From the article:
The researchers measured several hormones in blood samples from 90 men and 62 women who came to Barnes-Jewish Hospital with symptoms of COVID-19 and who had confirmed cases of the illness. For the 143 patients who were admitted to the hospital, the researchers measured hormone levels again at days 3, 7, 14 and 28, as long as the patients remained hospitalized over these time frames. In addition to testosterone, the investigators measured levels of estradiol, a form of estrogen produced by the body, and IGF-1, an important growth hormone that is similar to insulin and plays a role in maintaining muscle mass.
Among women, the researchers found no correlation between levels of any hormone and disease severity. Among men, only testosterone levels were linked to COVID-19 severity. A blood testosterone level of 250 nanograms per deciliter or less is considered low testosterone in adult men. At hospital admission, men with severe COVID-19 had average testosterone levels of 53 nanograms per deciliter; men with less severe disease had average levels of 151 nanograms per deciliter. By day three, the average testosterone level of the most severely ill men was only 19 nanograms per deciliter.
The lower the levels of testosterone, the more severe the disease. For example, those with the lowest levels of testosterone in the blood were at highest risk of going on a ventilator, needing intensive care or dying. Thirty-seven patients – 25 of whom were men – died over the course of the study.