Effects of Polygenic Risk Scores for Schizophrenia on Psychosocial Functioning Within and Across Diagnostic Boundaries

Janos Kalman*1, Sergi Papiol*1, Urs Heilbronner1, Dörthe Malzahn2, Jana Strohmaier3, Maren Lang3, Josef Frank3, Jens Treutlein3, Andrea Hofmann4, Franziska Degenhardt4,6, Stephanie H. Witt3, Sven Cichon3,6,7, Markus M. Nöthen4,6, Marcella Rietschel3, Thomas G. Schulze1,3,8 *These authors contributed equally

1Institute of Psychiatric Phenomics and Genomics, Ludwig-Maximilians-University Munich, Germany; 2Department of Genetic Epidemiology, University Medical Center, Georg-August-University, Göttingen, Germany; 3Department of Genetic Epidemiology in Psychiatry, Central Institute of Mental Health, Mannheim, Germany; 4Institute of Human Genetics, University of Bonn, Germany; 5Institute of Neuroscience and Medicine (INM-1), Research Centre Jülich (FZJ), Jülich, Germany; 6Department of Genomics, Life & Brain Center, University of Bonn, Bonn, Germany; 7Division of Medical Genetics, University Hospital Basel, University of Basel, Switzerland;

8Department of Psychiatry and Psychotherapy, University Medical Center, Georg-August-University, Göttingen, Germany

IntroductionBoth Schizophrenia (SCZ) and bipolar disorder (BD) are characterized by significant psychosocial impairment. There is a considerable gap between clinical and functional recovery in both conditions, with impaired daily functioning persisting in almost half of BD and the majority of SCZ patients even after reaching clinical remission (Harvey et al. 2011). Still, hardly any data is avaliable on the role of genetic factors in functional impairment as studies mostly investigated the role of clinical and environmental cha-racteristics. As there is a considerable overlap between the genetic back-ground of SCZ and BD, also supported by recent studies finding the poly-genic component of SCZ risk to be a significant predictor of BD (Ruderfer et al. 2015) we aimed to determine whether a higher polygenic burden on SCZ correlates with the level of (1) premorbid, (2) worst ever and (3) current level of functioning and the course of functioning over time in BD and SCZ patients.

MethodsParticipants

y Inpatients with a life-time diagnosis of SZ (297) and BD (516) (DSM-IV criteria) y Recruited at the Central Institute of Mental Health, Mannheim and the De-partment of Psychiatry, University of Bonn, Germany.

Measurements y (1) Premorbid functioning (defined as the highest GAF score present before illness onset),

y (2) The worst GAF score ever, i.e. the lowest GAF value ever present during an illness episode,

y (3) current GAF (pre-admission GAF): GAF score right before the current epi-sode for which the patient received clinical treatment at the time of inter-view.

Genotyping and polygenic scoring y Genotyping on the Illumina HH550/H610Q/H660W (Illumina, San Diego, USA). For more information on genotyping and quality control see Rietschel et al. 2012.

y Risk allele effect sizes were obtained using the PGC2 SCZ summary results (excluding our sample) as a discovery sample.

y SCZ polygenic scores were calculated for all individuals at high-resolution pTs (0.0005 steps) using PRSice (Eusden et al. 2015).

Statistical analyses y Linear regression was used to assess the effect of SCZ-PRS on premorbid, worst ever and current GAF scores and age at onset.

y GAF scores were adjusted for age at onset x sex (premorbid GAF) or duration of illness x sex (worst ever and current GAF) while age at onset was adjusted for sex.

y The effects of SCZ-PRS (at the best hit pT) on the longitudinal course of fun-ctioning of patients were assessed with a longitudinal non-parametric test. (Malzahn et al 2010) For this the sample was split into two and we compa-red (1) those above and under the median SCZ-PRS and (2) patients at the highest and lowest 33.33% of the SCZ-PRS distribution.

ReferencesHarvey et al. (2012) Schizophr Bull 38(6): 1318-26 Ruderfer et al. (2014) Mol Psychiatry 19(9): 1017-24 Rietschel et al. (2012) Mol Psychiatry 17(9): 906-17 Eusden et al. (2015) Bioinformatics 31 (9): 1466-1468 Malzahn et al. (2010) Genet Epidemiol 34: 469-78Figure 1. SCZ-PRS was significantly associated with both (1) premorbid

(R2=0.02) and (3) current level (R2=0.013) of functioning in BD patients,

Figure 2. No significant correlation was seen between SCZ-PRS and (1) pre-morbid,

ResultsResults of the LNPT analysis showed a significant difference in the GAF scores over all three assessment points between the BD patients with higher and lo-wer SCZ-PRS (p<0.015). When comparing BD patients at the lower and upper

DiscussionThis study was the first to investigate the influencing role of SCZ-PRS on the level of functioning and age at onset in SCZ and BD.The observed negative correlation between SCZ-PRS and premorbid functio-ning and age at onset in BD suggest, that having a higher polygenic load on SCZ, a condition with pronounced premorbid impairment in all levels of life, might contribute to a similar premorbid clinical picture in BD. The analysis of the longitudinal course of GAF gives further evidence for the influencing role of SCZ PRS on BD functioning as BD patients with higher SCZ polygenic load tended to have a more impaired functioning across all measurement points compared to those with less SCZ polygenic risk.We are currently working on a replication study in more than 700 longitudi-nally followed patients with SCZ and BD, featuring detailed GAF assessments to see if our results also stand in an independent population.

Schizophrenia (mean, SD) Bipolar Disorder (mean, SD)

Age 36.32 (11.24) 45.68 (12.91)

Sex (females %) 45% 55.8%

Age at onset 26.27 (8.5) 28.72 (11.92)

Duration of illness 10.04 (9.668) 16.96 (12.23)

Premorbid GAF 84.74 (10.869) 91.15 (9.38)

Worst ever GAF 25.33 (8.87) 29.21 (9.38)

Current GAF 63.37(15.87) 78.44 (14.86)

33.33% of the SCZ-PRS distribution, the observed group difference was even more pronounced (p<0.004).

Table 1. Demographic and clinical characteristics of the assessed population.

Figure 3. SCZ-PRS significantly predicted (1) age at onset in BD (though not after correction for multiple testing),

(though the association with current GAF lost significance after correcting for multiple testing). No association with (2) worst ever GAF was detected.

Conflict of InterestThere are no conflicts of interest.

(2) worst ever or (3) current GAF at any of the pTs.

but (2) not in SCZ.

Acknowledgements We thank all of the patients for participating in this study. This research was funded by the Deutsche Forschungsgemeinschaft (DFG): Klinische Forscher-gruppe (KFO) 241: TP1 (SCHU 1603/5-1), FKZ RO4076/1-1 and FKZ RO4076/3-1. Thomas G. Schulze was supported by the Lisa-Oehler-Foundation.