% confidence interval 1.19 to 2.08). For finasteride specifically the relative risk was 1.66 (1.20 to 2.30); for dutasteride it was 1.37 (0.81 to 2.32), which is not statistically significant, on a smaller body of trials [s2].

The reason that number should not be treated as settled comes from a separate analysis of how the trials reported harms at all. Across 34 clinical trials of finasteride for androgenetic alopecia, none had adequate safety reporting: 19 were partially adequate, 12 were inadequate, and three reported no adverse events whatsoever [s3]. Funnel plots were asymmetric, biased toward lower odds ratios for sexual adverse effects, which the authors read as evidence of systematic underdetection [s3]. No report assessed the adequacy of blinding. Just over half — 18 of 34 — disclosed conflicts of interest, and 19 (56%) received funding from the manufacturer [s3]. Safety evaluation lasted a year or less in 26 of the 34 trials [s3].

The generalisability finding in that paper is the one most relevant to a man considering the drug. Of 5,704 men in a clinical data repository prescribed finasteride at 1.25 mg a day or less for hair loss, only 31% would have met the inclusion criteria for the pivotal trials cited in the manufacturer's prescribing information, and only 33% took it for longer than a year [s3]. The authors' conclusion is that published trial reports provide insufficient information to establish the drug's safety profile [s3].

That is a different claim from "finasteride causes persistent sexual dysfunction", and a different claim from "the concern is unfounded". It says the studies that would answer the question were not designed to answer it.

What the over-the-counter products showed

The same network placed rosemary oil, melatonin, saw palmetto, procyanidin, ketoconazole, watercress and a marine complex alongside the conventional drugs. The authors note that many of the trials evaluating these agents enrolled substantially fewer participants than the large conventional trials, and that the mechanisms of most of them are far less studied — saw palmetto may inhibit 5-alpha reductase, procyanidin may act on oxidative stress and inflammation, and rosemary oil is attributed anti-oxidative, anti-inflammatory and antifungal effects without a settled account of which matters [s1].

The paper's framing of why it exists is unusually candid: consumers of hair-loss products are often laypeople who cannot distinguish gimmick from evidence, increased transparency about conventional drug side effects has pushed people toward alternatives marketed as free of them, and the placebo effect reinforces the switch [s1].

This article is informational and is not medical advice. Decisions about any of these drugs, including their risks, belong with a clinician.

Sources

• [s1] Relative Efficacy of Conventional Monotherapies and Select Nonconventional, Over-the-Counter Products for Male Androgenetic Alopecia: A Network Meta-Analysis Study — Journal of Cosmetic Dermatology, October 2025. https://doi.org/10.1111/jocd.70483
• [s2] Adverse Sexual Effects of Treatment with Finasteride or Dutasteride for Male Androgenetic Alopecia: A Systematic Review and Meta-analysis — Acta Dermato-Venereologica, published online 12 September 2018. https://doi.org/10.2340/00015555-3035
• [s3] Adverse Event Reporting in Clinical Trials of Finasteride for Androgenic Alopecia: A Meta-analysis — JAMA Dermatology, published online 1 April 2015. https://doi.org/10.1001/jamadermatol.2015.36