Sex hormones signal why virus hits men harder

While I agree T may be an important factor in the high death rate of men, some recent evidence suggests the correlation is opposite that posed in the article: men with low testosterone are at increased risk for poor COVID-19 outcomes. The article's analysis is based on studies with significant shortcomings that imply testosterone worsens infection rates and/or outcomes, and the conclusions fail to hold up under scrutiny. Moreover, opposing studies that directly investigated testosterone levels in men with COVID-19 are not presented. Understanding the differences between the studies discussed in the article and relevant contradictory studies is key to having a fuller understanding of T's influence on COVID-19 patients.

First, the evidence about on role of testosterone-driven TMRPSS2 and ACE2 expression rely on select studies that don't fully support the presented argument. As stated in the article, prostatic expression of TMRPSS2 is known to be driven by androgens, but the evidence elsewhere in the body is mixed. Second, while ACE2 may be necessary for SARS-CoV-2 infection, it may also play a protective role on the lungs and cardiovascular system (1). In fact, a recent preprint has shown ACE2 expression inversely correlate with COVID-19 deaths in rats (2). Finally, in men testosterone levels generally drop with age. If testosterone-driven ACE2 and TMPRSS2 expression are the reason for the asymmetric outcomes in men, one should ask why are older men dying at substantially higher rates than younger men?

Next, the author's argument that men on androgen depravation therapy (ADT) have lower disease severity is not supported by the data. The results of this retrospective study on SARS-CoV-2 infection rates of men with prostate cancer on ADT were presented, showing men on ADT were less likely to contract COVID-19. Whereas the discussion on different infection rates is supported by a large n, differences in clinical outcomes between non-ADT and ADT patients is based on 114 non-ADT patients and only 4 ADT patients. One could say over two times as many of ADT patients were admitted to the ICU: 25% (1/4) of ADT vs. 11% (13/114) of non-ADT. The reported study's argument than men on ADT have lower infection rates is well supported, but it is premature to draw far-reaching conclusions about clinical outcomes based on only 4 cases in the ADT group.

Finally, the article discusses two small studies claiming more hospitalized men have male pattern baldness than would be predicted by population estimates. While the baldness angle is interesting, the argument falls apart in a few ways. First, the association between an individual's male pattern baldness, that may have started decades ago, and his current testosterone levels is tenuous. Second, by the authors' own reference for population statistics, balding men are not over-represented. The study found 79% of hospitalized men, median age 62.5, had male pattern baldness (Hamilton-Norwood Scale ≥ 2). Their cited article on prevalence shows 80% of men age 60-64 have HNS ≥ 2 (3). Finally, there was no discussion on patients' use of antiandrogenic drugs commonly used to treat male pattern baldness.

Most critically, the article did a disservice to Science readers by failing to report two studies that directly measured testosterone in men at the time of hospital admission. These studies suggest that low testosterone levels predict poor prognosis in men with COVID-19. First, from a study by Rastrelli and colleagues, the median T of men who required ventilation was 29 ng/dL (normal range ~300 to 1100 ng/dL), whereas those who were discharged from the ICU had median T of 254 ng/dL at the time of ICU admission (4). In a separate study by Schroeder and colleagues, 89% of SARS-CoV-2-positive men admitted to the ICU had low total or bioavailable testosterone (5). More than 80% of men who died in this study had low bioavailable T. In both studies there was a significant inverse correlation between T level and multiple markers of poor COVID-19 prognosis.

SARS-CoV-1 is known to damage the testis, which highly express ACE2 (6). By examining the research from the above two studies, it is reasonable to conclude that low testosterone, either pre-existing or caused by SARS-CoV-2 infection, is possibly a major factor for the mortality rate in men being twice that of women. We must carefully and separately consider what factors increase risk of SARS-CoV-2 infection and factors that put one at risk for severe COVID-19 and death. While we may not know until we have results from ongoing clinical trials, giving men with severely low testosterone ADT may exacerbate the disease, with potentially deadly results. Rather, replacing testosterone with the goal of achieving physiologic T levels should be evaluated in clinical trials of COVID-19, particularly in men who are found to have T deficiency.

1. T. Yan, R. Xiao, G. Lin, Angiotensin-converting enzyme 2 in severe acute respiratory syndrome coronavirus and SARS-CoV-2: A double-edged sword? FASEB J 34, 6017-6026 (2020).
2. U. Bastolla, The differential expression of the ACE2 receptor across ages and gender explains the differential lethality of SARN-Cov-2 and SARS and suggests possible therapy. Pre-print, (2020).
3. G. Severi et al., Androgenetic alopecia in men aged 40-69 years: prevalence and risk factors. Br J Dermatol 149, 1207-1213 (2003).
4. G. Rastrelli et al., Low testosterone levels predict clinical adverse outcomes in SARS-CoV-2 pneumonia patients. Andrology Accepted, Pre-print, (2020).
5. M. Schroeder et al., The majority of male patients with COVID-19 present low testosterone levels on admission to Intensive Care in Hamburg, Germany: a retrospective cohort study. Pre-print, (2020).
6. W. D. Cardona Maya, S. S. Du Plessis, P. A. Velilla, SARS-CoV-2 and the testis: similarity with other viruses and routes of infection. Reprod Biomed Online 40, 763-764 (2020).