G. Paul Amminger1, Barnaby Nelson1, Connie Markulev1, Hok Pan Yuen1, Miriam R Schäfer1, Nilufar Mossaheb2, Monika Schlögelhofer2, Stefan Smesny3, Ian B Hickie4, Gregor E Berger5, Eric Y H Chen6, Lieuwe de Haan7, Dorien H Nieman7, Merete Nordentoft8, Anita Riecher-Rössler9, Swapna Verma10, Maximus Berger1, Andrew Thompson11, Alison R Yung12, Patrick D McGorry1; 1Orygen, The National Centre of Excellence in Youth Mental Health, Melbourne, Australia, 2Department of Psychiatry, Medical University of Vienna, Vienna, Austria, 3Department of Psychiatry, University Hospital, Jena, Germany, 4Brain and Mind Research Institute, University of Sydney, Sydney, Australia, 5Child and Adolescent Psychiatric Service of the Canton of Zurich, Zurich, Switzerland, 6Department of Psychiatry, University of Hong Kong, Hong Kong, 7Department of Psychiatry, Academic Medical Center, Amsterdam, the Netherlands, 8Psychiatric Centre Bispebjerg, Copenhagen, Denmark, 9Psychiatric University Clinics Basel, Basel, Switzerland, 10Institute of Mental Health, Singapore, Singapore, 11Division of Mental Health and Wellbeing, Warwick Medical School, University of Warwick, Coventry, England, 12Institute of Brain, Behaviour and Mental Health, University of Manchester, Manchester, England
The NEURAPRO multicentre RCT of long-chain omega-3 polyunsaturated fatty acids (n3PUFA) vs. placebo demonstrated no clinical benefits for n3PUFA (‘fish oil’) supplementation in individuals with At-Risk Mental State (ARMS) for psychosis. However, adherence in this trial was low and n3PUFA were also available outside the study through diet or non-study supplements. Therefore, we examined if pre-treatment n3PUFA levels measured in erythrocytes and their change during the trial were related to outcomes. Data from 285 of 304 (94%) NEURAPRO participants were analysed. PUFA levels were measured at baseline and after supplementation (6 months). The n3-index, a biomarker of n3PUFA status, was calculated as the combined relative abundance of eicosapentaeonic acid and docosahexaenoic acid. Outcome measures included psychosis transition, clinical improvement (CGI-I), general psychopathology (BPRS), depressive symptoms (MADS), negative symptoms (SANS) and functioning (SOFAS). The median n3-index in the sample at baseline was 3%, approximately 40% lower than in the (Australian) general population (5%). Adjusting for relevant baseline factors, the change (increase) of the n3-index from baseline to 6 months predicted better functional and symptomatic outcomes for the majority of applied measures at 6 and 12 months. Participants with an n3-index <3% at baseline were significantly more likely to show substantial clinical improvement from n3PUFA supplementation at 6 months. In contrast, in the group with n3-index ≥3%, n3PUFA supplementation was unrelated to outcomes. This analysis provides crucial evidence to support beneficial effects of fish oil supplementation in individuals with ARMS for psychosis, in particular in those with low n3PUFA levels at baseline.
