Elsevier

Schizophrenia Research

Volume 193, March 2018, Pages 168-172
Schizophrenia Research

Predictors of longer-term outcome in the Vienna omega-3 high-risk study

https://doi.org/10.1016/j.schres.2017.08.010Get rights and content

Abstract

Longer-term data on ω-3 polyunsaturated fatty acids (PUFA) for prevention of psychosis in (ultra high risk) UHR individuals have initially shown promising results.

This analysis aimed to assess clinical predictors of longer-term outcome in UHR individuals treated with ω-3 PUFAs versus placebo.

Data derived from an RCT in 81 UHR individuals treated with ω-3 PUFAs versus placebo for 12 weeks and follow-up assessment after a median of 6.7 years.

Baseline GAF, baseline PANSS global score, pre-to-post-intervention change in EPA (Eicosapentaenoic acid) level were significant predictors of transition to psychosis, PANSS negative score and baseline MADRS reached trend-levels. In the final multivariate Cox regression analysis change in EPA levels remained the only significant predictor.

Taking into account all other significant predictors, changes in EPA levels were found to be the single most significant predictor for transition to psychosis in a longer term observation of UHR individuals.

Introduction

Early psychosis research in the last 15 years has established criteria for the definition of high risk groups for the development of psychotic illnesses, therewith proposing a prospective concept for the originally retrospective “prodrome” idea (Nelson et al., 2013). According to meta-analytical data, these ultra-high risk criteria are associated with transition rates of around 22% to 36% after 1, respectively 3 years (Fusar-Poli et al., 2012).

Since transition rates have been reported as being in decline in specialized early intervention centres (Yung et al., 2007), additional potential clinical outcome predictors could serve to further stratify risk groups. Recently, isolated predictors for transition have been established by different groups in similar samples, amongst others: baseline global functioning (Cannon et al., 2008, Nelson et al., 2013, Thompson et al., 2011), decline in functioning alongside a genetic risk (Cannon et al., 2008, Thompson et al., 2011), history of substance abuse (Cannon et al., 2008), duration of symptoms (Nelson et al., 2013), sex (Walder et al., 2013), age (Amminger et al., 2006), attenuated psychotic symptoms (Ziermans et al., 2014), especially unusual thought content (Cannon et al., 2008, Thompson et al., 2011), paranoid ideations (Cannon et al., 2008) and formal thought disorder (Thompson et al., 2013), as well as negative symptoms (Lin et al., 2011). Low levels of nervonic acid were previously described by our group as being associated with a higher risk of transition to psychosis (Amminger et al., 2012). Other non-clinical predictors, such as neurocognitive- and related variables (Nieman et al., 2014, Seidman et al., 2016, Ziermans et al., 2014) have also shown good discrimination potential regarding transition status in young individuals. A staged probabilistic model approach performed in this sample using history of drug use, positive and negative and global symptoms, psychosocial functioning, total omega-3 levels and nervonic acid improved the prediction of transition within one year to over 70% (Clark et al., 2016).

Several early intervention strategies of indicated prevention have been investigated, amongst others low-dose antipsychotics, CBT (cognitive behavioural therapy) and ω-3 polyunsaturated fatty acids (PUFA) (Schmidt et al., 2015). The latter have provided promising results in one double-blind placebo-controlled study performed by our group (Amminger et al., 2010), while first results of the recent NEURAPRO multi-site replication trial did not replicate the original findings, notabene with generally low transition rates and substantial improvements in both group with high quality psychosocial treatment, which might at least partly have led to a ceiling effect (McGorry et al., 2016). As described in Amminger et al. (2010), treatment with ω-3 PUFAs over a period of 12 weeks proved to be effective in reducing 12-months transition rates in a UHR population compared to placebo. Transition rates in the ω-3 PUFAs were 4.9% at 12-months follow-up compared to 27.5% in the placebo group. A longer-term follow-up of the same population has shown sustained protective effects after a median of 6.7 years with transition rates of 9.8% in the ω-3 PUFAs and 40% in the placebo group (Amminger et al., 2015b). These results remain remarkable, and questions have arisen on the generalizability of the sample and its findings.

However, RCTs with ω-3 PUFAs bear the difficulty of controlling for non-study intake of ω-3 PUFAs, such as increases in fish and seafood consumption, or use of non-study ω-3 PUFAs supplements (James et al., 2014). Furthermore, tissue and erythrocyte levels of ω-3 PUFA can vary significantly between individuals. These potential pitfalls can be accounted for by analysing outcome independent of group allocation at baseline but with respect to changes in cell membrane ω-3 PUFA levels, which reflect all ω-3 intake regardless of its source.

Thus, the aim of this analysis was to test whether baseline levels or changes in ω-3 PUFA levels during the initial 12-week intervention phase of the study were predictive of longer-term outcome (i.e. transition to psychosis) while taking into account previously reported clinical predictors in a population of UHR individuals.

Section snippets

Sample

The initial sample was previously described in Amminger et al. (2010) and consisted of individuals between the ages of 13 and 25 with an increased risk for psychosis, meeting at least one of three operationally defined criteria (i.e., attenuated psychotic symptoms, brief limited intermittent psychotic symptoms, or a genetic risk with decreased functioning). Exclusion criteria were a previous history of a psychotic disorder or manic episode, a substance-induced psychotic disorder, acute suicidal

Results

When variables were tested separately in the univariate Cox regression analysis, the following variables were identified as significant predictors of transition to psychosis: baseline GAF (p = 0.004), baseline PANSS global scores (p = 0.022), baseline formal thought disorder (p = 0.007) and change in EPA levels (p = 0.038), while the following with barely missing significance level: baseline MADRS (p = 0.056) and baseline PANSS negative score (p = 0.053). None of the other variables, i.e. baseline PANSS

Discussion

The present analysis aimed to assess predictors of longer-term outcome in a sample of UHR individuals originally treated within a 12-weeks trial of ω-3 PUFA versus placebo supplementation. Baseline psychosocial functioning, negative symptoms, general psychopathology, formal thought disorder, depressive symptoms and changes in ω-3 PUFA EPA levels were significant predictors of outcome. However, after adjusting for one another only changes in EPA levels remained the main and single most

Role of the funding source

This work was supported by grant 03T-315 from the Stanley Medical Research Institute.

Contributors

Nilufar Mossaheb contributed to the trial itself, the analysis and the interpretation of the data, drafted the manuscript, revised it.

Paul Amminger contributed to the trial itself, to the concept and design of the trial, obtaining funding, as well as the analysis and the interpretation of the data and the revision of the manuscript.

Miriam Schäfer contributed to the trial itself, study supervision and in revising the manuscript.

Monika Schlögelhofer contributed to the trial itself, data

Conflicts of interest

The authors report no conflicts of interest.

Acknowledgments

  • -

    G.P.A. was supported by National Health and Medical Research Council (NHMRC) Australia Senior Research Fellowship (ID:1080963);

  • -

    P.D.M. was supported by NHMRC Senior Principal Research Fellowship (ID: 1060996).

  • -

    M.B. was supported by a JCU Postgraduate Research Scholarship.

References (36)

  • G.P. Amminger et al.

    Long-chain omega-3 fatty acids for indicated prevention of psychotic disorders: a randomized, placebo-controlled trial

    Arch. Gen. Psychiatry

    (2010)
  • G.P. Amminger et al.

    Decreased nervonic acid levels in erythrocyte membranes predict psychosis in help-seeking ultra-high-risk individuals

    Mol. Psychiatry

    (2012)
  • G.P. Amminger et al.

    Predictors of treatment response in young people at ultra-high risk for psychosis who received long-chain omega-3 fatty acids

    Transl. Psychiatry

    (2015)
  • G.P. Amminger et al.

    Longer-term outcome in the prevention of psychotic disorders by the Vienna omega-3 study

    Nat. Commun.

    (2015)
  • Association, A.A.P

    Diagnostic and Statistical Manual of Mental Disorders

    (1994)
  • G.E. Berger et al.

    Ethyl-eicosapentaenoic acid in first-episode psychosis: a randomized, placebo-controlled trial

    J. Clin. Psychiatry

    (2007)
  • T.D. Cannon et al.

    Prediction of psychosis in youth at high clinical risk: a multisite longitudinal study in North America

    Arch. Gen. Psychiatry

    (2008)
  • S.R. Clark et al.

    Towards indicated prevention of psychosis: using probabilistic assessments of transition risk in psychosis prodrome

    J. Neural Transm. (Vienna)

    (2015)
  • Cited by (7)

    • Relationship between allostatic load and clinical outcomes in youth at ultra-high risk for psychosis in the NEURAPRO study

      2020, Schizophrenia Research
      Citation Excerpt :

      Notably, several markers of immune activation, including IL-6 and TNFalpha, are highly correlated with the AL index in patients with schizophrenia and first-episode psychosis (Berger et al., 2018). In clinical trials of n-3 PUFA, increases in the n-3 PUFA eicosapentaenoic acid (EPA) are related to lower transition risk (Mossaheb et al., 2017), which is likely in part due to their anti-inflammatory properties, including their role as a precursor of EPA-derived anti-inflammatory molecules such as resolvins or their ability to downregulate the NFkappaB pathway (Kiecolt-Glaser et al., 2007; Rangel-Zuniga et al., 2015). In contrast to functioning, AL was not predictive of positive psychotic symptoms (BPRS) and transition to psychosis in the present study.

    • The NEURAPRO Biomarker Analysis: Long-Chain Omega-3 Fatty Acids Improve 6-Month and 12-Month Outcomes in Youths at Ultra-High Risk for Psychosis

      2020, Biological Psychiatry
      Citation Excerpt :

      Transition to psychosis by month 12 was predicted by higher baseline BPRS total and psychotic subscale scores. More severe positive symptoms and psychiatric symptoms in general have been previously linked with an increased psychosis risk in UHR patients (5,42). This biomarker analysis has several strengths, including the large sample size, the use of standardized assessment instruments, the consistency of findings across measures, and their robustness in multivariate analyses.

    View all citing articles on Scopus
    View full text