Exercise4Psychosis: A randomised control trial assessing the effect of moderate-to-vigorous exercise on inflammatory biomarkers and negative symptom profiles in men with first-episode psychosis

on

Introduction: First-Episode Psychosis (FEP) is a devastating mental health condition that commonly emerges during early adulthood, and is characterised by a disconnect in perceptions of reality.Current evidence suggests that inflammation and perturbed immune responses are involved in the pathology of FEP and may be associated specifically with negative symptoms.Exercise training is a potent anti-inflammatory stimulus that can reduce persistent inflammation, and can improve mood profiles in general populations.Therefore, exercise may represent a novel adjunct therapy for FEP.The aim of this study was to assess the effect of exercise on biomarkers of inflammation, negative symptoms of psychosis, and physiological health markers in FEP.Methods: Seventeen young males (26.67 ± 6.64 years) were recruited from Birmingham Early Intervention in Psychosis Services and randomised to a 6-week exercise programme consisting of two-to-three sessions per week that targeted 60-70 % heart-rate max (HRMax), or a treatment as usual (TAU) condition.Immune T-helper (Th-) cell phenotypes and cytokines, symptom severity, functional wellbeing, and cognition were assessed before and after 6-weeks of regular exercise.Results: Participants in the exercise group (n = 10) achieved 81.11 % attendance to the intervention, with an average exercise intensity of 67.54 % ± 7.75 % HRMax.This led to favourable changes in immune cell phenotypes, and a significant reduction in the Th1:Th2 ratio (− 3.86 %) compared to the TAU group (p = 0.014).After the exercise intervention, there was also a significant reduction in plasma IL-6 concentration (-22.17%) when compared to the TAU group (p = 0.006).IL-8, and IL-10 did not show statistically significant differences between the groups after exercise.Symptomatically, there was a significant reduction in negative symptoms after exercise (− 13.54 %, Positive and Negative Syndrome Scale, (PANSS) Negative) when compared to the TAU group (p = 0.008).There were no significant change in positive or general symptoms, functional outcomes, or cognition (all p > 0.05).Discussion: Regular moderate-to-vigorous physical activity is feasible and attainable in clinical populations.Exercise represents a physiological tool that is capable of causing significant inflammatory biomarker change and concomitant symptom improvements in FEP cohorts, and may be useful for treatment of symptom profiles that are not targeted by currently prescribed antipsychotic medication.

Introduction
First-Episode Psychosis (FEP) is a severe mental illness that presents during early adulthood, and is characterised by a disconnection from reality in one's perceptions and thoughts.In England, the United Kingdom, FEP affects 31.7 individuals per 100,000, and is particularly prevalent in young males, ethnic minority groups, and people of low socio-economic status (Kirkbride et al., 2012;Fearon et al., 2006).Individuals with FEP may experience positive symptoms (delusions, hallucinations, paranoia), negative symptoms (anhedonia, anergia, apathy), and difficulties with cognition and thought processing, all of which severely disrupt daily living and have crippling effects on individuals' quality of life (Eack and Newhill, 2007).FEP can represent an early clinical stage of schizophrenia (SCZ) and, as such, is a critical intervention point to promote recovery and prevention of progression to chronic forms of disease (McGorry et al., 2007;McGorry and Yung, 2003).Individuals experiencing FEP are typically prescribed second generation antipsychotic medication, which function to ameliorate positive symptoms via occupation of dopamine D2 receptors (Kaar et al., 2020).Concerningly, currently available medications largely fail to tackle or improve the negative symptoms of psychosis, and enduring negative symptomology is reported to be the driving force of the severe reductions in quality of life reported (Foussias et al., 2014).Given that FEP and SCZ pathology involves a dominant inflammatory component (Fraguas et al., 2019;Upthegrove et al., 2014;Upthegrove and Khandaker, 2020), and that persistent inflammation in FEP may contribute specifically to negative symptom presentation (Dunleavy et al., 2022), targeting immune responses as a means of adjunct treatment may hold therapeutic potential for alleviation of negative symptoms.
The immune system is comprised of an intricate network of cells, molecules, and signals that function synchronously to shield the human body from external stressors.Innate immunity is the initial response system, which is non-specific and rapid, targeted towards the elimination of pathogens via recognition of 'common' stressors or molecules (McComb et al., 2019).Acquired immunity is antigen specific and is slower to mobilise to stressors, operating on the basis of a 'memory' function, and is carried out by T-helper (Th) lymphocytes.These cells become active in response to stress and adopt unique phenotypes to carry out specific integrated immune responses, including the orchestration and secretion of various cytokines to combat stressors.
The 'two-hit' hypothesis of SCZ suggests that genetic perturbations increase the risk of poor immune responses to stress, and repeated stress exposure functions to bring about disease pathology in early adulthood (Müller, 2018).In FEP there appears to be an overactivation in immune responses, characterised by macrophage activation, elevated presence of proinflammatory Th1 and Th17 lymphocyte phenotypes, and increased circulation of proinflammatory cytokines including significantly elevated interleukin(IL)-1β, IL-6, IL-8, IL-12, Interferon-γ, Tumor Necrosis Factor-ɑ (TNF-ɑ), and Transforming Growth Factor-β (Fraguas et al., 2019;Upthegrove et al., 2014;Dunleavy et al., 2022;Miller et al., 2011;Orlovska-Waast et al., 2019).Large-scale genetic studies have even suggested that genetically predetermined elevations in IL-6 may have a causal role in SCZ pathology, alongside other psychiatric conditions (Williams et al., 2022).In early FEP pathology, there is also some evidence of early Th2 and Treg activation (Noto et al., 2019), which may be an attempt to mount a protective response against ongoing inflammation.It is possible that as disease pathology progresses, inflammation worsens and anti-inflammatory immune function fails, leading to persistent inflammation (Roomruangwong et al., 2020).Mechanistically, chronically elevated proinflammatory mediators can infiltrate the brain parenchyma to cause damage to cerebral tissue, and alterations in neurotransmission that may be responsible for the emergence of psychosis pathology.
Indeed, immune activation, cytokine perturbations, and inflammation have been linked with negative symptom pathology in FEP (Dunleavy et al., 2022;Goldsmith and Rapaport, 2020).Chronically activated immune cells and circulating cytokines can enter the CNS via leaky sections of the blood-brain barrier or saturable transporters, and cause activation of resident cerebral immune cells including microglia or astrocytes, leading to neuroinflammation (Banks, 2005;Comer et al., 2020).This has been linked with alteration in dopaminergic and glutamatergic signalling within the prefrontal areas that may precede the emergence of negative symptoms specifically (de Bartolomeis et al., 2022).Individuals presenting with negative symptoms often exhibit decreased activation of prefrontal brain areas associated with motivation, reward processing, and emotion, all of which have been correlated to elevated inflammatory biomarkers (Goldsmith and Rapaport, 2020;Goldsmith et al., 2019).In more chronic forms of SCZ pathology, there appears to be an intimate link between inflammation and 'deficit' features, which are defined as persistent and enduring negative symptoms that show no real trajectory of improvement (Garcia-Rizo et al., 2012;Goldsmith et al., 2018), and these negative symptoms continue indefinitely to significantly reduce social, occupational, and functional outcomes (Strauss et al., 2010).Therefore, adjunctive therapy tailored more specifically towards the reduction of inflammatory biomarkers may hold therapeutic potential for psychosis populations experiencing predominantly negative symptoms.
Exercise is a potent anti-inflammatory stimulus (Febbraio and Pedersen, 2002).Acute exercise bouts challenge homeostasis by generating reactive oxygen species and causing an upregulation in proinflammatory cytokines, which must be quenched by anti-inflammatory immune defences.Repeated sessions of exercise training over time causes increased anti-inflammatory immunity by preferential activation antiinflammatory (Th2, Treg) immune cells and an enhanced immune response.Regular exercise also causes 'myokine' release, which are proteins released by contracting muscle, possessing powerful antiinflammatory action (Pedersen and Febbraio, 2008).These mechanisms are suggested to reduce basal inflammation in both peripheral and central systems, and adaptation has been reported to occur in as little as 2 -12 weeks of training (Flynn et al., 2007).This adaptation also functions to reduce the extent of the initial proinflammatory exacerbation to external stressors in future, forming an immunocompetent training response to physical activity (Gleeson et al., 2011).In healthy populations, exercise has been suggested as capable of improving psychological wellbeing, and these changes in inflammation, cytokines, and immune cell phenotype are implicated as a potential mechanism of effect (Mikkelsen et al., 2017).Evidence from the physical activity and exercise literature suggests that an exercise intensity of 55-65 % of an individuals heart rate maximum (HRMax, defined as 220 -(0.6*age) is required to have a favourable biological effect on health (Tanaka et al., 2001;Karvonen and Vuorimaa, 1988).This translates to moderate intensity exercise, and is in line with physical activity guidelines set out by the American College of Sports Medicine (Medicine ACoS, 1980;Riebe et al., 2018).Indeed, moderate intensity exercise training has been reported to induce changes in immune-inflammatory biomarkers in the periphery of healthy cohorts.Therefore, exercise may be a useful adjunctive therapy for individuals experiencing FEP to tackle the dominant inflammatory component related to negative symptom presentation.
Indeed, in populations experiencing FEP and SCZ, exercise has been shown to be a feasible intervention that can retain FEP populations, and lead to favourable psychological changes that include improvements negative symptoms (Firth et al., 2015).Firth et al. (Firth et al., 2018), conducted a 10-week exercise project that achieved 107 min of moderate-to-vigorous physical activity per week, and retained 81 % of participants.Interestingly, this group also reported a significant reduction in negative symptomology after 10-weeks of exercise.Fisher et al. (Fisher et al., 2020), conducted a supervised 12-week exercise intervention, targeting 60-80 % HRMax and designed to facilitate participant preference in exercise modality and environment.This intervention design led to 83 % attendance to two exercise session per week, and exercise had protective effects against the worsening of ongoing C. Dunleavy et al. pathology in the control group (Fisher et al., 2020).Further to this, favourable biological consequences of exercise training in both FEP and healthy cohorts have been reported, though sufficient training intensity appears important (Gleeson et al., 2011).Ventura et al. (Ventura et al., 2021), reported that in FEP, attendance to a 10-week intervention comprised of mixed moderate-to-vigorous aerobic and resistance training was significantly correlated with reductions in circulating interleukin(IL)-6.In SCZ, 2-months of mixed training performed three times weekly, has also been reported to cause significant reductions in IL-6 (Lavratti et al., 2017).Within the study by Fisher et al. (Fisher et al., 2020), there was also a trend of a reduction in IL-6, Tumor Necrosis Factor-ɑ, and C-reactive protein, though this failed to reach statistical significance.
The favourable consequences of exercise training may be a consequence of high intensity training that reduces inflammation and increases the production of biomarkers that are beneficial towards brain health (see Fig. 1), however this is as yet untested.No previous study has addressed whether regular exercise can reduce inflammation in FEP cohorts, and whether a potential reduction in inflammation is related to improvement of negative symptoms.Intriguingly, only a few studies have attempted to characterise Th-cell profiles in FEP, and the effects of exercise on phenotypes is unknown.Understanding this potential association may provide new avenues for treatment for negative symptoms and may prevent crippling reductions in quality of life seen in FEP.
The primary aim of this study was to assess the effects of a 6-week regular exercise intervention on immune cell concentration and peripheral inflammatory biomarkers in FEP.Our secondary aims were to determine the effects of exercise on negative symptoms and investigate potential links between inflammatory biomarker profile alterations and severity of symptoms experienced.Further, given the well-established effects of exercise on brain health (Cotman et al., 2007), we aimed to investigate the effects of exercise on aspects of cognition.
We hypothesized that the exercise intervention would reduce indices of inflammation and increase the presence of anti-inflammatory cells and cytokines.Further, we expected to see a reduction in negative symptom severity after exercise, alongside improvements in functional outcomes.

Approvals
'Exercise4Psychosis' was a pilot randomised control trial, approved in the UK by the Northwest − Greater Manchester West Research Ethics Committee, and the HRA Health and Care Research Wales (HCRW) for the Birmingham Women and Children's' NHS Foundation Trust Early Intervention in Psychosis (EIP) Service on 22nd March 2022 (REC: 22/ NW/0056).This trial followed the CONSORT 2010 statement guidelines for the reporting of randomised controlled trials (Moher et al., 2012).

Trial design
Exercise4Psychosis was an unblinded, randomised controlled trial consisting of 6 weeks of personalised moderate-to-vigorous exercise training, targeted at 60-75 % of an individual's heart rate maximum.The period of data collection spanned the July 2022 -February 2023.F31.2, F32.3) were recruited from community Early Intervention Services in South/East Birmingham, UK.Patients were within their first 3 years of psychosis incidence as per their first presentation to a clinical setting, and taking prescribed antipsychotic medication.Eligibility criteria for inclusion in 'Exercise4Psychosis' were as follows: 1) Male sex, 2) aged 18 -38 years, 3) established diagnosis of FEP, 4) sufficiently recovered from acute episode and identified as stable by an NHS clinician (not hospitalised, low risk to self and others, absence of symptoms too severe to allow engagement in academic research), taking prescribed antipsychotic medication, 5) capacity to provide informed consent, 6) physically healthy (as identified by an NHS clinician, and general health questionnaire), in absence of disability that would preclude ability to exercise.Exclusion criteria were as follows: 1) inability to provide informed consent, 2) failure to agree to testing requirements, i.e., refusal to provide a blood sample, 3) significant risk to self or others as identified by clinical team, 4) immunological disorder or abnormality likely to confer inflammatory perturbations, i.e., autoimmune or immunodeficiency disorders.

Participant population
Recruitment was conducted by the primary researcher (CD), who was integrated within Birmingham NHS EIP Services' and liaised with clinical care teams to refer any potentially suitable service users to participate in the 'Exercise4Psychosis' program.
All referrals were screened against prespecified inclusion criteria, after which they were invited to an initial meeting at the University of Birmingham to provide informed consent, complete baseline measures, and provide a blood sample.Patients were than randomly allocated to either: 1) exercise, or 2) treatment as usual (TAU) groups.Only male participants were recruited as inflammatory biomarker and immune cell concentrations were identified as the primary outcome measure.Variations in inflammatory cytokine and associated hormone concentrations across the menstrual cycle would be likely to conceal potential exerciseinduced effects in a study of this size if both male and female participants were included (O'Brien et al., 2007;Angstwurm et al., 1997;Konecna et al., 2000).

Sample size
This project aimed to recruit 30 patients with FEP to the exercise intervention.Exercise4Psychosis was a proof-of-concept experimental study designed to assess the effects of a short-term exercise intervention on inflammatory biomarkers and symptom profiles in FEP.This sample size was selected in accordance with an audit of United Kingdom Clinical Research Network (UKCRN) pilot and feasibility studies that indicated pilot sample sizes of 30-36 as feasible (Billingham et al., 2013).This study could inform statistical power and sample size calculations for clinical efficacy of future, larger scale studies assessing the effects of exercise as adjunct treatment for psychosis.

Randomisation
After providing informed consent and subsequent eligibility screening, participants were randomised to either: 1) exercise, or 2) TAU groups by an independent researcher using an online block randomisation tool (https://www.randomizer.org).After randomisation, participants were given a Garmin VivoSmart 3 or 4 TM heart rate monitor (Garmin, USA) and instructed to wear the monitor for 1-week, to provide information on each participant's baseline activity.Group allocation was concealed from participants and the research team until after completion of the baseline measures, habitual baseline activity, and return of the activity tracking watch.

Exercise intervention
Individual exercise sessions were developed and delivered by a trained Sport and Exercise Scientist from the School of Sport, Exercise, and Rehabilitation Sciences, at the University of Birmingham.Each session was 60 min in duration and supervised in entirety, comprised of a warm-up, exercise activity, and cool down.The researcher also took part in each exercise session to provide motivation to attend and engage.
The exercise intervention was 6-weeks in duration.This was selected to maximise the potential opportunity for physiological change, whilst minimising time commitments and disruption to participants' daily routines.During the exercise intervention, participants were invited to the University of Birmingham to exercise 2 -3 times per week for a total of 180 min, above the American College of Sports Medicine (ACSM) recommendation of 150 min of moderate-to-vigorous physical activity per week as a means of improving physiological fitness (Garber et al., 2011).Exercise sessions were targeted at 60 -75 % of an individual participants'HRMax (0.6 or 0.75*(220 -(0.6*age) as this exceeds the minimum intensity of exercise required in order to provoke improvements in physiological health (Karvonen and Vuorimaa, 1988).The research team developed a pre-specified list of exercise activities that were sufficient to increase % HRMax to the desired intensity based on previous evidence in FEP populations (Fisher et al., 2020).This was employed to ensure high activity levels during exercise sessions, whilst also allowing for participant choice and preference in modality.Garmin VivoSmart 3 or 4 TM heart rate monitors were used to track real-time energy expenditure, heart rate during session, exercise intensity minutes, and steps taken.Fig. 3 provides some examples of exercise activities that were presented to participants for selection during exercise sessions.
The control group were asked to maintain their regular habitual activity for the duration of the 6 week intervention, which was tracked via Garmin VivoSmart 3 or 4 TM heart rate monitors.

Primary outcomesinflammatory biomarkers
The primary outcome for this exercise intervention was immune cell phenotype and inflammatory cytokine concentration after 6-weeks of exercise.Blood samples (30 mL) were obtained at 0-weeks (pre-intervention) and 6-weeks (post-intervention) via venepuncture of the cubital/cephalic/basilic vein at the antecubital fossa.Whole blood samples were collected in BD Vaccutainer® MAP K2 EDTA/ Clot Activator (CA) Tubes (BD Biosciences, USA).The samples were then centrifuged at 500xG for 45 min (room temperature, 21 • C), plasma or PBMC separations aliquoted, and stored at − 80 • C before subsequent analysis.

T-helper cell quantification
Concentration of immune cell subsets withinin peripheral blood mononuclear cell (PBMC) populations were determined via flow cytometry method, using commercially available kits as per the manufacturer's instructions.For all experiments, a CytoFLEX BB2008 (Beckman Coulter Life Sciences, USA) was used, and analysis conducted on CytExpert (Beckman Coulter Life Sciences, USA).PBMCs were first thawed and assessed for viability using a Nexcelom Cellometer Auto 2000 (Nexcelom, USA), before seeding into flasks for exposure to stimulation cocktails, and processing for analysis.
For quantification of Th1, Th2, Th17, and Treg, commercially available T-helper cell phenotyping kits were used (BD Pharmingen TM Human Th1/Th2/Th17 Phenotyping Kit, BD Biosciences, AB_2869360, USA; BD Pharmingen TM Human Th17/Treg phenotyping Kit, BD Biosciences, AB_2869366, USA).Briefly, PBMC suspensions were stimulated with 10 ng/mL IL-1β for 4 h (R&D Systems, USA) in the presence of protein transport inhibitors, GolgiPlug/GolgiStop TM (BD Biosciences, USA) to cause intracellular cytokine accumulation, which were then antibody-tagged for identification of specific T-cell subset.UltraComp eBeads™ Compensation Beads (ThermoFisher, USA) were used to account for fluorescence overlap that may be present across fluorochromes C. Dunleavy et al. used to detect the different T-helper cell phenotypes.The full flow cytometry preparation and gating strategy can be found in the Supplementary materials.

Inflammatory cytokine quantification
Plasma and serum concentration of IL-6, IL-8, and IL-10 were quantified via commercially available Enzyme-Linked ImmunoSorbent Assay (ELISA) kits (Human IL-6 Quantikine HS ELISA Kit (HS600C) Human IL-8/CXCL8 Quantikine HS ELISA Kit (HS800), and Human IL-10 Quantikine HS ELISA Kit (HS100C) from R&D Systems (USA).Concentration of each cytokine was quantified as per the manufacturer's instructions.Peripheral cytokine measurements were quantified in triplicate, and were conducted at baseline (0 weeks), and endpoint (6weeks) of the exercise intervention for participants in the exercise and TAU groups.

Positive and negative Syndrome Scale
Symptomology experienced in FEP was measured using the Positive and Negative Syndrome Scale (PANSS) questionnaire (Kay et al., 1987).This is a semi-structured clinical interview that assesses a spectrum of positive, negative, and cognitive symptomology, assessed, and was scored by the primary researcher.

Negative symptoms
The Temporal Experience of Pleasure Scale (TEPS) is an 18-item questionnaire implemented to assesses two aspects of self-reported motivation (or lack thereof), including anticipatory pleasure, and consummatory pleasure (Gard et al., 2006).

Functional wellbeing
The World Health Organisation Quality of Life (WHOQOL), and Assessment of Disability Scale (WHODAS) questionnaires were implemented to assess aspects of participants' views on their quality of life, and self-assessment of their subjective disability rating.For the WHO-QOL a higher score indicates a better quality of life, however for the WHODAS, a higher score indicates increased subjective severity of disability.
All symptom severity and wellbeing assessments were completed at 0 weeks (pre-intervention), and 6-weeks (post-intervention) for both exercise and control groups.

Cognition
The Forward and Reverse Digit Span tests were administered to assess attention and working memory.The 'forward' test comprised sequences of numbers that were read out by the primary researcher which were increasing in length and had to be verbally recited by the participant.During the 'backward' test, participants were required to recite the sequences in reverse order.

Physiological measures
The International Physical Activity Questionnaire (IPAQ) was a selfreport questionnaire used to assess the extent of physical activity in four aspects of daily life and estimates total physical activity in Metabolic Equivalent of Task (MET) minutes.Garmin VivoSmart 3/4 TM heart rate monitors (Garmin, USA) were also used to obtain an objective measure of physical activity levels.Height (cm) and weight (kg) were measured and used to calculate Body Mass Index (BMI, = kg/m 2 ), and indices of waist circumference were also measured (cm).

Statistical analyses
Statistical analyses were conducted using IBM SPSS (version 29.0.0,2023) and GraphPad Prism 9 software (version 9.5.1, 2023).At baseline, unpaired t-tests were used to determine any differences in outcome markers between groups.To assess the effects of exercise on biomarkers, psychopathy, and wellbeing, paired t-tests and repeated measures ANOVAs were conducted.Regression analyses were conducted to determine the relationship between exercise, biomarker concentrations, and symptom severity measures.All biomarker analyses were completed in triplicate, and the average value was taken.Normality of data was assessed using the Shapiro-Wilk Normality test, and alpha significance level was determined at 0.05.Any outliers were identified and discarded via Robust Regression and Outlier Removal (ROUT) method (Q = 1 %).Significance values were set at 'p = 0.05′.

Baseline characteristics, demographic data, and habitual activity
Out of a possible caseload of 300 from the EI services in Birmingham, 102 individuals met the inclusion criteria.Seventeen individuals experiencing FEP accepted the study invitation, were recruited, and randomised to either the exercise group (n = 10), or non-exercising TAU group (n = 7).The full recruitment process, inclusions, and exclusions are displayed in Fig. 2.There was no significant differences in demographic details or habitual activity (p = >0.05),except for total step count.The exercise group had a significantly higher daily step count at baseline, before group allocation was revealed (t = 2.99, p = 0.009, see Table 1).
At baseline, there was no significant difference in positive, negative, or general psychopathology symptoms experienced between the exercise and TAU groups, measured via the PANSS (p > 0.05).Further, there was no significant difference in TEPS anticipatory or consummatory pleasure, functional outcomes (WHOQOL, WHODAS), cognition (digit span), or self-reported physical activity levels between the exercise and TAU groups at baseline (all p > 0.05, see Table 2).
Both the exercise group (+396.75min, t = 3.09, p = 0.001) and TAU group (+262.95min, t = 2.1, p = 0.03) significantly overestimated their habitual physical activity at baseline via IPAQ, when compared to values obtained by the Garmin VivoSmart activity tracker.There was no significant difference between exercise and TAU groups for self-reported METminutes, or actual intensity minutes at baseline (p > 0.05).

Attendance, exercise modality, and intensity
The exercise intervention group achieved a retention rate of 90 % (n = 9), and overall, the study achieved a retention rate of 88.24 % (n = 15).Within the exercise group, the overall attendance to the exercise sessions was 81.11 %.The dropout rate was identical in the exercise and TAU group (n = 1).The average number of sessions completed out of a possible 18, was 14.6 ± 3.06 sessions.The average attendance to two sessions per week was 88.33 %, and this dropped to 51.67 % for attendance to a third session per week.
Within the exercise group, the most popular exercise modality was  3. The average exercise intensity minutes achieved per week was 152.29 ± 43.47 min, which was significantly higher than baseline habitual activity minutes (47.95 min, t = 9.99, p < 0.001), and was in line with recommendations set out by ACSM (Medicine ACoS, 1980).

Exercise-Induced changes in T-cell phenotype
There was no significant difference in Th1, Th2, Th17, or Treg phenotypes between the exercise or TAU group at baseline (all p > 0.05).

Th1:Th2 ratio
The Th1:Th2 ratio was calculated to provide insight on the inflammatory state of each participant, with a larger value indicating higher Th1 concentration.At baseline, there was no significant difference in Th1:Th2 ratio between the exercise and TAU group (p > 0.05).After the exercise intervention, there was a significant reduction in the Th1:Th2 ratio within the exercise group (− 2.39 %), compared to an increase in the TAU group (+1.32 %), in a group-by-time interaction (F(1,13) = 9.19, p = 0.01).

Inflammatory cytokines
At baseline, the exercise group had significantly higher IL-6 concentration (2.5 pg/ml) when compared to the TAU group (1.75 pg/ml, p = 0.045).The exercise group also had significantly lower IL-8 concentration (8.39 pg/ml) than the TAU group (12.15 pg/ml, t = 3.3, p = 0.046).There was no significant difference in IL-10 concentration between the exercise and TAU groups at baseline (p > 0.05).
There was a significant reduction in IL-6 concentration in the exercise group after the exercise intervention (-22.17%), in a group-by-time interaction (F(1,13) = 10.69,p < 0.01).IL-8 and IL-10 were not changed significantly after 6-weeks of exercise (all p > 0.05).When exploring the changes in cytokines in the exercise and TAU groups, there was a significant reduction in IL-6 post exercise training (-22.17%), compared to a respective increase in IL-6 in the TAU group (11.41 %).Similarly, there was an increase in IL-8 after exercise (21.43 %) compared to a respective decrease in IL-8 within the TAU group (− 7.38 %) which suggests exercise affects these cytokines differently to TAU.IL-10 increased in both exercise (21.43 %) and TAU (23.46 %) groups across the 6-week intervention.Fig. 5 displays all changes in biomarker concentrations pre-topost exercise.

Psychopathology measures
After the exercise intervention, there was a significant reduction in negative symptom score in the exercise group, in a group-by-time interaction (F(1, 13) = 5.92, p = 0.03).6-weeks of regular exercise caused a reduction in the PANSS Negative subscale (− 13.53 %), compared to a respective increase in the TAU group (8.23 %).There was no significant change in positive symptoms, or general psychopathology after the exercise intervention (p > 0.05).Fig. 6 displays all changes in symptom severity as per the PANSS interview.There was no significant change in the self-report TEPS total score, or the anticipatory and consummatory subscales (all p > 0.05).

Functional outcomes
Therw was no significant effect of exercise on functional outcomes (all p > 0.05).In the exercise group, there was an improvement in each WHODAS domain (mobility, self-care, getting along with others, life activities, participating in society), except for domain 1 (communication) post intervention, however these changes failed to reach statistical significance.In the TAU group, domain 2 (mobility) improved after 6weeks, and there was an worsening of scores in the remaining 4 domains.
In the exercise group, all 4 WHOQOL domains (physical health, psychological health, social relationships, environmental health) and total score improved, and in the TAU group, 3 out of the 4 quality of life domains increased and one reduced.However, these changes did not reach statistical significance (all p > 0.05).
There was no significant change in IPAQ self-reported physical activity after the 6-week exercise intervention (p > 0.05).The exercise group still significantly overestimated their physical activity level immediately post exercise intervention, when compared to actual minutes achieved per week during the intervention (152.29 intensity minutes, vs 708.2 self-report MET minutes, (t(8) = 2.56, p = 0.03), displayed in Fig. 7.

Cognition
No significant change in cognition scores was recorded via the Digit Span test, after 6-weeks of exercise (p > 0.05).

Relationship between inflammation symptom severity
No significant relationship was reported for inflammatory biomarkers and negative symptom severity at baseline, endpoint, or for rate of change (all p > 0.05).

Biomarkers of physiological health
After the exercise intervention there was no significant change in weight (KG), waist circumference, and BMI (all > 0.05).In the exercise group there was a reduction in weight (− 1.52 %), whereas in the TAU group there was an increase in weight (+0.38 %) after 6-weeks.

Overview
This randomised controlled study demonstrated that 6-weeks of regular moderate-to-vigorous physical activity can lead to significant changes in immune biomarkers and symptom profiles that are central to FEP pathophysiology.In this intervention, standardised exercise bouts targeted at 60-75 % of an individuals' HRMax led to a significant shift in immune cell phenotypes towards anti-inflammatory action and Mean and standard deviation (SD).BPM, Beats per minute; CM, centimetres; kcals, calories; y/n, yes/no.alteration of the proinflammatory state in a FEP cohort.Further, this study found significant reductions in basal circulating IL-6 concentration after regular exercise in FEP.These changes occurred simultaneously alongside a significant reduction in negative symptoms.The results from this study suggest that exercise could be a useful physiological tool for manipulating biomarkers related to FEP pathology, and these changes may be important for alterations in symptom profiles.

Physical activity as an anti-inflammatory treatment
In the current study there was a trend of an inflammatory immunophenotype in FEP characterised by elevated circulating Th1 and Th17 cells, and low Th2 and Treg cells.When calculated as a ratio, the immunophenotype of the cohort was skewed towards Th1 activation, which is indicative of a heightened inflammatory state (Kim et al., 2004).Similar research conducted elsewhere suggests that circulating immune cell phenotypes are altered in both FEP and SCZ, and the Th1: Th2 ratio is imbalanced (Kim et al., 2004;Drexhage et al., 2011;Miller et al., 2013).Our study suggested that exercise can alter the Th-cell composition, and we reported a post-exercise reduction in Th1 and Th17 cells, alongside an increase in Treg cells, and an increase in Th2 cells that bordered significance.Importantly, regular exercise caused a significant reduction in the Th1:Th2 ratio, indicative of a preferential anti-inflammatory shift within the exercise group.This change was not present in the TAU group.As far as we are aware, this is the first study to assess the effects of exercise on immune cell phenotypes in FEP.Generally, regular physical activity causes redistribution of immune cells, and has been reported in healthy populations to cause a shift in Th-    phenotype characterised by elevated Th2 and Treg presence, which is generally interpreted as anti-inflammatory (Shaw et al., 2018).These results are in line with this and suggest that regular exercise in populations experiencing FEP has the capacity to change imbalanced immunophenotypes and may be useful for diseases underpinned by lowgrade inflammation.This project also reported significant changes in circulating cytokine concentration.At baseline the average IL-6 concentration within this entire FEP cohort was 2.20 pg/ml.In larger existing FEP datasets, a mean IL-6 concentration of 0.7 pg/ml has been reported, and therefore may used as a threshold for evidence of inflammation (Foley et al., 2023).This confirms that the both the exercise (2.50 pg/ml) and TAU group (1.75 pg/ml) displayed evidence of an inflammatory signature in peripheral blood at baseline.These findings are in line with the Th1skewed ratio of immune cells across both exercise and TAU groups.6weeks of exercise caused a significant reduction of IL-6 in the exercise group.These results are in line with exercise interventions conducted elsewhere in FEP and SCZ populations (Ventura et al., 2021;Fisher et al., 2020).However, the current intervention produced significant IL-6 reductions in FEP after just 6 weeks, whereas most other studies have engaged participants in longer periods of exercise.Some groups do report conflicting results regarding IL-6 (Bigseth et al., 2023), and methodological variation and disease chronicity may be important for this.Significant changes in IL-6 after 6 weeks may be a consequence of a higher target exercise intensity, frequency, and supervisory aspects of the current intervention that facilitated engagement and promoted biomarker change.
Although IL-6 is subject to fluctuation based on various external factors, both exercise and TAU groups display evidence of inflammation that is substantially higher than the average IL-6 concentrations reported elsewhere in FEP, which is indicative of an inflammatory signature of this cohort (Foley et al., 2023).Therefore, exercise induced changes in IL-6 may well be protective against the harmful effect associated with continued psychosis that were present in the TAU group, including an elevating IL-6 concentration.Exercise-induced 'myokine' IL-6 production is exponential, increases as much as ~ 100 fold immediately after physical activity, and has anti-inflammatory functions (Petersen and Pedersen, 2005).Therefore, repeated bouts of exercise release are suggested to reduce the peak IL-6 production upon stimulation, as well basal level of IL-6 present in peripheral systems (Pedersen  and Febbraio, 2008), forming an adaptation to physical activity that could be responsible for the reduction of IL-6 seen in the present study.Given the suggestion of IL-6 as a causal factor for brain abnormalities in mental illnesses including FEP (Williams et al., 2022), a reduction in circulating IL-6 in the exercise group may be beneficial for disease status and severity of illness.
In contrast to IL-6, there was no significant change in IL-8 or IL-10 after the exercise intervention.There are no other studies in FEP or SCZ that we are aware of that measure these cytokines after a period of regular exercise.In healthy populations, IL-8 has been reported to remain relatively unaltered by exercise training, and is suggested to exert local angiogenic effects after release from muscle (Docherty et al., 2022).IL-10 is suggested to increase post exercise and represents an anti-inflammatory response to training, however a combination of training modalities (aerobic and resistance training), and working at higher intensities appear to be important factors for causing significant changes in IL-10 (Docherty et al., 2022;Islam et al., 2021).This may explain lack of changes in the current study, as the average heart rate may not have reached a high enough intensity to alter IL-10 concentration and may also be influenced by participant choice of exercise mode.Further studies that assess exercise related changes in these cytokines in healthy populations and FEP are needed.

Exercise and symptom severity
Significant reductions in negative symptoms were recorded following exercise training.Prescribing exercise as therapy to reduce negative symptoms has been effective in previous studies (Fisher et al., 2020;Sabe et al., 2020;Kim et al., 2023), however most of these studies are feasibility designs, and are conducted in established SCZ.Within FEP, Fisher et al. (Fisher et al., 2020), saw a protective effect of exercise against increasing negative symptoms after 12-weeks of exercise, with participants averaging 51.29 min per exercising day.Firth et al. (Firth et al., 2018), reported significant reductions in negative symptoms after a 10-week intervention with an average of 107 min of exercise per week.The results presented herein represent the highest exercise 'dose' when compared to the aforementioned studies, with participants undertaking an average of 152.29 min per week across 6 weeks.Following this, a significant reduction in negative symptoms was reported.Interestingly, the exercise group had higher negative symptoms at baseline alongside elevated IL-6 and a higher Th1:Th2 ratio, all of which significantly decreased after exercise training.This may further implicate inflammation as a potential mechanism for negative symptom presentation and severity specifically (Goldsmith and Rapaport, 2020).Elsewhere, training related decreases in inflammatory markers including IL-6 have been significantly correlated with a reduction in depressive symptoms, and the prescribed dose of exercise was also noted as important in causing these changes (Ventura et al., 2021;Miller and Raison, 2016).
This trial we reported no significant relationship between inflammatory biomarkers and negative symptoms at baseline, postintervention, or as of rate of change.In context of this pilot study, it is important to note the modest sample size, meaning it is likely underpowered for the detection of genuine relationships between inflammation and negative symptoms.Further to this, the literature suggests that inflammation may impact specific constructs of negative symptoms, for example motivational anhedonia (Upthegrove and Khandaker, 2020).Therefore, generalising negative symptom severity to a total symptom score may conceal the relationship between these two variables.Similarly, given the pleiotropic role of each immune cell type and cytokine (Petersen and Pedersen, 2005), generalisation of function towards 'proand-anti-inflammatory' categories may not be sensitive enough to determine the cause-and-effect relationship between specific inflammatory perturbations and aspects of negative symptoms.Future research should seek to define an explicit proinflammatory signature in FEP spanning from genomic profiling to immune cell phenotypes and associated cytokines, which may be useful to interrogate the relationship with specific constructs of negative symptoms.

Strengths and limitations
This study emphasizes that exercise is a feasible treatment for individuals experiencing FEP.This trial implemented an intervention structure that was centred on participant preferences, allowing for flexibility, and facilitating social interaction.These factors are indispensable for exercise interventions in FEP populations (Firth et al., 2016), may have driven the high attendance and adherence to exercise sessions, and are a major strength of the current project.Further, this study adds to the literature surrounding immune function in FEP, and the capacity of exercise to manipulate symptoms and biomarkers of disease pathology.However, this study does also have some limitations, for example, sample size.'Exercise4Psychosis' commenced in July 2022 and concluded in February 2023, and targeted individuals within three years of psychotic symptom onset.This means that most of the participants involved in the study experienced their first psychotic episode during the COVID-19 pandemic.The nature of the COVID-19 pandemic and post-pandemic procedures related to lack of face-to-face clinical contact and removal of social inclusion activities may have affected the willingness of a clinical cohort to engage in experimental research.Therefore, a relatively small sample size may limit generalisability of findings to wider populations experiencing FEP.Additionally, the nature of implementing an exercise intervention includes regular social interaction over a prolonged period.Due to the scope of the study, we were unable to provide the TAU group with similar social interaction across the 6 weeks, and cannot determine the potential effect of social interaction on measures of symptoms and wellbeing.Relatedly, it is impossible to enforce blinding in an exercise trial, and participants were aware of their group allocation which leads to the possibility of expectation bias and behaviour alteration.Finally, all training sessions and outcomes were determined by the primary researcher (C.D.).Participant answers may be subject to change on the basis of familiarity with the researcher since baseline measurements, which could both extend or reduce the extent of symptom severity described based on degree of trust, or intervention expectation bias.Future research should employ external researchers to conduct all measurements.

Future directions and conclusion
Overall, this project describes a potential inflammatory signature that may be present in FEP.The study highlights the feasibility of personalised exercise training as a physiological therapy to manipulate peripheral immune cell profiles, cytokine concentrations, and symptom profiles experienced in FEP.This study shows a trial design of 2-3 exercise sessions over a 6-week period, and implementation of exercise modalities that reach over 60 % HRMax are useful for manipulating biomarkers related to disease.Social group activities, for example group football, appear important in driving both attendance and intensity of exercise sessions, which may be related to changes in biomarkers.
It is important to note that Exercise4Psychosis represents a small pilot study, and only males were included in attempt to assess the effect of exercise on cytokine and symptom profiles in FEP.This is a developing field of research and studies investigating female psychobiological profiles is still lacking.Larger scale studies should strive to include both biological sexes to provide information on pathology and potential treatment options for females as well as males.Future research should seek to further define the inflammatory biotype of FEP through identification of cellular, protein, and genetically defined inflammatory signatures.This could aid in the stratification of patients into 'inflamed' subgroups for precision therapy during early intervention, whereby personalised exercise programs may be useful.Importantly, FEP is not unique in its' inflammatory pathophysiology, which is shared across multiple clinical cohorts experiencing mental illness, and personalised exercise training could represent a useful transdiagnostic therapy for alleviating symptom profiles in clinical cohorts characterised by inflammation.
circuit training, followed by group football, high intensity interval training (HIIT), indoor/outdoor cycling, treadmill/outdoor running, and badminton.The breakdown of exercise modality is shown in Fig. 3.The exercise intervention reached the target exercise intensity for all exercise modalities (60-75 % HRMax) with an average of 67.65 % ± 7.75 % HRMax.Group football sessions achieved the highest average exercise intensity (76.27 % HRMax, 145.18 bpm ± 20.58 bpm), and exceeded the target exercise intensity.The HRMax achieved for each exercise modality is displayed in Fig. 4. On average, group football sessions burned significantly more calories (560.46kcals)than circuit training (− 186.54kcals, p = 0.038), HIIT (− 210.18kcals, p = 0.02), and cycling (− 242.20kcals, p = 0.013), apart from running (− 133.28kcals, p = 0.067).The full breakdown of exercise quality is presented in Table

Fig. 2 .
Fig. 2. Flow chart displaying the inclusion process for the exercise project, from the initial caseload to end-point assessment.

Fig. 4 .
Fig. 4. Average heart rate (beats per minute) per exercise activity.Red bar indicates average HRMax for the exercise group, and the green zone represents the 60-75% HRMax based on the max HR of the exercise group (the target exercise intensity for intervention).

Fig. 7 .
Fig. 7. Self-reported physical activity at baseline, and immediately post-intervention, compared to actual intensity minutes achieved per week across the intervention.

Table 1
Demographic details and baseline characteristics of participants randomised to exercise and control groups.

Table 2
Baseline symptom severity, functional wellbeing, and cognition scores.

Table 3
Exercise modality and intensity per exercise activity.

Table 4
Peripheral blood mononuclear cell immunophenotypes in the exercise and control group, before and after the exercise intervention.