In schizophrenia, chronic fatigue syndrome- and fibromyalgia-like symptoms are driven by breakdown of the paracellular pathway with increased zonulin and immune activation-associated neurotoxicity.

Background: A meaningful part of schizophrenia patients suffer from physiosomatic symptoms (formerly named psychosomatic) which are reminiscent of chronic fatigue syndrome and fibromyalgia (FF) and are associated with signs of immune activation and increased levels of tryptophan catabolites (TRYCATs). Aims: To examine whether FF symptoms in schizophrenia are associated with breakdown of the paracellular pathway, zonulin, lowered natural IgM responses to oxidative specific epitopes (OSEs); and whether FF symptoms belong to the behavioral-cognitive-physical-psychosocial-(BCPS)-worsening index consisting of indices of a general cognitive decline (G-CoDe), symptomatome of schizophrenia, and quality of life (QoL)-phenomenome. Methods: FF symptoms were assessed using the Fibromyalgia and Chronic Fatigue Rating scale in 80 schizophrenia patients and 40 healthy controls and serum cytokines/chemokines, IgA levels to TRYCATs, IgM to OSEs, zonulin and transcellular/paracellular (TRANS/PARA) molecules were assayed using ELISA methods. Results: A large part (42.3%) of the variance in the total FF score was explained by the regression on the PARA/TRANS ratio, pro-inflammatory cytokines, IgM to zonulin, IgA to TRYCATs (all positively) and IgM to OSEs (inversely). There were highly significant correlations between the total FF score and G-CoDe, symtopmatome, QoL phenomenome and BCPS-worsening score. FF symptoms belong to a common core shared by G-CoDe, symtopmatome, and QoL phenomenome. Discussion: The physio-somatic symptoms of schizophrenia are driven by various pathways including increased zonulin, breakdown of the paracellular tight-junctions pathway, immune activation with induction of the TRYCAT pathway, and consequent neurotoxicity. It is concluded that FF symptoms are part of the phenome of schizophrenia and BCPS-worsening as well.


Introduction
Recently, we developed a new method, namely nomothetic network psychiatry, to model major psychiatric disorders including schizophrenia and major depressive and bipolar disorder based on causome, AOP (adverse outcome pathways), and phenome data (Maes et al., 2020b;2021a;Stoyanov and Maes, 2021). Applied to schizophrenia, we constructed a new reliable and cross-validated, data-driven model, which assembles the effects of risk (causome) and resilience (protectome) features, different AOPs (neuro-immune pathways), cognitome (i.e., the aggregate of neuro-cognitive impairments), symptomatome (i.e., the aggregate of all symptoms of schizophrenia), and phenomenome (i.e., the self-experience of illness features by the subjects) features (Maes et al., 2020b).
We conceptualized the symptomatome of schizophrenia as the first factor extracted from newly developed scores of psychosis, hostility, excitation, mannerism (PHEM), negative symptoms, psychomotor retardation (PMR), and formal thought disorders (FTD) (Almulla et al., 2020b;Al-Hakeim et al., 2020). We conceptualized the cognitome as the first factor extracted from two neuropsychological batteries, namely the Cambridge Neuropsychological Test Automated Battery and the Consortium to Establish a Registry for Alzheimer's disease and including test scores reflecting disorders in executive functions, strategy use, attention set-shifting, visual sustained attention, rule acquisition, emotional recognition, semantic and episodic memory, and recall (Maes and Kanchanatawan et al., 2021). These factor scores reflect the severity of the General Cognitive Decline (G-CoDe) in schizophrenia (Maes and Kanchanatawan et al., 2021).
The phenomenome of schizophrenia was conceptualized as the first factor extracted from selfrated scores of physical health, psychological health, and social and environmental responsible behaviors as scored with the World Health Organization Quality of Life Instrument-Abbreviated . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
Recently, we reported that a meaningful part of schizophrenia patients suffer from physiosomatic symptoms (formerly named psychosomatic symptoms) which are reminiscent of chronic fatigue syndrome and fibromyalgia (Kanchanatawan et al 2017;Almulla et al., 2020a;Mousa et al., 2021). We assessed these symptoms using the Fibromyalgia and Chronic Fatigue Rating (FF) Scale, including fatigue, muscle tension, cramps and pain, a flu-like malaise, headache, and autonomic and cognitive symptoms (Kanchanatawan et al 2017;Almulla et al., 2020a;Mousa et al., 2021). Moreover, the FF symptoms were strongly correlated with PHEM and negative symptoms, PMR, and FTD (Almulla et al., 2020a). The total FF score loaded highly on the same latent vector, which was extracted from these typical schizophrenia symptoms, suggesting that FF symptoms are another manifestation of the symptomatome of that illness (Almulla et al., 2020a). Furthermore, FF symptoms were strongly associated with impairments in semantic and episodic memory, working memory, executive functions, spatial planning, paired associative learning, and attention (Kanchanatwan et al., 2017;Almulla et al., 2020a). In another study, we found that health-related (HR)-Qol in schizophrenia was best predicted by severity and anxiety and FF symptoms and that these symptoms were more important than psychosis and negative symptoms in predicting a lowered HR-QoL (Kanchanatawan et al., 2019). Nevertheless, there are no data whether FF symptoms also contribute to the BCPS-worsening index.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 11, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 There is now also evidence that the symptomatome, cognitome, Qol-phenomenome, and the BCPS-index are strongly and positively associated with different AOPs, namely with a) biomarkers of immune activation including increased levels of pro-inflammatory cytokines/chemokines such as interleukin-6 (IL-6), MIF-1, tumor necrosis factor (TNF)-α, IL-1β, and CCL11 (or eotaxin); b) activation of the tryptophan catabolite (TRYCAT) pathway as assessed with IgA responses to diverse TRYCATs; c) breakdown of the paracellular gut pathway with increased levels of zonulin; d) increased neurotoxicity as measured with the above cytotoxic cytokines/chemokines coupled with biomarkers of oxidative stress including lipid hydroperoxides (LOOH), malondialdehyde (MDA), and advanced oxidation protein products (AOPP); and e) increased levels of lipopolysaccharides of Gram-negative bacteria. Moreover, we found also negative association between the above clinical features of schizophrenia and natural IgM to oxidatively specific epitopes (OSEs), including MDA and azelaic acid, which generally have immune-regulatory properties (Maes et al., 2021b).
Nevertheless, there are no data whether FF symptoms in schizophrenia are associated with indicants of the paracellular gut pathway, increased levels of zonulin and neurotoxicity, and lowered levels of protective IgM to OSEs.
Hence, the present study was conducted to a) examine whether FF symptoms in schizophrenia are associated with breakdown of the paracellular pathway as assessed with IgM responses to paracellular / transcellular molecules, IgM responses to zonulin and OSEs, immune . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 11, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 activation and neurotoxicity indices, and IgA responses to TRYCATs; and b) examine whether FF symptoms belong to the BCPS-worsening index and, consequently, constitute another manifestation of the BCPS-worsening in schizophrenia.

Participants
This is a case-control study which recruited Thai participants, both men and women aged between 18 and 65 years old. We recruited schizophrenia outpatients at the Department of Psychiatry, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand and normal controls from the same catchment area (Bangkok, Thailand). The patients complied with the diagnostic criteria for schizophrenia according to DSM-IV-TR criteria. They were all in a clinically stable phase of illness for at least one year and were treated with the same antipsychotic maintenance treatment during that period. The healthy controls were healthy individuals, and they were recruited by word of mouth as friends or family of staff members and friends of the patients.
Schizophrenia patients were excluded when they showed a lifetime or current diagnosis of axis-I DSM-IV-TR or DSM-5 psychiatric disorders, including psycho-organic disorders, bipolar disorder, major depressive disorder, and substance use disorders (all except nicotine dependence).
We excluded controls with any lifetime/current DSM-IV-TR / DSM-5 axis 1 diagnoses of psychiatric illness and/or with a positive family history of psychosis. Controls and patients were excluded when they showed: a) systemic (auto)immune or inflammatory illness, such as rheumatoid arthritis, psoriasis, diabetes, chronic obstructive pulmonary disease, and inflammatory bowel disease; b) neuroinflammatory or neurodegenerative disorders including Alzheimer's and Parkinson's disorder, stroke and multiple sclerosis; c) lifetime (or current) treatment with . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted May 11, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 immunomodulatory drugs such as methotrexate and glucocorticoids; d) recent use (last 6 months) of therapeutical dosages of antioxidants or omega-3 supplements; and e) alcohol abuse. We also excluded pregnant or lactating women.
All controls and patients as well as their guardians (parents or other close family members of the patients) gave written informed consent prior to participation in this study. The study was conducted according to Thai and international ethics and privacy laws. Approval for the study was  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 11, 2021. ;https://doi.org/10.1101/2021.05.09.21256897 doi: medRxiv preprint (Kanchanatawan et al., 2017. We computed the sum of selected physiosomatic FF symptoms, namely FF1 + FF2 + FF3 + FF9 + FF10 + FF11 + FF12 thereby deleting the cognitive and affective FF symptoms from the FF physiosomatic score. To compute the severity of the symptomatome of schizophrenia, we assessed the Brief Psychiatric Rating Scale (BPRS) (Overall and Gorham, 1962) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 11, 2021. shifting total errors adjusted (IED_TEA) (CERAD, 1986;CANTAB, 2018). As explained previously, we extracted the first LV from these different test scores and this first PC reflects the G-CoDe (Maes and Kanchanatawan, 2021). Finally, we have extracted the first PC from the symptomatome, G-CoDe, and the QoL-phenomenome factor scores which, therefore, reflects the BCPC-worsening index (Maes et al., 2021b). Consequently, we have binned the BCPS-worsening index of the patients into two groups using the median-split method thereby forming two patient groups, namely patients with a low (SCZ-BCPS) and a more severe (SCZ+BCPS) worsening index.
DSM-IV-TR criteria were used to make the diagnosis of tobacco use disorder (TUD). Body mass index (BMI) was computed as body weight (in kg) divided by body length (in meter) squared.

Assays.
Fasting blood was sampled at 8.00 a.m. and serum and plasma Eppendorfs were frozen at -80 o C until thawed for assay. Blood samples of controls and patients were taken at the same time, using exactly the same procedures. Each set of assays was performed in the same run by the same operator who was blinded to the clinical results. Serum levels of IL-6, IL-10, TNF-α, MIP-1, CCL11, and the soluble IL-1 receptor antagonist (sIL-1RA) (R&D Systems, Inc, Minneapolis, MN, USA) were assayed employing the Bio-Plex ® 200 System (Bio-Rad Laboratories, Inc.) [22,24]. AOPP was assayed in a microplate reader (EnSpire, Perkin Elmer, USA) at a wavelength of 340 nm. LOOH was assayed by chemiluminescence in a Glomax Luminometer (TD 20/20), in the dark, at 30 °C for 60 min. MDA was assayed through complexation with two molecules of thiobarbituric acid and high-performance liquid chromatography (HPLC Alliance e2695, Waters', Barueri, SP, Brazil). Indirect ELISA tests were used to measure IgM directed against conjugated . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
Zonulin, occludin, E-cadherin, talin and vinculin were purchased from Bio-Synthesis® Pseudomonas putida. were conducted using ELISA techniques as described by Roomruanwong et al. (2018). Consequently, we computed a z unit-weighted composite score summing up the 5 z transformed values of IgA to the 5 Gram-negative bacteria (IgA LPS). This score reflects plasma LPS load following increased gut permeability or leaky gut.
Based on these assays we computed a number of z composite scores: a) z score of PA (zPA)+ z3HK + zXA -zAA -zKA, reflecting the noxious / generally more protective TRYCAT ratio (Kanchanatawan et al., 2019); b) sum of 4 cytokines which are increased during immune responses (4 cytokines index) as sum a z scores of zIL-6 + zTNF-α + zIL10 + zMIP-1; c) the neurotoxicity index was computed as zTNF-α + zIL-6 + zMIP + z(zXA+PA+3HK) + zLOOH + zMDA + zAOPP. CCL11 was not included in this index to allow for the additional effects of this . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Statistics
We employed analysis of variance (ANOVA) to check differences in scale variables between groups and analysis of contingency tables (Χ 2 -test) or Fisher's exact probability test to check associations between categorical variables. Associations between two scale variables were assessed using Pearson's product moment correlation coefficients. In order to correct for type 1 error due to multiple comparisons we employed a p correction for false discovery rate (FDR).
Multiple regression analysis was employed to assess the significant predictors of the FF scores, and we employed the R 2 value as an estimate of the effect size. We allowed for the effects of sex, age, education, and BMI when examining the effects of biomarkers on FF scores and significant background variables were included in the final model. All regression analyses were checked for multicollinearity. We used principal component analysis to extract PCs from a data set and latent vectors (LV) were considered to be reliable when the explained variance was > 50.0% and all variables loaded highly (>0.66) on the first PC. The adequacy for factorization was checked with the Kaiser-Meyer-Olkin (KMO) test and Bartlett's test of sphericity. Statistical analyses were performed using IBM SPSS Windows version 25. Tests were 2-tailed, and an alpha level of 0.05 indicated a statistically significant effect. Table 1 shows the sociodemographic data in schizophrenia patients divided into those with and without an increased BCPS-worsening index and normal controls. There were no significant . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 11, 2021. ; https://doi.org/10.1101/2021.05.09.21256897 doi: medRxiv preprint differences in age, BMI, nicotine dependence, between the three study groups. There were significantly more males in the schizophrenia group than in controls. The unemployment rate increased from controls → SCZ-BCPS → SCZ+BCPS. The total FF and FF physiosomatic, symptomatome, and BCPS scores were significantly different between the three study groups and increased from controls → SCZ-BCPS → SCZ+BCPS. The G-CoDe and QoL phenomenome decreased from controls to SCZ-BCPS to SCZ+BCPS. These differences in clinical indices remained significant after FDR p correction. CCL11 and IgA TRYCATs were significantly higher in schizophrenia than in controls.
The 4 cytokines, immune activation, neurotoxicity, and MITOTOX indices were significantly different between the three subgroups and increased from controls → SCZ-BCPS → SCZ+BCPS.
The IgM protectome was significantly lower in SCZ+BCPS than in controls, and IgM PARA/TRANS significantly higher in SCZ+BCPS than in the two other groups. The differences in biomarkers remained significant after p correction. The differences in FF scores, clinical indices, and biomarkers were not significantly affected by the drug state of the patients (even without FDR p correction), namely antidepressants n=28, risperidone n=34, olanzapine n= 5, quetiapine n=4, clozapine n=10, haloperidol n=12, perphenazine n=21, chlorpromazine n=5. There were no differences in the drug state (or chlorpromazine equivalents) between SCZ-BCPS and SCZ+BCPS patients. Table 2 shows the correlations (after FDR p correction) between both FF scores and clinical and biomarker data. Both the total and physiosomatic FF scores were significantly associated with the symptomatome, BCPS-worsening score, CCL11, IgA TRYCATs, and 4 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 11, 2021. ; https://doi.org/10.1101/2021.05.09.21256897 doi: medRxiv preprint cytokines, immune activation, neurotoxicity, and MITOTOX indices, the IgM TRANS/PARA ratio and zonulin.     Associations with QoL phenomenome. Table 4 shows the outcome of multiple regression analyses with the WHO-QoL PC as dependent variable and both FF scores, the symptomatome, G-Code and biomarkers as explanatory variables. Table 4, regression #1 shows that 50.3% of the variance in the WHO-QoL PC was . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 11, 2021. ; https://doi.org/10.1101/2021.05.09.21256897 doi: medRxiv preprint explained by the total FF score, symptomatome, and G-CoDe. Figure 2 shows the partial regression of the QoL phenomenome on the total FF score. The second regression shows that 50.1% in the variance in the WHO-QoL score was explained by FF physiosomatic score, symptomatome, and G-CoDe.
Given the strong associations between the FF scores and the other clinical scores (see Table   2), we have performed a PC analysis to determine whether the total FF score (and FF physiosomatic score) belong to the same PC as the symptomatome, G-CoDe, and WHO-Qol PC.
We found that the first PC extracted from these 4 variables explained 69.8% of the variance and that all variables loaded highly on the first PC, namely total FF score: 0.840; symptomatome: 0.859; G-CoDe: -0.801; and WHO-Qol PC (-0.840). The KMO value (0.815) and Bartlett's test of sphericity (X 2 =201.52, df=6, p<0.001) showed the factoriability of the data matrix was adequate.
Using the FF physiosomatic score instead of the total FF score yielded comparable results, namely a loading of 0.830 on the first PC extracted from the 4 variables, which explained 69.3% of the variance and showed a KMO of 0.815. These results show that both FF scores belong to the same PC as the BCPS-worsening score. Consequently, we have computed the PC scores extracted from the total FF score and the three other clinical scores which reflects the BCBS + FF worsening index. Table 4, regression #3 shows that 42.3% of the variance in this BCBS+FF worsening score was explained by 3 biomarkers, namely CCL11, IgM PARA/TRANS, and immune activation index.

Discussion
Clinical aspects . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 11, 2021. ; https://doi.org/10.1101/2021.05.09.21256897 doi: medRxiv preprint The first major finding of this study is that schizophrenia is characterized by increased total FF scores and the severity of the physiosomatic symptoms of the FF and that the both FF scores are strongly associated with the BCPS-worsening index and its components, namely the symptomatome, G-CoDe, and QoL-phenomenome. Also, in other study groups, we found that FF symptoms as assessed with the FF scale, were part of the symptomatome of schizophrenia and belonged to a same latent factor as the PHEM and negative symptoms of schizophrenia (Almulla et al., 2020a;Mousa et al., 2020). This indicates that FF symptoms and physiosomatic symptoms are key manifestations of the symptomatome of schizophrenia. Moreover, the current study found that the total FF score and the severity of physiosomatic symptoms better predicted lowered quality of life that the symptomatome comprising PHEM and negative symptoms, PMR and FTD, and the G-CoDe as well.
Our finings that the total FF and physiosomatic FF scores were highly significantly associated with the BCPS-worsening index indicate that those FF symptoms are another feature of BCPS-worsening and should routinely be screened in schizophrenic patients to estimate the worsening of the illness. The results also suggest that FF symptoms and physiosomatic symptoms are driven, at least in part, by the same pathways that underpin aberrations in the connectome and neuroplasticity and which cause BCPS-worsening, namely "the prefronto-striato-thalamic, prefronto-parietal, prefronto-temporal, and dorsolateral prefrontal cortex, hippocampus and amygdala (Orellana et al., 2013;Orellana and Slachevsky, 2013;Maes et al., 2021b;Maes and Kanchanatawan, 2021).

Biomarker correlates: focus on leaky gut
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 11, 2021. ; https://doi.org/10.1101/2021.05.09.21256897 doi: medRxiv preprint The second major finding of this study is that the total FF and physiosomatic FF scores are strongly associated with the IgM PARA/TRANS ratio and increased zonulin levels. Previously, Interestingly, in the present study, IgM to zonulin was significantly associated with FF symptoms and physiosomatic symptoms, whereas previously no significant associations between zonulin and PHEM, negative symptoms, and memory deficits were detected in schizophrenia (Maes et al., It is interesting to note that the severity of FF symptoms in ME/CFS is significantly associated with indicants of leaky gut including increased IgA and IgM responses to the same Gram-negative bacteria as determined in the present study (Maes et al., 2007). Moreover, treatment with specific antioxidants, which are known to improve leaky gut (zinc, glutamine, and N-acetylcysteine) improves these indicants of leaky gut in association with a significant improvement in the severity of FF symptoms (Maes and Leunis, 2008). Part of the schizophrenia patients, especially those with first episode schizophrenia, show leaky gut and increased bacterial translocation an, therefore, those patients may benefit from treatments for leaky gut (Maes et al., 2021b).

Biomarker correlates: focus on immune activation and neurotoxicity
The third major finding of the present study is that a considerable part of the variance in the total FF score and the severity of physiosomatic FF symptoms is explained not only by the PARA/TRANS ratio and IgM to zonulin, but also by immune activation as indicated by increased . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
These cytokines/chemokines are all increased during an immune-inflammatory response with IL-6 and TNF-α displaying inflammatory, CCL11 and MIP-1 chemotaxic, and IL-10 immuneregulatory effects (Maes et al., 2021b). These results extend those of previous reports which showed that a larger part of the variance in the FF score in schizophrenia is accompanied by increased levels of CCL11 and high mobility group box 1 protein (HMGB1) (Mousa et al., 2020). It is important to note that the severity of FF symptoms in ME/CFS is significantly associated with signs of immune activation including increased levels of pro-inflammatory . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
Interestingly, the combined impact of leaky gut, immune activation and natural IgM biomarkers was more important than that of neurotoxicity indices, although the MITOTOX index comprising the neurotoxic cytokines/chemokines and TRYCATs coupled with LPS and the oxidative stress damage was strongly associated with the physiosomatic symptoms. Previously, we have discussed that the symptomatome, G-CoDe, Qol-phenome, and the BCPS-worsening index are significantly associated with indices of increased neurotoxicity as indicated by increased levels of neurotoxic cytokines including IL-6, TNF-α, CCL11, MIP-1, increased levels of neurotoxic TRYCATs, including PA, XA, and 3HK, and signs of oxidative/chlorinative stress as indicated by increased LOOH, MDA, and AOPP levels (Maes et al., 2020a;2021b).

Mechanistic explanations.
This study showed that FF symptoms belong to the same common core shared by G-CoDe, and the symptomatome and, therefore, we may suggest that at least part of the variability in FF symptoms is determined by the same pathways which underpin cognitive deficits and the symptomatic phenome. The mechanistic explanation is that the symptomatome and cognitome of schizophrenia are probably caused by pathways which cause alterations in gray and white matter . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
We have described previously how these same peripheral pathways including leaky gut, immune activation, TRYCAT pathway activation, and nitro-oxidative stress may cause central changes

Limitations.
Our results should be interpreted with regard to its limitations. Firstly, this is a case-control study, and this design does not allow to draw solid conclusions on causality. Secondly, it would have been more interesting if we had measured M1, Th-1, Th-2, and T regulatory profiles as well as the connectome using functional MRI.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Conclusions
The combined effects of leaky gut, immune and TRYCAT pathway activation and lowered protection due to lowered natural IgM levels drive in part the physiosomatic symptoms in schizophrenia. Physiosomatic symptoms are part of the phenome of schizophrenia and the BCPSworsening index, and they strongly predict a lowered quality of life.

Disclosure of potential conflicts of interests
The authors report no conflict of interest with any commercial or other association in connection with the submitted article.
Research involving human participants.
Approval for the study was obtained from the Institutional Review Board of Chulalongkorn University, Bangkok, Thailand.
Informed consent.
All controls and patients and the guardians of patients gave written informed consent before participation in our study.
Data Access Statement.
The dataset generated during and/or analyzed during the current study will be available from the corresponding author upon reasonable request and once the dataset has been fully exploited by the authors.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 11, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 Funding This research has been supported by the Asahi Glass Foundation, Chulalongkorn University Centenary Academic Development Project.

Author's contributions
All the contributing authors have participated in the manuscript. BK and MM designed the study.
BK recruited patients and completed diagnostic interviews and rating scales measurements. MM carried out the statistical analyses. AV, SS and DB performed the analyses. All authors contributed to the interpretation of the data and writing of the manuscript.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 11, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 11, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 Vascular Barrier in Patients with Deficit Schizophrenia. Neurotox. Res. 36 (2) Targets   . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 11, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted May 11, 2021. ; https://doi.org/10.1101/2021.05.09.21256897 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)