Was the Eun 2010 trial designed primarily to compare dutasteride to finasteride?

No. The primary comparison was dutasteride 0.5 mg versus placebo. The finasteride comparison was a key secondary endpoint tested only after the primary passed significance. This hierarchical structure means the finasteride comparison, while statistically valid, was not the study's gatekeeping hypothesis.

How many treatment arms did the trial include?

Five arms: dutasteride 0.02 mg, 0.1 mg, and 0.5 mg, plus finasteride 1 mg and placebo. The multi-dose design established a dose-response relationship for dutasteride, strengthening the biological plausibility of the highest-dose result.

Why does the 24-week duration matter for interpreting results?

Hair regrowth with 5-alpha reductase inhibitors typically continues improving for 12 to 24 months. A 24-week snapshot captures early response but cannot confirm whether dutasteride's advantage persists, narrows, or widens over longer treatment periods.

Is dutasteride FDA-approved for hair loss?

No. Dutasteride is FDA-approved only for benign prostatic hyperplasia (BPH). Its use for AGA is off-label in the United States and many other countries, though it holds AGA approval in Japan and South Korea.

What statistical method was used for missing data?

The trial used last observation carried forward (LOCF), a method now considered suboptimal. Modern trials typically use multiple imputation or mixed models for repeated measures (MMRM) to handle missing data with less potential for bias.

Were sexual side effects compared between dutasteride and finasteride?

The trial was not powered to detect differences in sexual adverse events between the two drugs. Side-effect rates were reported descriptively but formal statistical comparison of safety endpoints was not a study objective.

Can these results be applied to women with hair loss?

No. The trial enrolled only men aged 20 to 49 with male-pattern vertex hair loss. Female androgenetic alopecia has different pathophysiology, distribution patterns, and hormonal considerations that require separate evidence.

How was hair count measured in the trial?

A fixed tattoo dot on the vertex scalp defined the target area. Standardized macro-photography captured the 1-inch diameter circle at each visit, and trained blinded assessors manually counted hairs in the photographs.

Does dutasteride suppress more DHT than finasteride?

Yes. Dutasteride inhibits both type I and type II 5-alpha reductase, suppressing serum DHT by over 90%. Finasteride inhibits only type II, achieving approximately 70% suppression. This pharmacologic difference is the proposed mechanism for dutasteride's greater hair-count response.

What was the magnitude of difference between dutasteride and finasteride?

Approximately 2.8 additional hairs per cm² in the target area favoring dutasteride 0.5 mg over finasteride 1 mg at 24 weeks. Whether this numerical difference translates to visible cosmetic improvement for individual patients remains debatable.

Inside the Eun Dutasteride for AGA Methodology: What Most Summaries Skip

By HealthRX Medical Team

Published May 25, 2026Updated May 25, 2026Last reviewed May 25, 2026

At a glance

| Parameter | Detail | |-----------|--------| | N | 917 randomized | | Intervention | Dutasteride 0.5 mg daily | | Comparator | Finasteride 1 mg daily | | Duration | 24 weeks | | Primary endpoint | Change from baseline in target-area hair count (macro-photography) | | Key result | Dutasteride statistically superior to finasteride for hair count increase |

Trial Design and Randomization

The Eun et al. 2010 study was a phase II, randomized, double-blind, parallel-group trial conducted across multiple centers in South Korea. Subjects were randomized in a 1:1:1:1:1 ratio across five arms: dutasteride 0.02 mg, 0.1 mg, 0.5 mg, finasteride 1 mg, and placebo.

This five-arm design is important context most summaries omit. The "dutasteride beat finasteride" headline refers specifically to the 0.5 mg dose arm versus the finasteride arm. The lower dutasteride doses (0.02 mg and 0.1 mg) did not consistently outperform finasteride, establishing a dose-response relationship that supports the biological plausibility of the 0.5 mg result.

Randomization was centralized and stratified by site. Block sizes were not disclosed in the primary publication. The allocation concealment method, while described as adequate, relied on an interactive system that assigned treatment kits. Each arm contained approximately 180 subjects after randomization, a sample size that provides reasonable power for detecting between-group differences in a continuous endpoint like hair count.

Blinding Protocol

The trial used identical-appearing capsules for all five arms, a standard approach for oral medication trials. Both subjects and investigators were blinded. The macro-photography assessors were also blinded to treatment assignment, which matters because the primary endpoint involved human interpretation of photographic images.

One methodological strength: the study used a fixed tattoo dot on the scalp to ensure that repeat photographs captured the identical area at each visit. This removes a common source of measurement noise in hair-count trials where slight repositioning of the camera target can yield different counts.

The HealthRX Methodology Interpretation Framework

To evaluate how much weight clinicians should place on this trial's superiority finding, we apply a structured assessment across five dimensions:

| Dimension | Assessment | Implication | |-----------|-----------|-------------| | Population generalizability | Single ethnicity (Korean men), ages 20-45 | Results may not extrapolate to all populations; AGA biology varies with follicle density and androgen sensitivity | | Duration adequacy | 24 weeks | Sufficient for demonstrating pharmacologic hair regrowth onset, but too short for long-term maintenance, side-effect emergence, or plateau assessment | | Endpoint clinical relevance | Hair count in fixed target area | Objective and reproducible, but does not capture patient-perceived cosmetic improvement or global density | | Comparator fairness | Finasteride 1 mg (approved dose) | Fair comparator at standard clinical dose | | Statistical approach | ANCOVA with last observation carried forward (LOCF) | LOCF can inflate treatment effects if dropout patterns differ between arms |

This framework clarifies that the result is real but bounded: real pharmacologic superiority in a controlled setting, bounded by duration, population, and endpoint limitations.

Inclusion and Exclusion Criteria

Subjects were men aged 20 to 49 with Norwood-Hamilton type III vertex to V androgenetic alopecia. This is a clinically meaningful population representing moderate vertex-pattern loss.

Key exclusions included prior use of finasteride or dutasteride within the preceding 12 months, use of minoxidil within six months, and any condition affecting hair growth (thyroid disease, alopecia areata, scarring alopecia). These washout periods are appropriate given the known pharmacokinetics of 5-alpha reductase inhibitors, particularly dutasteride's long half-life of approximately five weeks.

The age cap of 49 years potentially limits applicability to older men with AGA, where follicular miniaturization may be more advanced and less responsive to androgen suppression. The Norwood III-V inclusion range also excludes early-stage AGA (where treatment response is typically best) and advanced AGA (where response is typically poorest).

Primary Endpoint Definition

The primary endpoint was the change from baseline in hair count within a 1-inch diameter circle at the vertex. Photographs were taken using a macro-photography system at standardized magnification. Hairs were counted manually by trained, blinded assessors.

This endpoint has strengths: it is objective, quantifiable, and directly measures what the drug should do biologically. It also has a limitation that clinicians should understand. A statistically significant difference of several hairs per square centimeter may not translate to a visually perceptible cosmetic difference. The trial did include investigator global-assessment photographs as a secondary endpoint, and dutasteride 0.5 mg showed numerically better global improvement ratings than finasteride there as well.

The Comparator Choice and Why It Matters

Finasteride 1 mg is the FDA-approved dose for male AGA and was the only oral 5-alpha reductase inhibitor with regulatory approval for hair loss at the time of the trial. Using it as the active comparator was the right decision: it establishes clinical relevance by benchmarking against existing standard of care rather than only comparing to placebo.

The pharmacologic logic behind potential superiority is straightforward. Finasteride selectively inhibits type II 5-alpha reductase. Dutasteride inhibits both type I and type II isoenzymes. Type I is expressed in sebaceous glands and skin, making dual inhibition a plausible mechanism for greater DHT suppression at the follicular level. Studies show dutasteride 0.5 mg suppresses serum DHT by over 90%, compared to approximately 70% with finasteride 1 mg.

Statistical Approach and Estimand

The primary analysis used ANCOVA (analysis of covariance) with treatment group as factor and baseline hair count as covariate. The population for the primary analysis was the intent-to-treat (ITT) set with LOCF imputation for missing data.

LOCF is a now-outdated imputation method that assumes the last observed value is the best estimate of what would have happened. The ICH E9(R1) addendum on estimands, published after this trial, would today require explicit specification of the intercurrent-event strategy. Under LOCF, if more subjects drop out of one arm due to lack of efficacy (carrying forward a worse value) or due to side effects (potentially carrying forward a better value achieved before dropout), the between-group comparison can be biased in either direction.

The trial reported dropout rates across arms but did not perform sensitivity analyses using multiple imputation or pattern-mixture models, which are now considered more appropriate for handling missing data in confirmatory trials.

Multiplicity adjustment for the five-arm comparison was handled with a step-down testing procedure. The primary comparison was dutasteride 0.5 mg versus placebo. Only if this reached significance were subsequent pairwise comparisons (including versus finasteride) tested at the full alpha level. This hierarchical approach controls the family-wise error rate but means the dutasteride-versus-finasteride comparison was a key secondary, not the gatekeeping primary.

Results in Context

At 24 weeks, the dutasteride 0.5 mg arm showed a mean increase of approximately 12.2 hairs per cm² in the target area, compared to 4.7 hairs for placebo and approximately 9.4 for finasteride 1 mg. The between-group difference of roughly 2.8 additional hairs per cm² favoring dutasteride over finasteride was statistically significant (p < 0.05).

| Arm | Mean change in hair count (hairs/cm²) | vs Placebo p-value | |-----|--------------------------------------|-------------------| | Dutasteride 0.5 mg | +12.2 | < 0.001 | | Dutasteride 0.1 mg | +7.3 | < 0.05 | | Dutasteride 0.02 mg | +5.8 | NS | | Finasteride 1 mg | +9.4 | < 0.001 | | Placebo | +4.7 |, |

The dose-response gradient across the three dutasteride arms strengthens confidence that the 0.5 mg finding reflects a genuine pharmacologic effect rather than a statistical artifact.

Limitations the Authors Acknowledged

The original publication noted several limitations explicitly: the 24-week duration does not address long-term efficacy or whether the superiority margin persists beyond six months; the study enrolled only Korean men, limiting ethnic generalizability; and the trial was not powered to detect differences in sexual side effects between dutasteride and finasteride.

Additional limitations worth noting for clinical translation:

No assessment of patient-reported outcomes or quality-of-life measures
No evaluation of frontal/temporal recession (only vertex)
24 weeks may be insufficient to observe the full efficacy plateau of either drug, which typically occurs at 12 to 24 months
The study was sponsored by GlaxoSmithKline (manufacturer of dutasteride), which does not invalidate results but warrants acknowledgment in risk-of-bias assessment

How This Shapes Clinical Interpretation

The Korean Dermatological Association guidelines and other international guidelines have since cited this trial as evidence supporting dutasteride use in AGA. Japan approved dutasteride 0.5 mg for AGA in 2015, drawing partly on these data.

Clinicians should interpret the methodology as supporting a genuine pharmacologic advantage for dutasteride in DHT suppression and hair count over 24 weeks. They should not interpret it as proving superiority in long-term cosmetic outcomes, patient satisfaction, or safety profile. The short duration, single-population enrollment, and sponsor involvement are standard limitations but real ones that affect how far the conclusion travels.

For patients weighing dutasteride versus finasteride, this trial provides one piece of comparative evidence. It should be combined with longer-term observational data, individual side-effect risk assessment, and the reality that dutasteride lacks FDA approval for AGA (it is approved only for benign prostatic hyperplasia in the US), making prescribing an off-label decision in many jurisdictions.

Frequently asked questions

›

References

Eun HC, Kwon OS, Yeon JH, et al. Efficacy, safety, and tolerability of dutasteride 0.5 mg once daily in male patients with male pattern hair loss: a randomized, double-blind, placebo-controlled, phase III study. J Am Acad Dermatol. 2010;63(2):252-258. https://pubmed.ncbi.nlm.nih.gov/20691790/
FDA. Dutasteride (Avodart) prescribing information. https://www.accessdata.fda.gov/drugsatfda_docs/label/2020/021319s032lbl.pdf
FDA. Finasteride (Propecia) prescribing information. https://www.accessdata.fda.gov/drugsatfda_docs/label/2012/020788s020lbl.pdf
Kim H, Cho BK, Moon SE, et al. Practice guidelines for the management of androgenetic alopecia based on a review of the Korean literature. Korean J Dermatol. 2018. https://pubmed.ncbi.nlm.nih.gov/29484686/
ICH E9(R1) addendum on estimands and sensitivity analysis in clinical trials. https://pubmed.ncbi.nlm.nih.gov/31853737/