Inflammatory evidence for the psychosis continuum model

Accepted Manuscript Title: INFLAMMATORY EVIDENCE FOR THE PSYCHOSIS CONTINUUM MODEL Author: Ragni H. Mørch Ingrid Dieset Ann Færden Sigrun Hope Monica Aas Mari Nerhus Erlend S. Gardsjord Inge Joa Gunnar Morken Ingrid Agartz P˚al Aukrust Srdjan Djurovic Ingrid Melle Thor Ueland Ole A. Andreassen PII: DOI: Reference:

S0306-4530(16)30041-5 http://dx.doi.org/doi:10.1016/j.psyneuen.2016.02.011 PNEC 3210

To appear in: Received date: Revised date: Accepted date:

7-12-2015 11-2-2016 12-2-2016

Please cite this article as: Morch, Ragni H., Dieset, Ingrid, Faerden, Ann, Hope, Sigrun, Aas, Monica, Nerhus, Mari, Gardsjord, Erlend S., Joa, Inge, Morken, Gunnar, Agartz, Ingrid, Aukrust, P˚al, Djurovic, Srdjan, Melle, Ingrid, Ueland, Thor, Andreassen, Ole A., INFLAMMATORY EVIDENCE FOR THE PSYCHOSIS CONTINUUM MODEL.Psychoneuroendocrinology http://dx.doi.org/10.1016/j.psyneuen.2016.02.011 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Ragni Mørch

INFLAMMATORY EVIDENCE FOR THE PSYCHOSIS CONTINUUM MODEL Running title : -A cross-sectional study of inflammation and the psychosis continuum model. Ragni H. Mørcha,b, Ingrid Dieseta,b, Ann Færdena,b, Sigrun Hopea,c, Monica Aasa,b, Mari Nerhus a,b, Erlend S. Gardsjorda,b, Inge Joad, Gunnar Morkene,f, Ingrid Agartza,g, Pål Aukrusth,i,j,k, Srdjan Djurovicl,m, Ingrid Mellea,b, Thor Uelandh,j,k, Ole A. Andreassena,b a

KG Jebsen Centre for Psychosis Research, University of Oslo and Oslo University Hospital, Oslo, Norway b Division of Mental Health and Addiction, Oslo University Hospital, Ullevål, Oslo, Norway c Department of Neuro Habilitation, Oslo University Hospital Ullevål, Oslo, Norway d Centre for Clinical Research in Psychosis, Psychiatric Division, Stavanger University Hospital, Stavanger, Norway e The Department of Psychiatry, St. Olav University Hospital of Trondheim, Trondheim, Norway f Department of Neuroscience, Norwegian University of Science and Technology g Department of Psychiatric Research, Diakonhjemmet Hospital, Oslo, Norway h Research Institute of Internal Medicine, Oslo University Hospital Rikshospitalet, Oslo, Norway i Section of Clinical Immunology and Infectious Diseases, Oslo University Hospital, Rikshospitalet, Norway, j Institute of Clinical Medicine, University of Oslo, Oslo, Norway, k K.G. Jebsen Inflammatory Research Center, University of Oslo, Oslo, Norway l Department of Medical Genetics, Oslo University Hospital, Oslo, Norway m NORMENT, KG Jebsen Centre for Psychosis Research, Department of Clinical Science, University of Bergen, Bergen, Norway Correspondence to: Ragni H. Mørch M.D. Oslo University Hospital Division of Mental Health and Addiction Psychosis Research Unit/TOP Ullevål Hospital, building 49 P.O. Box 4956 Nydalen N- 0424 Oslo Norway Phone: + 47 23 02 73 34, Fax: + 47 23 02 73 33 E-mail: [email protected] Author notes OAA, ID and RHM designed the study. The other co-authors participated in data collection and quality control. RHM and ID analyzed the data. RHM, ID and OAA wrote the first draft of the manuscript. All co-authors participated in revising the manuscript and approved the final version. 1

Ragni Mørch

Highlights 

Inflammatory markers, cytokines are implicated in pathophysiology of severe mental disorders



We measured levels of inflammatory markers in patients (n=992) and healthy controls (n=638)



Levels of sTNF-R1 and IL-1Ra were increased in patients compared to controls



The schizophrenia group showed higher levels than the affective group providing evidence for the psychosis continuum model



Antipsychotic medication was not associated with the inflammatory markers

2

Ragni Mørch

ABSTRACT Background: Inflammation and immune activation have been implicated in the pathophysiology of severe mental disorders. Previous studies of inflammatory markers, however, have been limited with somewhat inconsistent results.

Aims: We aimed to determine the effect sizes of inflammatory marker alterations across diagnostic groups of the psychosis continuum and investigate association to antipsychotic medications.

Methods: Plasma levels of soluble tumor necrosis factor receptor 1 (sTNF-R1), interleukin 1 receptor antagonist (IL-1Ra), osteoprotegerin (OPG), and von Willebrand factor (vWf) were measured in patients (n=992) with schizophrenia spectrum (SCZ, n=584), schizoaffective disorder (SA, n=93), affective spectrum disorders (AFF, n=315), and healthy controls (HC, n=638).

Results: Levels of sTNF-R1 (p=1.8x10-8, d=0.23) and IL-1Ra (p=0.002, d=0.16) were increased in patients compared to HC. The SCZ group had higher levels of sTNF-R1 (p=8.5x10-8, d= 0.27) and IL-1Ra (p=5.9x10-5, d=0.25) compared to HC, and for sTNF-R1 this was also seen in the SA group (p=0.01, d=0.3) and in the AFF group (p= 0.002, d= 0.12). Further, IL-1Ra (p=0.004, d= 0.25) and vWf (p=0.02, d=0.21) were increased in the SCZ compared to the AFF group. There was no significant association between inflammatory markers and use of antipsychotic medication. Conclusion: We demonstrate a small increase in sTNF-R1 and IL-1Ra in patients with severe mental disorders supporting a role of inflammatory mechanisms in disease pathophysiology. The increase was more pronounced in SCZ compared to AFF supporting a continuum psychosis model related to immune factors.

Keywords: schizophrenia; bipolar disorder; psychosis continuum model; antipsychotics; inflammation

3

Ragni Mørch

1. INTRODUCTION

Recent studies have shown clinical, epidemiological and genetic overlap among the severe mental disorders schizophrenia and bipolar disorder, indicating shared disease mechanisms (Lee et al., 2013; Maier et al., 2006; McGrath et al., 2008). This has led to an increased interest in the psychosis continuum model with schizophrenia and bipolar disorder at opposite ends of the scale with schizoaffective disorder in between (Craddock et al., 2009; Craddock and Owen, 2010; Keshavan et al., 2011). The pathophysiology of these disorders is largely unknown, but several lines of evidence implicate a role for immune activation and inflammation (Bergink et al., 2014; Drexhage et al., 2010; Potvin et al., 2008; Tomasik et al., 2014). Several studies demonstrate increased systemic levels of inflammatory markers in psychotic disorders (Munkholm et al., 2013; Potvin et al., 2008). We have previously reported increased levels of soluble tumor necrosis factor receptor 1 (sTNF-R1), osteoprotegerin (OPG), and von Willebrand factors (vWf), in patients with schizophrenia and bipolar disorder (Hope et al., 2010; Hope et al., 2009) as compared to healthy controls (HC). We have also found elevated levels of sTNF-R1 and IL-1Ra associated with psychotic symptoms and cognitive abilities (Hope et al., 2015), OPG levels associated with affective symptoms in bipolar disorder (Hope et al., 2011), and a relationship between vWf levels and brain morphology (Dieset et al., 2014). The inflammatory markers are reflecting aspects of the immune system which have in single disease studies found to be altered in psychotic disorders, including our previous work (Hope et al., 2011; Hope et al., 2010; Hope et al., 2009). This allows us to study similarities and differences across the psychosis continuum. The markers in the present study represent different aspects of possible immune mechanisms; the TNF and IL-1 systems represent pro-inflammatory cytokine activity pathways, vWf is a marker of endothelial cell activation and OPG is a marker of calcium related vascular inflammation. sTNF-R1 is implicated in many diseases; representing the pro-inflammatory TNF-system, it is associated with endothelial cell activation and leakage (Aukrust et al., 2011). vWf is a wellknown marker of endothelial cell activation and inflammation (Denis, 2002), OPG is a soluble decoy receptor in the TNF-superfamily and a marker of calcium related vascular inflammation (Corallini et al., 2008), and IL-1Ra is known as an inflammatory marker representing the IL-1 system which has also been suggested to influence the permeability of the brain endothelial barrier (Michael et al., 2015). The markers are collectively good indicators of endothelial cell activation and inflammation which have been suggested as disease mechanisms in severe mental disorders. It has been speculated that endothelial 4

Ragni Mørch

alterations can cause blood-brain-barrier (BBB) dysfunction and thus be related to severe mental disorders. In addition, the markers for this study are collectively robust markers that are stable in plasma and can easily be measured with reliable results (Gu et al., 2009; Kreuzer et al., 1996). Taken together, these previous findings indicate overlapping immunepathophysiological mechanisms across severe mental disorders, in line with recent reviews (Baumeister et al., 2014). However, markers of inflammation and immune activation have not yet been evaluated across the whole psychosis continuum. Further, while there is evidence suggesting that inflammatory markers are influenced by antipsychotic medication (Miller et al., 2011; Tourjman et al., 2013); few studies have taken a potential medication effect into account when analyzing inflammatory markers in these patients(Upthegrove et al., 2014). In the present study we re-evaluate these circulating inflammatory markers in a sample size around 3 times the original, including a spectrum of diagnoses reflecting the psychosis continuum represented by schizophrenia, schizoaffective and affective diagnostic groups as well as healthy controls. As there are several demographic differences across the psychosis spectrum, the larger sample size allows us to maintain sufficient power to adjust for a range of possible confounders including an evaluation of the impact of medication.

2. METHODS 2.1 Setting and participants The current study is part of the ongoing Thematically Organized Psychosis (TOP) Study which includes patients from hospitals in the Oslo region, as well as Stavanger, Trondheim and South-East regional hospitals in Norway (Hope et al., 2011; Hope et al., 2015; Simonsen et al., 2011; Tesli et al., 2014). The main criterion of inclusion in the TOP Study is a diagnosis of a severe mental disorder including schizophrenia spectrum disorder, bipolar spectrum disorder or major depressive disorder with psychotic features according to DSM-IV. In addition, the participants have to be between 18-65 years and able to give a written informed consent. They all undergo the same protocol, which includes a clinical, psychiatric, and physical examination, neuropsychological testing, and collection of blood samples. The healthy controls were randomly selected from statistical records (www.ssb.no) from the same catchment areas as the patients. Healthy control subjects were aged between 18-60 years. 5

Ragni Mørch

They were included if they had no history of severe mental disorders, head injury, neurological disorders, illicit drug use, close relatives with severe mental disorder or medical problems that somehow could interfere with brain function (i.e. hypothyroidism, uncontrolled hypertension and diabetes) (Dieset et al., 2012a; Dieset et al., 2012b; Hope et al., 2011; Hope et al., 2015; Hope et al., 2009; Hope et al., 2013). All participants have given their written consent and the study was approved by the Norwegian Scientific Ethical Committees and the Norwegian Data Protection Agency and conducted in accordance with the Declaration of Helsinki. 2.2 Sample A total of 992 patients with severe mental disorder and 638 healthy controls were included in the current study and the patients were divided into three groups along the psychosis continuum, in line with previously defined categories(Craddock and Owen, 2010; Tesli et al., 2014). Schizophrenia (n=378), schizophreniform disorder (n=31), and other psychoses (n=175; brief psychotic episode, delusional disorder, psychosis not otherwise specified) were combined in the “schizophrenia group” (SCZ, n=584), schizoaffective disorder was included in the “schizoaffective group” (SA, n=93), and bipolar 1 disorder (n=179), bipolar 2 disorder (n=73), bipolar disorder not otherwise specified (n=13), and major depressive disorder (n=50), were combined in the “affective group” (AFF, n= 315). All patients with a diagnosis of a major depressive disorder according to DSM-IV had psychotic symptoms along with their depressive episode. Participants with cancer, autoimmune or inflammatory diseases, ongoing infections (CRP above 20 mg/L) or receiving treatment with possible immune modulating or suppressant drugs were excluded (n=108). 2.3 Assessments The patients were interviewed and sociodemographic history, medical history, substance use, psychiatric symptoms, medication and potential side effects were recorded. They all underwent diagnostic interviews based on Structured Clinical Interview in DSM-IV axis I Disorders (SCID) and symptom assessments. Diagnostic evaluation was performed by trained psychologists and psychiatrists, all of whom participated regularly in diagnostic meetings supervised by professors in psychiatry. The reliability was very good regarding diagnostic assessment, and the overall agreement for the DSM-IV diagnostic categories tested was 82% and the overall Kappa 0.77 (95% CI: 0.60–0.94) (Simonsen et al., 2011).

6

Ragni Mørch

The healthy controls were interviewed for history of severe mental disorder themselves or in their family, and assessed with Primary Care Evaluation of Mental Disorders (PRIME MD). The clinical assessment of the patients and healthy controls participating in this study is described in more detail elsewhere (Dieset et al., 2012a; Djurovic et al., 2010). All participants were screened with history taking and routine blood tests, including weight, height, body mass index (BMI), smoking habits, alcohol consumption and use of illicit drugs in the 2 weeks period prior to blood sampling, in addition all patients underwent a physical examination.

2.4 Inflammatory markers Plasma levels of sTNF-R1, IL-1Ra, OPG (R&D systems, Stillwater, MN, USA) and vWf (Dako, Glostrup, Denmark) were measured in duplicate by enzyme immunoassays (EIA) in the laboratory at Research Institute of Internal Medicine, Oslo University Hospital with intraand inter-assay coefficients of variation <10% for all EIAs. These markers had been measured previously in a sub-sample of the current sample (n=465 participants) (Hope et al., 2009). These plasma samples were re-analyzed in the current study. For immunoassays blood was taken using EDTA vials and the plasma was isolated the next working day and stored at -80°C. Blood samples were drawn in the morning (average at 1000h) or in the afternoon (average at 1500h), with the majority of the patients in the morning and the majority of controls in the afternoon. Most participants were fasting during blood collection, but only before the morning blood sampling.

2.5 Medication The information regarding prescribed medications including antipsychotics, antidepressants, mood stabilizers and lithium used by patients were obtained by clinical interview and hospital records. Patients were instructed not to take their morning dose of prescribed medication prior to blood sampling according to standard protocol. Serum concentrations of medications were analyzed at the Department of Clinical Pharmacology, St. Olav’s Hospital, Trondheim (Jonsdottir et al., 2009). The reference range for each drug has been derived at the laboratory based on their extensive database and long experience with measuring psychotropic drugs. When considering serum concentrations of psychotropic medicines the concentration/dose ratio was used as this gives the best picture of drug intake (Jonsdottir et al., 2009). By using the concentration/dose ratio the serum concentration is divided by daily dosage and thus corresponds to the serum concentration per mg of the medication taken daily by the individual. 7

Ragni Mørch

By using this measure, the values can be compared directly between subjects independently of different drug treatment regimens. We also calculated “defined daily dosages” (DDD) according to the World Health Organization (WHO) principles. The DDD is the assumed average maintenance dose per day for a drug used for its main indication in adults and provide a fixed unit of measurement independent of dosage form (http://www.whocc.no/atc_ddd_index/). Drug consumption data presented in DDDs only give a rough estimate of consumption and not an exact picture of actual use. In our sample 929 patients were using psychotropic drugs (n=268 monotherapy) and we investigated if the different psychotropic drug groups (antipsychotics, antidepressants, Lithium and mood stabilizers) correlated with the inflammatory markers. As a large number of patients (n=211) were treated with olanzapine (n=99 monotherapy), we also investigated specifically if serumconcentrations of olanzapine correlated with the inflammatory markers as olanzapine is a widely used antipsychotic globally, and is effective in the treatment of both psychotic disorders and manic episodes.

2.6 Statistical Analyses All statistical analyses were done using the SPSS software package for Windows, version 22.0 (SPSS Chicago. USA). All analyses were two-tailed with a level of significance set at p˂0.05. Data normality was assessed using the Kolmogorov–Smirnov test. Preliminary checks were done to rule out any violations of assumptions of the analyses. The inflammatory markers, BMI and age had a skewed distribution and were logarithmic transformed before further analyses. We used analysis of variance model (ANOVA) and Chi-square to investigate group differences in demographic variables. We performed bivariate correlations using Pearson coefficients in patient and healthy control groups separately to investigate potential known confounders according to literature and previous studies (Hope et al., 2010; Hope et al., 2009) and investigated age, sex, ethnicity, smoking, BMI, alcohol consumption, use of illicit drugs, time of blood sampling and use of different medication groups (antipsychotics, mood stabilizers, lithium and antidepressants) in relation to the inflammatory markers. We identified BMI as a confounder for sTNF-R1 and thus included BMI as a covariate in the subsequent analysis of covariance (ANCOVA) for sTNF-R1. Age and sex were identified only as confounders for OPG, but due to literature we included age and sex in the subsequent analyses for all inflammatory markers. We investigated correlations of core symptoms assessed with the Positive and Negative Syndrome Scale total (PANSS), Young Mania Rating Scale (YMRS), Inventory of Depressive Symptoms (IDS) in the total patient group and the 8

Ragni Mørch

inflammatory markers. In addition, we tested for association between use of different medication groups (cumulative DDD of antipsychotics, mood stabilizers, lithium and antidepressants) and serum concentration of olanzapine and inflammatory markers by doing bivariate correlations with Pearson coefficients. In order to rule out any potential diagnostic effects on relationships between inflammatory markers and medication, we repeated these analyses in the separate diagnostic groups of the psychosis continuum. We also did additional sub-analyses with use of the different medication groups as categorical covariates in the ANCOVA model. Furthermore, we investigated if there were any differences in correlations between first and second generation antipsychotics, and the inflammatory markers or for those on monotherapy, and the inflammatory markers. We applied an ANCOVA model with Bonferroni correction for multiple comparisons, to evaluate the differences in inflammatory markers between patients and healthy controls, and across the 3 groups of the psychosis continuum (schizophrenia, schizoaffective, and affective groups) and healthy controls. The inflammatory markers were used as dependent variables, patients and healthy controls as fixed factor in the first analysis, and the psychosis continuum groups (SCZ, SA and AFF) and healthy controls (HC) as fixed factors in the second analyses. For the inflammatory markers different covariates were used in the ANCOVA model depending on the results from the previous correlation analyses in which possible confounders were identified. Lastly, we did sub-analyses in which we excluded both patients and healthy controls with glucose higher than 7mmol/L to rule out any effect of fasting variations. In order to rule out possible confounding effect of the inflammatory markers in relation to time of blood sampling, we investigated if the levels varied at different times during the day by making 3 time categories (morning, mid-day and afternoon) for patients and healthy controls separately by conducting ANOVA, and found no significant effects except for sTNF-R1 in morning and afternoon samples in HC group. We did further sub-analysis with ANCOVA with matched morning samples of HC (n=142) and patients (n=273) controlling for age, sex and BMI and the results remained the same with significantly higher levels in patients compared to HC. Thus, time of blood sampling was not included as a confounder in the main statistical analyses. When calculating Cohen’s d we used online effect size calculators (http://www.uccs.edu/~lbecker/). Effect size is considered small, medium or large depending on if the levels are in between 0.1-0.3, 0.3-0.5 or more than 0.5, respectively.

9

Ragni Mørch

3. RESULTS 3.1 Clinical characteristics The clinical characteristics of the study population are shown in Table 1. The mean age of the HC and the AFF group were nearly the same, while the SCZ and SA groups were slightly younger. There were more men in the SCZ and the HC groups than in the AFF and SA groups. Overall, more patients smoked, had non-European ethnicity, were overweight and had used illicit drugs in the previous 2 weeks prior to inclusion compared to HC. Finally, the SCZ and SA groups consumed less alcohol than HC. 3.2 Inflammatory markers across the diagnostic groups and healthy controls We found significantly increased levels of sTNF-R1 (p=1.8 x10-8, d= 0.23) and IL-1Ra (p=0.002, d= 0.16), but not for OPG and vWf, in the total patient group compared to HC after adjusting for confounders (i.e., age, sex, BMI; see Statistical methods) and multiple comparisons (Table 2). Results of inflammatory markers between psychosis continuum groups and HC after adjusting for confounders and multiple comparisons are presented in Table 2 and figure 1A-D. Levels of sTNF-R1 (p=8.5x10-8, d =0.27) and IL-1Ra (p=5.9x10-5, d=0.25), but not OPG and vWf levels, were significantly higher in the SCZ group compared to HC. sTNF-R1 levels were significantly higher in the SA group (p=0.01, d=0.3) and in the AFF group (p=0.002, d=0.12) compared to HC, while there were no significant findings regarding IL-1Ra, vWf or OPG in these diagnostic groups. Levels of IL-1Ra (p=0.004, d=0.25) and vWf (p=0.02, d=0.21) were different between the SCZ and the AFF group. There were no other significant differences between the diagnostic groups. In the sub-analyses, in which we excluded all participants (n=37) with glucose higher than 7mmol/L, the main results remained the same. The correlation analyses of core symptoms and inflammatory markers did not reveal any strong associations, only a weak association (r=0.090.10) between OPG and PANSS total and IDS, which did not remain significant after correction of multiple testing. The results are presented in supplemental file (Table 4).

10

Ragni Mørch

3.3 Medication and inflammatory markers The bivariate analyses showed no significant correlations between DDD or concentration/dose ratio of antipsychotics, lithium or serum-concentrations of olanzapine with any of the inflammatory markers (Table 3). We found trend level correlations between antidepressants and IL-1Ra (r=-0.10, p=0.09) and mood stabilizers and vWf (r=-0.14, p=0.07). The sub-analyses did not reveal any significant correlations between inflammatory markers and the DDDs of psychotropic drug groups (antipsychotics, antidepressants, Lithium, mood stabilizers) or serum concentration of olanzapine in patients or in the separate diagnostic groups (schizophrenia, schizoaffective and affective groups) of the psychosis continuum. Further, there were no significant correlations between inflammatory markers and antipsychotic monotherapy (including olanzapine) DDDs or serum concentration. In addition, there were no significant differences between first and second generation antipsychotic agents in relation to the inflammatory markers in the bivariate correlation analysis. Further, when all the different medication groups were used as categorical covariates in addition to age and sex in the ANCOVA model, the main results with no significant associations did not change.

4. DISCUSSION 4.1 Main finding The main finding of the current study was a robust increase of sTNF-R1 and IL-1Ra levels in patients with severe mental disorders, while we did not replicate previous findings of elevated vWf or OPG in this larger study population. The highest levels of these inflammatory markers were found in SCZ, followed by SA and AFF groups, but the effect size estimates were small (Cohen’s d =0.1-0.3). Both sTNF-R1 and IL-1Ra were increased in SCZ, while only sTNFR1 was increased in the SA and AFF groups. We did not find any significant association between antipsychotics and inflammatory markers. These findings support a continuum psychosis model related to inflammation.

4.2 Psychosis Continuum Model Our results of different patterns of inflammatory markers indicate that inflammation is related to the severity of psychotic features along the psychosis continuum model. This is supported 11

Ragni Mørch

by previous findings of an association between inflammatory markers and disease severity in psychotic disorders (Hope et al., 2011). Further, recent genetic evidence suggest different associations of the HLA region in bipolar disorder and schizophrenia (Andreassen et al., 2015) which could represent risk factors underlying the pattern of inflammatory markers reported here. Others demonstrate overlap and differences between schizophrenia, schizoaffective and affective spectrum disorders along the psychosis continuum with brain volumetric reductions (Rimol et al., 2010), neurocognitive impairment (Simonsen et al., 2011), functional brain imaging studies investigating brain networks (Brandt et al., 2014), neurodevelopmental adversity (Akabaliev et al., 2014), polygenic risk scores (Tesli et al., 2014) and molecular validation of the schizophrenia spectrum (Ruderfer et al., 2014). The results from these different domains indicate a psychosis continuum in line with a more dimensional approach to psychiatric diagnoses as the RDoC initiative also proposes (Cuthbert, 2015). Even though alterations of inflammatory markers in severe mental disorders have most consistently been demonstrated during acute disease phases, some studies have shown that aberrations are persistent during more stable illness phases (Hope et al., 2013; Pedrini et al., 2012). A metaanalysis has shown that sTNF-R1 may be a trait that does not vary much during acute phases, during treatment and/or in remission of disease (Miller et al., 2011) and this is in line with the present results regarding sTNF-R1. Patients included in the current study were usually in a more stable phase in their illness as they must be able to provide an informed consent. Thus, we cannot rule out that the levels of inflammatory markers could be higher in the more acute phases of illness. However, high levels of inflammatory markers have been suggested to be a trait influenced by genetic factors (Rafiq et al., 2007; Vistoropsky et al., 2010) and/or influenced by duration of illness (Dickerson et al., 2016). Our correlation analysis of the symptom assessments with PANSS total, IDS and YMRS did not reveal any significant correlations with the inflammatory markers except for OPG. Schizophrenia consists of many dimensions, signs and symptoms which may not be reflected in these assessments and longitudinal studies of both acute and more chronic phases of illness would probably provide more insight into the role of inflammatory markers in severe mental disorders and to what extent they are related to psychotic dimensions. Taken together, the present findings support a role of the immune-system along the psychosis continuum that should be followed up with further studies.

12

Ragni Mørch

4.3 Inflammatory Markers To the best of our knowledge, the present study provides the largest sample to date investigating inflammatory markers across the psychosis spectrum in one analysis. In addition, our sample is well characterized allowing adjustment for confounders, and has a high diagnostic reliability further strengthening the results. The current findings of increased sTNF-R1 in psychotic disorders are in line with previous results (Bergink et al., 2014; Drexhage et al., 2010; Goldstein et al., 2009; Hope et al., 2009; Hope et al., 2013; Munkholm et al., 2013; Potvin et al., 2008). The increase in IL-1Ra, which was not found in our previous smaller study, may support a role of IL-1 related pathways in schizophrenia (Hope et al., 2011; Hope et al., 2013). Although vWf tended to be higher in the patient group, we did not confirm our previous finding of raised levels of vWf and OPG in patients with severe psychiatric disorders, underscoring the need for investigating large samples as the present before making firm conclusion. Some of the present results are inconsistent with the literature regarding the AFF group where more profound abnormalities in inflammatory markers have been reported (Baumeister et al., 2014; Goldstein et al., 2009; Hope et al., 2011; Munkholm et al., 2013). Possibly some of the discrepancies are due to random variation found in studies of small sample sizes. Patient characteristics may also play a role, as the current sample mainly included patients in a stable phase. Moreover, not all these previous studies have controlled for all relevant confounders. Increasing evidence support an abnormal inflammation and immune activation in patients with schizophrenia with resulting aberrant levels of cytokines or other inflammatory markers (Baumeister et al., 2014; Hope et al., 2009; Hope et al., 2013; Potvin et al., 2008), but the exact pathophysiological mechanisms are not known. The inflammatory markers chosen for the present study represent different aspects of immune mechanisms as described in the introduction. It is also possible that other parts of the immune system are involved, as aberrant levels of other inflammatory markers such as interleukin-6 (IL-6), interleukin-1β (IL-1β), soluble interleukin-2 receptor (sIL-2R) and brain-derived neurotrophic factor (BDNF) have been reported in patients with severe mental disorders. This underscores the complex role of immune pathways in these disorders, and further studies are needed to disentangle the different mechanisms. Although we have no mechanistic data, the present findings of enhanced levels of sTNF-R1 and IL-1Ra, as markers of activity in the TNF and IL-1 systems, may support a role for activation of these pathways in the pathogenesis of the disorders. TNF is implicated in many illnesses like rheumatoid arthritis, inflammatory bowel disease, 13

Ragni Mørch

Alzheimer’s disease and cancer, and has also been associated with endothelial cell activation and endothelial leakage, neurotoxicity and neuroplasticity (Golan et al., 2004; Stellwagen and Malenka, 2006; Viviani et al., 2007). TNF interacts with dopaminergic and serotonergic pathways that are the main targets for medication used for psychiatric illnesses(Werkman et al., 2006) and it may be involved in oxidative stress processes in hippocampal regions(Grinberg et al., 2013). IL-1 related pathways have also been shown to interact with the same serotonergic and dopaminergic pathways as TNF (Tsao et al., 2008). IL-1Ra has been proposed as a possible treatment biomarker in schizophrenia patients as the levels were elevated in the acute phases and lower after treatment with antipsychotics (Miller et al., 2011). sTNF-R1 has been proposed as a trait biomarker that remains elevated in psychosis patients despite treatment and changing illness phases (Miller et al., 2011). The current findings suggest that sTNF-R1 may have a similar role in affective disorders. Several studies on neuroinflammation have found a relationship between inflammatory markers and white matter pathology in patients with schizophrenia which may contribute to structural and functional disconnectivity (Najjar and Pearlman, 2015; Prasad et al., 2015). Neuroimaging studies have also found an association between carrier status for functional single nucleotide polymorphism in the interleukin-1β gene and abnormal white matter volumes in schizophrenia patients compared to controls (Najjar and Pearlman, 2015). Further, there is accumulating evidence that points toward communication pathways between peripheral immune system and brain involving inflammatory markers like cytokines and endothelial cells in the BBB (Khandaker and Dantzer, 2015). Further studies should clarify if these molecules are not only markers, but also mediators and if they could give information in relation to prognosis and treatment effect in these patients.

4.4 Medication We had a unique opportunity to investigate the relationship between amount of drugs used, in the form of DDD and serum concentrations. We did not find any significant correlations between inflammatory markers and the use of psychotropic medication in this large sample, but some studies have suggested that antipsychotic medication increase the levels of inflammatory markers (Drzyzga et al., 2006), while others indicate that antipsychotics and mood stabilizers may have anti-inflammatory effects (Kato et al., 2011; Tourjman et al., 2013). Olanzapine and clozapine might influence levels of inflammatory markers due to 14

Ragni Mørch

adverse effects increasing the risk for metabolic syndrome (Kluge et al., 2009), and some have proposed that studies of inflammatory markers should take medication effects into account (Baumeister et al., 2014). However, the relationship between inflammatory markers and medication in psychiatric disorders is not yet fully understood. Different studies have focused on different inflammatory markers in patients with severe mental disorders and the sample size has often been relatively small. In the present study we analyzed the effect of both dosages and serum concentrations of antipsychotics, mood stabilizers, antidepressants and/or lithium and inflammatory markers, but found no impact of these medications on inflammation. Our results support that these medications, including olanzapine, do not influence the activation of inflammatory pathways. Our large sample allowed us to control for confounders like age, sex and BMI which is in line with a previous meta-analysis of IL-1Ra and TNF-α (Tourjman et al., 2013). Despite the lack of large randomized studies, the present and previous clinical findings indicate that much of the apparent association between inflammatory markers and antipsychotic medication is driven by the disorder itself. Better longitudinal studies before and after initiation of medication will help to address this question.

4.5 Limitations There are some limitations to this study that need to be addressed. Firstly, the naturalistic and cross-sectional design makes it difficult to draw firm conclusions regarding causality. Although the sample is well described and we have controlled for a range of confounding factors, we cannot rule out unknown factors that could have influenced the results. Some argue that comorbid alcohol dependence should be excluded from immune-studies (Baumeister et al., 2014), but as we found no association to cytokine levels we did not exclude those patients. Secondly, the measurements of inflammatory markers in peripheral blood may not necessarily reflect the inflammatory activity in the brain. However, studies have shown high correlation between biomarkers in the CSF and the periphery (Tomasik et al., 2014), and there is accumulating evidence indicating communication across the blood brain barrier (Kipnis et al., 2012; Louveau et al., 2015). Finally, as plasma was not isolated immediately after blood collection, we cannot exclude ex vivo activation of inflammatory pathways and thus the markers. However, since samples were processed similarly in patients and controls, this effect would be random and more likely attenuate any differences between the study groups. 15

Ragni Mørch

4.6 Conclusion The current results support previous findings of increased levels of the inflammatory markers sTNF-R1 and IL-1Ra in severe mental disorders. Moreover, in this large sample we also show that levels of inflammatory activity are in line with the psychosis continuum model with the strongest increase in schizophrenia and a relatively lesser increase in schizoaffective and affective disorders respectively. Longitudinal studies are needed to investigate if the markers of inflammation can be used as potential biomarkers and to evaluate if the levels are state and/or trait dependent. Further, experimental studies as well as clinical proof-of-concept studies are needed to determine the potential role of anti-inflammatory agents in the management of these disorders, potentially representing a new treatment strategy in patients with severe mental disorders.

Declaration of interest OAA has received Speaker’s honorarium from Otsuka, Janssen, BMS and GSK. All other authors report no conflicts of interest.

Acknowledgements The authors would like to thank the participants of the study for their contribution, and the clinicians who were involved in patient recruitment and clinical assessments.

16

Ragni Mørch

References Akabaliev, V.H., Sivkov, S.T., Mantarkov, M.Y., 2014. Minor physical anomalies in schizophrenia and bipolar I disorder and the neurodevelopmental continuum of psychosis. Bipolar Disord. 16, 633-641. Andreassen, O.A., Harbo, H.F., Wang, Y., Thompson, W.K., Schork, A.J., Mattingsdal, M., Zuber, V., Bettella, F., Ripke, S., Kelsoe, J.R., Kendler, K.S., O'Donovan, M.C., Sklar, P., McEvoy, L.K., Desikan, R.S., Lie, B.A., Djurovic, S., Dale, A.M., 2015. Genetic pleiotropy between multiple sclerosis and schizophrenia but not bipolar disorder: differential involvement of immune-related gene loci. Mol Psychiatry. 20, 207-214. Aukrust, P., Sandberg, W.J., Otterdal, K., Vinge, L.E., Gullestad, L., Yndestad, A., Halvorsen, B., Ueland, T., 2011. Tumor necrosis factor superfamily molecules in acute coronary syndromes. Ann Med. 43, 90-103. Baumeister, D., Russell, A., Pariante, C.M., Mondelli, V., 2014. Inflammatory biomarker profiles of mental disorders and their relation to clinical, social and lifestyle factors. Soc Psychiatry Psychiatr Epidemiol. 49, 841-849. Bergink, V., Gibney, S.M., Drexhage, H.A., 2014. Autoimmunity, inflammation, and psychosis: a search for peripheral markers. Biol Psychiatry. 75, 324-331. Brandt, C.L., Eichele, T., Melle, I., Sundet, K., Server, A., Agartz, I., Hugdahl, K., Jensen, J., Andreassen, O.A., 2014. Working memory networks and activation patterns in schizophrenia and bipolar disorder: comparison with healthy controls. Br J Psychiatry. 204, 290-298. Corallini, F., Rimondi, E., Secchiero, P., 2008. TRAIL and osteoprotegerin: a role in endothelial physiopathology? Front Biosci. : a journal and virtual library 13, 135-147. Craddock, N., O'Donovan, M.C., Owen, M.J., 2009. Psychosis genetics: modeling the relationship between schizophrenia, bipolar disorder, and mixed (or "schizoaffective") psychoses. Schizophr Bull. 35, 482-490. Craddock, N., Owen, M.J., 2010. The Kraepelinian dichotomy - going, going... but still not gone. Br J Psychiatry. 196, 92-95. Cuthbert, B.N., 2015. Research Domain Criteria: toward future psychiatric nosologies. Dialogues Clin Neurosci. 17, 89-97. Denis, C.V., 2002. Molecular and cellular biology of von Willebrand factor. Int J Hematol. 75, 3-8. Dickerson, F., Stallings, C., Origoni, A., Schroeder, J., Katsafanas, E., Schweinfurth, L., Savage, C., Khushalani, S., Yolken, R., 2016. Inflammatory Markers in Recent Onset Psychosis and Chronic Schizophrenia. Schizophr Bull. 42, 134-141. Dieset, I., Djurovic, S., Tesli, M., Hope, S., Mattingsdal, M., Michelsen, A., Joa, I., Larsen, T.K., Agartz, I., Melle, I., Rossberg, J.I., Aukrust, P., Andreassen, O.A., Ueland, T., 2012a. Up-regulation of NOTCH4 gene expression in bipolar disorder. Am J Psychiatry. 169, 12921300. 17

Ragni Mørch

Dieset, I., Haukvik, U.K., Melle, I., Rossberg, J.I., Ueland, T., Hope, S., Dale, A.M., Djurovic, S., Aukrust, P., Agartz, I., Andreassen, O.A., 2014. Association between altered brain morphology and elevated peripheral endothelial markers - Implications for psychotic disorders. Schizophr Res. 161, 222-228. Dieset, I., Hope, S., Ueland, T., Bjella, T., Agartz, I., Melle, I., Aukrust, P., Rossberg, J.I., Andreassen, O.A., 2012b. Cardiovascular risk factors during second generation antipsychotic treatment are associated with increased C-reactive protein. Schizophr Res. 140, 169-174. Djurovic, S., Gustafsson, O., Mattingsdal, M., Athanasiu, L., Bjella, T., Tesli, M., Agartz, I., Lorentzen, S., Melle, I., Morken, G., Andreassen, O.A., 2010. A genome-wide association study of bipolar disorder in Norwegian individuals, followed by replication in Icelandic sample. J Affect Disord. 126, 312-316. Drexhage, R.C., Knijff, E.M., Padmos, R.C., Heul-Nieuwenhuijzen, L., Beumer, W., Versnel, M.A., Drexhage, H.A., 2010. The mononuclear phagocyte system and its cytokine inflammatory networks in schizophrenia and bipolar disorder. Expert Rev Neurother. 10, 5976. Drzyzga, L., Obuchowicz, E., Marcinowska, A., Herman, Z.S., 2006. Cytokines in schizophrenia and the effects of antipsychotic drugs. Brain Behav Immun. 20, 532-545. Golan, H., Levav, T., Mendelsohn, A., Huleihel, M., 2004. Involvement of tumor necrosis factor alpha in hippocampal development and function. Cereb Cortex. 14, 97-105. Goldstein, B.I., Kemp, D.E., Soczynska, J.K., McIntyre, R.S., 2009. Inflammation and the phenomenology, pathophysiology, comorbidity, and treatment of bipolar disorder: a systematic review of the literature. J Clin Psychiatry. 70, 1078-1090. Grinberg, Y.Y., Dibbern, M.E., Levasseur, V.A., Kraig, R.P., 2013. Insulin-like growth factor-1 abrogates microglial oxidative stress and TNF-alpha responses to spreading depression. J Neurochem. 126, 662-672. Gu, Y., Zeleniuch-Jacquotte, A., Linkov, F., Koenig, K.L., Liu, M., Velikokhatnaya, L., Shore, R.E., Marrangoni, A., Toniolo, P., Lokshin, A.E., Arslan, A.A., 2009. Reproducibility of serum cytokines and growth factors. Cytokine 45, 44-49. Hope, S., Dieset, I., Agartz, I., Steen, N.E., Ueland, T., Melle, I., Aukrust, P., Andreassen, O.A., 2011. Affective symptoms are associated with markers of inflammation and immune activation in bipolar disorders but not in schizophrenia. J Psychiatr Res. 45, 1608-1616. Hope, S., Hoseth, E., Dieset, I., Morch, R.H., Aas, M., Aukrust, P., Djurovic, S., Melle, I., Ueland, T., Agartz, I., Ueland, T., Westlye, L.T., Andreassen, O.A., 2015. Inflammatory markers are associated with general cognitive abilities in schizophrenia and bipolar disorder patients and healthy controls. Schizophr Res. 165, 188-194. Hope, S., Melle, I., Aukrust, P., Agartz, I., Lorentzen, S., Steen, N.E., Djurovic, S., Ueland, T., Andreassen, O.A., 2010. Osteoprotegerin levels in patients with severe mental disorders. J Psychiatry Neurosci.: JPN 35, 304-310. 18

Ragni Mørch

Hope, S., Melle, I., Aukrust, P., Steen, N.E., Birkenaes, A.B., Lorentzen, S., Agartz, I., Ueland, T., Andreassen, O.A., 2009. Similar immune profile in bipolar disorder and schizophrenia: selective increase in soluble tumor necrosis factor receptor I and von Willebrand factor. Bipolar Disord. 11, 726-734. Hope, S., Ueland, T., Steen, N.E., Dieset, I., Lorentzen, S., Berg, A.O., Agartz, I., Aukrust, P., Andreassen, O.A., 2013. Interleukin 1 receptor antagonist and soluble tumor necrosis factor receptor 1 are associated with general severity and psychotic symptoms in schizophrenia and bipolar disorder. Schizophr Res. 145, 36-42. Jonsdottir, H., Friis, S., Horne, R., Pettersen, K.I., Reikvam, A., Andreassen, O.A., 2009. Beliefs about medications: measurement and relationship to adherence in patients with severe mental disorders. Acta Psychiatr Scand. 119, 78-84. Kato, T.A., Monji, A., Mizoguchi, Y., Hashioka, S., Horikawa, H., Seki, Y., Kasai, M., Utsumi, H., Kanba, S., 2011. Anti-Inflammatory properties of antipsychotics via microglia modulations: are antipsychotics a 'fire extinguisher' in the brain of schizophrenia? Mini Rev Med Chem. 11, 565-574. Keshavan, M.S., Morris, D.W., Sweeney, J.A., Pearlson, G., Thaker, G., Seidman, L.J., Eack, S.M., Tamminga, C., 2011. A dimensional approach to the psychosis spectrum between bipolar disorder and schizophrenia: the Schizo-Bipolar Scale. Schizophr Res. 133, 250-254. Khandaker, G.M., Dantzer, R., 2015. Is there a role for immune-to-brain communication in schizophrenia? Psychopharmacology in press. DOI: 10.1007/s00213-015-3975-1 Kipnis, J., Gadani, S., Derecki, N.C., 2012. Pro-cognitive properties of T cells. Nat Rev Immunol. 12, 663-669. Kluge, M., Schuld, A., Schacht, A., Himmerich, H., Dalal, M.A., Wehmeier, P.M., HinzeSelch, D., Kraus, T., Dittmann, R.W., Pollmacher, T., 2009. Effects of clozapine and olanzapine on cytokine systems are closely linked to weight gain and drug-induced fever. Psychoneuroendocrinology 34, 118-128. Kreuzer, K.A., Rockstroh, J.K., Sauerbruch, T., Spengler, U., 1996. A comparative study of different enzyme immunosorbent assays for human tumor necrosis factor-alpha. J Immunol Methods. 195, 49-54. Lee, S.H., Ripke, S., Neale, B.M., Faraone, S.V., Purcell, S.M. et al., 2015. Structural and functional features of central nervous system lymphatic vessels. Nature 523, 337-341. Maier, W., Zobel, A., Wagner, M., 2006. Schizophrenia and bipolar disorder: differences and overlaps. Curr Opin Psychiatry. 19, 165-170. McGrath, J., Saha, S., Chant, D., Welham, J., 2008. Schizophrenia: a concise overview of incidence, prevalence, and mortality. Epidemiologic reviews 30, 67-76. Michael, B.D., Griffiths, M.J., Granerod, J., Brown, D., Keir, G., Wnek, M., Cox, D.J., Vidyasagar, R., Borrow, R., Parkes, L.M., Solomon, T., 2015. The interleukin-1 balance is 19

Ragni Mørch

associated with clinical severity, blood-brain barrier permeability, neuroimaging changes and outcome in encephalitis. J Infect Dis. in press. DOI: 10.1093/infdis/jiv771 Miller, B.J., Buckley, P., Seabolt, W., Mellor, A., Kirkpatrick, B., 2011. Meta-analysis of cytokine alterations in schizophrenia: clinical status and antipsychotic effects. Biol Psychiatry. 70, 663-671. Munkholm, K., Brauner, J.V., Kessing, L.V., Vinberg, M., 2013. Cytokines in bipolar disorder vs. healthy control subjects: a systematic review and meta-analysis. J Psychiatr Res. 47, 1119-1133. Najjar, S., Pearlman, D.M., 2015. Neuroinflammation and white matter pathology in schizophrenia: systematic review. Schizophr Res. 161, 102-112. Pedrini, M., Massuda, R., Fries, G.R., de Bittencourt Pasquali, M.A., Schnorr, C.E., Moreira, J.C., Teixeira, A.L., Lobato, M.I., Walz, J.C., Belmonte-de-Abreu, P.S., Kauer-Sant'Anna, M., Kapczinski, F., Gama, C.S., 2012. Similarities in serum oxidative stress markers and inflammatory cytokines in patients with overt schizophrenia at early and late stages of chronicity. J Psychiatr Res. 46, 819-824. Potvin, S., Stip, E., Sepehry, A.A., Gendron, A., Bah, R., Kouassi, E., 2008. Inflammatory cytokine alterations in schizophrenia: a systematic quantitative review. Biol Psychiatry. 63, 801-808. Prasad, K.M., Upton, C.H., Nimgaonkar, V.L., Keshavan, M.S., 2015. Differential susceptibility of white matter tracts to inflammatory mediators in schizophrenia: an integrated DTI study. Schizophr Res. 161, 119-125. Rafiq, S., Stevens, K., Hurst, A.J., Murray, A., Henley, W., Weedon, M.N., Bandinelli, S., Corsi, A.M., Guralnik, J.M., Ferruci, L., Melzer, D., Frayling, T.M., 2007. Common genetic variation in the gene encoding interleukin-1-receptor antagonist (IL-1RA) is associated with altered circulating IL-1RA levels. Genes Immun. 8, 344-351. Rimol, L.M., Hartberg, C.B., Nesvag, R., Fennema-Notestine, C., Hagler, D.J., Jr., Pung, C.J., Jennings, R.G., Haukvik, U.K., Lange, E., Nakstad, P.H., Melle, I., Andreassen, O.A., Dale, A.M., Agartz, I., 2010. Cortical thickness and subcortical volumes in schizophrenia and bipolar disorder. Biol Psychiatry. 68, 41-50. Ruderfer, D.M., Fanous, A.H., Ripke, S., McQuillin, A., Amdur, R.L., Gejman, P.V., O'Donovan, M.C., Andreassen, O.A., Djurovic, S., Hultman, C.M., Kelsoe, J.R., Jamain, S., Landen, M., Leboyer, M., Nimgaonkar, V., Nurnberger, J., Smoller, J.W., Craddock, N., Corvin, A., Sullivan, P.F., Holmans, P., Sklar, P., Kendler, K.S., 2014. Polygenic dissection of diagnosis and clinical dimensions of bipolar disorder and schizophrenia. Mol Psychiatry. 19, 1017-1024. Simonsen, C., Sundet, K., Vaskinn, A., Birkenaes, A.B., Engh, J.A., Faerden, A., Jonsdottir, H., Ringen, P.A., Opjordsmoen, S., Melle, I., Friis, S., Andreassen, O.A., 2011. Neurocognitive dysfunction in bipolar and schizophrenia spectrum disorders depends on history of psychosis rather than diagnostic group. Schizophr Bull. 37, 73-83.

20

Ragni Mørch

Stellwagen, D., Malenka, R.C., 2006. Synaptic scaling mediated by glial TNF-alpha. Nature 440, 1054-1059. Tesli, M., Espeseth, T., Bettella, F., Mattingsdal, M., Aas, M., Melle, I., Djurovic, S., Andreassen, O.A., 2014. Polygenic risk score and the psychosis continuum model. Acta Psychiatr Scand. 130, 311-317. Tomasik, J., Rahmoune, H., Guest, P.C., Bahn, S., 2014. Neuroimmune biomarkers in schizophrenia. Schizophr Res. in press. DOI: 10.1016/j.schres.2014.07.025 Tourjman, V., Kouassi, E., Koue, M.E., Rocchetti, M., Fortin-Fournier, S., Fusar-Poli, P., Potvin, S., 2013. Antipsychotics' effects on blood levels of cytokines in schizophrenia: a meta-analysis. Schizophr Res. 151, 43-47. Tsao, C.W., Lin, Y.S., Cheng, J.T., Lin, C.F., Wu, H.T., Wu, S.R., Tsai, W.H., 2008. Interferon-alpha-induced serotonin uptake in Jurkat T cells via mitogen-activated protein kinase and transcriptional regulation of the serotonin transporter. J Psychopharmacol. 22, 753760. Upthegrove, R., Manzanares-Teson, N., Barnes, N.M., 2014. Cytokine function in medication-naive first episode psychosis: a systematic review and meta-analysis. Schizophrenia research 155, 101-108. Vistoropsky, Y., Ermakov, S., Toliat, M.R., Trofimov, S., Altmuller, J., Malkin, I., Nurnberg, P., Livshits, G., 2010. Genetic determinants of circulating levels of tumor necrosis factor receptor II and their association with TNF-RII gene polymorphisms. Cytokine 51, 28-34. Viviani, B., Gardoni, F., Marinovich, M., 2007. Cytokines and neuronal ion channels in health and disease. Int Rev Neurobio.l 82, 247-263. Werkman, T.R., Glennon, J.C., Wadman, W.J., McCreary, A.C., 2006. Dopamine receptor pharmacology: interactions with serotonin receptors and significance for the aetiology and treatment of schizophrenia. CNS Neurol Disord Drug Targets. 5, 3-23.

21

Ragni Mørch

Figure 1 A-D. Psychosis continuum and inflammatory markers These figures illustrate the differences between the diagnostic groups along the psychosis continuum for the inflammatory markers 1A) Soluble Tumor Necrosis Factor Receptor 1 (sTNF-R1), 1B) Interleukin 1 Receptor antagonist (IL-1Ra), 1C) von Willebrand Factor (vWf), 1D) Osteoprotegerin (OPG). Mean± standard error of mean of log transformed values is presented. *p< 0.05, ** p<0.01, ***p<0.001. Log=logarithmic, SCZ=schizophrenia group, SA=schizoaffective group, AFF=affective group, HC=healthy controls

22

Ragni Mørch

Table 1. Demographic and Clinical Characteristics of Study Sample

SCZ= Schizophrenia group, SA=Schizoaffective group, AFF= Affective group, HC= Healthy

Parameter

All patients (n=992)

SCZ (n=584)

SA (n=93)

AFF (n=315)

% (n)

% (n)

% (n)

HC (n=638)

Post Hoc

% (n)

% (n) Sex (male)**

53 (526)

71 (415)

49 (46)

49 (154)

54.5 (348) 97.6 (623)

SCZ,HC> SA,AFF HC>SCZ, SA,AFF> SCZ

Ethnicity (European)**

74.2 (736)

70.5 (412)

69 (64)

82.5 (260)

Tobacco (use/day)** Cannabis (use/2weeks) Illicit substances (use/2weeks) Antipsychotics Lithium* Antidepressant s Antiepileptics

47.1 (467)

53 (262)

67.9 (53)

48.3 (152)

8.6 (55)

5.2 (4)

7.2 (21)

0.4 (1)

9.2 (45)

5.2 (4)

8.6 (25)

0.4 (1)

SA>SCZ, AFF>HC SCZ,SA, AFF>HC N.S.

7.2 (62)

7.4 (37)

8.7 (74)

64.9 (644) 6.1 (61) 27.5 (273)

72 (418) 0.7 (4) 23.3 (136)

76.3 (71) 6.5 (6) 34.4 (32)

49.2 (155) 16.2 (51) 33.3 (105)

-

N.S. AFF>SCZ N.S.

18.5 (184) Mean (SD) 31.5 (10.5)

8.9 (52) Mean (SD)

27 (25) Mean (SD) 32.1 (10.9)

33.9 (107) Mean (SD) 33.6 (11.9)

Mean (SD) 33 (9.3)

N.S.

SCZ BMI** (3.5) HC 61.9 47.9 (12.3) SCZ,SA> PANSS total ** 57.5 (17.0) 62.4 (16.9) (17.2) AFF 4.8 (5.1) 5.3 (5.1) 4.9 (4.9) 4.0 (5.2) - SCZ>AFF YMRS total * 18.5 (12.8) 17.9 (12.8) 22.8 18.3 (12.5) SA>AFF, IDS total * (13.0) SCZ controls, PANSS=Positive and Negative Syndrome Scale, YMRS= Young Mania Rating Scale, IDS=Inventory of Depressive Symptoms, n=number, SD=standard deviation, N.S.= non-significant, IU = international units, BMI=body mass index, *p<0.05, ** p<0.01 (ANOVA, post hoc Tukey for continuous variables or Chi-square for categorical variables). Age**

30.3 (9.4)

-

23

Ragni Mørch

Table 2. Inflammatory markers across diagnostic groups of the psychosis continuum (ANCOVA)

Inflammatory marker

sTNF-R11

All patients

SCZ

SA

AFF

HC

(n=992)

(n=584)

(n=93)

(n=315)

(n=638)

Mean (SD)

Mean (SD)

Mean (SD)

Mean (SD)

Mean (SD)

1.98 (0.82)**

1.99 (0.82)**

2.08 (1.01)* 1.88(0.74)*

1.80 (0.70)

ANCOV A

SCZ, SA,

* IL-1Ra2

392.1(859.2)** 433.6 (937.1)**

vWf2

101.3 (89.4)

103.6 (79.6)*

AFF>HC

400.2

311.6(655.3

382.2(957.8

(933.9)

)

)

108.9 (90.3) 94.9(100.4)

SCZ>AF F, HC

98.7 (92.9)

SCZ>AF F

OPG2

1.41 (0.41)

1.38 (0.44)

1.42 (0.69)

1.47 (0.52)

1.40 (0.41)

N.S.

SCZ= Schizophrenia group, SA=Schizoaffective group, AFF= Affective group, HC= Healthy controls group, sTNF-R1= soluble tumor necrosis factor receptor 1, IL-1Ra=interleukin 1receptor antagonist, vWf=von Willebrand factor, OPG=osteoprotegerin, n=number, SD=standard deviation, N.S.= non-significant, *p<0.05, ** p<0.01 after Bonferroni correction; analyses performed with log-transformed variables, the mean± standard deviation presented is not logarithmic transformed, 1Adjusted for age, sex and BMI, 2,Adjusted for age and sex, ANCOVA= Analysis of covariance.

24

Ragni Mørch

Table 3. Association between medication and inflammatory markers

AP (n=641 )

AP Monotherap y (n=268 )

Lithiu m (n=61)

Antidepressan ts (n=272)

Mood Stabilize rs (n=183)

sOlanzapi ne (n=212)

sOlanzapin e Monothera py (n=99)

Typical AP (n=25)

Atypic al AP (n=616)

0.00 0.99

0.06 0.56

-0.22 0.29

-0.04 0.33

-0.02 0.75

-0.08 0.46

0.01 0.95

-0.01 0.80

TNFR1 r p-value

-0.01 0.88

-0.04 0.49

0.07 0.57

-0.03 0.69

IL1Ra r p-value

0.03 0.49

-0.05 0.42

0.18 0.16

-0.10 0.09

vWf r p-value

0.02 0.62

0.001 0.99

0.03 0.83

0.03 0.64

-0.14 0.07

0.02 0.82

-0.09 0.36

0.18 0.4

-0.07 0.12

OPG r p-value

0.01 0.84

-0.06 0.31

0.08 0.52

0.04 0.44

-0.08 0.25

0.01 0.83

0.06 0.60

0.22 0.30

-0.005 0.91

0.01 0.90

-0.12 0.10

sTNF-R1=soluble tumor necrosis factor receptor 1, IL-1Ra=interleukin 1 receptor antagonist, vWf=von Willebrand factor, OPG=osteoprotegerin, n=number, r= Pearson correlation, s= serum-concentration, AP= antipsychotic, Typical= 1st generation antipsychotic, Atypical= 2nd generation antipsychotic. In sub-analyses of the association of medication and inflammatory markers, the main results remained the same after adjustment for confounders and multiple comparisons.

25