BackgroundIdentifying biological alterations in patients with depression, particularly those that differ between responders and non-responders, is of interest to clinical practice. Biomarker candidates involve neuroactive steroids, including pregnenolone (PREG) and allopregnanolone (ALLO). However, alterations in PREG and ALLO associated with treatment response are understudied. This study's main aim was to evaluate the effects of antidepressant treatment, clinical response, and treatment duration on PREG and ALLO in depression. Materials and MethodsIn a 4-week, open-label trial, participants were allocated randomly to the venlafaxine (n = 27) or mirtazapine (n = 30) group. Urine concentrations of PREG and ALLO were assessed through gas chromatography-mass spectrometry. Participants collected night urine between 10:30 p.m. and 8:00 a.m. Two primary outcomes were analyzed. Firstly, the effect of treatment (mirtazapine or venlafaxine), clinical response (operationalized through the Hamilton Depression Rating Scale), and time (baseline compared to 28 days) on the urine concentrations of PREG or ALLO in depression. Finally, the effect of clinical response and time on the urine concentration of PREG or ALLO, independently of the antidepressant given (mirtazapine or venlafaxine). Linear mixed models were carried out. ResultsThere was no significant difference in PREG and ALLO concentrations between baseline and 28 days in responders and non-responders when investigating the venlafaxine or the mirtazapine group. However, we found a significant reduction of urine PREG concentration after 28 days of treatment in responders who received either venlafaxine or mirtazapine (estimate = −0.56; p = 0.016; 95CI [−1.003; −0.115]; Cohen’s d = −0.61). ConclusionsOur main results indicate that responders in depression show reduced urinary PREG concentrations after 4-weeks of therapy, independently of the antidepressant used. More studies are needed to confirm these findings.