Hostname: page-component-586b7cd67f-2plfb Total loading time: 0 Render date: 2024-11-22T05:13:06.422Z Has data issue: false hasContentIssue false
  • English
  • Français

Longitudinal course of inflammatory-cognitive subgroups across first treatment severe mental illness and healthy controls

Published online by Cambridge University Press:  02 October 2024

Linn Sofie Sæther Open the ORCID record for Linn Sofie Sæther [Opens in a new window] ,
Thor Ueland Open the ORCID record for Thor Ueland [Opens in a new window] ,
Beathe Haatveit Open the ORCID record for Beathe Haatveit [Opens in a new window] ,
Anja Vaskinn Open the ORCID record for Anja Vaskinn [Opens in a new window] ,
Camilla Bärthel Flaaten Open the ORCID record for Camilla Bärthel Flaaten [Opens in a new window] ,
Christine Mohn Open the ORCID record for Christine Mohn [Opens in a new window] ,
Monica B. E.G. Ormerod Open the ORCID record for Monica B. E.G. Ormerod [Opens in a new window] ,
Pål Aukrust Open the ORCID record for Pål Aukrust [Opens in a new window] ,
Ingrid Melle Open the ORCID record for Ingrid Melle [Opens in a new window]  and
Nils Eiel Steen Open the ORCID record for Nils Eiel Steen [Opens in a new window]
...Show all authors
Linn Sofie Sæther*
Affiliation:
Section for Clinical Psychosis Research, Division of Mental Health and Addiction, Oslo University Hospital, Oslo, Norway Department of Psychology, University of Oslo, Oslo, Norway
Thor Ueland
Affiliation:
Research Institute of Internal Medicine, Oslo University Hospital, Rikshospitalet, Oslo, Norway Faculty of Medicine, University of Oslo, Norway Thrombosis Research Center (TREC), Division of internal medicine, University hospital of North Norway, Tromsø Norway
Beathe Haatveit
Affiliation:
Section for Clinical Psychosis Research, Division of Mental Health and Addiction, Oslo University Hospital, Oslo, Norway
Anja Vaskinn
Affiliation:
Centre for Research and Education in Forensic Psychiatry, Oslo University Hospital, Oslo, Norway Centre for Precision Psychiatry, Division of Mental Health and Addiction, Institute of Clinical Medicine, University of Oslo, Oslo, Norway
Camilla Bärthel Flaaten
Affiliation:
Section for Clinical Psychosis Research, Division of Mental Health and Addiction, Oslo University Hospital, Oslo, Norway
Christine Mohn
Affiliation:
National Centre for Suicide Research and Prevention, Institute of Clinical Medicine, University of Oslo, Oslo, Norway
Monica B. E.G. Ormerod
Affiliation:
Section for Clinical Psychosis Research, Division of Mental Health and Addiction, Oslo University Hospital, Oslo, Norway Institute of Clinical Medicine, University of Oslo, Oslo Norway
Pål Aukrust
Affiliation:
Research Institute of Internal Medicine, Oslo University Hospital, Rikshospitalet, Oslo, Norway Faculty of Medicine, University of Oslo, Norway Section of Clinical Immunology and Infectious Diseases, Oslo University Hospital, Rikshospitalet, Oslo, Norway
Ingrid Melle
Affiliation:
Section for Clinical Psychosis Research, Division of Mental Health and Addiction, Oslo University Hospital, Oslo, Norway Institute of Clinical Medicine, University of Oslo, Oslo Norway
Nils Eiel Steen
Affiliation:
Section for Clinical Psychosis Research, Division of Mental Health and Addiction, Oslo University Hospital, Oslo, Norway Centre for Precision Psychiatry, Division of Mental Health and Addiction, Institute of Clinical Medicine, University of Oslo, Oslo, Norway Department of Psychiatric Research, Diakonhjemmet Hospital, Oslo, Norway
Ole A. Andreassen
Affiliation:
Centre for Precision Psychiatry, Division of Mental Health and Addiction, Institute of Clinical Medicine, University of Oslo, Oslo, Norway Section for Precision Psychiatry, Division of Mental Health and Addiction, Oslo University Hospital, Norway
Torill Ueland
Affiliation:
Section for Clinical Psychosis Research, Division of Mental Health and Addiction, Oslo University Hospital, Oslo, Norway Department of Psychology, University of Oslo, Oslo, Norway
*
Corresponding author: Linn Sofie Sæther; Email: [email protected]
Rights & Permissions [Opens in a new window]

Abstract

Background

While inflammation is associated with cognitive impairment in severe mental illnesses (SMI), there is substantial heterogeneity and evidence of transdiagnostic subgroups across schizophrenia (SZ) and bipolar (BD) spectrum disorders. There is however, limited knowledge about the longitudinal course of this relationship.

Methods

Systemic inflammation (C-Reactive Protein, CRP) and cognition (nine cognitive domains) was measured from baseline to 1 year follow-up in first treatment SZ and BD (n = 221), and healthy controls (HC, n = 220). Linear mixed models were used to evaluate longitudinal changes separately in CRP and cognitive domains specific to diagnostic status (SZ, BD, HC). Hierarchical clustering was applied on the entire sample to investigate the longitudinal course of transdiagnostic inflammatory-cognitive subgroups.

Results

There were no case-control differences or change in CRP from baseline to follow-up. We confirm previous observations of case-control differences in cognition at both time-points and domain specific stability/improvement over time regardless of diagnostic status. We identified transdiagnostic inflammatory-cognitive subgroups at baseline with differing demographics and clinical severity. Despite improvement in cognition, symptoms and functioning, the higher inflammation – lower cognition subgroup (75% SZ; 48% BD; 38% HC) had sustained inflammation and lower cognition, more symptoms, and lower functioning (SMI only) at follow-up. This was in comparison to a lower inflammation – higher cognition subgroup (25% SZ, 52% BD, 62% HC), where SMI participants showed cognitive functioning at HC level with a positive clinical course.

Conclusions

Our findings support heterogenous and transdiagnostic inflammatory-cognitive subgroups that are stable over time, and may benefit from targeted interventions.

Keywords

bipolar disordercognitionheterogeneityInflammationlongitudinalSchizophreniaSubgroups
Type
Original Article
Information
Psychological Medicine , Volume 54 , Issue 12 , September 2024 , pp. 3519 - 3529
Check for updates
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
Copyright © The Author(s), 2024. Published by Cambridge University Press

Introduction

Cognitive impairment is a central feature of severe mental illnesses (SMI), such as schizophrenia (SZ) and bipolar (BD) spectrum disorders (McCleery & Nuechterlein, Reference McCleery and Nuechterlein2019; Stainton et al., Reference Stainton, Chisholm, Griffiths, Kambeitz-Ilankovic, Wenzel, Bonivento and Wood2023). While highly prevalent, there is considerable heterogeneity in cognitive symptoms, ranging from mild to severe (Catalan et al., Reference Catalan, McCutcheon, Aymerich, Pedruzo, Radua, Rodríguez and Fusar-Poli2024; Haatveit et al., Reference Haatveit, Westlye, Vaskinn, Flaaten, Mohn, Bjella and Ueland2023; Lee et al., Reference Lee, Cernvall, Borg, Plavén-Sigray, Larsson, Erhardt and Cervenka2024; Van Rheenen et al., Reference Van Rheenen, Lewandowski, Tan, Ospina, Ongur, Neill, Gurvich and Burdick2017; Wenzel et al., Reference Wenzel, Badde, Haas, Bonivento, Van Rheenen, Antonucci and Kambeitz-Ilankovic2023). Numerous studies have identified transdiagnostic cognitive subgroups that are associated with different neurobiological characteristics, as well as clinical- and functional outcomes (Bora et al., Reference Bora, Verim, Akgul, Ildız, Ceylan, Alptekin and Akdede2023; Cowman et al., Reference Cowman, Holleran, Lonergan, O'Connor, Birchwood and Donohoe2021; Lewandowski, Reference Lewandowski2020; Vaskinn et al., Reference Vaskinn, Haatveit, Melle, Andreassen, Ueland and Sundet2020; Wenzel et al., Reference Wenzel, Haas, Dwyer, Ruef, Oeztuerk, Antonucci and Kambeitz-Ilankovic2021). For instance, cognitive subgroups with severe impairment typically have more symptoms and lower functioning (Miskowiak et al., Reference Miskowiak, Kjærstad, Lemvigh, Ambrosen, Thorvald, Kessing and Fagerlund2023; Vaskinn et al., Reference Vaskinn, Haatveit, Melle, Andreassen, Ueland and Sundet2020), brain abnormalities as assessed by magnetic resonance imaging (de Zwarte et al., Reference de Zwarte, Brouwer, Agartz, Alda, Alonso-Lana, Bearden and van Haren2020; Wenzel et al., Reference Wenzel, Badde, Haas, Bonivento, Van Rheenen, Antonucci and Kambeitz-Ilankovic2023, Reference Wenzel, Haas, Dwyer, Ruef, Oeztuerk, Antonucci and Kambeitz-Ilankovic2021; Wolfers et al., Reference Wolfers, Doan, Kaufmann, Alnæs, Moberget, Agartz and Marquand2018; Woodward & Heckers, Reference Woodward and Heckers2015), and higher levels of systemic inflammation (Pan, Qian, Qu, Tang, & Yan, Reference Pan, Qian, Qu, Tang and Yan2020; Watson et al., Reference Watson, Giordano, Suckling, Barnes, Husain, Jones and Joyce2023). Evidence further suggests that cognitive functioning remains relatively stable throughout the illness course in both SZ and BD (Bora & Özerdem, Reference Bora and Özerdem2017; Catalan et al., Reference Catalan, McCutcheon, Aymerich, Pedruzo, Radua, Rodríguez and Fusar-Poli2024; Ehrlich et al., Reference Ehrlich, Ryan, Burdick, Langenecker, McInnis and Marshall2022; Flaaten et al., Reference Flaaten, Melle, Bjella, Engen, Åsbø, Wold and Ueland2022, Reference Flaaten, Melle, Bjella, Engen, Åsbø, Wold and Ueland2023a, Reference Flaaten, Melle, Gardsjord, Bjella, Engen, Vaskinn and Ueland2023b; Samamé, Cattaneo, Richaud, Strejilevich, & Aprahamian, Reference Samamé, Cattaneo, Richaud, Strejilevich and Aprahamian2022; Watson, Harrison, Preti, Wykes, & Cella, Reference Watson, Harrison, Preti, Wykes and Cella2022). Developing successful personalized treatments is contingent on increasing our understanding of the causes and maintenance of cognitive impairment in SMI.

Current pharmacotherapies targeting symptom relief in SMI have limited effects on cognition, which may have a different underlying pathophysiology (Howes, Bukala, & Beck, Reference Howes, Bukala and Beck2024; McCutcheon, Keefe, & McGuire, Reference McCutcheon, Keefe and McGuire2023). Evidence suggests immune- and inflammatory-related abnormalities, which are well documented across the psychosis spectrum (Andreassen, Hindley, Frei, & Smeland, Reference Andreassen, Hindley, Frei and Smeland2023; Benros, Eaton, & Mortensen, Reference Benros, Eaton and Mortensen2014; Goldsmith, Rapaport, & Miller, Reference Goldsmith, Rapaport and Miller2016; Steen et al., Reference Steen, Rahman, Szabo, Hindley, Parker, Cheng and Andreassen2023; Webster, Reference Webster, Savitz and Yolken2023), are associated with cognitive impairment (Jovasevic et al., Reference Jovasevic, Wood, Cicvaric, Zhang, Petrovic, Carboncino and Radulovic2024; Morozova et al., Reference Morozova, Zorkina, Abramova, Pavlova, Pavlov, Soloveva and Chekhonin2022; Rosenblat et al., Reference Rosenblat, Brietzke, Mansur, Maruschak, Lee and McIntyre2015; Wang, Meng, Liu, An, & Hu, Reference Wang, Meng, Liu, An and Hu2022). Dysregulated systemic levels of inflammatory markers have been observed in first-episode and chronic stages of SMI (Halstead et al., Reference Halstead, Siskind, Amft, Wagner, Yakimov, Liu and Warren2023; Perry et al., Reference Perry, Upthegrove, Kappelmann, Jones, Burgess and Khandaker2021), including in medication naïve patients (Dunleavy, Elsworthy, Upthegrove, Wood, & Aldred, Reference Dunleavy, Elsworthy, Upthegrove, Wood and Aldred2022; Fernandes et al., Reference Fernandes, Steiner, Bernstein, Dodd, Pasco, Dean and Berk2016a, Reference Fernandes, Steiner, Molendijk, Dodd, Nardin, Gonçalves and Berk2016b; van den Ameele et al., Reference van den Ameele, van Diermen, Staels, Coppens, Dumont, Sabbe and Morrens2016). The most extensively studied and reliable marker of systemic inflammation in SMI is C-Reactive Protein (CRP), in part due its low-cost and global accessibility at routine medical laboratories (Clyne & Olshaker, Reference Clyne and Olshaker1999; Ullah et al., Reference Ullah, Awan, Aamir, Diwan, de Filippis, Awan and De Berardis2021). CRP levels fluctuate in response to change in inflammatory status and may be used to infer whether low-grade systemic inflammation is associated with cognitive impairment. In fact, increased levels of CRP have been consistently reported in SZ and BD relative to healthy controls, and previously found to be modestly associated with clinical- and cognitive characteristics (Fernandes et al., Reference Fernandes, Steiner, Bernstein, Dodd, Pasco, Dean and Berk2016a; Reference Fernandes, Steiner, Molendijk, Dodd, Nardin, Gonçalves and Berk2016b; Fond, Lançon, Auquier, & Boyer, Reference Fond, Lançon, Auquier and Boyer2018; Halstead et al., Reference Halstead, Siskind, Amft, Wagner, Yakimov, Liu and Warren2023; Jacomb et al., Reference Jacomb, Stanton, Vasudevan, Powell, O'Donnell, Lenroot and Weickert2018; Johnsen et al., Reference Johnsen, Fathian, Kroken, Steen, Jørgensen, Gjestad and Løberg2016; Lestra, Romeo, Martelli, Benyamina, & Hamdani, Reference Lestra, Romeo, Martelli, Benyamina and Hamdani2022; Millett et al., Reference Millett, Perez-Rodriguez, Shanahan, Larsen, Yamamoto, Bukowski and Burdick2021; Misiak et al., Reference Misiak, Stańczykiewicz, Kotowicz, Rybakowski, Samochowiec and Frydecka2018; Patlola, Donohoe, & McKernan, Reference Patlola, Donohoe and McKernan2023).

It is increasingly clear that only a subset of individuals with SMI show signs of increased systemic inflammation (Bishop, Zhang, & Lizano, Reference Bishop, Zhang and Lizano2022; Chen, Tan, & Tian, Reference Chen, Tan and Tian2024; Miller & Goldsmith, Reference Miller and Goldsmith2019), partly explaining mixed or weak associations between inflammatory markers and cognition in case-control studies (Bora, Reference Bora2019; Miller & Goldsmith, Reference Miller and Goldsmith2019; Morrens et al., Reference Morrens, Overloop, Coppens, Loots, Van Den Noortgate, Vandenameele and De Picker2022). This is also in line with genetic findings of mixed effect directions, which includes higher load of increasing and decreasing genetic variants for CRP in SMI (Hindley et al., Reference Hindley, Drange, Lin, Kutrolli, Shadrin, Parker and Andreassen2023). Similar to findings on cognitive subgroups (Bora et al., Reference Bora, Verim, Akgul, Ildız, Ceylan, Alptekin and Akdede2023; Cowman et al., Reference Cowman, Holleran, Lonergan, O'Connor, Birchwood and Donohoe2021; Lewandowski, Reference Lewandowski2020; Wenzel et al., Reference Wenzel, Badde, Haas, Bonivento, Van Rheenen, Antonucci and Kambeitz-Ilankovic2023, Reference Wenzel, Haas, Dwyer, Ruef, Oeztuerk, Antonucci and Kambeitz-Ilankovic2021), the higher-inflammation subtype is associated with more adverse neurobiological and clinical outcomes, and is associated with lower cognitive functioning (Boerrigter et al., Reference Boerrigter, Weickert, Lenroot, O'Donnell, Galletly, Liu and Weickert2017; Fillman et al., Reference Fillman, Weickert, Lenroot, Catts, Bruggemann, Catts and Weickert2016; Lizano et al., Reference Lizano, Kiely, Mijalkov, Meda, Keedy, Hoang and Bishop2023a, Reference Lizano, Lutz, Xu, Rubin, Paskowitz, Lee and Bishop2020; Millett et al., Reference Millett, Perez-Rodriguez, Shanahan, Larsen, Yamamoto, Bukowski and Burdick2021; Nettis et al., Reference Nettis, Pergola, Kolliakou, O'Connor, Bonaccorso, David and Mondelli2019; Zhang et al., Reference Zhang, Lizano, Guo, Xu, Rubin, Hill and Bishop2022). A common observation is that a larger proportion of individuals with SMI compared to control participants, belong to a higher-inflammation subtype (Boerrigter et al., Reference Boerrigter, Weickert, Lenroot, O'Donnell, Galletly, Liu and Weickert2017; Fillman et al., Reference Fillman, Weickert, Lenroot, Catts, Bruggemann, Catts and Weickert2016; Lizano et al., Reference Lizano, Lutz, Xu, Rubin, Paskowitz, Lee and Bishop2020). Including both SMI and control participants when using unsupervised clustering techniques allows for evaluation of similarities and differences across phenotypes, regardless of diagnostic status.

Recent evidence from machine learning suggests higher accuracy of case-control prediction when both cognition and inflammatory markers are evaluated together (Fernandes et al., Reference Fernandes, Karmakar, Tamouza, Tran, Yearwood, Hamdani and Leboyer2020). Using hierarchical clustering, we recently identified a transdiagnostic subgroup with cognitive impairment and higher inflammation using different immune and inflammatory marker panels (Sæther et al., Reference Sæther, Ueland, Haatveit, Maglanoc, Szabo, Djurovic and Ueland2023, Reference Sæther, Szabo, Akkouh, Haatveit, Mohn, Vaskinn and Ueland2024). This subgroup also had more symptoms and lower functioning, compared to a subgroup with milder impairments and lower inflammation. The clinical relevance of these subgroups remains to be determined, and longitudinal studies are essential to address if these subgroups are trait or state phenomenon. Longitudinal studies on subgroups based on cognition suggest stability over time for both SZ and BD (Ehrlich et al., Reference Ehrlich, Ryan, Burdick, Langenecker, McInnis and Marshall2022; Flaaten et al., Reference Flaaten, Melle, Bjella, Engen, Åsbø, Wold and Ueland2022; Lim et al., Reference Lim, Smucny, Barch, Lam, Keefe and Lee2021). Longitudinal studies of inflammatory markers, including CRP, are in general scarce, and most of them focus on the effects of antipsychotic treatment in SZ cohorts only (Fathian et al., Reference Fathian, Gjestad, Kroken, Løberg, Reitan, Fleichhacker and Johnsen2022; Feng, McEvoy, & Miller, Reference Feng, McEvoy and Miller2020; Meyer et al., Reference Meyer, McEvoy, Davis, Goff, Nasrallah, Davis and Lieberman2009). Evidence based on a few studies suggests a diminished correlation between CRP and cognition after 6 weeks of admittance to hospital with acute psychosis (Johnsen et al., Reference Johnsen, Fathian, Kroken, Steen, Jørgensen, Gjestad and Løberg2016), and an early drop in CRP level may predict improved cognitive functioning after 6 months (Fathian et al., Reference Fathian, Løberg, Gjestad, Steen, Kroken, Jørgensen and Johnsen2019). To our knowledge, no previous study has evaluated temporal characteristics of subgroups based on both inflammation and cognition in SMI and controls.

The current study is an extension of our previous work with partially overlapping samples (Sæther et al., Reference Sæther, Ueland, Haatveit, Maglanoc, Szabo, Djurovic and Ueland2023, Reference Sæther, Szabo, Akkouh, Haatveit, Mohn, Vaskinn and Ueland2024), and aimed to elucidate the longitudinal course of systemic inflammation and cognition in first treatment SMI (SZ = 133, BD = 88), and healthy controls (n = 220). This study used data from the decades long TOP-study in Norway, with the overall aim to investigate biological, psychological, and environmental factors underlying development and maintenance of SMI (see i.e. Ormerod et al., Reference Ormerod, Ueland, Frogner Werner, Hjell, Rødevand, Sæther and Steen2022; Rødevand et al., Reference Rødevand, Steen, Elvsåshagen, Quintana, Reponen, Mørch and Andreassen2019; Simonsen et al., Reference Simonsen, Sundet, Vaskinn, Birkenaes, Engh, Faerden and Andreassen2011). The TOP-study has collected baseline and follow-up data through the first year of adequate treatment of SMI (~12 months later), which includes measurement of systemic inflammation assessed with CRP, and cognition with nine core domains including fine-motor speed, psychomotor processing speed, mental processing speed, attention, verbal learning, verbal memory, semantic fluency, working memory and cognitive control. We first investigated the specific trajectories of CRP levels and cognitive domains associated with diagnostic status (SZ, BD, HC), using separate linear mixed models. Based on our findings from previous overlapping samples, we expect domain-specific stability or improvement over the first year of treatment in SMI and HC (Demmo et al., Reference Demmo, Lagerberg, Aminoff, Hellvin, Kvitland, Simonsen and Ueland2017; Engen et al., Reference Engen, Simonsen, Melle, Færden, Lyngstad, Haatveit and Ueland2019; Haatveit et al., Reference Haatveit, Vaskinn, Sundet, Jensen, Andreassen, Melle and Ueland2015). The trajectory of CRP levels from baseline to 1 year follow-up in first treatment SZ and BD is, however, unknown. Based on a similar approach to our previous work (Sæther et al., Reference Sæther, Szabo, Akkouh, Haatveit, Mohn, Vaskinn and Ueland2024, Reference Sæther, Ueland, Haatveit, Maglanoc, Szabo, Djurovic and Ueland2023), we used hierarchical clustering to identify transdiagnostic inflammatory-cognitive subgroups using CRP and a cognitive composite score at baseline. The subgroups were assessed longitudinally across demographic, clinical, and cognitive measures.

Methods

Sample

This study is part of the ongoing Thematically Organized Psychosis (TOP)-study. Participants meeting the Diagnostic Manual of Mental Disorders (DSM)-IV criteria for schizophrenia or bipolar spectrum disorders are continuously recruited from in- and out-patient psychiatric units in the larger Oslo area. Healthy controls (HC) from the same catchment area are randomly chosen using statistical records and invited by letter. Exclusion criteria for all participants are: (1) age <18 or >65, (2) moderate/severe head injury, (3) severe somatic/neurological disorder, (4) not fluent in a Scandinavian language, (5) IQ<70. HC are excluded in the case of drug dependency, history of mental illness, or relatives with SMI. Any participant (SMI and HC) with signs of acute infection at baseline and/or follow-up (CRP>10 mg/L) was excluded.

This study included SMI participants who at baseline was within the first 12 months of starting their first adequate treatment of SZ or BD spectrum disorder, while in a stable illness phase. We opted to use ‘first treatment’ as a classified for both SMI groups, as ‘first episode’ can be especially challenging to establish in BD where correct diagnosis and treatment may be preceded by several mood episodes that are not recognized as part of BD by either the patient or the health care system. Adequate treatment was here defined as treatment with antipsychotic or mood stabilizing medication, not antidepressant since they have minor effects on BD disorders. The patients were recruited as soon as possible after the start of treatment, however, the enrollment in the study was dependent on their ability to give informed consent. Participants had to have follow-up assessment 6 months to 1.5 year later (mean = 400 days), with relatively complete cognitive assessment at both time points, and blood samples taken at both time points. Baseline assessments were conducted between 2004–2020, and follow-up assessments between 2005–2021. The final sample included n = 133 SZ spectrum (schizophrenia = 76, schizophreniform = 13, schizoaffective = 8, psychosis not otherwise specified = 36), n = 88 BD spectrum (bipolar I = 53, bipolar II = 30, bipolar not otherwise specified = 5) and n = 220 healthy controls. Due to selection criteria the retention rate for this study was not possible to determine. However, the retention rate for one-year follow-up of cognitive assessment in the TOP-study has previously been reported to be 53–66%, with little or no difference in clinical or demographic characteristics between those eligible for follow-up v. completers (Demmo et al., Reference Demmo, Lagerberg, Aminoff, Hellvin, Kvitland, Simonsen and Ueland2017; Engen et al., Reference Engen, Simonsen, Melle, Færden, Lyngstad, Haatveit and Ueland2019). All participants provided informed consent and the study was approved by the Regional Ethics Committee.

Clinical assessments

The Structured Clinical Interview for DSM-IV axis 1 disorders (SCID-I) (First, Spitzer, Gibbon, & Williams, Reference First, Spitzer, Gibbon and Williams1995) was administered by trained clinical psychologists or physicians. The Positive and Negative Syndrome Scale (PANSS) was used to assess symptoms according to the five-factor model including positive, negative, disorganized/concrete, excited, and depressed symptoms (Kay, Fiszbein, & Opler, Reference Kay, Fiszbein and Opler1987; Wallwork, Fortgang, Hashimoto, Weinberger, & Dickinson, Reference Wallwork, Fortgang, Hashimoto, Weinberger and Dickinson2012). Manic symptoms were assessed with the Young Mania Rating Scale (YMRS) (Young, Biggs, Ziegler, & Meyer, Reference Young, Biggs, Ziegler and Meyer1978). Level of functioning was assessed with the split version of the Global Assessment of Functioning scale (GAF F, GAF S; Pedersen, Hagtvet, and Karterud, Reference Pedersen, Hagtvet and Karterud2007). Duration of untreated psychosis (DUP) was estimated as time of onset from psychotic symptoms until start of first adequate treatment. The average time between physical examination (blood sampling, height/weight), and cognitive assessment was 4.2 days for baseline and 5.3 days at follow-up. The defined daily dose (DDD) of psychopharmacological treatment (antipsychotics, antidepressants, antiepileptics and lithium) was determined according to World Health Organization guidelines (https://www.whocc.no/atc_ddd_index). Somatic medication use (yes/no) in the SMI group is provided in online Supplementary Table S1.

Cognitive assessments

Trained clinical psychologists or research personnel administered one of two test batteries: Battery 1 (from 2004–2012) or Battery 2 (from 2012). The test batteries included different tests of equivalent cognitive functions, as well as some identical measures. Thus, to ensure the highest possible N, corresponding tests from the two batteries were standardized separately (Z-scores) before combining to cover nine cognitive domains: Fine-motor speed, psychomotor processing speed, mental processing speed, attention, verbal learning, verbal memory, semantic fluency, working memory and cognitive control. We have previously shown robust between-battery correspondence of test performance for SZ, BD, and HC (Sæther et al., Reference Sæther, Szabo, Akkouh, Haatveit, Mohn, Vaskinn and Ueland2024). The cognitive batteries consisted of tests from the MATRICS Consensus Cognitive Battery (MCCB) (Nuechterlein et al., Reference Nuechterlein, Green, Kern, Baade, Barch, Cohen and Marder2008), Halstead-Reitan (Klove, Reference Klove1963), the Wechsler Adult Intelligence Scale (WAIS-III) (Wechsler, Reference Wechsler1997), Delis Kaplan Executive Functioning System (D-KEFS) (Delis, Kaplan, & Kramer, Reference Delis, Kaplan and Kramer2001), the California Verbal Learning Test (CVLT-II) (Delis, Kramer, Kaplan, & Ober, Reference Delis, Kramer, Kaplan and Ober1987), and the Hopkins Verbal Learning Test-Revised (HVLT-R) (Benedict, Schretlen, Groninger, & Brandt, Reference Benedict, Schretlen, Groninger and Brandt1998). We assessed intellectual functioning with the Matrix Reasoning and Vocabulary subtests from the Wechsler Abbreviated Scale of Intelligence (WASI) (Wechsler, Reference Wechsler2011). See online Supplementary Table S2 for an overview of tests.

Blood sampling

Blood was sampled from the antecubital vein in EDTA vials and stored at 4 °C overnight before transport to the hospital central laboratory the next day. The samples (2 × 9 ml EDTA tubes) were centrifuged at 1800 g for 15 min, and isolated plasma was stored at −80 °C in multiple aliquots. Blood samples were analysed for CRP by a particle enhanced immunoturbidimetric method with a Cobas 8000 instrument (Roche Diagnostics, Basel Switzerland) at the Department of Medical Biochemistry, Oslo University Hospital, Oslo, Norway.

Statistical procedure

Data preprocessing, sample, and clinical characteristics

Data preprocessing, statistical analyses and visualization of results were conducted in the R- environment (https://www.r-project.org/; v.4.2.0, main R-packages reported in Supplementary Methods 1). Cognitive data was standardized (Z-scores) based on the HC group mean and standard deviation (s.d.) at baseline, and CRP was log-10 transformed. A cognitive composite score was computed as the mean score across cognitive domains for participants with baseline data in at least five cognitive domains. Sample and clinical characteristics were compared across groups using Kruskal–Wallis rank sum test and pairwise permutation (n = 10 000) based t tests for continuous variables, and chi-squared tests for categorical variables. All analyses were adjusted for multiple comparisons using Bonferroni correction. An overview of the number of observations for CRP and all cognitive domains at baseline and follow-up, as well as descriptive statistics for these can be found in online Supplementary Table S3-S4. Correlations between CRP and cognitive domains at baseline and follow-up are found in online Supplementary Fig. S1.

Linear mixed models

Linear mixed models were used to analyze group-level changes over time separately for CRP and each cognitive domain in order to account for individual variability and repeated measures within subjects. We included sex and age as covariates as they may impact cognition in the cognitive model, and sex, age, and BMI as covariates in the CRP model as they may influence CRP (in the CRP model). We used the following formula for cognitive data:

$$Yij = ( {\beta_0 + b_{0i}} ) + \beta _1 \times {\rm timepoin}{\rm t}_{ij} + \beta _2 \times {\rm grou}{\rm p}_{ij} + \beta _3 \times {\rm timepoin}{\rm t}_{ij} \times {\rm grou}{\rm p}_{ij} + \beta _4 \times {\rm se}{\rm x}_{ij} + \beta _5{\rm xag}{\rm e}_{ij} + e_{ij}$$

where Yij is the cognitive score for participant i = 1…441 at time j = 0…1, β signifies fixed effects, b random effects (random intercept for each unique ID), and e the residual error term. The same model structure was used for CRP, with the addition of BMI as a covariate.

Hierarchical clustering

We used hierarchical clustering to identify subgroups based on inflammation and cognition in a subsample of participants with available CRP and a cognitive composite score at baseline (SZ = 121, BD = 87, HC = 216). In brief, we (1) generated a Euclidian distance matrix, (2) evaluated the optimal linkage method based on the agglomerative coefficient (average, single, complete, Ward's), (3) determined the optimal number of clusters by inspecting the average silhouette index, (4) tested the presence of clusters using a previously described data simulation procedure (Dinga et al., Reference Dinga, Schmaal, Penninx, van Tol, Veltman, van Velzen and Marquand2019), and (5) evaluated the stability of the cluster solution using a resampling procedure (bootstrapping). A Jaccard similarity index for clustering stability was computed with an index >0.7 (70%) was considered stable. We compared the subgroups on inflammation, cognition, sample (age, sex, education, IQ, BMI), clinical, and functional characteristics, at baseline and follow-up using Welch's t tests (effect sizes: Cohen's d). In the case of sustained subgroup differences in any of the sample/clinical/functional characteristics, we investigated the effect of time, and potential subgroup differences in change over time (e.g. change scores, ΔY = Y1-Y0), using Wilcoxon signed-rank tests. All comparisons were corrected for multiple comparisons (Bonferroni).

Code availability

Main analysis code/scripts are available at: https://osf.io/ek68q/

Results

Sample and clinical demographics

Sample and clinical characteristics at baseline are provided in Table 1. See online Supplementary Table S5 for clinical characteristics at follow-up.

Table 1. Sample and clinical characteristics at baseline

a Mean (s.d.); n (%).

b Kruskal–Wallis rank sum test; Pearson's Chi-squared test.

c Pairwise two-sample permutation test (for 3 groups).

SZ, schizophrenia; BD, bipolar disorder; HC, healthy controls; WASI, Wechsler Abbreviated Scale of Intelligence; BMI, body mass index; PANSS, Positive and Negative Syndrome Scale; YMRS, Young Mania Rating Scale; GAF, Global Assessment of Functioning scale; DDD, defined daily dosage; ns, non-significant.

Note: WASI IQ scores may be slightly overestimated due to properties of the Norwegian WASI, which uses US norms (see Siqveland, Dalsbø, Harboe, and Leiknes, Reference Siqveland, Dalsbø, Harboe and Leiknes2014).

Inflammation and cognition over time comparing diagnostic status

As seen in Fig. 1, temporal assessment using linear mixed models suggested stable levels of CRP over time. There was no difference between SMI groups or HC at baseline or follow-up, with a positive relationship between BMI and CRP (online Supplementary Table S6). There was no association between the number of days between assessments and change in CRP level for the SMI group (r = 0.05, p = 0.467), suggesting limited effect of shorter or longer duration in treatment on CRP. For cognitive measures (online Supplementary Fig. S2), we confirm previous findings from studies using overlapping samples (Demmo et al., Reference Demmo, Lagerberg, Aminoff, Hellvin, Kvitland, Simonsen and Ueland2017; Engen et al., Reference Engen, Simonsen, Melle, Færden, Lyngstad, Haatveit and Ueland2019; Flaaten et al., Reference Flaaten, Melle, Gardsjord, Bjella, Engen, Vaskinn and Ueland2023b, Reference Flaaten, Melle, Bjella, Engen, Åsbø, Wold and Ueland2023a, Reference Flaaten, Melle, Bjella, Engen, Åsbø, Wold and Ueland2022; Haatveit et al., Reference Haatveit, Vaskinn, Sundet, Jensen, Andreassen, Melle and Ueland2015), i.e. regardless of time-point, the cognitive scores remained attenuated in SMI, with SZ on average scoring ~1 s.d. and BD ~0.5 s.d. lower than HC. BD, however, had similar performance to HC on attention and semantic fluency at both time-points. Further, for all groups there was improvement in fine-motor speed, psychomotor speed, verbal learning, and cognitive control over time, whereas stability was observed for the remaining domains (mental speed, verbal memory, attention, semantic fluency, working memory). There was a significant time by group interaction for working memory, indicating improved performance over time for BD relative to HC. See online Supplementary Table S6 for extended model output.

Figure 1. Inflammation (C-Reactive Protein, CRP) at baseline and follow-up between HC, BD and SZ.

Boxplots (interquartile range separated by median line), density plot (kernel density estimate) and lines between mean scores (including error bars: ± SEM) shows no difference among HC, BD, and SZ at baseline or follow-up and indicates stability of CRP-levels over time.

Subgroups based on inflammation and cognition

Evaluation of hierarchical clustering on CRP and the cognitive composite score revealed a 2-cluster solution to be optimal, with a favourable agglomerative coefficient (0.99) when using Ward's linkage method (online Supplementary Fig. S3). The simulation procedure resulted in a significant silhouette index (p < 0.001), rejecting the null hypothesis that the data comes from a single Gaussian distribution (online Supplementary Fig. S4). The cluster assignment was robust for both clusters following bootstrapping, with 81% (cluster 1) and 74% (cluster 2) overlap. As seen in Fig. 2A, the first cluster captured a subgroup (n = 209, SZ = 30 [25%], BD = 45 [52%], HC = 134 [62%]) characterized by a higher proportion of HC, lower inflammation and higher cognition (see Table 2), compared to the second subgroup (n = 215, SZ = 91 [75%], BD = 42 [48%], HC = 82 [38%]) which had a larger proportion of the SZ group, higher inflammation and lower cognition (chi square p < 0.001, d = 0.5–1.9). We additionally performed hierarchical clustering on the SMI group alone and found that the same inflammation-cognition pattern emerged, albeit characterized by even higher CRP levels and lower composite score in the higher inflammation – lower cognition subgroup, which also included predominantly SZ (online Supplementary Table S7).

Figure 2. Inflammatory-cognitive subgroups at baseline and follow-up.

Panel A shows the distribution of SZ, BD, and HC in each cluster/subgroup. Panel B shows cluster differences in CRP (log10 transformed) and the cognitive composite score (Z-scores) at baseline and follow-up, with boxplots (interquartile range separated by median line), density plot (kernel density estimate), and line from mean scores (error bars: ± SEM). Panel C shows PANSS factors that were significantly different across clusters at both time points, and panel D shows differences across level of functioning (GAF symptom and GAF Function). Panel C and D is SMI only (line from mean, error bars: ± SEM).

Table 2. Subgroup comparisons on sample and clinical characteristics at baseline and follow-up

a Mean (s.d.); n (%).

b CI = Confidence Interval, 95%.

c Welch Two Sample t test; Pearson's Chi-squared test (Bonferroni corrected p-values).

WASI, Wechsler Abbreviated Scale of Intelligence; CRP, C-reactive Protein; BMI, body mass index; PANSS, Positive and Negative Syndrome Scale; YMRS, Young Mania Rating Scale; GAF, Global Assessment of Functioning scale; DUI, Days of untreated illness; DDD, defined daily dosage.

Note: Subgroup 1 = lower inflammation – higher cognition; Subgroup 2 = higher inflammation – lower cognition.

Characteristics of inflammatory-cognitive subgroups at baseline and follow-up

As seen in Fig. 2B, the subgroup pattern was consistent over time, with higher inflammation and lower cognition in the second subgroup relative to the first also at follow-up (d = 0.4–1.3, Table 2). Relative to the first subgroup, the higher inflammation – lower cognition subgroup had shorter education and lower IQ (all p < 0.001, d = 0.5–0.9), but they did not differ in age, sex, or BMI. The higher inflammation – lower cognition subgroup had lower scores on all cognitive domains both at baseline (d = 0.8–1.4) and follow-up (d = 0.5–1.1) compared to the lower inflammation – higher cognition subgroup (all p < 0.001, online Supplementary Fig. S5). Compared to the other subgroup, participants with SMI in the higher inflammation – lower cognition subgroup had more positive, negative, and disorganized symptoms (Fig. 2C), as well as lower functioning (GAFS and GAFF; Figure 2D), at both time points (d baseline = 0.5–0.7, d follow−up = 0.4–0.5). Regardless of group, there was a significant improvement in the cognitive composite score (p < 0.001), and all symptoms and functioning scores (all p < 0.001), except for disorganized symptoms which remained stable. The level of CRP however, remained stable (p = 0.623). Analysis of change scores revealed a slightly higher gain in cognitive performance from baseline to follow-up in the second subgroup compared to the first (p < 0.001, Wilcoxon effect size, r = 0.2(small)). There was no difference in change scores between the subgroups on any symptoms or functioning measures.

Discussion

This study evaluated the longitudinal course of inflammation and cognition in a large sample of first treatment SZ and BD, and a HC cohort. While there were case-control differences in CRP at baseline or follow-up, we identified two transdiagnostic inflammatory-cognitive subgroups with differing levels of clinical and functional characteristics. The higher inflammation – lower cognition subgroup (predominantly SZ) had more symptoms and lower functioning at both time-points, compared to the lower inflammation – higher cognition subgroup. While inflammation, cognition, symptoms, and functioning remained stable or improved over time for both subgroups, the higher inflammation – lower cognition group still scored well below the other subgroup at follow-up. The fact that SZ, BD, and HC were represented in both subgroups shows that heterogeneity is characteristic for both inflammation and cognition. Our findings suggest transdiagnostic inflammatory-cognitive subgroups that are stable across time. This indicates that the inflammatory-cognitive association may be more trait- than state-related.

The main finding is that inflammatory-cognitive subgroups based on CRP as a measure of inflammation and a cognitive composite score, is stable over one year in first treatment SMI and HC. These findings also confirm the inflammatory-cognitive subgroup pattern that we previously identified using broad panels of inflammatory and immune-related markers and cognitive domains (Sæther et al., Reference Sæther, Ueland, Haatveit, Maglanoc, Szabo, Djurovic and Ueland2023, Reference Sæther, Szabo, Akkouh, Haatveit, Mohn, Vaskinn and Ueland2024). Importantly, while cognition, symptoms, and level of functioning generally improved over the first year of treatment for SMI participants, we observed stable differences between the subgroups at both time-points, with the higher inflammation – lower cognition subgroup having worse cognition, higher inflammation, more symptoms, and lower functioning. Results from clinical trials suggest that add-on anti-inflammatory treatments are more effective in SMI patients exhibiting higher inflammation (Jeppesen et al., Reference Jeppesen, Christensen, Pedersen, Nordentoft, Hjorthøj, Köhler-Forsberg and Benros2020; Nettis et al., Reference Nettis, Lombardo, Hastings, Zajkowska, Mariani, Nikkheslat and Mondelli2021). Similarly, cognitive remediation may be particularly efficacious for patients with significant cognitive impairments, although those with milder impairments also benefit (Vita et al., Reference Vita, Barlati, Ceraso, Nibbio, Ariu, Deste and Wykes2021; Wykes, Huddy, Cellard, McGurk, & Czobor, Reference Wykes, Huddy, Cellard, McGurk and Czobor2011). Given the between-subgroup stability in characteristics (inflammatory, cognitive, clinical) over time, these subgroups could be ideal candidates for personalized interventions. For the more impaired subgroup this could include cognitive remediation combined with anti-inflammatory add-on treatments, as the latter may also have beneficial effects on cognition (Jeppesen et al., Reference Jeppesen, Christensen, Pedersen, Nordentoft, Hjorthøj, Köhler-Forsberg and Benros2020). One could speculate that HC in the impaired subgroup constitute a vulnerable group, particularly since low-grade inflammation is also a risk factor in the general population for developing autoimmune-, cardiovascular-, and neurodegenerative disease (Furman et al., Reference Furman, Campisi, Verdin, Carrera-Bastos, Targ, Franceschi and Slavich2019). It is worth noting that 36% of the SMI group showed a similar pattern to the HC group (i.e. those in the lower inflammation – higher cognition group) with a positive clinical trajectory. This group may benefit from other interventions that should also focus on cognitive strengths (Allott et al., Reference Allott, Steele, Boyer, de Winter, Bryce, Alvarez-Jimenez and Phillips2020).

Although we need external replication of the clustering pattern to be certain, our findings suggest that immune-cognition associations follow a relatively simple high-low pattern that is observed across diagnostic categories and HC status. The same high-low pattern emerged when performing clustering on the SMI group alone. This is perhaps not surprising, as similar high-low patterns are observed in separate clustering studies on cognition (i.e. Vaskinn et al., Reference Vaskinn, Haatveit, Melle, Andreassen, Ueland and Sundet2020) and inflammation (i.e. Lizano et al., Reference Lizano, Lutz, Xu, Rubin, Paskowitz, Lee and Bishop2020). However, we cannot exclude the possibility that more complex subgroup patterns could emerge with different clustering strategies, larger samples, and/or more inflammatory markers, as suggested by recent machine learning approaches (Lalousis et al., Reference Lalousis, Schmaal, Wood, Reniers, Cropley, Watson and Upthegrove2023). Regardless, it is noteworthy that in this study and in our previous studies (Sæther et al., Reference Sæther, Ueland, Haatveit, Maglanoc, Szabo, Djurovic and Ueland2023, Reference Sæther, Szabo, Akkouh, Haatveit, Mohn, Vaskinn and Ueland2024), the subgroups seem to differ primarily in the magnitude rather than different patterns of cognitive functioning, inflammation, and clinical severity. This is further strengthened by the observation that even though the subgroups differ on these measures, they follow a similar longitudinal trajectory.

One could speculate that parallel and interacting processes in the brain and the immune system during development are important sources of individual variance in immune-cognition patterns at later stages. Cytokines expressed in the brain have important neuromodulatory functions that are involved in shaping neural circuits during neurodevelopment (Salvador, de Lima, & Kipnis, Reference Salvador, de Lima and Kipnis2021). Further, it is possible that immune and inflammatory dysregulation during this time, which is more common among clinical high-risk groups compared to healthy peers (Misiak et al., Reference Misiak, Bartoli, Carrà, Stańczykiewicz, Gładka, Frydecka and Miller2021), could have a long-term impact on brain functioning and cognition. Immune-cognitive associations could be bidirectional, as cognitive impairment in SMI has been linked to poor decision-making regarding physical health (Whitson et al., Reference Whitson, O'Donoghue, Hester, Baldwin, Harrigan, Francey and Allott2021), possibly contributing to, or exacerbating, low-grade inflammatory states. Similarly, low-grade systemic inflammation could influence the permeability of the blood-brain barrier (Futtrup et al., Reference Futtrup, Margolinsky, Benros, Moos, Routhe, Rungby and Krogh2020; Lizano, Pong, Santarriaga, Bannai, & Karmacharya, Reference Lizano, Pong, Santarriaga, Bannai and Karmacharya2023b), activate immunocompetent glial cells and contribute to neuroinflammation (Almeida, Nani, Oses, Brietzke, & Hayashi, Reference Almeida, Nani, Oses, Brietzke and Hayashi2019; Bishop et al., Reference Bishop, Zhang and Lizano2022), ultimately affecting cognitive functioning.

As shown in previous studies with overlapping samples (Demmo et al., Reference Demmo, Lagerberg, Aminoff, Hellvin, Kvitland, Simonsen and Ueland2017; Engen et al., Reference Engen, Simonsen, Melle, Færden, Lyngstad, Haatveit and Ueland2019; Flaaten et al., Reference Flaaten, Melle, Gardsjord, Bjella, Engen, Vaskinn and Ueland2023b, Reference Flaaten, Melle, Bjella, Engen, Åsbø, Wold and Ueland2023a, Reference Flaaten, Melle, Bjella, Engen, Åsbø, Wold and Ueland2022; Haatveit et al., Reference Haatveit, Vaskinn, Sundet, Jensen, Andreassen, Melle and Ueland2015), our analyses comparing diagnostic status show domain-specific stability or improvement in cognitive functioning from baseline to follow-up. This is in line with longitudinal findings in SMI from other groups (Bora & Özerdem, Reference Bora and Özerdem2017; Catalan et al., Reference Catalan, McCutcheon, Aymerich, Pedruzo, Radua, Rodríguez and Fusar-Poli2024; Torgalsbøen, Mohn, Larøi, Fu, & Czajkowski, Reference Torgalsbøen, Mohn, Larøi, Fu and Czajkowski2023). A similar course of improvement in both SMI and HC may indicate practice effects, which is known for some of the cognitive tests used in this study (Beglinger et al., Reference Beglinger, Gaydos, Tangphao-Daniels, Duff, Kareken, Crawford and Siemers2005). In terms of subgroups, we observed that while the higher inflammation – lower cognition subgroup had a slight improvement in cognition, they still performed significantly lower than the lower inflammation – higher cognition subgroup at follow-up. Sustained cognitive impairment is strongly associated with poor functional outcomes (Cowman et al., Reference Cowman, Holleran, Lonergan, O'Connor, Birchwood and Donohoe2021), underscoring the need to develop and implement effective treatments for cognitive impairment in SMI.

Our data did not suggest case-control differences in CRP levels at baseline or follow-up. While meta-analyses have reported consistent evidence of elevated CRP in SMI compared to HC, it may be higher during acute manic or psychotic episodes (Fernandes et al., Reference Fernandes, Steiner, Bernstein, Dodd, Pasco, Dean and Berk2016a, Reference Fernandes, Steiner, Molendijk, Dodd, Nardin, Gonçalves and Berk2016b; Fond et al., Reference Fond, Lançon, Auquier and Boyer2018; Halstead et al., Reference Halstead, Siskind, Amft, Wagner, Yakimov, Liu and Warren2023; Lestra et al., Reference Lestra, Romeo, Martelli, Benyamina and Hamdani2022). However, participants in the TOP-study have been evaluated in euthymic/milder symptom states. We accounted for age, sex, and BMI which has been shown to attenuate CRP findings on psychiatric symptoms (Figueroa-Hall et al., Reference Figueroa-Hall, Xu, Kuplicki, Ford, Burrows, Teague, Sen and Paulus2022). These covariates are not always included in studies reported by meta-analyses (Fernandes et al., Reference Fernandes, Steiner, Bernstein, Dodd, Pasco, Dean and Berk2016a, Reference Fernandes, Steiner, Molendijk, Dodd, Nardin, Gonçalves and Berk2016b; Fond et al., Reference Fond, Lançon, Auquier and Boyer2018; Halstead et al., Reference Halstead, Siskind, Amft, Wagner, Yakimov, Liu and Warren2023; Lestra et al., Reference Lestra, Romeo, Martelli, Benyamina and Hamdani2022). Further, inflammatory markers in SMI are typically in the smaller effect size range (Carvalho et al., Reference Carvalho, Solmi, Sanches, Machado, Stubbs, Ajnakina and Herrmann2020; Miller & Goldsmith, Reference Miller and Goldsmith2020). This may pose a challenge for detecting case-control differences, as only a subset of individuals with SMI show elevated levels of inflammation (Bishop et al., Reference Bishop, Zhang and Lizano2022; Miller & Goldsmith, Reference Miller and Goldsmith2019). Nonetheless, the higher inflammation – lower cognition subgroup suggests some interaction with CRP and cognition, particularly in SZ participants that were overrepresented in this subgroup. This also aligns with previous findings that individuals with SMI in high-inflammatory subgroups have lower cognitive performance (Fillman et al., Reference Fillman, Weickert, Lenroot, Catts, Bruggemann, Catts and Weickert2016; Lizano et al., Reference Lizano, Kiely, Mijalkov, Meda, Keedy, Hoang and Bishop2023a, Reference Lizano, Lutz, Xu, Rubin, Paskowitz, Lee and Bishop2020). Our findings suggest that there are trait-related cognitive-immune subgroups in SMI, which seems independent of state dependent fluctuations of immune markers.

There are some limitations to consider. While CRP is an inexpensive and accessible marker of systemic inflammation, it cannot provide further insight about specific inflammatory pathways or mechanisms that might be related to cognitive impairment. Unfortunately, the only marker consistently re-measured in the TOP-study was CRP. However, CRP is a reliable and established down-stream marker of systemic inflammation, covering several inflammatory pathways. Moreover, in contrast to measurement of several cytokines, CRP measurement is available in all hospitals and can be used in clinical practice for monitoring of patients. Nonetheless, studies should include a broader spectrum of markers, preferably those relevant for cognition (see i.e. Patlola et al., Reference Patlola, Donohoe and McKernan2023; Sæther et al., Reference Sæther, Ueland, Haatveit, Maglanoc, Szabo, Djurovic and Ueland2023, Reference Sæther, Szabo, Akkouh, Haatveit, Mohn, Vaskinn and Ueland2024), in longitudinal designs. Although we found stability in inflammatory-cognitive subgroups over time, the study was unable to establish whether inflammation and lower cognition simply co-occurs or has a causal relationship. There are also other factors that potentially could influence both cognition and inflammation that were not accounted for in this study, i.e. clinical relapse, poor diet, disturbed sleep, stress, and drug abuse, which should be addressed in future studies. Strengths of this study lie in the longitudinal design, the large sample of first treatment SMI and the inclusion of HC, as well as the robust evaluation of the clustering solution with stability analyses. However, our findings should be replicated using independent samples.

Conclusion

Results from our study suggest that transdiagnostic inflammatory-cognitive subgroups defined at baseline are stable over time. Individuals with SMI in the higher inflammation – lower cognition subgroup had sustained symptoms and lower functioning, suggesting a specific phenotype that may benefit from personalized treatments targeting both inflammation and cognition.

Supplementary material

The supplementary material for this article can be found at https://doi.org/10.1017/S003329172400206X.

Acknowledgements

We wish to thank all participants for their valuable contribution to the TOP-study and staff for their contribution to data collection and curation. Funding for this project was provided by the South-Eastern Norway Regional Health Authority (grant #2020089, #2023031) and Research Council of Norway (#223273, #326813). We wish to acknowledge Sigma2 (the National Infrastructure for High Performance Computing and Data Storage in Norway), Services for Sensitive Data (TSD) at the University of Oslo, and the Department of Medical Biochemistry at Oslo University Hospital.

Competing interests

OOA is a consultant to cortechs.ai and precision health, and has received speakers honorarium from Janssen, Otsuka, and Lundbeck. Remaining authors have no conflicts of interest to declare.

References

Allott, K., Steele, P., Boyer, F., de Winter, A., Bryce, S., Alvarez-Jimenez, M., & Phillips, L. (2020). Cognitive strengths-based assessment and intervention in first-episode psychosis: A complementary approach to addressing functional recovery? Clinical Psychology Review 79, 101871. https://doi.org/10.1016/j.cpr.2020.101871CrossRefGoogle ScholarPubMed
Almeida, P.G.C., Nani, J.V., Oses, J.P., Brietzke, E., & Hayashi, M.A.F. (2019). Neuroinflammation and glial cell activation in mental disorders. Brain, Behavior, and Immunity - Health 2, 100034. https://doi.org/10.1016/j.bbih.2019.100034CrossRefGoogle ScholarPubMed
Andreassen, O.A., Hindley, G.F.L., Frei, O., & Smeland, O.B. (2023). New insights from the last decade of research in psychiatric genetics: Discoveries, challenges and clinical implications. World Psychiatry 22, 4–24. https://doi.org/10.1002/wps.21034CrossRefGoogle ScholarPubMed
Beglinger, L.J., Gaydos, B., Tangphao-Daniels, O., Duff, K., Kareken, D.A., Crawford, J., … Siemers, E.R. (2005). Practice effects and the use of alternate forms in serial neuropsychological testing. Archives of Clinical Neuropsychology 20, 517–529. https://doi.org/10.1016/j.acn.2004.12.003CrossRefGoogle ScholarPubMed
Benedict, R.H.B., Schretlen, D., Groninger, L., & Brandt, J. (1998). Hopkins verbal learning test – revised: Normative data and analysis of inter-form and test-retest reliability. Clinical Neuropsychology 12, 43–55. https://doi.org/10.1076/clin.12.1.43.1726CrossRefGoogle Scholar
Benros, M.E., Eaton, W.W., & Mortensen, P.B. (2014). The epidemiological evidence linking autoimmune diseases and psychosis. Biological Psychiatry 75, 300–306. https://doi.org/10.1016/j.biopsych.2013.09.023CrossRefGoogle ScholarPubMed
Bishop, J.R., Zhang, L., & Lizano, P. (2022). Inflammation subtypes and translating inflammation-related genetic findings in schizophrenia and related psychoses: A perspective on pathways for treatment stratification and novel therapies. Harvard Review of Psychiatry 30, 59–70. https://doi.org/10.1097/HRP.0000000000000321CrossRefGoogle ScholarPubMed
Boerrigter, D., Weickert, T.W., Lenroot, R., O'Donnell, M., Galletly, C., Liu, D., … Weickert, C.S. (2017). Using blood cytokine measures to define high inflammatory biotype of schizophrenia and schizoaffective disorder. Journal of Neuroinflammation 14, 188. https://doi.org/10.1186/s12974-017-0962-yCrossRefGoogle ScholarPubMed
Bora, E. (2019). Peripheral inflammatory and neurotrophic biomarkers of cognitive impairment in schizophrenia: A meta-analysis. Psychological Medicine, 49, 1971–1979. https://doi.org/10.1017/S0033291719001685CrossRefGoogle ScholarPubMed
Bora, E., & Özerdem, A. (2017). Meta-analysis of longitudinal studies of cognition in bipolar disorder: Comparison with healthy controls and schizophrenia. Psychological Medicine 47, 2753–2766. https://doi.org/10.1017/S0033291717001490CrossRefGoogle ScholarPubMed
Bora, E., Verim, B., Akgul, O., Ildız, A., Ceylan, D., Alptekin, K., … Akdede, B.B. (2023). Clinical and developmental characteristics of cognitive subgroups in a transdiagnostic sample of schizophrenia spectrum disorders and bipolar disorder. European Neuropsychopharmacology 68, 47–56. https://doi.org/10.1016/j.euroneuro.2022.12.005CrossRefGoogle Scholar
Carvalho, A.F., Solmi, M., Sanches, M., Machado, M.O., Stubbs, B., Ajnakina, O., … Herrmann, N. (2020). Evidence-based umbrella review of 162 peripheral biomarkers for major mental disorders. Translational Psychiatry 10, 152. https://doi.org/10.1038/s41398-020-0835-5CrossRefGoogle ScholarPubMed
Catalan, A., McCutcheon, R.A., Aymerich, C., Pedruzo, B., Radua, J., Rodríguez, V., … Fusar-Poli, P. (2024). The magnitude and variability of neurocognitive performance in first-episode psychosis: A systematic review and meta-analysis of longitudinal studies. Translational Psychiatry 14, 1–9. https://doi.org/10.1038/s41398-023-02718-6CrossRefGoogle ScholarPubMed
Chen, S., Tan, Y., & Tian, L. (2024). Immunophenotypes in psychosis: Is it a premature inflammaging disorder? Molecular Psychiatry 1–15. https://doi.org/10.1038/s41380-024-02539-zGoogle ScholarPubMed
Clyne, B., & Olshaker, J.S. (1999). The C-reactive protein. Journal of Emergency Medicine 17, 1019–1025. https://doi.org/10.1016/S0736-4679(99)00135-3CrossRefGoogle ScholarPubMed
Cowman, M., Holleran, L., Lonergan, E., O'Connor, K., Birchwood, M., & Donohoe, G. (2021). Cognitive predictors of social and occupational functioning in early psychosis: A systematic review and meta-analysis of cross-sectional and longitudinal data. Schizophrenia Bulletin 47, 1243–1253. https://doi.org/10.1093/schbul/sbab033CrossRefGoogle ScholarPubMed
Delis, D.C., Kaplan, E., & Kramer, J.H. (2001). Delis–Kaplan Executive Function System. https://doi.org/10.1037/t15082-000CrossRefGoogle Scholar
Delis, D.C., Kramer, J.H., Kaplan, E., & Ober, B.A. (1987). California Verbal Learning Test – Second Edition. https://doi.org/10.1037/t15072-000CrossRefGoogle Scholar
Demmo, C., Lagerberg, T.V., Aminoff, S.R., Hellvin, T., Kvitland, L.R., Simonsen, C., … Ueland, T. (2017). Course of neurocognitive function in first treatment bipolar I disorder: One-year follow-up study. Psychiatry Research 249, 286–292. https://doi.org/10.1016/j.psychres.2016.12.048CrossRefGoogle ScholarPubMed
de Zwarte, S.M.C., Brouwer, R.M., Agartz, I., Alda, M., Alonso-Lana, S., Bearden, C.E., … van Haren, N.E.M. (2020). Intelligence, educational attainment, and brain structure in those at familial high-risk for schizophrenia or bipolar disorder. Human Brain Mapping 43, 414–430. https://doi.org/10.1002/hbm.25206CrossRefGoogle ScholarPubMed
Dinga, R., Schmaal, L., Penninx, B.W.J.H., van Tol, M.J., Veltman, D.J., van Velzen, L., … Marquand, A.F. (2019). Evaluating the evidence for biotypes of depression: Methodological replication and extension of Drysdale et al. (2017). NeuroImage: Clinical 22, 101796. https://doi.org/10.1016/j.nicl.2019.101796CrossRefGoogle Scholar
Dunleavy, C., Elsworthy, R.J., Upthegrove, R., Wood, S.J., & Aldred, S. (2022). Inflammation in first-episode psychosis: The contribution of inflammatory biomarkers to the emergence of negative symptoms, a systematic review and meta-analysis. Acta Psychiatrica Scandinavica 146, 6–20. https://doi.org/10.1111/acps.13416CrossRefGoogle Scholar
Ehrlich, T.J., Ryan, K.A., Burdick, K.E., Langenecker, S.A., McInnis, M.G., & Marshall, D.F. (2022). Cognitive subgroups and their longitudinal trajectories in bipolar disorder. Acta Psychiatrica Scandinavica 146, 240–250. https://doi.org/10.1111/acps.13460CrossRefGoogle ScholarPubMed
Engen, M.J., Simonsen, C., Melle, I., Færden, A., Lyngstad, S.H., Haatveit, B., … Ueland, T. (2019). Cognitive functioning in patients with first-episode psychosis stratified by level of negative symptoms: A 1-year follow-up study. Psychiatry Research 281, 112554. https://doi.org/10.1016/j.psychres.2019.112554CrossRefGoogle ScholarPubMed
Fathian, F., Gjestad, R., Kroken, R.A., Løberg, E.-M., Reitan, S.K., Fleichhacker, W.W., … Johnsen, E. (2022). Association between C-reactive protein levels and antipsychotic treatment during 12 months follow-up period after acute psychosis. Schizophrenia Research 241, 174–183. https://doi.org/10.1016/j.schres.2022.01.049CrossRefGoogle ScholarPubMed
Fathian, F., Løberg, E.-M., Gjestad, R., Steen, V.M., Kroken, R.A., Jørgensen, H.A., & Johnsen, E. (2019). Associations between C-reactive protein levels and cognition during the first 6 months after acute psychosis. Acta Neuropsychiatrica 31, 36–45. https://doi.org/10.1017/neu.2018.25CrossRefGoogle ScholarPubMed
Feng, T., McEvoy, J.P., & Miller, B.J. (2020). Longitudinal study of inflammatory markers and psychopathology in schizophrenia. Schizophrenia Research 224, 58–66. https://doi.org/10.1016/j.schres.2020.10.003CrossRefGoogle ScholarPubMed
Fernandes, B.S., Karmakar, C., Tamouza, R., Tran, T., Yearwood, J., Hamdani, N., … Leboyer, M. (2020). Precision psychiatry with immunological and cognitive biomarkers: A multi-domain prediction for the diagnosis of bipolar disorder or schizophrenia using machine learning. Translational Psychiatry 10, 1–13. https://doi.org/10.1038/s41398-020-0836-4CrossRefGoogle ScholarPubMed
Fernandes, B. S., Steiner, J., Bernstein, H.-G., Dodd, S., Pasco, J.A., Dean, O.M., … Berk, M. (2016a). C-reactive protein is increased in schizophrenia but is not altered by antipsychotics: Meta-analysis and implications. Molecular Psychiatry 21, 554–564. https://doi.org/10.1038/mp.2015.87CrossRefGoogle Scholar
Fernandes, B.S, Steiner, J., Molendijk, M.L., Dodd, S., Nardin, P., Gonçalves, C.-A., … Berk, M. (2016b). C-reactive protein concentrations across the mood spectrum in bipolar disorder: A systematic review and meta-analysis. The Lancet. Psychiatry 3, 1147–1156. https://doi.org/10.1016/S2215-0366(16)30370-4CrossRefGoogle ScholarPubMed
Figueroa-Hall, L.K., Xu, B., Kuplicki, R., Ford, B.N., Burrows, K., Teague, T.K., Sen, S., … Paulus, M.P. (2022). Psychiatric symptoms are not associated with circulating CRP concentrations after controlling for medical, social, and demographic factors. Translational Psychiatry 12, 1–12. https://doi.org/10.1038/s41398-022-02049-yCrossRefGoogle Scholar
Fillman, S.G., Weickert, T.W., Lenroot, R.K., Catts, S.V., Bruggemann, J.M., Catts, V.S., & Weickert, C.S. (2016). Elevated peripheral cytokines characterize a subgroup of people with schizophrenia displaying poor verbal fluency and reduced Broca's area volume. Molecular Psychiatry 21, 1090–1098. https://doi.org/10.1038/mp.2015.90CrossRefGoogle ScholarPubMed
First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. (1995). Structured clinical interview for DSM-IV axis I disorders: Patient edition (SCID-P, version 2.0). New York: Biometrics Research Department, New York State Psychiatric Institute.Google Scholar
Flaaten, C.B., Melle, I., Bjella, T., Engen, M.J., Åsbø, G., Wold, K. F., … Ueland, T. (2022). Domain-specific cognitive course in schizophrenia: Group- and individual-level changes over 10 years. Schizophrenia Research: Cognition 30, 100263. https://doi.org/10.1016/j.scog.2022.100263Google ScholarPubMed
Flaaten, C.B., Melle, I., Bjella, T., Engen, M.J., Åsbø, G., Wold, K.F., … Ueland, T. (2023a). Long-term course of cognitive functioning in bipolar disorder: A ten-year follow-up study. Bipolar Disorders, 26, 136–147. https://doi.org/10.1111/bdi.13364CrossRefGoogle ScholarPubMed
Flaaten, C.B., Melle, I., Gardsjord, E., Bjella, T., Engen, M.J., Vaskinn, A., … Ueland, T. (2023b). Course of intellectual functioning in schizophrenia and bipolar disorder: A 10-year follow-up study. Psychological Medicine 53, 2662–2670. https://doi.org/10.1017/S0033291721004645CrossRefGoogle ScholarPubMed
Fond, G., Lançon, C., Auquier, P., & Boyer, L. (2018). C-Reactive protein as a peripheral biomarker in schizophrenia. An updated systematic review. Frontiers in Psychiatry 9, 392. https://doi.org/10.3389/fpsyt.2018.00392CrossRefGoogle ScholarPubMed
Furman, D., Campisi, J., Verdin, E., Carrera-Bastos, P., Targ, S., Franceschi, C., … Slavich, G. M. (2019). Chronic inflammation in the etiology of disease across the life span. Nature Medicine 25, 1822–1832. https://doi.org/10.1038/s41591-019-0675-0CrossRefGoogle ScholarPubMed
Futtrup, J., Margolinsky, R., Benros, M.E., Moos, T., Routhe, L.J., Rungby, J., & Krogh, J. (2020). Blood-brain barrier pathology in patients with severe mental disorders: A systematic review and meta-analysis of biomarkers in case-control studies. Brain, Behavior, and Immunity - Health 6, 100102. https://doi.org/10.1016/j.bbih.2020.100102CrossRefGoogle ScholarPubMed
Goldsmith, D.R., Rapaport, M.H., & Miller, B.J. (2016). A meta-analysis of blood cytokine network alterations in psychiatric patients: Comparisons between schizophrenia, bipolar disorder and depression. Molecular Psychiatry 21, 1696–1709. https://doi.org/10.1038/mp.2016.3CrossRefGoogle ScholarPubMed
Haatveit, B., Vaskinn, A., Sundet, K.S., Jensen, J., Andreassen, O.A., Melle, I., & Ueland, T. (2015). Stability of executive functions in first episode psychosis: One year follow up study. Psychiatry Research 228, 475–481. https://doi.org/10.1016/j.psychres.2015.05.060CrossRefGoogle ScholarPubMed
Haatveit, B., Westlye, L.T., Vaskinn, A., Flaaten, C.B., Mohn, C., Bjella, T., … Ueland, T. (2023). Intra- and inter-individual cognitive variability in schizophrenia and bipolar spectrum disorder: An investigation across multiple cognitive domains. Schizophrenia 9, 1–9. https://doi.org/10.1038/s41537-023-00414-4CrossRefGoogle ScholarPubMed
Halstead, S., Siskind, D., Amft, M., Wagner, E., Yakimov, V., Liu, Z.S.J., … Warren, N. (2023). Alteration patterns of peripheral concentrations of cytokines and associated inflammatory proteins in acute and chronic stages of schizophrenia: A systematic review and network meta-analysis. The Lancet. Psychiatry 10, 260–271. https://doi.org/10.1016/S2215-0366(23)00025-1CrossRefGoogle ScholarPubMed
Hindley, G., Drange, O.K., Lin, A., Kutrolli, G., Shadrin, A.A., Parker, N., … Andreassen, O.A. (2023). Cross-trait genome-wide association analysis of C-reactive protein level and psychiatric disorders. Psychoneuroendocrinology 157, 106368. https://doi.org/10.1016/j.psyneuen.2023.106368CrossRefGoogle ScholarPubMed
Howes, O.D., Bukala, B.R., & Beck, K. (2024). Schizophrenia: From neurochemistry to circuits, symptoms and treatments. Nature Reviews Neurology 20, 22–35. https://doi.org/10.1038/s41582-023-00904-0CrossRefGoogle ScholarPubMed
Jacomb, I., Stanton, C., Vasudevan, R., Powell, H., O'Donnell, M., Lenroot, R., … Weickert, T.W. (2018). C-Reactive protein: Higher during acute psychotic episodes and related to cortical thickness in schizophrenia and healthy controls. Frontiers in Immunology 9, 2230. https://doi.org/10.3389/fimmu.2018.02230CrossRefGoogle ScholarPubMed
Jeppesen, R., Christensen, R.H.B., Pedersen, E.M.J., Nordentoft, M., Hjorthøj, C., Köhler-Forsberg, O., & Benros, M.E. (2020). Efficacy and safety of anti-inflammatory agents in treatment of psychotic disorders - A comprehensive systematic review and meta-analysis. Brain, Behavior, and Immunity 90, 364–380. https://doi.org/10.1016/j.bbi.2020.08.028CrossRefGoogle ScholarPubMed
Johnsen, E., Fathian, F., Kroken, R.A., Steen, V.M., Jørgensen, H.A., Gjestad, R., & Løberg, E.-M. (2016). The serum level of C-reactive protein (CRP) is associated with cognitive performance in acute phase psychosis. BMC Psychiatry 16, 60. https://doi.org/10.1186/s12888-016-0769-xCrossRefGoogle ScholarPubMed
Jovasevic, V., Wood, E.M., Cicvaric, A., Zhang, H., Petrovic, Z., Carboncino, A., … Radulovic, J. (2024). Formation of memory assemblies through the DNA-sensing TLR9 pathway. Nature, 628, 145–153. https://doi.org/10.1038/s41586-024-07220-7CrossRefGoogle ScholarPubMed
Kay, S.R., Fiszbein, A., & Opler, L.A. (1987). The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophrenia Bulletin 13, 261–276. https://doi.org/10.1093/schbul/13.2.261CrossRefGoogle ScholarPubMed
Klove, H. (1963). Clinical neuropsychology. Medical Clinics of North America 47, 1647–1658. https://doi.org/10.1016/S0025-7125(16)33515-5CrossRefGoogle ScholarPubMed
Lalousis, P.A., Schmaal, L., Wood, S.J., Reniers, R.L.E.P, Cropley, V.L., Watson, A., … Upthegrove, R. (2023). Inflammatory subgroups of schizophrenia and their association with brain structure: A semi-supervised machine learning examination of heterogeneity. Brain, Behavior, and Immunity 113, 166–175. https://doi.org/10.1016/j.bbi.2023.06.023CrossRefGoogle Scholar
Lee, M., Cernvall, M., Borg, J., Plavén-Sigray, P., Larsson, C., Erhardt, S., … Cervenka, S. (2024). Cognitive function and variability in antipsychotic drug–naive patients with first-episode psychosis: A systematic review and meta-analysis. JAMA Psychiatry, 81, 468–476. https://doi.org/10.1001/jamapsychiatry.2024.0016CrossRefGoogle ScholarPubMed
Lestra, V., Romeo, B., Martelli, C., Benyamina, A., & Hamdani, N. (2022). Could CRP be a differential biomarker of illness stages in schizophrenia? A systematic review and meta-analysis. Schizophrenia Research 246, 175–186. https://doi.org/10.1016/j.schres.2022.06.026CrossRefGoogle ScholarPubMed
Lewandowski, K.E. (2020). Genetically, developmentally, and clinically distinct cognitive subtypes in schizophrenia: A tale of three trajectories. The American Journal of Psychiatry 177, 282–284. https://doi.org/10.1176/appi.ajp.2020.20020132CrossRefGoogle ScholarPubMed
Lim, K., Smucny, J., Barch, D.M., Lam, M., Keefe, R.S.E., & Lee, J. (2021). Cognitive subtyping in schizophrenia: A latent profile analysis. Schizophrenia Bulletin 47, 712–721. https://doi.org/10.1093/schbul/sbaa157CrossRefGoogle ScholarPubMed
Lizano, P., Kiely, C., Mijalkov, M., Meda, S.A., Keedy, S.K., Hoang, D., … Bishop, J.R. (2023a). Peripheral inflammatory subgroup differences in anterior Default Mode network and multiplex functional network topology are associated with cognition in psychosis. Brain, Behavior, and Immunity 114, 3–15. https://doi.org/10.1016/j.bbi.2023.07.014CrossRefGoogle ScholarPubMed
Lizano, P., Lutz, O., Xu, Y., Rubin, L.H., Paskowitz, L., Lee, A.M., … Bishop, J.R. (2020). Multivariate relationships between peripheral inflammatory marker subtypes and cognitive and brain structural measures in psychosis. Molecular Psychiatry 26, 3430–3443. https://doi.org/10.1038/s41380-020-00914-0CrossRefGoogle ScholarPubMed
Lizano, P., Pong, S., Santarriaga, S., Bannai, D., & Karmacharya, R. (2023b). Brain microvascular endothelial cells and blood-brain barrier dysfunction in psychotic disorders. Molecular Psychiatry 28, 3698–3708. https://doi.org/10.1038/s41380-023-02255-0CrossRefGoogle ScholarPubMed
McCleery, A., & Nuechterlein, K.H. (2019). Cognitive impairment in psychotic illness: Prevalence, profile of impairment, developmental course, and treatment considerations. Dialogues in Clinical Neuroscience 21, 239–248. https://doi.org/10.31887/DCNS.2019.21.3/amccleeryCrossRefGoogle ScholarPubMed
McCutcheon, R.A., Keefe, R.S.E., & McGuire, P.K. (2023). Cognitive impairment in schizophrenia: Aetiology, pathophysiology, and treatment. Molecular Psychiatry, 28, 1902–1918. https://doi.org/10.1038/s41380-023-01949-9CrossRefGoogle Scholar
Meyer, J.M., McEvoy, J.P., Davis, V.G., Goff, D.C., Nasrallah, H.A., Davis, S.M., … Lieberman, J.A. (2009). Inflammatory markers in schizophrenia: Comparing antipsychotic effects in phase 1 of the clinical antipsychotic trials of intervention effectiveness study. Biological Psychiatry 66, 1013–1022. https://doi.org/10.1016/j.biopsych.2009.06.005CrossRefGoogle ScholarPubMed
Miller, B.J., & Goldsmith, D.R. (2019). Inflammatory biomarkers in schizophrenia: Implications for heterogeneity and neurobiology. Biomarkers in Neuropsychiatry 1, 100006. https://doi.org/10.1016/j.bionps.2019.100006CrossRefGoogle Scholar
Miller, B.J., & Goldsmith, D.R. (2020). Evaluating the hypothesis that schizophrenia is an inflammatory disorder. Focus 18, 391–401. https://doi.org/10.1176/appi.focus.20200015CrossRefGoogle ScholarPubMed
Millett, C.E., Perez-Rodriguez, M., Shanahan, M., Larsen, E., Yamamoto, H.S., Bukowski, C., … Burdick, K.E. (2021). C-reactive protein is associated with cognitive performance in a large cohort of euthymic patients with bipolar disorder. Molecular Psychiatry 26, 4096–4105. https://doi.org/10.1038/s41380-019-0591-1CrossRefGoogle Scholar
Misiak, B., Bartoli, F., Carrà, G., Stańczykiewicz, B., Gładka, A., Frydecka, D., … Miller, B. J. (2021). Immune-inflammatory markers and psychosis risk: A systematic review and meta-analysis. Psychoneuroendocrinology, 127, 105200. doi: 10.1016/j.psyneuen.2021.105200CrossRefGoogle Scholar
Misiak, B., Stańczykiewicz, B., Kotowicz, K., Rybakowski, J.K., Samochowiec, J., & Frydecka, D. (2018). Cytokines and C-reactive protein alterations with respect to cognitive impairment in schizophrenia and bipolar disorder: A systematic review. Schizophrenia Research 192, 16–29. https://doi.org/10.1016/j.schres.2017.04.015CrossRefGoogle ScholarPubMed
Miskowiak, K.W., Kjærstad, H.L., Lemvigh, C.K., Ambrosen, K.S., Thorvald, M.S., Kessing, L.V., … Fagerlund, B. (2023). Neurocognitive subgroups among newly diagnosed patients with schizophrenia spectrum or bipolar disorders: A hierarchical cluster analysis. Journal of Psychiatry Research 163, 278–287. https://doi.org/10.1016/j.jpsychires.2023.05.025CrossRefGoogle ScholarPubMed
Morozova, A., Zorkina, Y., Abramova, O., Pavlova, O., Pavlov, K., Soloveva, K., … Chekhonin, V. (2022). Neurobiological highlights of cognitive impairment in psychiatric disorders. International Journal of Molecular Sciences 23, 1217. https://doi.org/10.3390/ijms23031217CrossRefGoogle ScholarPubMed
Morrens, M., Overloop, C., Coppens, V., Loots, E., Van Den Noortgate, M., Vandenameele, S., … De Picker, L. (2022). The relationship between immune and cognitive dysfunction in mood and psychotic disorder: A systematic review and a meta-analysis. Molecular Psychiatry, 27, 3237–3246. https://doi.org/10.1038/s41380-022-01582-yCrossRefGoogle ScholarPubMed
Nettis, M.A., Lombardo, G., Hastings, C., Zajkowska, Z., Mariani, N., Nikkheslat, , … Mondelli, V. (2021). Augmentation therapy with minocycline in treatment-resistant depression patients with low-grade peripheral inflammation: Results from a double-blind randomised clinical trial. Neuropsychopharmacology 46, 939–948. https://doi.org/10.1038/s41386-020-00948-6CrossRefGoogle ScholarPubMed
Nettis, M.A., Pergola, G., Kolliakou, A., O'Connor, J., Bonaccorso, S., David, , … Mondelli, V. (2019). Metabolic-inflammatory status as predictor of clinical outcome at 1-year follow-up in patients with first episode psychosis. Psychoneuroendocrinology 99, 145–153. https://doi.org/10.1016/j.psyneuen.2018.09.005CrossRefGoogle ScholarPubMed
Nuechterlein, K.H., Green, M.F., Kern, R.S., Baade, L.E., Barch, D.M., Cohen, J.D., … Marder, S.R. (2008). The MATRICS Consensus Cognitive Battery, part 1: Test selection, reliability, and validity. American Journal of Psychiatry 165, 203–213. https://doi.org/10.1176/appi.ajp.2007.07010042CrossRefGoogle ScholarPubMed
Ormerod, M. B. E. G. S., Ueland, T., Frogner Werner, M. C., Hjell, G., Rødevand, L., Sæther, L. S., … Steen, N. E. (2022). Composite immune marker scores associated with severe mental disorders and illness course. Brain, Behavior, and Immunity - Health, 24, 100483. doi: 10.1016/j.bbih.2022.100483CrossRefGoogle Scholar
Pan, L.-H., Qian, M., Qu, W., Tang, Q., & Yan, Y. (2020). Serum C-reactive protein in patients with deficit schizophrenia and the relationship with cognitive function. Neuropsychiatric Disease and Treatment 16, 2891–2897. https://doi.org/10.2147/NDT.S284149CrossRefGoogle ScholarPubMed
Patlola, S.R., Donohoe, G., & McKernan, D.P. (2023). The relationship between inflammatory biomarkers and cognitive dysfunction in patients with schizophrenia: A systematic review and meta-analysis. Progress in Neuro-Psychopharmacology and Biological Psychiatry 121, 110668. https://doi.org/10.1016/j.pnpbp.2022.110668CrossRefGoogle ScholarPubMed
Pedersen, G., Hagtvet, K.A., & Karterud, S. (2007). Generalizability studies of the global assessment of functioning–split version. Comprehensive Psychiatry 48, 88–94. https://doi.org/10.1016/j.comppsych.2006.03.008CrossRefGoogle ScholarPubMed
Perry, B.I., Upthegrove, R., Kappelmann, N., Jones, P.B., Burgess, S., & Khandaker, G.M. (2021). Associations of immunological proteins/traits with schizophrenia, major depression and bipolar disorder: A bi-directional two-sample Mendelian randomization study. Brain, Behavior, and Immunity 97, 176–185. https://doi.org/10.1016/j.bbi.2021.07.009CrossRefGoogle ScholarPubMed
Rødevand, L., Steen, N. E., Elvsåshagen, T., Quintana, D. S., Reponen, E. J., Mørch, R. H., … Andreassen, O. A. (2019). Cardiovascular risk remains high in schizophrenia with modest improvements in bipolar disorder during past decade. Acta Psychiatrica Scandinavica, 139, 348–360. https://doi.org/10.1111/acps.13008CrossRefGoogle ScholarPubMed
Rosenblat, J.D., Brietzke, E., Mansur, R.B., Maruschak, N.A., Lee, Y., & McIntyre, R.S. (2015). Inflammation as a neurobiological substrate of cognitive impairment in bipolar disorder: Evidence, pathophysiology and treatment implications. Journal of Affective Disorders 188, 149–159. https://doi.org/10.1016/j.jad.2015.08.058CrossRefGoogle ScholarPubMed
Sæther, L.S., Ueland, T.., Haatveit, B., Maglanoc, L.A., Szabo, A., Djurovic, S., … Ueland, T. (2023). Inflammation and cognition in severe mental illness: Patterns of covariation and subgroups. Molecular Psychiatry 28, 1284–1292. https://doi.org/10.1038/s41380-022-01924-wCrossRefGoogle ScholarPubMed
Sæther, L.S., Szabo, A., Akkouh, I.A., Haatveit, B., Mohn, C., Vaskinn, A., & …Ueland, T. (2024). Cognitive and inflammatory heterogeneity in severe mental illness: Translating findings from blood to brain. Brain, Behavior, and Immunity 118, 287–299. https://doi.org/10.1016/j.bbi.2024.03.014CrossRefGoogle ScholarPubMed
Salvador, A.F., de Lima, K.A., & Kipnis, J. (2021). Neuromodulation by the immune system: A focus on cytokines. Nature Reviews Immunology 21, 526–541. https://doi.org/10.1038/s41577-021-00508-zCrossRefGoogle ScholarPubMed
Samamé, C., Cattaneo, B. L., Richaud, M. C., Strejilevich, S., & Aprahamian, I. (2022). The long-term course of cognition in bipolar disorder: A systematic review and meta-analysis of patient-control differences in test-score changes. Psychological Medicine, 52(2), 217–228. doi: 10.1017/S0033291721004517CrossRefGoogle ScholarPubMed
Simonsen, C., Sundet, K., Vaskinn, A., Birkenaes, A. B., Engh, J. A., Faerden, A., … Andreassen, O. A. (2011). Neurocognitive dysfunction in bipolar and schizophrenia spectrum disorders depends on history of psychosis rather than diagnostic group. Schizophrenia Bulletin, 37, 73–83. doi: 10.1093/schbul/sbp034CrossRefGoogle ScholarPubMed
Siqveland, J., Dalsbø, T.K., Harboe, I., & Leiknes, K.A. (2014). Psychometric evaluation of the Norwegian version of the wechsler abbreviated scale of intelligence. Oslo, Norway: Norwegian Knowledge Center for the Health Services. https://www.fhi.no/en/publ/2014/psychometric-evaluation-of-the-norwegian-version-of-the-wechsler-abbreviate/Google Scholar
Stainton, A., Chisholm, K., Griffiths, S.L., Kambeitz-Ilankovic, L., Wenzel, J., Bonivento, C., … Wood, S.J. (2023). Prevalence of cognitive impairments and strengths in the early course of psychosis and depression. Psychological Medicine 53, 5945–5957. https://doi.org/10.1017/S0033291723001770CrossRefGoogle Scholar
Steen, N.E., Rahman, Z., Szabo, A., Hindley, G.F.L., Parker, N., Cheng, W., … Andreassen, O.A. (2023). Shared genetic loci between schizophrenia and white blood cell counts suggest genetically determined systemic immune abnormalities. Schizophrenia Bulletin 49, 1345–1354. https://doi.org/10.1093/schbul/sbad082CrossRefGoogle ScholarPubMed
Torgalsbøen, A.-K., Mohn, C., Larøi, F., Fu, S., & Czajkowski, N. (2023). A ten-year longitudinal repeated assessment study of cognitive improvement in patients with first-episode schizophrenia and healthy controls: The Oslo schizophrenia recovery (OSR) study. Schizophrenia Research 260, 92–98. https://doi.org/10.1016/j.schres.2023.08.008CrossRefGoogle ScholarPubMed
Ullah, I., Awan, H.A., Aamir, A., Diwan, M.N., de Filippis, R., Awan, S., … De Berardis, D. (2021). Role and perspectives of inflammation and C-reactive protein (CRP) in psychosis: An economic and widespread tool for assessing the disease. International Journal of Molecular Sciences 22, 13032. https://doi.org/10.3390/ijms222313032CrossRefGoogle Scholar
van den Ameele, S., van Diermen, L., Staels, W., Coppens, V., Dumont, G., Sabbe, B., & Morrens, M. (2016). The effect of mood-stabilizing drugs on cytokine levels in bipolar disorder: A systematic review. Journal of Affective Disorders 203, 364–373. https://doi.org/10.1016/j.jad.2016.06.016CrossRefGoogle ScholarPubMed
Van Rheenen, T.E., Lewandowski, K.E., Tan, E.J., Ospina, L.H., Ongur, D., Neill, E., Gurvich, C., … Burdick, K.E. (2017). Characterizing cognitive heterogeneity on the schizophrenia-bipolar disorder spectrum. Psychological Medicine 47, 1848–1864. https://doi.org/10.1017/S0033291717000307CrossRefGoogle ScholarPubMed
Vaskinn, A., Haatveit, B., Melle, I., Andreassen, O.A., Ueland, T., & Sundet, K. (2020). Cognitive heterogeneity across schizophrenia and bipolar disorder: A cluster analysis of intellectual trajectories. Journal of the International Neuropsychological Society 26, 860–872. https://doi.org/10.1017/S1355617720000442CrossRefGoogle ScholarPubMed
Vita, A., Barlati, S., Ceraso, A., Nibbio, G., Ariu, C., Deste, G., & Wykes, T. (2021). Effectiveness, core elements, and moderators of response of cognitive remediation for schizophrenia. JAMA Psychiatry 78, 1–12. https://doi.org/10.1001/jamapsychiatry.2021.0620CrossRefGoogle ScholarPubMed
Wallwork, R.S., Fortgang, R., Hashimoto, R., Weinberger, D.R., & Dickinson, D. (2012). Searching for a consensus five-factor model of the Positive and Negative Syndrome Scale for schizophrenia. Schizophrenia Research 137, 246–250. https://doi.org/10.1016/j.schres.2012.01.031CrossRefGoogle ScholarPubMed
Wang, Y., Meng, W., Liu, Z., An, Q., & Hu, X. (2022). Cognitive impairment in psychiatric diseases: Biomarkers of diagnosis, treatment, and prevention. Frontiers in Cellular Neuroscience 16, 1046692. https://doi.org/10.3389/fncel.2022.1046692CrossRefGoogle ScholarPubMed
Watson, A.J., Giordano, A., Suckling, J., Barnes, T.R.E, Husain, B., Jones, P.B., … Joyce, E.M. (2023). Cognitive function in early-phase schizophrenia-spectrum disorder: IQ subtypes, brain volume and immune markers. Psychological Medicine 53, 2842–2851, 35–48. https://doi.org/10.1017/S0033291721004815CrossRefGoogle ScholarPubMed
Watson, A. J., Harrison, L., Preti, A., Wykes, T., & Cella, M. (2022). Cognitive trajectories following onset of psychosis: A meta-analysis. The British Journal of Psychiatry, 221(6), 714–721. https://doi.org/10.1192/bjp.2022.131CrossRefGoogle ScholarPubMed
Webster, M.J. (2023). Infections, inflammation, and psychiatric illness: Review of postmortem evidence. In Savitz, J., Yolken, R.H. (Eds.), Microorganisms and mental health (Vol. 61, pp. 35–48). Cham: Springer. https://doi.org/10.1007/7854_2022_362CrossRefGoogle Scholar
Wechsler, D. (1997). Wechsler Adult Intelligence Scale – third edition. https://doi.org/10.1037/t49755-000CrossRefGoogle Scholar
Wechsler, D. (2011). Wechsler Abbreviated Scale of Intelligence – second edition. https://doi.org/10.1037/t15171-000CrossRefGoogle Scholar
Wenzel, J., Haas, S.S., Dwyer, D.B., Ruef, A., Oeztuerk, O.F., Antonucci, L.A., … Kambeitz-Ilankovic, L. (2021). Cognitive subtypes in recent onset psychosis: Distinct neurobiological fingerprints? Neuropsychopharmacology 46, 1475–1483. https://doi.org/10.1038/s41386-021-00963-1CrossRefGoogle ScholarPubMed
Wenzel, J., Badde, L., Haas, S.S., Bonivento, C., Van Rheenen, T.E., Antonucci, L.A., … Kambeitz-Ilankovic, L. (2023). Transdiagnostic subgroups of cognitive impairment in early affective and psychotic illness. Neuropsychopharmacology, 49, 573–583. https://doi.org/10.1038/s41386-023-01729-7CrossRefGoogle ScholarPubMed
Whitson, S., O'Donoghue, B., Hester, R., Baldwin, L., Harrigan, S., Francey, S., … Allott, K. (2021). Cognitive ability and metabolic physical health in first-episode psychosis. Schizophrenia Research: Cognition, 24, 100194. https://doi.org/10.1016/j.scog.2021.100194Google ScholarPubMed
Wolfers, T., Doan, N.T., Kaufmann, T., Alnæs, D., Moberget, T., Agartz, I., … Marquand, A.F. (2018). Mapping the heterogeneous phenotype of schizophrenia and bipolar disorder using normative models. JAMA Psychiatry 75, 1146–1155. https://doi.org/10.1001/jamapsychiatry.2018.2467CrossRefGoogle ScholarPubMed
Woodward, N.D., & Heckers, S. (2015). Brain structure in neuropsychologically defined subgroups of schizophrenia and psychotic bipolar disorder. Schizophrenia Bulletin 41, 1349–1359. https://doi.org/10.1093/schbul/sbv048CrossRefGoogle ScholarPubMed
Wykes, T., Huddy, V., Cellard, C., McGurk, S.R., & Czobor, P. (2011). A meta-analysis of cognitive remediation for schizophrenia: Methodology and effect sizes. American Journal of Psychiatry 168, 472–485. https://doi.org/10.1176/appi.ajp.2010.10060855CrossRefGoogle ScholarPubMed
Young, R.C., Biggs, J.T., Ziegler, V.E., & Meyer, D.A. (1978). A rating scale for mania: Reliability, validity, and sensitivity. British Journal of Psychiatry 133, 429–435. https://doi.org/10.1192/bjp.133.5.429CrossRefGoogle ScholarPubMed
Zhang, L., Lizano, P., Guo, B., Xu, Y., Rubin, L.H., Hill, S.K., … Bishop, J.R. (2022). Inflammation subtypes in psychosis and their relationships with genetic risk for psychiatric and cardiometabolic disorders. Brain, Behavior, and Immunity - Health 22, 100459. https://doi.org/10.1016/j.bbih.2022.100459CrossRefGoogle ScholarPubMed
Figure 0

Table 1. Sample and clinical characteristics at baseline

Figure 1

Figure 1. Inflammation (C-Reactive Protein, CRP) at baseline and follow-up between HC, BD and SZ.Boxplots (interquartile range separated by median line), density plot (kernel density estimate) and lines between mean scores (including error bars: ± SEM) shows no difference among HC, BD, and SZ at baseline or follow-up and indicates stability of CRP-levels over time.

Figure 2

Figure 2. Inflammatory-cognitive subgroups at baseline and follow-up.Panel A shows the distribution of SZ, BD, and HC in each cluster/subgroup. Panel B shows cluster differences in CRP (log10 transformed) and the cognitive composite score (Z-scores) at baseline and follow-up, with boxplots (interquartile range separated by median line), density plot (kernel density estimate), and line from mean scores (error bars: ± SEM). Panel C shows PANSS factors that were significantly different across clusters at both time points, and panel D shows differences across level of functioning (GAF symptom and GAF Function). Panel C and D is SMI only (line from mean, error bars: ± SEM).

Figure 3

Table 2. Subgroup comparisons on sample and clinical characteristics at baseline and follow-up

Supplementary material: File

Sæther et al. supplementary material 1

Sæther et al. supplementary material
Download Sæther et al. supplementary material 1(File)
File 21 MB
Supplementary material: File

Sæther et al. supplementary material 2

Sæther et al. supplementary material
Download Sæther et al. supplementary material 2(File)
File 21.3 KB
Supplementary material: File

Sæther et al. supplementary material 3

Sæther et al. supplementary material
Download Sæther et al. supplementary material 3(File)
File 50 KB