BackgroundExperimental studies demonstrate beneficial immunological and hemodynamic effects of estradiol in animal models of sepsis. This raises the question whether estradiol contributes to sex differences in the incidence and outcomes of sepsis in humans. Yet, total estradiol levels are elevated in sepsis patients, particularly nonsurvivors. Bioavailable estradiol concentrations have not previously been reported in septic patients. The bioavailable estradiol concentration accounts for aberrations in estradiol carrier protein concentrations that could produce discrepancies between total and bioavailable estradiol levels. We hypothesized that bioavailable estradiol levels are low in septic patients and sepsis nonsurvivors.MethodsWe conducted a combined case-control and prospective cohort study. Venous blood samples were obtained from 131 critically ill septic patients in the medical and surgical intensive care units at the University of Rochester Medical Center and 51 control subjects without acute illness. Serum bioavailable estradiol concentrations were calculated using measurements of total estradiol, sex hormone-binding globulin, and albumin. Comparisons were made between patients with severe sepsis and control subjects and between hospital survivors and nonsurvivors. Multivariable logistic regression analysis was also performed.ResultsBioavailable estradiol concentrations were significantly higher in sepsis patients than in control subjects (211 [78–675] pM vs. 100 [78–142] pM, p < 0.01) and in sepsis nonsurvivors than in survivors (312 [164–918] pM vs. 167 [70–566] pM, p = 0.04). After adjustment for age and comorbidities, patients with bioavailable estradiol levels above the median value had significantly higher risk of hospital mortality (OR 4.27, 95 % CI 1.65–11.06, p = 0.003). Bioavailable estradiol levels were directly correlated with severity of illness and did not differ between men and women.ConclusionsContrary to our hypothesis, bioavailable estradiol levels were elevated in sepsis patients, particularly nonsurvivors, and were independently associated with mortality. Whether estradiol’s effects are harmful, beneficial, or neutral in septic patients remains unknown, but our findings raise caution about estradiol’s therapeutic potential in this setting. Our findings do not provide an explanation for sex-based differences in sepsis incidence and outcomes.Electronic supplementary materialThe online version of this article (doi:10.1186/s13054-016-1525-9) contains supplementary material, which is available to authorized users.