UTN-number:
U1111-1304-2750
ClinicalTrials.gov:
NCT06431009
Summations
-
• The RMS cohort is a large-scale, prospective, longitudinal cohort of representative first-episode schizophrenia spectrum disorder (SSD) with deep-phenotyped data on biological, psychological and socio-economic measures.
-
• The vision of the cohort is high generalisability and representativeness as recruitment is at all psychiatric hospitals in the Central Denmark Region and all patients with a first-episode SSD will be offered participation.
-
• Participants are included from the age of 15 years, bridging adult and adolescent patients.
Perspectives
-
• Multidimensional integration for aetiological insight: By combining genetic, molecular, inflammatory, clinical and psychosocial data with longitudinal registry information, this study enables comprehensive analyses across the lifespan covering biological and environmental domains to elucidate the complex aetiology and progression of SSD.
-
• Translational potential for personalised psychiatry: The deeply phenotyped and representative cohort supports the development of predictive models and biomarkers for diagnosis, prognosis and individualised treatment, bridging the gap between biological mechanisms and clinical application in SSD research.
Introduction
Schizophrenia spectrum disorders (SSD) affect 2–4% of the population and have a major impact on the individual and their families (Pulay et al., Reference Pulay, Stinson, Dawson, Goldstein Rë, Chou, Huang, Saha, Smith, Pickering, Ruan, Hasin and Grant2009; Jauhar et al., Reference Jauhar, Johnstone and McKenna2022; Tandon, Reference Tandon, Nasrallah, Akbarian, DeLisi, Gaebel, Green, Gur, Heckers, Kane, Malaspina, Meyer-Lindenberg, Murray, Owen, Smoller, Yassin and Keshavan2024). SSD have immense direct (i.e. for hospitalisations and nursing homes) and indirect costs for the society, primarily due to lost productivity as individuals become ill very early in life (Vestergaard et al., Reference Vestergaard, Rasmussen, Stallknecht, Olsen, Skipper, Sørensen and Christiansen2020). The most frequent and severe SSD is schizophrenia, affecting 1–1.5% of the population (Pedersen et al., Reference Pedersen, Mors, Bertelsen, Waltoft, Agerbo, McGrath, Mortensen and Eaton2014; Tandon, Reference Tandon, Nasrallah, Akbarian, DeLisi, Gaebel, Green, Gur, Heckers, Kane, Malaspina, Meyer-Lindenberg, Murray, Owen, Smoller, Yassin and Keshavan2024) with an early age-of-onset (≈20 years) and the majority of patients requiring long-term pharmacological and psychosocial treatment (Rohde et al., Reference Rohde, Højlund, Gasse, Hauser, Petersen, Christensen, Benros, Hallas, Mors and Köhler-Forsberg2022). Despite improving prognostic chances, the most recent meta-analysis suggested 1 in 4 patients recover and 60% showed moderate or better outcomes (Molstrom et al., Reference Molstrom, Nordgaard, Urfer-Parnas, Handest, Berge and Henriksen2022). Furthermore, individuals with schizophrenia on average have 10–15 years shorter life expectancy, which is due to both somatic diseases, including cardiovascular diseases and diabetes, and external causes such as suicide (Plana-Ripoll et al., Reference Plana-Ripoll, CBøcker, Agerbo, Holtz, Erlangsen, Canudas-Romo, Andersen, Charlson, Christensen, Erskine, Ferrari, Iburg, Momen, Mortensen, Nordentoft, Santomauro, Scott, Whiteford, Weye, McGrath and Laursen2019).
The aetiology and prognosis of SSDs is complex, multifactorial and still not fully understood, with both biological, psychological and social aspects interacting and contributing to the individual risk of developing an SSD (Jauhar et al., Reference Jauhar, Johnstone and McKenna2022; Tandon, Reference Tandon, Nasrallah, Akbarian, DeLisi, Gaebel, Green, Gur, Heckers, Kane, Malaspina, Meyer-Lindenberg, Murray, Owen, Smoller, Yassin and Keshavan2024) and affecting the prognosis for an individual with a newly diagnosed SSD. The illness trajectory can vary from good to chronic courses characterised by treatment-resistant symptoms, low functional and cognitive capacity and frequent hospitalisations (Molstrom et al., Reference Molstrom, Nordgaard, Urfer-Parnas, Handest, Berge and Henriksen2022). In addition to the defining features of the disorders, psychotic and negative symptoms, patients frequently suffer from cognitive impairments and other symptoms, such as depression, anxiety and reduced sleep quality.
Inflammation is a suspected biological factor which may contribute to the aetiology and prognosis of SSD (Kohler et al., Reference Köhler, Petersen, Mors, Mortensen, Yolken, Gasse and Benros2017; Gallego et al., Reference Gallego, Blanco, Husain-Krautter, Madeline Fagen, Moreno-Merino, del Ojo-Jiménez, Ahmed, Rothstein, Lencz and Malhotra2018; Kohler-Forsberg et al., Reference Kohler-Forsberg, Petersen, Gasse, Mortensen, Dalsgaard, Yolken, Mors and Benros2018; Orlovska-Waast et al., Reference Orlovska-Waast, Köhler-Forsberg, Brix, Nordentoft, Kondziella, Krogh and Benros2019; Quidé et al., Reference Quidé, Bortolasci, Spolding, Kidnapillai, Watkeys, Cohen-Woods, Berk, Carr, Walder and Green2019; Gangadin et al., Reference Gangadin, Enthoven, van Beveren, Laman and Sommer2024). Patients with SSD have higher inflammatory markers compared to healthy controls, both in peripheral blood (Song et al., Reference Song, Fan, Song, Zhang, Zhang, Li, Gao, Harrington, Ziedonis and Lv2013; Upthegrove et al., Reference Upthegrove, Manzanares-Teson and Barnes2014) and cerebrospinal fluid (Gallego et al., Reference Gallego, Blanco, Husain-Krautter, Madeline Fagen, Moreno-Merino, del Ojo-Jiménez, Ahmed, Rothstein, Lencz and Malhotra2018; Jeppesen et al., Reference Jeppesen, Orlovska-Waast, Sørensen, Christensen and Benros2022). Inflammatory and immunological diseases have been associated with an increased risk of developing an SSD (Kohler et al., Reference Köhler, Petersen, Mors, Mortensen, Yolken, Gasse and Benros2017; Jeppesen et al., Reference Jeppesen and Benros2019; Perry et al., Reference Perry, Zammit, Jones and Khandaker2021) and higher inflammatory markers correlate with worse treatment response (Fond et al., Reference Fond, dAlbis, Jamain, Tamouza, Arango, Fleischhacker, Glenthoj, Leweke, Lewis, McGuire, Meyer-Lindenberg, Sommer, Winter-van Rossum, Kapur, Kahn, Rujescu and Leboyer2015; Enache et al., Reference Enache, Nikkheslat, Fathalla, Morgan, Sôn, Drake, Deakin, Walters, Lawrie, Egerton, MacCabe and Mondelli2021).
Regarding psychological and social factors, it is well known that adverse childhood experiences (ACEs), such as abuse and neglect, represent important risk factors for developing an SSD and negatively affect the prognosis and level of functioning (Varese et al., Reference Varese, Smeets, Drukker, Lieverse, Lataster, Viechtbauer, Read, van Os and Bentall2012; Quidé et al., Reference Quidé, Bortolasci, Spolding, Kidnapillai, Watkeys, Cohen-Woods, Berk, Carr, Walder and Green2019; Dragioti et al., Reference Dragioti, Radua, Solmi, Arango, Oliver, Cortese, Jones, Il Shin, Correll and Fusar-Poli2022, Inyang et al., Reference Inyang, Gondal, Abah, Minnal Dhandapani, Manne, Khanna, Challa, Kabeil and Mohammed2022, Li et al., Reference Li, Jinxiang, Shu, Yadong, Ying, Meng, Ping, Xiao and Yixiao2022). A recent umbrella review found ACEs to be the strongest potentially modifiable risk factor in schizophrenia, attributable to 37.8% of cases (Dragioti et al., Reference Dragioti, Radua, Solmi, Arango, Oliver, Cortese, Jones, Il Shin, Correll and Fusar-Poli2022). Another recent systematic review supported this but emphasised that the correlation is affected by other environmental, genetic and biological factors (Baldwin et al., Reference Baldwin, Wang, Karwatowska, Schoeler, Tsaligopoulou, Munafò and Pingault2023). Of particular importance is a clear interaction between inflammation and ACEs, with several studies emphasising that individuals with an SSD, who have experienced ACEs, show a more pro-inflammatory profile compared to patients without ACEs (Dennison et al., Reference Dennison, McKernan, Cryan and Dinan2012; Aas et al., Reference Aas, Dieset, Hope, Hoseth, Mørch, Reponen, Steen, JFæra, Ueland, Pål, Agartz, Andreassen and Melle2017; Chase et al., Reference Chase, Melbourne, Rosen, McCarthy-Jones, Jones, Feiner and Sharma2019, Quidé et al., Reference Quidé, Bortolasci, Spolding, Kidnapillai, Watkeys, Cohen-Woods, Berk, Carr, Walder and Green2019; Stanton et al., Reference Stanton, Denietolis, Goodwin and Dvir2020, Quidé et al., Reference Quidé, Bortolasci, Spolding, Kidnapillai, Watkeys, Cohen-Woods, Carr, Berk, Walder and Green2021; Den Noortgate et al., Reference Den Noortgate, Morrens, Foiselle and De Picker2025).
Finally, SSD have a clear genetic component, with schizophrenia being among the most heritable mental disorders with an estimated twin heritability of 70–80% (Hilker et al., Reference Hilker, Helenius, Fagerlund, Skytthe, Christensen, Werge, Nordentoft and Glenthøj2018). However, the genetic architecture is highly complex with more than 300 genome-wide significant loci identified for schizophrenia in the most recent large-scale genome-wide association study (GWAS) (Trubetskoy et al., Reference Trubetskoy, Pardiñas, Qi, Panagiotaropoulou, Awasthi, Bigdeli, Bryois, Chen, Dennison, Hall, Lam, Watanabe, Frei, Ge, Harwood, Koopmans, Magnusson, Richards, Sidorenko and O’Donovan2022). These variants individually confer very small effects but collectively highlight key biological pathways, including those related to synaptic function and immune processes, such as the major histocompatibility complex (MHC) (29, 30) (Singh et al., Reference Singh, Poterba, Curtis, Akil, Al Eissa, Barchas, Bass, Bigdeli, Breen, Bromet, Buckley, Bunney, Bybjerg-Grauholm, Byerley, Chapman, Chen, Churchhouse, Craddock, Cusick and Daly2022). Together, these findings underscore the polygenic and heterogeneous nature of the disorder.
Noteworthy, genetic predisposition does not act in isolation. Environmental exposures, such as ACEs, may interact or contribute additively with the underlying genetic vulnerability to the risk and illness course. ACEs are associated not only with a more pro-inflammatory profile but also with long-lasting epigenetic alterations, which may mediate the biological embedding of early life stress. These interactions highlight the complex and dynamic interplay between genetic liability and environmental influences in the aetiology of SSDs, further emphasising the complexity (Houtepen et al., Reference Houtepen, Hardy, Maddock, Kuh, Anderson, Relton, Suderman and Howe2018; Tang et al., Reference Tang, Howe, Suderman, Relton, Crawford and Houtepen2020; Van Den Oord et al., Reference van den Oord, Copeland, Zhao, Xie, Aberg and van den Oord2022).
Hence, it is well established that many different factors contribute to and interact in the development of SSDs. However, most studies on the aetiology and prognosis of SSD were limited by several important aspects. Previous studies were either rather small, did not have information on important measures (e.g. no data on biological measures, sleep or cognition), had no long-term follow-up data (Isele et al., Reference Isele and Angst1985; Mason et al., Reference Mason, Harrison, Glazebrook, Medley and Croudace1996; Golay et al., Reference Golay, Ramain, Abrahamyan Empson, Mebdouhi, Elowe, Solida and Conus2022; Molstrom et al., Reference Molstrom, Nordgaard, Urfer-Parnas, Handest, Berge and Henriksen2022), included patients at different disease stages (e.g. chronic and FES) or had a transdiagnostic inclusion approach (e.g. also included bipolar disorder). Indeed, the largest clinical cohort of patients with SSD included approximately 500 patients with most cohort studies often including less than 200. Consequently, a particular limitation of previous cohort studies (Johnstone et al., Reference Johnstone, Macmillan, Frith, Benn and Crow1990; Lieberman et al., Reference Lieberman, Alvir, Woerner, Degreef, Bilder, Ashtari, Mayerhoff, Geisler, Loebel, Hinrichsen, Szymanski, Chakos, Koreen, Borenstein, Kane, Bogerts and Levy1992; Mason et al., Reference Mason, Harrison, Glazebrook, Medley, Dalkin and Croudace1995; Robinson et al., Reference Robinson, Woerner, McMeniman, Mendelowitz and Bilder2004; He et al., Reference He, Migliorini, Huang, Wang, Zhou, Chen, Xiao, Wang, Wang, Harvey and Hou2022; Jeong et al., Reference Jeong, Kim, Lee, Kim, Yu, Won, Lee, Kim, Kang, Kim, Chung and Lee2022) concerns low statistical power and generalisability. Furthermore, only six prospective studies have follow-up ≥10 years (Petersen et al., Reference Petersen, Jeppesen, Thorup, Abel, Øhlenschlæger, TØstergaard, Krarup, Jørgensen and Nordentoft2005; Hegelstad et al., Reference Hegelstad Wten, Larsen, Børn, Evensen, Haahr, Joa, Johannesen, Langeveld, Melle, Opjordsmoen, Rossberg, Rund Børn, Simonsen, Sundet, Vaglum, Friis and McGlashan2012; Morgan et al., Reference Morgan, Lappin, Heslin, Donoghue, Lomas, Reininghaus, Onyejiaka, Croudace, Jones, Murray, Fearon, Doody and Dazzan2014; Kotov et al., Reference Kotov, Fochtmann, Li, Tanenberg-Karant, Constantino, Rubinstein, Perlman, Velthorst, Fett, Carlson and Bromet2017; Chan et al., Reference Chan, Hui, Chang, Lee and Chen2019; Wold et al., Reference Wold, Kreis, Åsbø, Flaaten, Widing, Engen, Lyngstad, Johnsen, Ueland, Simonsen and Melle2024), although SSD in most cases are life-long disorders. Therefore, there is a clear need for large cohort studies on representative patients with SSD to further study the role of both modifiable risk factors, such as ACEs and inflammation, as well as genetics and epigenetics, in the aetiology and long-term prognosis of SSD.
Objectives
This paper describes the Region Midt Schizophrenia Cohort (termed RMS Cohort) and gives a brief overview of the current status. We aim to establish continuous recruitment for a large representative cohort of patients with a first-episode SSD in the Central Denmark Region (CDR, ‘Region Midt’ in Danish). This will result in a large cohort with detailed information on clinical (e.g. psychotic and negative symptoms, level of functioning, cognition, sleep), psychological (e.g. ACEs) and biological factors (e.g. blood sampling enabling detailed analyses). We aim for long-term follow-up of participants (i.e. 5, 10, 15, 20 years). By merging with the Danish nationwide registers, information on important details such as dispensed medications, hospitalisations and other psychiatric and somatic diagnoses from birth and onwards will be included. Hence, the RMS cohort will enable large-scale studies on several important measures in the aetiology and long-term prognosis of SSD.
Methods and analysis
Design and participants
Study design
This is an investigator-initiated multi-centre non-interventional clinical cohort study with pre-defined longitudinal follow-up visits. In order to participate, patients have to be diagnosed with an SSD (ICD-10: F20-29) by a clinician independent of the present study protocol.
Patient and public involvement statement
Patients were not involved in the initial design of the study. A person with lived experience has been hired soon after recruitment initiation to enable a patient-perspective to the recruitment including material used in the study. We have conducted interviews with several participants to improve recruitment and study design from a patient perspective and to optimise the study design for the long-term follow-up after 2, 3, 5 and 10 years. To qualify the interpretation of our future findings, included patients will also be invited to giving feedback on the presentation of results and conclusions.
Participants and target population
In the CDR, each year approximately 700 individuals are diagnosed with an SSD (hereof approximately 300 with schizophrenia (F20), 80 with schizotypal disorder (F21), 90 with acute psychosis (F23), 30 with schizoaffective disorder (F25) and 200 with non-organic psychosis (F29)). Recruitment takes place at all psychiatric hospitals in the CDR (i.e. in Aarhus, Randers, Horsens, Silkeborg, Viborg and Gødstrup) at both inpatient and outpatient departments through direct contact or contact by mail or telephone. We expect to establish continuous recruitment of 100 patients/year. The largest clinical cohort on patients with SSD included 547 patients (Petersen et al., Reference Petersen, Jeppesen, Thorup, Abel, Øhlenschlæger, TØstergaard, Krarup, Jørgensen and Nordentoft2005). Yet, as SSDs are highly heterogenous, we aim to recruit a larger sample, enabling more distinct patient subgroups and more robust estimates. Hence, the RMS cohort will represent one of the largest cohorts of patients with SSD.
The expected number of dropouts is difficult to estimate. Using the Danish OPUS study (Austin et al., Reference Austin, Mors, Secher, Hjorthøj, Albert, Bertelsen, Jensen, Jeppesen, Petersen, Randers, Thorup and Nordentoft2013) as a representative example, we expect that approximately 90% will participate in the 3-month follow-up visit, 80% in the 1-year follow-up visit and 70% in the 2-year follow-up visit.
Patient screening
Screening of eligible patients is performed by research assistants who are employed part-time at a clinical department at one of the psychiatric hospitals in the CDR and part-time in a research position. If a patient has been diagnosed with an SSD (ICD-10: F20-29) at a psychiatric department within the recent 3 months, the patient is offered the possibility to participate. To ensure consistency in the screening process all individuals are included within 3 months of their introductory appointment in the outpatient psychosis clinic, which is where the diagnosis is confirmed by experienced clinicians. In order to enable broad inclusion, recruitment is possible at all psychiatric in-and outpatient departments in the CDR.
Inclusion criteria
-
1. Age ≥15 years
-
2. Diagnosed within the previous 3 months with first-episode SSD (ICD-10: F20-29)
-
3. Able to give informed oral and written consent in Danish or English.
Exclusion criteria
-
1. Any coercive measure at the time of informed consent including patients in forensic psychiatry. That is, patients with a previous involuntary admission are eligible as soon as the admission becomes voluntary or they are discharged.
-
2. In a clinical condition where the treating clinician evaluates that the patient is not able to attend the research study.
Patient selection
Any patient aged ≥15 years diagnosed with first-episode SSD and without any coercive measure at the time of inclusion is relevant for inclusion in the study.
Patient withdrawal
Patients are free to withdraw from the study at any time without providing reason(s) for withdrawal and without prejudice to further treatment. Data from patients who discontinued participation will be included in the data processing unless the patient explicitly requests exclusion from this.
Timetable
All psychiatric departments in the CDR participate in the inclusion of patients. The anticipated timetable for the study is:
Inclusion of adolescents aged 15–18 years
The present study includes adolescents (i.e. minors aged 15–18 years), since SSDs develop early in life. The mean age of onset for schizophrenia is 20 years and a substantial number of individuals are diagnosed before the age of 18 years (Pedersen et al., Reference Pedersen, Mors, Bertelsen, Waltoft, Agerbo, McGrath, Mortensen and Eaton2014; Kahn et al., Reference Kahn Ré, Sommer, Murray, Meyer-Lindenberg, Weinberger, Cannon, O’Donovan, Correll, Kane, van Os and Insel2015). Such an early onset is often associated with a poorer prognosis (Kahn et al., Reference Kahn Ré, Sommer, Murray, Meyer-Lindenberg, Weinberger, Cannon, O’Donovan, Correll, Kane, van Os and Insel2015), emphasising the need for research in these young individuals. We restrict inclusion to children aged between 15 and 18 years of age in order to only include children who are rather well developed (e.g. compared to children aged 10 years) and because most cases of schizophrenia with onset before the age of 18 years occur in the age range 15–18 years (Pedersen et al., Reference Pedersen, Mors, Bertelsen, Waltoft, Agerbo, McGrath, Mortensen and Eaton2014). According to the Danish Health Research Ethics Committee, participation of adolescents in health research requires the informed consent of both the individual participant and his or her parents.
Inclusion of healthy controls
We will include healthy controls to allow for age-and sex-matched comparisons for several of the analyses, for example, the genetic analyses. Healthy controls will be recruited via social media according to the following in-and exclusion criteria:
-
• Aged ≥15 years.
-
• No history of a mental disorder requiring psychopharmacological treatment or treatment in a psychiatric hospital, established via The Scheduled Clinical Assessment in Neuropsychiatry (SCAN) interview.
-
• No history of psychotic disorders, major depressive disorder or bipolar disorder among first-degree relatives.
-
• No current psychoactive substance use disorders (except nicotine).
-
• No previous serious head trauma, neurological or medical disorder that could affect brain functioning.
-
• An estimated IQ >70 based on prior history or testing.
Procedures
Pre-inclusion evaluation – screening visit
All participants are provided oral and written information about the study, including the procedures involved in the study. If the participant meets all inclusion criteria and no exclusion criteria and a SCAN interview has ensured a correct diagnosis or excluded mental disorders in healthy controls, an appointment for the inclusion visit is set up.
Clinical assessment – baseline and follow-up visits for patients
Please see Figure 1 for an overview of the study design. At the inclusion visit, the patient participates in an interview covering the following rating scales (see Table 1 for details) conducted by a trained research assistant (i.e. either nurses, psychologists or doctors with clinical experience): 6-item Positive and Negative Syndrome Scale (PANSS-6), Clinical Global Impression Severity Scale (CGI-S), Global Assessment of Functioning Scale (GAF) and ‘Udvalg for Kliniske Undersøgelser’ side effect scale (UKU). Furthermore, patients will fill out the self-reported WHO Adverse Childhood Experience International Questionnaire (ACE-IQ), Aarhus Side effect Assessment Questionnaire (ASAQ), Insomnia Severity Index (ISI), Pittsburgh Sleep Quality Index (PSQI), Alcohol Use Disorders Identification Test (AUDIT), Drug Use Disorders Identification Test (DUDIT), the Fagerström Test for Nicotine Dependence (FTND) and Schizophrenia Quality of Life Scale (SQLS), the Deliberate Self-Harm Inventory (DSHI), the bodily distress syndrome (BDS) Checklist and the Whiteley-6 checklist. In addition, to assess cognition, we perform the Matrics Consensus Cognitive Battery (MCCB), which measures seven cognitive domains that are typically affected in SSD. Some scales are repeated at the physical follow-up visits after 1, 2, 3, 12 and 24 months (see Table 1 for an overview of the timing of the assessments). Furthermore, participants fill out the Consensus Sleep Diary to be able to compare self-reported data on sleep to the objectively measured sleep via actigraphy. Finally, at the online questionnaires at week 2, 6, 10 and 26, participants fill out the ISI and the symptom self-rating scale for schizophrenia (4S). Weight, height, waist circumference, blood pressure and heart rate are measured. Current antipsychotic, other psychotropic and somatic pharmacological treatment is drawn from the electronic medical journal at the inclusion visit, 3 months follow-up, 1, 2, 3, 5 and 10 years follow-up. At the same visits medication adherence is assessed with the Brief Adherence Rating Scale (BARS).
Procedures for participants in the trial
a Performed within 2 weeks of the inclusion visit.
* Send to the participants online via RedCAP.
^ Self-report scales, filled out during a visit while the research assistant is filling out other information or send online.
1 Filled out by the research assistant independent of the participant, for example, while the participant fills out self-report forms.
The consent form includes a paragraph where participants agree to being contacted for long-term follow-up visits after 5 and 10 years.
Blood samples are collected at baseline and after 3, 12 and 24 months of follow-up from an antecubital vein to establish a research biobank. Samples from the regional hospitals (i.e. Randers, Viborg, Horsens, Silkeborg, Gødstrup) are transported to Aarhus. For immediate isolation of peripheral blood mononuclear cells (PBMCs), blood is drawn into lithium–heparin tubes, and the isolated PBMCs are cryopreserved and stored in liquid nitrogen. Ethylenediaminetetraacetic acid (EDTA) tubes are used for collection of whole blood for plasma preparation and subsequent DNA extraction. An additional sample is collected in a clot activator tube for serum preparation. Plasma, serum and whole blood/DNA isstored at-80C. The analyses of the collected blood tests will be performed when a sufficient number of blood tests have been collected and in as few batches as possible to minimise the risk for measurement bias.
Data from clinical blood tests will be recorded (e.g. on C-reactive protein (CRP), Haemoglobin A1c (HbA1c), high-density lipoprotein, low-density lipoprotein, triglycerides and organ markers including thyroid stimulating hormone, creatinine and alanine aminotransferase).
At the first visit, participants are handed out a GENEActiv 1.2 actigraph (Activinsights, Kimbolton, England), which is worn at their non-dominant wrist for 3 months. Data from the actigraph will be used as a proxy for sleep, mobility and activity. For measurements at 40 Hz battery life of these actigraphs is approximately 36 days, with battery change occurring at the clinical visits after 1 and 2 months.
All clinical visits are conducted by trained investigators at the psychiatric hospitals in Aarhus, Randers, Horsens, Gødstrup, Silkeborg and Viborg. All investigators are trained in SCAN interviews including all clinical assessments, rating scales and the eliciting of information from study participants in a uniform reproducible manner. Regular co-ratings are performed at least every three months with experienced expert raters (e.g. on the PANSS-6, UKU, MCCB).
Non-interventional study
Participation in the study will not affect the participants’ present or future treatment. Patients will receive pharmacological and non-pharmacological treatment according to treatment guidelines at the local hospitals, that is, patients follow treatment-as-usual and participation in the RMS cohort is independent of previous and present treatment.
Inclusion visit for healthy controls
Healthy controls will participate in one visit, where we will take a blood test and conduct the following scales: GAF, ACE-IQ, ISI, PSQI, AUDIT, DUDIT, FTND, DSHI, BDS Checklist and the Whiteley-6 checklist. Weight, height, waist circumference, blood pressure and heart rate will be measured. Sociodemographic data as well as medication use will be recorded. The blood test will be handled and prepared in the same way as the blood test for patients.
Molecular phenotyping
DNA will be extracted from blood samples collected at baseline and genome-wide genotyping performed using high-density SNP arrays. In addition, the cohort will provide opportunities to investigate the contribution of rare genetic variation. This may include exome or genome sequencing in subsets of participants, or targeted sequencing in follow-up projects, to complement common variant analyses.
Blood-derived DNA will also be used for DNA methylation (DNAm) profiling with high-throughput technologies to explore epigenetic signatures.
Peripheral blood mononuclear cells (PBMCs) will be isolated for immunophenotyping using flow cytometry (FACS) and single-cell technologies, including scRNA-seq and scATAC-seq, to characterise immune cell states, gene expression and chromatin accessibility at high resolution.
Plasma and serum samples will be subjected to large-scale proteomic profiling. Proteomic platforms will be updated over time to reflect methodological advances.
Together, these molecular data layers will form a biobank resource for integrated analyses of genetic, epigenetic, immune and proteomic mechanisms in SSD.
Merging with Danish nationwide registers
The clinical RMS data will be merged with the Danish nationwide registers, covering the Danish Psychiatric Central Research Register (Mors et al., Reference Mors, Perto and Mortensen2011), the Danish National Prescription Register (Pottegard et al., Reference Pottegård, Schmidt, Wallach-Kildemoes, Sørensen, Hallas and Schmidt2017), the Danish National Patient Register (Schmidt et al., Reference Schmidt, Schmidt, Sandegaard, Ehrenstein, Pedersen and Sørensen2015), the Danish Civil Registration System (Pedersen, Reference Pedersen2011) and the Danish National Education and Workforce Registers (Jensen et al., Reference Jensen and Rasmussen2011). Merging of the clinical research data and the register-based data will occur at a secured server at Statistics Denmark, which is a well-established procedure (Posselt et al., Reference Posselt, Albert, Nordentoft and Hjorthøj2021; Kofod et al., Reference Kofod, Elfving, Nielsen, Mors and Köhler-Forsberg2022). This will enable to link important information on the participants from birth and onwards.
Study outcomes
Primary outcome measures
The primary outcomes will explore how ACEs, genetic risk profiles and molecular markers (including immune, epigenetic, transcriptomic, proteomic and metabolic data) are associated with illness course and prognosis in SSD. Analyses will examine how these factors relate to case status, symptom severity (PANSS-6), cognitive performance (MCCB), relapse, hospitalisation rates and treatment response. The aim is to identify integrated molecular and environmental signatures that can inform diagnosis, prognosis and personalised treatment strategies. These outcomes will be considered hypothesis-generating and adjusted for multiple testing.
Secondary outcome measures
Secondary endpoints include:
-
• Sleep:
-
• Mean (SD) total sleep time for 3 months as measured by actigraphy in users of sleep medication compared to non-users.
-
• Mean (SD) sleep quality for three months measured by ISI and PSQI in users of sleep medication compared to non-users.
-
-
• Cognition:
-
• Mean (SD) scores for the 10 tests in the MCCB at baseline and follow-up.
-
-
• Cardiometabolic:
-
• Association between ACEs with cardiometabolic measures (e.g. weight, HbA1c, cholesterol)
-
-
• Long-term outcomes via Danish registers
-
• Associations between symptoms (via PANSS-6), cognition (via MCCB), level of functioning (via GAF) and sleep (ISI, PSGI, actigraphy) with long-term outcomes on relapse/hospitalisation (via National Hospital Register), antipsychotic medication use (via National prescription Register) and psychosocial functioning (via National Education and Workforce Registers).
-
Statistical analyses
All clinical data will be presented by means and standard deviations or medians and interquartile ranges. All data will be checked for normal distribution and, if deemed necessary for statistical analyses, be transformed if no normal distribution is present (e.g. log-transformation for inflammatory markers, where we expect a non-normal distribution).
We will perform multivariable linear or logistic regression analyses between the variable of interest (e.g. the number of ACEs experienced) and the outcome of interest (e.g. PANSS-6 at follow-up). For example, the analyses on ACEs will be performed as multivariable linear regression analyses using the number of reported ACEs as a continuous variable (i.e. 0, 1, 2, 3, etc. ACEs) and as logistic regression analyses for dichotomous variables (i.e. yes = 1 or no = 0 for the specific ACEs).
We will perform multimodal regression and causal mediation analyses to study potential correlations between the exposure (e.g. ACEs, multi-polygenic score or immune system pathways) with the outcomes of interest (e.g. PANSS-6, MCCB, sleep, level of functioning, hospitalisation rates) and whether a measure may represent a mediator on a pathway, for example, inflammation on the pathway between ACEs and SSDs.
We will perform cross-sectional association analyses with PGSs for (i) psychiatric disorders and (ii) comorbid and related traits (e.g. cognition, sleep, cardiometabolic, immune traits), as well as with epigenetic markers, to map their correlation with patient characteristics at baseline and follow-up visits. We will also study how patient PGSs together with longitudinal changes in epigenetic markers correlate with outcomes at follow-up visits. PGSs will include both disorder-specific (e.g. schizophrenia, depression, bipolar disorder) and domain-specific scores (e.g. cognition, sleep, cardiometabolic traits). Multi-PGS approaches and multivariate models will be applied to investigate how different genetic liabilities contribute jointly to early illness course. Where sequencing data are available, rare variant burden testing will complement common variant analyses.
To maximise the value of our comprehensive molecular phenotyping, we will employ machine learning techniques and neural network models to integrate genetic, epigenetic, environmental (e.g. ACEs), transcriptomic, proteomic and clinical data. By combining these diverse data types, we aim to identify multimodal signatures that predict clinical outcomes and trajectories, such as symptom severity, long-term prognosis and treatment response. This integrative approach will allow us to uncover complex, non-linear relationships between molecular markers, environmental factors and clinical phenotypes and thereby refine risk stratification strategies for personalised treatment.
External GWAS summary statistics will be used to perform genetic correlation analyses and Mendelian randomisation, enabling tests of causal relationships between exposures (e.g. inflammation, ACEs) and SSD outcomes. Mendelian randomisation (MR) analyses will primarily be conducted in a two-sample framework using large external GWAS datasets, while within-cohort MR will be considered exploratory given expected sample size limitations. Finally, findings will be replicated and meta-analysed through participation in international consortia such as the Psychiatric Genomics Consortium (PGC), ensuring robustness and comparability across independent cohorts.
Ethics and dissemination
The study has been approved by the institutional review board in the Central Denmark Region (‘Videnskabsetisk Komite for Region Midtjylland’, 1-10-72-147-23) and registered at WHO (UTN-number: U1111-1304-2750) and at Clinicaltrials.gov (NCT06431009). Study findings, independent of whether they are positive, negative or inconclusive, will be published in international peer-reviewed scientific journals. The Helsinki declaration will be followed during all steps of the study. Authorship will be based on the ICJME criteria.
Status
Research assistants have been trained and data collection has started at all sites, which exceeds the original plan. Data collection started in February 2024 and by February 23rd, 2026, 131 patients have agreed to participate (109 with schizophrenia, median age 25 years (IQR 8), 54% females), Figure 1. Retention rates at the follow-up visits after 1, 2, 3 and 12 months are 86%, 80%, 78% and 75%, respectively. Five participants have withdrawn their consent. Not all patients qualified for inclusion have been asked to participate in the beginning of the project, due to initial stage delays.
An overview of visits, assessments and preliminary data including retention rates for the RMS cohort. Abbreviations: ACEs: adverse childhood experiences, ASAQ: Aarhus Side effect Assessment Questionnaire, PANSS-6: 6-item Positive and Negative Syndrome Scale, RMS: Region Midt Schizophrenia, SCAN: The Scheduled Clinical Assessment in Neuropsychiatry, UKU: Udvalg for Kliniske Undersøgelser (side effect scale).
Figure 1. Long description
Panel A: The diagram begins with a schizophrenia diagnosis, followed by a series of visits and assessments over a two-year period. Initial baseline assessments include SCAN, PANSS-6, UKU, ASAQ, sleep, functioning, and ACEs. Follow-up visits occur at weeks 4, 8, and 12, involving blood tests, rating scales, and actigraphy. Online scales are assessed at weeks 2, 6, and 10. At week 26, an online questionnaire is conducted. Another follow-up visit with rating scales is scheduled at week 52. Panel B: Preliminary data indicates that 110 patients were recruited across seven sites, with a median age of 25.5 years and 54 percent being female. Retention rates are high, with 86 percent at one month, 79 percent at two months, 76 percent at three months, and 76 percent at twelve months. Blood tests were taken for 93 percent at baseline, 92 percent after three months, and 84 percent after twelve months. Actigraphy data was collected for 74 percent for three months. Cognitive tests were performed on 70 percent of participants, with 80 percent filling out WHO-ACE-IQ, where 92 percent had experienced more than one ACE, with a mean of 5.2 and a standard deviation of 2.9.
Perspectives
The overall aim of this study is to investigate the contribution of genetics, biological and psychosocial factors to the aetiology and prognosis of SSD in a large, deeply phenotyped cohort. By integrating data from national registries and medical charts, the present cohort will be able to picture the lifetime exposure to many important risk factors, enabling comprehensive analyses on the aetiology and prognosis of SSDs (Figure 2).
An overview of all data and their coverage of the lifespan of patients with SSD in the RMS cohort, using immunological characterisation as an example for biological measures.
Importantly, the integration of molecular data layers (genetics, epigenetics, immunological markers, proteomic and transcriptomic profiles) with clinical, environmental and register-based information will enable analyses across biological and psychosocial domains. This multidimensional approach will not only improve understanding of disease mechanisms but also support the development of predictive models and novel biomarkers for diagnosis, prognosis and personalised treatment in SSD.
Studies on ACEs and correlation with clinical phenotype
One of the main aims with the RMS cohort is to study the impact of ACEs on the aetiology and prognosis of SSDs. The RMS cohort aims to contribute to the literature within ACEs and schizophrenia by providing a deep-phenotyped comprehensive dataset on well-characterised patients with detailed information on ACEs and many important aspects, such as symptoms, cognition, level of functioning, drug abuse or biological measures.
In addition, the RMS cohort aims to provide comprehensive analyses on several clinically important areas, such as cognition, psychopathology or sleep.
Cognition
Approximately three quarters of patients with schizophrenia display impairments in at least two cognitive domains (Heinrichs et al., Reference Heinrichs and Zakzanis1998). At a group level, the disturbances are global (Blanchard et al., Reference Blanchard and Neale1994) and are thought to play a causal role in the development of SSD (Toulopoulou et al., Reference Toulopoulou, Zhang, Cherny, Dickinson, Berman, Straub, Sham and Weinberger2019) with most patients showing cognitive impairments before the onset of psychosis (Mortensen et al., Reference Mortensen, Sørensen, Jensen, Reinisch and Mednick2005). After onset, the impairments are relatively stable (Reichenberg et al., Reference Reichenberg, Weiser, Rapp, Rabinowitz, Caspi, Schmeidler, Knobler, Lubin, Nahon, Harvey and Davidson2005); however, a subgroup experiences progressive cognitive decline (Hoff et al., Reference Hoff, Svetina, Shields, Stewart and DeLisi2005). The cognitive impairments are a strong predictor of functional prognosis (Green, Reference Green2006) and related to treatment compliance (Green et al., Reference Green, Kern and Heaton2004) and clinical prognosis (Moritz et al., Reference Moritz, Andresen, Perro, Schickel, Krausz and Naber2002). (Kim et al., Reference Kim, Oh, Hong and Choi2011; Karamaouna et al., Reference Karamaouna, Zouraraki and Giakoumaki2021).
Research into the cognitive impairments seen in SSD has been characterised by relatively small cohorts, short follow-up periods and methodological weaknesses, for example, lack of consensus on which cognitive domains should be assessed with which tests. Some of these limitations have been countered by the development of the MCCB, which today is often highlighted as the ‘golden standard’ (Bismark et al., Reference Bismark, Thomas, Tarasenko, Shiluk, Rackelmann, Young and Light2018).
To address limitations from previous research, the RMS cohort will use the MCCB to perform several cognitive assessments during long-term follow-up while having detailed data on other important measures potentially affecting cognitive performance (e.g. sleep, drug abuse). The RMS cohort will thereby contribute to a deeper understanding of the aetiology of cognitive impairments in SSD. For instance, by studying the interactive effects between genetics and ACEs on specific cognitive domains and the development of cognition after the first SSD diagnosis, we aim to get a better understanding of the different cognitive courses and which factors contribute to progressive cognitive decline.
Sleep
Sleep problems are frequent in FES (Ma et al., Reference Ma, Song, Xu, Tian and Chang2018; Reeve et al., Reference Reeve, Sheaves and Freeman2019; Ayers et al., Reference Ayers, McCall and Miller2023) and are associated with low quality of life (Ong et al., Reference Ong, Tan, Shahwan, Satghare, Cetty, Ng, Tang, Verma, Chong and Subramaniam2020) and a 3-fold increased risk of suicidal ideation (Ayers et al., Reference Ayers, McCall and Miller2023). Cognitive behavioural therapy has proven effective (Pant and Mishra, Reference Pant and Mishra2022), but most patients with sleep problems are treated pharmacologically. Sleep problems in schizophrenia have been studied in cross-sectional designs or in cohorts with short follow-up of 1–2 weeks. The three months with actigraphy and sleep quality questionnaires in the RMS cohort will represent the longest investigation of sleep problems in SSD, coupled with a week-by-week investigation of the usage of sleep-promoting drugs. The RMS cohort will represent the longest and most comprehensive investigation of sleep and activity patterns combined with use of sleeping aids (i.e. on-and off-label medications, herbal medicine, recreational drugs). The long-term aims are (1) to inform intervention trials, such as efficacy of sleeping medications head-to-head and (2) to investigate sleep instability as a biomarker for schizophrenia severity, possibly aiding patients and professionals in earlier detection of clinical relapse.
Studies on the biology of schizophrenia
Inflammation and omics
The RMS cohort enables a comprehensive, systems-level investigation of immune and molecular mechanisms by combining longitudinal sampling with high-dimensional single-cell analyses (e.g. Fluorescence-activated cell sorting (FACS), single cell RNA-sequencing, single cell Assay for Transposase-Accessible Chromatin-sequencing) and broad molecular profiling via serum proteomics. This integrated multi-omic approach will allow for the delineation of dynamic immune cell states, gene regulatory landscapes and systemic molecular signatures, and how these relate to clinical trajectories and environmental exposures. This design offers a unique opportunity to elucidate molecular pathways and to identify biologically anchored predictors of clinical course and treatment responses in SSD.
Genetics/epigenetics
The RMS cohort presents a unique opportunity to advance genetic research in schizophrenia by providing a well-characterised sample of individuals at various stages of the disease. By leveraging detailed clinical phenotyping, longitudinal follow-up and molecular data, the cohort allows for a deeper understanding of how genetic factors relate to clinical profiles, disease trajectories and treatment responses.
A key advantage of the RMS cohort is its integration of clinical data, genetic profiles, epigenetic markers, blood markers and environmental factors, such as ACEs, all of which contribute to a comprehensive understanding of SSD. Combined with state-of-the-art genomic techniques and multi-omics approaches, the cohort’s longitudinal design enables the exploration of how genetic predispositions interact with environmental influences over time and how genetic factors influence molecular pathways and biomarkers associated with schizophrenia. Additionally, it will allow us to determine how genetic biomarkers, both in isolation and in combination with other molecular data, may improve risk stratification and trait specification - thereby enabling personalised treatment strategies.
Strengths and limitations of a real-world representative cohort
The broad inclusion and limited exclusion criteria (i.e. also including patients with concurrent drug abuse, low level of functioning, reduced cognitive abilities, comorbidities, suicidal thoughts) will ensure a representative and generalisable cohort. The inclusion of first-episode patients and the planned long follow-up yields a unique opportunity for studying prognostic factors in schizophrenia.
Concerning limitations, although we aim for broad inclusion of representative patients, participation in the cohort entails extra visits and demands. Hence, there is a risk that the most severely ill patients will not be able to participate, which can lead to selection bias towards a higher level of functioning. Second, the representative nature of the RMS cohort will likely result in a rather heterogeneous cohort, which necessitates a high number of participants to ensure sufficient statistical power. Third, despite the multi-centre design, we will not be able to screen all eligible patients in the CDR. Fourth, several proposed outcomes measures are secondary and/or hypothesis-generating, which will be clearly emphasised.
Long-term aims
We aim to continue recruitment and extend the present set-up to follow participants for 20 years with phenotyped data and biological samples. This long-term follow-up and the combination with the comprehensive Danish register data will make the RMS cohort a unique research tool to study an array of important aspects.
Basis for future studies
By having established a recruitment pipeline, the RMS cohort will serve as a recruitment basis for other future studies and collaborations. For example, the detailed knowledge of participants in the RMS cohort has enabled the initiation of an entirely new PET-MR study on brain metabolism, where we recruit antipsychotic-naïve FES patients and measure glucose and ketone body concentrations in the brain compared to healthy controls and before and after antipsychotic treatment (Zemmour et al., Reference Zemmour, Samson, Fortier, Parent, Leus, Grignon, Carrier, Whittingstall, Danielsen, Köhler-Forsberg, Hansen, Agarwal, Andreazza, Ebdrup, Hahn and Cunnane2025). The RMS cohort enables recruitment of these difficult-to-recruit patients and by participating in both the PET study and the RMS cohort, this will enable additional data on potential biological mechanisms and pathways in SSDs. Furthermore, the RMS cohort is intended to establish research collaborations with national and international institutions and private entities.
Comparison with other large schizophrenia cohorts
Due to the complexity of SSD, it is imperative to replicate research findings in independent cohorts. Therefore, we aim to establish international collaborations with consortia, such as the FEP-GEN consortium, and other research groups recruiting similar cohorts, such as the US ‘EPINET,’ French Minds and FACE.
Acknowledgements
We thank Sissel Holm Madsen, Mathias Stæhr and Hanne Barckmann for their contributions in recruiting participants for the cohort in Aarhus, Gødstrup and Silkeborg, respectively.
Author contributions
OKF developed the idea and is the grant holder. OKF, NFB and AL coordinated the study and wrote the first draft of the manuscript. The remaining authors contributed with data collection and/or intellectual input. All authors have read and commented on drafts and have approved the final manuscript. All author(s) meet criteria for authorship as recommended by the International Committee of Medical Journal Editors (ICMJE).
Funding statement
This work was supported by Strategic Research Funding from the Central Denmark Region, a PhD-scholarship from Aarhus University (for MD Nina FB Fuglsang) and an unrestricted grant from Boehringer-Ingelheim (BI). This was a collaborative research study where BI was involved in the design, analysis or interpretation of the results but was not the regulatory sponsor. BI was given the opportunity to review the manuscript for medical and scientific accuracy as it relates to BI substances, as well as intellectual property considerations.
Competing interests
OKF reports honoraria for lectures or advisory boards from Lundbeck Pharma A/S, Teva Pharmaceuticals and Boehringer-Ingelheim and consultant work for WCG International. NFB reports honoraria for lectures from Lundbeck Pharma A/S.
Ethical standards
The protocol has been approved by the ethics committee of the Central Denmark Region (reference: 1-10-72-147-23). Study findings will be published in peer-reviewed journals independent of whether they are positive, negative or inconclusive and will be presented at national and international conferences.
Artificial intelligence contributions
Artificial Intelligence was used for language refinement in a few sections of the abstract and introduction.
Open access
