22. The Effects of Antidepressant use on Clinical Outcomes in the NEURAPRO-E Trial

Abstract Background: The NEURAPRO-E trial is a randomized controlled multicenter trial of long-chain omega-3 polyunsaturated fatty acids in 304 young people at clinical risk of psychosis. The trial failed to replicate the earlier single-center trial in 81 participants (the Vienna omega-3 study). This analysis aims to investigate if antidepressant (AD) use during NEURAPRO-E influenced clinical outcomes. Methods: Secondary analysis of a RCT of omega-3 PUFAs versus placebo involving a total cohort of 304 participants. Randomization was stratified by total score on the Montgomery-Asberg Depression Rating Scale (MADRS score ≥21 vs. <21) as both depression and ADs may impact on UHR symptoms and illness progression. We determined the influence of AD use on following outcome measures: time to transition to psychotic disorder; measures of psychiatric symptoms and psychosocial functioning including the MADRS, BPRS, SANS and SOFAs. Since omega-3 supplementation versus placebo did not impact on outcomes the original group allocations were ignored for this analysis and participants were treated as a single cohort. Results: Three hundred and four treatment-seeking individuals were enrolled in this study. Thirty-two (10.5%) of 304 patients discontinued the intervention prematurely (i.e., prior to 6 months). Forty-six (15.1%) patients were unable to be contacted and one became pregnant, meaning that in total 79 (26%) participants were lost to follow-up. A MADRS score ≥21 at baseline was observed in 134 (44.1%) individuals. ADs were used in 189 (62.2%) of the participants. The progression to psychosis rate in individuals who received an AD (13.2%; 25 of 189) was more than twice as high as in individuals without AD treatment (6.1%; 7 of 115). Kaplan-Meier survival analysis indicated a group difference at trend level (χ2 = 3.237, P = .072, log rank test). After stratification for MADRS scores, KM survival analyses indicated a higher transition to psychosis rate in individuals who received ADs in the stratum with low depressive symptoms (14% vs. 4%; χ2 = 3.679, P = .055, log rank test), while in the stratum with high depressive symptoms, transition to psychosis rates were 13% and 10% for groups with and without AD use, respectively (χ2 = .226, P = .635, log rank test). Furthermore, repeated measures analysis indicated that MADRS depressive symptoms significantly improved between baseline and 12-month follow-up in groups with and without AD use in both strata of severity of depressive symptoms (all P < .001). Repeated measures analyses for the other secondary outcomes also showed no significant clinical advantage for AD use compared to no AD use in groups with either high or low depressive symptoms. Conclusion: This is the first analysis on the effects of ADs in a large cohort of UHR individuals. No statistically significant effects of ADs on outcomes were observed. The findings are limited by fact that ADs were provided as concomitant medication upon clinical discretion.