A new study suggests that male hormones do not explain worse COVID outcomes in men.

It’s been clear since very early on in the pandemic that men fare worse with COVID than women.

This is true for both hospitalization rates and death, as this graph, which looks at the ratio of deaths in men to women over time across various states shows.

That dashed line is equal mortality rates. You can see that rates are about 25% higher in men pretty much regardless of state or time. The question, of course, is why?

Could it be the testosterone?

Whenever biological differences are measured between men and women, all eyes turn to the sex hormones first. After all, testosterone levels are around 20-fold higher in men than women and hormones have profound biologic effects.

Could men’s high testosterone levels make them more susceptible to COVID-19-related mortality? Researchers led by Sandheep Dhindsa at the St. Louis University School of Medicine decided to find out.

If testosterone were the problem, they reasoned, men with lower testosterone would fare better when infected with COVID-19.

So they dug into their electronic health record system to look at individuals who were infected with COVID-19 who also happened to have testosterone levels measured. They created three groups: Those with normal testosterone, those with low testosterone (and no – I am not going to call it “low T” because I’m not a hack), and those who were receiving testosterone supplementation.

And here are the basic results – 45% of the men with low testosterone were hospitalized for their COVID-19 infection, compared to 12% of those with normal testosterone and 16% of those on testosterone supplements. A similar pattern was seen when looking at outcomes like ICU use and overall mortality, although small numbers limit the statistical significance.

That’s a pretty compelling argument that testosterone is not so bad for you, COVID-19 wise. But I know what you’re thinking. There are lots of things associated with low testosterone that ALSO lead to bad COVID outcomes – older age being an obvious one.

After adjusting for age, race, ethnicity, BMI, the use of immunosuppression and a score based on comorbid conditions, the results held up – low testosterone was still a significant risk factor for hospitalization for COVID-19.

How can we square this with the observation that men do worse with COVID then women? This study makes it look like testosterone is protective, and women, you know, don’t have much of it.

One explanation is simple selection bias – remember this was a study of men who had testosterone levels measured for some reason. For example, 32 men in the low testosterone group were receiving androgen deprivation therapy for prostate cancer – 56% of them were hospitalized – that may not be well-captured by adjustment for an overall comorbidity score.

Another possibility is that the problem was never testosterone to begin with. Men and women behaved differently during the pandemic. Women were more likely to wear masks, wash their hands, and comply with social distancing recommendations. Surveys have shown that, during the pandemic, men were less likely to be concerned about their own health and the health of others. And of course, men may have different occupational and social exposures than women.

In short, the testosterone hypothesis is probably too simple. What makes a man a man? Well, we are more than our testosterone levels, and our risk of bad COVID-19 outcomes stems from a host of factors – some biological, some social, some environmental. But probably, it’s not just the hormones.

A version of this commentary first appeared on Medscape.com.