A tweeted rebuttal to yet another analysis of finasteride adverse events

A thread originally posted on Twitter is reproduced below. For an in-depth version, see the essay Context matters. See also a critique of three previous papers with similar designs, findings and conclusions.

Twitter thread

Nguyen et al, 2022 is the fourth analysis of adverse events of finasteride to appear since 2018. All four studies play up external & coincidental factors, discounting a causal role of the drug itself.

This thread points out limitations and blind spots of the study. Main point: It treats all AEs as interchangeable — a data cloud abstracted from its sources. Despite an appearance of rigor and precision, the discarding of context raises questions and concerns.


  • The study overlooks the role of stigma in mediating AE reporting
  • ‘Stimulated reporting’ normally describes short-term effects. The proportion of finasteride reports with sexual AEs has been fairly stable for a decade.
  • Study omits limitations of dutasteride comparison
  • Overlooks age as mediating factor
  • Exhibits bias for status quo prior to 2012
  • Exhibits bias against agency and judgment of patients & the public

Sexual AEs are not like headache or fever; they are embarrassing & inappropriate to discuss. Awareness is needed to overcome this barrier. Also, patient and doctor awareness are needed to link a hair loss drug to sexual AEs.

Stimulated reporting effects in the literature are notable but short-lived. Studies have shown a return to pre-stimulus baseline within months. By contrast, the proportion of finasteride reports with sexual AEs has been fairly stable since 2012. (dois: 10.32872/cpe.v1i1.29642; 10.1016/j.jad.2006.04.026)

Return to baseline in about 5 mos.
Return to baseline in ~8–9 mos.
The 5-year moving average (green dotted line) shows that the proportion of cases with sexual AEs increased from 22% to 57% between 2006–2012, then remained fairly stable from 2013–2021.

Differences of dutasteride & finasteride limit validity of AE comparison. Mechanism & pharmacokinetics different (inhibition of 5-AR types 1 & 2 vs. primarily type 2; half-life: 5 weeks vs. 5–6 hours). Dutasteride is only approved for hair loss in a handful of countries (S. Korea & Japan?).

Volume of dutasteride AE reports is thin. Rough estimate of ~15 dutasteride reports/year for men age 18–44 in VigiBase.

Impact of age on AE reporting: DHT inhibition could differentially affect old vs. young men. Baseline sexual function differs in young vs. old. Sexual AEs may be weighted differently by young vs. old. Use for hair loss or prostate conditions may affect weighting of sexual AEs.

Nguyen et al. identify ‘stimulation’ after 2012. Without public awareness, sexual AEs may be drastically underreported. Post-2012 change might not reflect a reporting bias, but rather mitigates prior underreporting.

The argument for ‘stimulated reporting’ is consistent with a view that patients are passive, uninformed and unreliable; that the public cannot be trusted with greater awareness. Any response to such awareness is not to be taken seriously.

It is in the tradition of ‘doctor knows best,’ where physicians assert control over what counts as real in health matters.

In summary: Nguyen et al, 2021 plays up external & coincidental factors to discount sexual AE reporting trends. Statistical analysis lends an appearance of scientific authority, but omits social context and counterarguments which could limit findings & interpretation.

Paper effectively endorses status quo. Asserts authority of physician perspective over that of patients regarding their own health.

An alternative explanation: finasteride is hazardous to sexual function, especially in young men. Public awareness of this hazard has steadily increased over past two decades, despite efforts of Merck and physicians to downplay it.

Changes in reporting over time might not reflect bias, but rather expose medicine’s Achilles heel: intolerance of information that challenges its authority. /end

View the thread online: