Rebuttals to denialists

There’s a cottage industry of physicians casting doubt on 5-alpha syndrome. Their arguments commit logical errors, misread literature and betray a physician-centric attitude.

The harms of finasteride have long been obfuscated by industry and physicians. Merck hid unfavorable outcomes in clinical trials1 as well as communications with the Swedish drugs regulator in the 2000s.2

In the 2010s, new warnings were added to the U.S. drug label and Merck faced litigation over alleged harms of finasteride. Physicians then took to the literature to defend the drug’s safety.3 Their arguments overlooked social context, misread literature, omitted inconvenient facts, ignored industry influence, disparaged patients, and made empty calls for more research. Rebuttals to these arguments—which are still in circulation—are provided below.

See also: ‘Safe and well-tolerated’: Arguments used to discredit reports of persistent adverse effects

Quick reference

Doubting physician: “Finasteride is safe based on trials and experience”

Rebuttal: The quality of Merck’s safety reporting is questionable and difficult to assess. Social taboos and embarrassment prevent reporting of sexual concerns. Jump to details

“Patients are experiencing a nocebo effect”

No evidence for a post-drug nocebo effect has been found. Jump to details

“Patients are influenced by news, social media, advocacy & more”

Post-finasteride cases were observed and reported in the 1990s and early 2000s, before risks were publicized. Jump to details

“Patients have pre-existing conditions”

Cases without pre-existing conditions have been described, including one in Merck’s clinical trial and a postmarketing case disclosed by Merck. Jump to details

“Patients have depression or another mental illness”

Genital abnormalities and numbness are not symptoms of common psychiatric disorders. Jump to details

“Patients lack expertise to report adverse events accurately”

Symptoms reported by patients and health care providers are similar, although psychological symptoms are more common in reports from health care providers. Research has shown health care providers discount or overlook certain adverse events. Jump to details

“Studies on 5-alpha syndrome are of low quality”

Studies deemed “high-quality” have industry bias. Other findings are preliminary because the condition is variable, funding is scarce, and the research community is small. Jump to details

“There is no causal link between finasteride and an alleged syndrome”

A causal link has not been adequately investigated. Jump to details

“Controlled trials are needed”

Physicians who call for a trial seem to have no intention of carrying it out. Jump to details

“The disease is ill-defined”

The condition is not well-defined because it has not been adequately studied. Many diseases have previously been “ill-defined,” for example bacterial infections. Jump to details

“The disease lacks biological plausibility”

The disease has not been adequately studied. The burden should be on researchers to explore a biological basis rather than expecting the disease to match expectations. Jump to details

“There is a history of controversy surrounding the disease”

Many diseases have been “controversial,” such as long Covid. The presence of disagreement neither supports nor refutes a disease’s validity. Jump to details

“I’ve never seen it in my practice”

The syndrome develops after discontinuing the drug and involves embarrassing symptoms, which may prevent patients from reporting back to the doctor. Moreover, physicians may have received genuine reports, but denied a link to finasteride. Jump to details

Detailed responses

“The drug is safe and well-tolerated per clinical trials and clinician experience”

Rebuttal: In Merck’s Phase III extension trial, there was a case of post-finasteride sexual dysfunction which Merck did not disclose to FDA. By 2006, the drug maker had received at least 100 reports of this outcome. Adverse outcomes in other trials were concealed, for example the PLESS trial where the finasteride group had double the rate of unrecovered sexual dysfunction compared to controls.

The Phase III trial excluded patients with pre-existing conditions that could confound results, so the above-mentioned case of persistent sexual dysfunction cannot be explained as a pre-existing condition.4

Online polls have many limitations, but their anonymity is an advantage. In an analysis which pooled the results of 7 informal polls, the rate of side effects was nearly 6 times higher than the official rate.

Clinicians may not hear about post-finasteride experiences because it develops after treatment ends. Patients may not attribute adverse experiences to the drug at all. Even if they do, they may be shy about discussing use of a medication for hair loss, and unwilling to broach sexual dysfunction and psychiatric concerns with a doctor—perhaps especially a dermatologist who does not specialize in these areas. Moreover, the dermatologist may signal that he or she is not open to discussion of side effects.

“Patients are experiencing a nocebo effect”

Rebuttal: Many authors have suggested post-finasteride syndrome is a nocebo effect, a phenomenon where side effects arise from negative expectations. They invariably cite a 2007 study in which some patients were advised in advance that finasteride might have sexual side effects, while others were not. The group which received advance notice reported side effects at a higher rate.5 In fact, this study is not relevant to 5-alpha syndrome because the reports occurred while patients were using the drug. In a literature search, no evidence for a post-drug nocebo effect could be found, regarding finasteride or any other drug. The explanation of 5-alpha syndrome as a nocebo effect appears to have no basis in evidence.

Even if there were evidence for a post-drug nocebo effect, it would not suffice for certain cases. For example, the patient with post-finasteride sexual dysfunction after leaving Merck’s extension trial would not have had access to websites, media coverage or warnings that would alert him to this possibility. A self-report of post-finasteride sexual dysfunction was posted in the year 2000. Merck received 100 reports of unrecovered sexual dysfunction after stopping finasteride in the early 2000s, when public information about risks was scarce.

“Media coverage, internet forums, an advocacy group and litigation have led patients to report adverse events”

Rebuttal: An uptick of reports over a period of years raises eyebrows among physicians.6 While this is interpreted as an artificial effect, there is no consideration of what might have suppressed reports before that point. For example, taboos and embarrassment over male sexual dysfunction may have prevented expression of these concerns. Only with increased forum activity, advisories, news and the like would men attribute the drug to their sexual dysfunction and feel emboldened to report it. In this scenario, awareness does not generate artificial reports, but unlocks experiences that were previously masked by taboos. With publicity, more physicians might also correctly attribute patient reports to finasteride, whereas before they might have denied any link.

Moreover, the stimulated reporting argument could not explain the case of post-finasteride sexual dysfunction in Merck’s extension trial, other cases in its trials, nor postmarketing reports Merck received in the early 2000s. The first known online report of persistent sexual dysfunction linked to finasteride dates from September 2000. A Yahoo group called Propecia Side Effects was founded in 2003.

“Patients have pre-existing psychological and sexual disorders”

Rebuttal: Countering this argument, a board-certified urologist and sexual medicine specialist tweeted: “I have personally seen in my practice otherwise completely healthy young men who come in for ED and the only medication they had taken was Finasteride for hair loss.”

The case in Merck’s Phase III extension trial also casts doubt on this argument, because the company excluded men with “[a] history of any illness or condition that might have confounded the results of the study or posed additional risk.”7

A case Merck disclosed to the Swedish drugs regulator in 2006 cannot be dismissed as a pre-existing condition (emphasis added):

[A] 22 year old male emotionally healthy male with no prior history of sexual dysfunction was placed on therapy with finasteride for the treatment of early male pattern balding. Three to four months later the patient began to experience complete loss of sexual drive, including loss of spontaneous erections… Therapy with finasteride was discontinued. Eight months later, the patient reported he continued to experience the same symptoms with no sign of any spontaneous resolution.”8

“Patients are experiencing depression, anxiety or another mental illness”

Rebuttal: Symptoms of depression, anxiety, panic or suicidality do not rule out the presence of another illness. For example, non-psychiatric conditions such as chronic pain, migraines and stroke have an association with depression, but most physicians do not claim that they are merely depression.

Most physicians making these arguments do not have expertise in psychiatry, and have not sought a psychiatric assessment of patients reporting 5-alpha syndrome.

Finally, a common symptom of 5-alpha syndrome is genital numbness, which is not a symptom of depression, anxiety or panic disorder.

“Patients lack expertise in reporting their adverse events”

Rebuttal: Turning this objection around, physician reporting of adverse events has its own limitations, including underreporting,9,10 underestimation of severity and limited scope. Dr. Ethan Basch wrote (emphasis added):

Current drug-labeling practice for adverse events is based on the implicit assumption that an accurate portrait of patients’ subjective experiences can be provided by clinicians’ documentation alone. Yet a substantial body of evidence contradicts this assumption, showing that clinicians systematically downgrade the severity of patients’ symptoms, that patients’ self-reports frequently capture side effects that clinicians miss, and that clinicians’ failure to note these symptoms results in the occurrence of preventable adverse events.11

In a commentary entitled “Should we trust patient-reported outcomes?” Dr. Marie-Christine Nizzi wrote (emphasis added):

The expertise bias holds that physicians are best equipped to speak to patients’ quality of life because their medical knowledge grants them the relevant kind of expertise, which the patients lack. This one-sided view of what constitutes relevant expertise belongs to an outdated model of medicine. At its best, the healthcare relationship is not a tutelage but a partnership, where the physician’s expertise in biological processes is guided by the patient’s expertise in their own thoughts, feelings, and sensations.12

In an analysis of adverse events of finasteride, patient reports included more sexual adverse events while healthcare providers included more psychiatric adverse events. This reflects the deep divide between physicians and patients: patients have a direct experience of sexual adverse events while physicians are prone to interpret the patient’s report in psychological terms. Since sexuality arises from thoughts, feelings and sensations, patients would be a more reliable source than the physician who stands outside the patient’s experience.

“Studies are of low quality”

Rebuttal: Merck reportedly spent $450 million on the development of Propecia.13 Clinical trials spawned dozens of publications, giving the impression of a massive and incontrovertible evidence base. Yet the safety methodology has been questioned,14 and adverse event rates may have been understated.

Meanwhile, funding to support research on drug harms is scarce. Some early research on 5-alpha syndrome was funded by the Post-Finasteride Syndrome Foundation, with contributions from patients and their relatives. Recruiting for studies is challenging because the disease carries stigma. There is heterogeneity in the patient population—for example, time since stopping the drug and symptom profile.

Qualitative, low-budget studies are not necessarily “low quality,” but different from an industry-funded trial. They should be judged based on the context, not on a rigid hierarchy of evidence where studies costing millions of dollars overrule all other forms of evidence. It is entirely expected that early studies on a disease would be descriptive and exploratory. This is why case reports are published in journals.

Rebuttal: Dermatologists and regulators have not undertaken the research needed to assess a causal link. A prospective trial could take a decentralized approach which relies on mobile devices. Another option is a systematic review of evidence. Khan et al noted six criteria for establishing causation in systematic reviews: strength of association, consistency, temporality, specificity, biological gradient, plausibility, coherence and analogy, and experimental evidence.15

Even without such studies, a wide range of evidence is available to assess causation: cases Merck did not disclose until a judge ordered the release of documents in 2021; obfuscated cases from clinical trials; more than 20 years of adverse event data; animal studies; case reports and series; physician perspectives; and media reports going back to 2010. An assessment should account for the influence of industry and physician interests on the medical literature, especially since permanent harms of a medication are at issue.16

“A randomized controlled trial should be conducted”

Rebuttal: It is common to call for prospective trials in conclusions. Gray and Semla wrote: “We need placebo controlled trials using validated questionnaires and long term follow-up after treatment to examine persistence of symptoms.”17 But the researchers who make these calls do not seem to have any intention to carry out those trials or raise funds to support such studies. Leliefeld et al made a realistic observation that a randomized, prospective trial is “unlikely to happen due to a lack of funding.”18

Given the lack of funding, researchers could suggest other types of studies such as a decentralized trial, perhaps in partnership with telehealth companies or health systems.

“The proposed disease ‘post-finasteride syndrome’ is ill-defined”

Rebuttal: Any problem worth studying is not well-defined (for example, cosmology or long Covid). Once problems are well-defined, they can be put in an encyclopedia and investigators can move on to other problems which are not well-defined.

Post-finasteride syndrome has not been adequately studied. The onus should be on researchers to study it, rather than for the disease to define itself. This is a hard problem. The speaker implies she dislikes that it’s hard, and questions its validity rather than investigating.

“No mechanism for permanent adverse effects can be conceived. The disease lacks biological plausibility”

Rebuttal: The disease has not been adequately studied. The burden should be on researchers to explore a biological basis rather than expecting the disease to meet their prior expectations.

Meanwhile, it turns out that the most common symptoms and dysfunctions of 5-alpha syndrome rely on organs and tissues dependent on steroid signaling, for example the reproductive system, hypothalamus and eyes. Finasteride interferes with these pathways. There is evidence of penile damage in men with the syndrome, while dihydrotestosterone (DHT), the androgen suppressed by finasteride, is critical for early development of the male genitals. It is entirely plausible that DHT may be involved in maintenance of penile tissue in adult men. Moreover, low DHT is associated with inflammation.

It is easier to express doubt than to acknowledge a lack of understanding, and then investigate.

“The syndrome has been controversial”

Rebuttal: The mere presence of controversy does not support or rule out any position. The origins of Covid, climate change and the germ theory of disease have all been controversial. This means there was disagreement. This argument amounts to a defense of the status quo, and perhaps also interests which benefit from it.

“I’ve never seen it in my practice”

Rebuttal: The physician assumes her knowledge and experience are authoritative, while overlooking factors that might prevent reporting 5-alpha syndrome back to the physician. It develops after discontinuation of the drug. It involves embarrassing symptoms that violate social taboos. The affected individual may not attribute his problems to the drug. He may feel it is inappropriate to report sexual or psychiatric problems to a dermatologist. Or, the physician may hear a genuine report but deny a link to finasteride. In this case, she has seen it in her practice, but denied its validity. Finally, physicians may resist acknowledging treatment-related harms because of concerns about liability.

  1. See The lost men, and Since 1994, Merck has been aware of unresolved sexual dysfunction… ↩︎
  2. Merck’s response to the Swedish drugs regulator. ↩︎
  3. See: Responding to a literature of doubt, ‘Safe and well-tolerated’, Response to Dr. Ralph Trüeb and Temperature check. ↩︎
  4. Merck Research Laboratories. Drug Approval Package: Propecia NDA #020788; Trial 087. Drugs@FDA. December 19, 1997. ↩︎
  5. Mondaini N, Gontero P, Giubilei G, et al. Finasteride 5 mg and sexual side effects: how many of these are related to a nocebo phenomenon? J Sex Med. 2007. doi:10.1111/j.1743-6109.2007.00563.x ↩︎
  6. Harrell MB, Ho K, Te AE, Kaplan SA, Chughtai B. An evaluation of the federal adverse events reporting system data on adverse effects of 5-alpha reductase inhibitors. World J Urol. 2020. doi:10.1007/s00345-020-03314-9 ↩︎
  7. Merck Research Laboratories. Drug Approval Package: Propecia NDA #020788; Trial 087. Drugs@FDA. December 19, 1997. ↩︎
  8. Merck Research Laboratories Worldwide Product Safety. Periodic Safety Report for: Finasteride, 1 mg tablet and 0.2 mg tablet, MSD. November 30, 2006. View at ↩︎
  9. Hazell L, Shakir SA. Under-reporting of adverse drug reactions: a systematic review. Drug Saf. 2006. doi:10.2165/00002018-200629050-00003 ↩︎
  10. Lopez-Gonzalez E, Herdeiro MT, Figueiras A. Determinants of under-reporting of adverse drug reactions: a systematic review. Drug Saf. 2009. doi:10.2165/00002018-200932010-00002 ↩︎
  11. Basch E. The missing voice of patients in drug-safety reporting. N Engl J Med. 2010. doi:10.1056/NEJMp0911494 ↩︎
  12. Nizzi MC. Should we trust patient-reported outcomes? AJOB Neurosci. 2021 Apr-Sep. doi:10.1080/21507740.2021.1904040 ↩︎
  13. Morrow DJ. New Baldness Drug Is Older Product at a Premium PriceNew York Times. January 20, 1998. ↩︎
  14. Belknap SM, Aslam I, Kiguradze T, et al. Adverse event reporting in clinical trials of finasteride for androgenic alopecia: a meta-analysis. JAMA Dermatol. 2015. doi:10.1001/jamadermatol.2015.36 ↩︎
  15. Khan KS, Ball E, Fox CE, et al
    Systematic reviews to evaluate causation: an overview of methods and application
    BMJ Evidence-Based Medicine 2012. doi:10.1136/ebmed-2011-100287 ↩︎
  16. Abramson J. Sickening: How Big Pharma Broke American Health Care and How We Can Repair It. Mariner Books; 2022. Amazon ↩︎
  17. Gray SL, Semla TP. Post-finasteride syndrome. BMJ. 2019. doi:10.1136/bmj.l5047 ↩︎
  18. Leliefeld HHJ, Debruyne FMJ & Reisman Y. The post-finasteride syndrome: possible etiological mechanisms and symptoms. Int J Impot Res. 2023. doi:10.1038/s41443-023-00759-5 ↩︎