Systematic Review and Meta-Analysis of Damage Associated Molecular Patterns HMGB1 and S100B in Schizophrenia

Michael Mackey; Laurena Holleran; Gary Donohoe; Declan P. McKernan

doi:10.30773/pi.2022.0173

Psychiatry Investig > Volume 19(12); 2022 > Article

Mackey, Holleran, Donohoe, and McKernan: Systematic Review and Meta-Analysis of Damage Associated Molecular Patterns HMGB1 and S100B in Schizophrenia

Original Article

Psychiatry Investigation 2022;19(12):981-990.

Published online: December 22, 2022

DOI: https://doi.org/10.30773/pi.2022.0173

Systematic Review and Meta-Analysis of Damage Associated Molecular Patterns HMGB1 and S100B in Schizophrenia

Michael Mackey^1,², Laurena Holleran²

, Gary Donohoe²

, Declan P. McKernan¹

¹Pharmacology & Therapeutics, School of Medicine, NUI Galway, Galway, Ireland

²School of Psychology, NUI Galway, Galway, Ireland

Correspondence: Declan P. McKernan, PhD Pharmacology & Therapeutics, School of Medicine, NUI Galway, Galway H9 TK33, Ireland Tel: +353-91-493826, Fax: +353-91-525586 E-mail: declan.mckernan@nuigalway.ie

Received June 22, 2022 Revised August 17, 2022 Accepted August 24, 2022

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Objective

Immune system dysregulation is hypothesised to be central to the aetiopathogenesis of schizophrenia; however, the role of sterile inflammation remains unclear. Damage associated molecular patterns are key initiators of sterile inflammation and are detectable in peripheral blood.

Methods

A defined systematic search of the Web of Science, PubMed, and Scopus was performed to identify adult case-control studies published between January 1990 and June 2022. Three studies consisting of 242 cases and 83 controls met inclusion for the systematic review and meta-analysis of HMGB1 while twenty-eight studies consisting of 1,544 cases and 1,248 healthy controls were included for S100B.

Results

A significant standardised mean difference in peripheral S100B and HMGB1 concentrations was detected between cases and controls. S100B subgroup analysis determined the largest significant effect size for unmedicated individuals diagnosed with schizophrenia.

Conclusion

This study provides evidence that peripheral S100B and HMGB1 concentrations are elevated in individuals diagnosed with schizophrenia when compared with healthy controls. These results should be interpreted with caution as significant heterogeneity was present during meta-analysis of S100B in the entire sample and in sub-group analysis. The persistence of significant heterogeneity throughout subgroup analysis indicates that the current diagnostic groupings may be a barrier to understanding human behaviours and emotions.

Keywords: Schizophrenia, Inflammation, Danger associated molecular patterns, Meta-analysis

INTRODUCTION

Schizophrenia is defined as a complex neuropsychiatric disorder affecting approximately 1% of the global population [1]. The disorder includes delusions, hallucinations, and disorganised speech as positive symptoms, while flattened affect, anhedonia, alogia, avolition, and deficits in social functioning are collectively termed negative symptoms [2,3]. Immune system alterations are intricately interlinked with neurodevelopmental [4,5], neurodegenerative [6,7], genetic [8,9], and biopsychosocial models [10,11] of schizophrenia. Multiple metaanalyses have uncovered increases in pro-inflammatory gene transcripts and protein levels including tumour necrosis factor α, interleukin 1β (IL-1β), IL-6 and C-reactive protein in the peripheral blood and post-mortem brain samples of individuals diagnosed with schizophrenia compared to healthy controls [7,12]. Maternal inflammation prior to birth [13], or a history of autoimmune disorder14 are associated with increased risk of developing schizophrenia. Inflammation can also occur in the absence of infection. Sterile inflammation is initiated when damaged or dying cells release endogenous molecules termed damage associated molecular patterns (DAMPs), that ligate with DAMP sensing receptors and activate the innate immune response [15,16]. High mobility group box 1 (HMGB1) and S100 calcium binding protein B (S100B) are well characterised DAMPs that serve diagnostic and prognostic functions in other disorders including cancer17 and traumatic brain injury [18].

S100B is a calcium binding protein localised in the cytosol of glial cells and astrocytes of the central nervous system [19], and both immune and non-immune cells of the peripheral nervous system [20]. S100B is important for a plethora of homeostatic activities including cell cycle progression and differentiation [21,22], calcium homeostasis [23], activation of calcium dependent enzymes [24], and microtubule assembly and disassembly of the cytoskeleton [25]. S100B is released by damaged or dying cells and interacts with the receptor for advanced glycation end products (RAGE) [26], conferring neuroprotective or neurotoxic effects on neurons in a concentration dependent manner [27]. HMGB1 is a chromatin binding protein and key epigenetic regulator that modulates the binding ability of regulatory elements such as nucleosomes and transcription factors for some genes associated with inflammation [28]. HMGB1 is expressed in all nucleated cells and remains within the nucleus under physiological conditions [29]. Damaged and dying cells release HMGB1 when the structural integrity of cells is diminished [30]. Extracellular HMGB1 interacts with a plethora of receptors including triggering receptor in myeloid cells 1, RAGE, toll-like receptor 2 (TLR2), and TLR4 to execute a number of functions pertaining to pro-inflammatory cytokine release [31]. cell survival and cell death mechanisms [32,33].

The sterile immune response is critical for physiological functioning; however, chronic and sustained levels of inflammation can result in altered cellular functioning, cellular damage and death, which is hypothesised to contribute to the symptomology of schizophrenia [6]. This research aimed to quantify differences in peripheral DAMP concentrations in adults diagnosed with schizophrenia and healthy controls by conducting a systematic review and meta-analysis.

METHODS

Search criteria

A systematic search of the Web of Science, PubMed, and Scopus databases was conducted to identify relevant papers that were published between January 1990 and June 2022 using a combination of the following key search terms: (“schizophrenia” OR “schizoaffective”) AND (“HMGB1” or “high mobility Group Box 1” OR “high mobility group protein 1” OR “S100” OR “S-100” OR “S100B”). Additional articles were identified by searching bibliographical references from articles that were retrieved through the original search.

Study selection criteria

Studies identified by the initial search were screened and selected for inclusion according to the following criteria: 1) longitudinal or cross-sectional, peer-reviewed original articles, 2) article is written in English, 3) case-control studies, 4) quantitative measures of S100B and HMGB1 reported, 5) individuals diagnosed with schizophrenia in accordance with the Diagnostic and Statistical Manual of Mental Disorders, fourth or fifth edition, or World Health Organisations International Statistical Classification of Disease versions ten or eleven, and 6) sample is isolated from peripheral blood. Studies were excluded if: 1) individuals were diagnosed with comorbid neuropsychiatric disorders, 2) individuals were diagnosed with a chronic inflammatory, neurological, or autoimmune disorder, 3) individuals under the age of 18 were included, and 4) arithmetic mean and standard deviation for measures of S100B and/or HMGB1 were not reported or could not be extracted from the article.

Data extraction

Data pertaining to age, sex, concentration of peripheral S100B and/or HMGB1, neuropsychiatric diagnostic method, quantitative assay used, sample location, medication usage, psychopathological state, and medical history were collected where applicable for both the control and schizophrenia cohorts. In the case of longitudinal studies that included repeated blood draws and quantitative measurement of S100B and HMGB1, the values reported from the initial blood draw were included. In the case where the same patient cohort samples were published in multiple journal articles, the first article to appropriately report relevant measures was included, with all other articles using identical samples excluded. When the sample cohort of individuals diagnosed with schizophrenia was further divided into subpopulations and no information for the entire schizophrenia group was reported, the data from the subpopulations was pooled to generate a pooled mean and standard deviation. Quantitative data that was reported in alternative units was converted to ng/L for S100B and ng/mL for HMGB1 to allow comparison. When the arithmetic mean and standard error (SE) of the mean was not reported, the following calculation as used to convert the SE of the mean to standard deviation: standard deviation=(SE of the mean)×(√n) [34]. The online tool WebPlotDigitizer (Pacifica, CA, USA) was used to extract data reported in graphic form only.

Statistical analysis

A meta-analysis was conducted for the thirty studies included in the systematic review of S100B. The standardised mean difference between individuals diagnosed with schizophrenia and controls served as the main outcome measure, while the conservative random effects model was fitted to the data to account for inter-study variance [35]. The effect size was reported as Hedges’ g value with 95% confidence intervals (CIs) [36]. Statistical heterogeneity was measured using Hedges estimator and reported as tau2 [36], the Q-test for heterogeneity [37], and the I² statistic [38]. Subgroup analysis was performed to explore possible causes of heterogeneity. Publication bias was measured using the Fail-Safe N [39], Begg and Mazumdar’s rank correlation [40], Egger’s regression [41], and trim and fill number of Studies [42]. The statistical analysis software jamovi (V1.6) [43] was employed to conduct all statistical analysis.

Data reporting

This systematic review and meta-analysis were conducted in accordance with the most recent version of the Preferred Reporting Items for Systematic reviews and Meta-Analyses guidelines [44].

RESULTS

Characteristics of included studies

The literature search identified 555 potentially relevant publications, of which 30 articles satisfied the inclusion criteria (Figure 1). One article was applicable for inclusion in both S100B and HMGB1 analysis [45]. Two other articles were suitable for inclusion in the HMGB1 group [46,47] and 28 articles were applicable for inclusion in the S100B group which are systematically reported in Supplementary Table 1 (in the online-only Data Supplement). Quantitative analysis of HMGB1 concentration was performed on serum samples from participants using an ELISA assay for all three studies. Kozłowska et al. [45], (2021), measured clinical symptom severity using the Scale for assessment of Negative Symptoms (SANS), and Calgary Depression Scale for Schizophrenia (CDSS), while Mousa et al. [46], (2021) and Yilmaz et al. [47], (2021) used the Positive and Negative Symptom Scale (PANSS).

In total twenty-eight studies consisting of 1,544 individuals diagnosed with schizophrenia and 1,248 healthy controls were included in the systematic review and meta-analysis of S100B. The characteristics of individuals included is illustrated in Figure 2 and reported in Supplementary Table 1 (in the online-only Data Supplement). Ten studies included individuals who were exclusively unmedicated, ten studies included both individuals who were medicated and individuals who were not medicated, while the remaining eight studies included individuals who were all actively taking medication. Twenty-one studies recruited inpatients exclusively, while only three studies recruited only outpatients. A further two studies recruited both inpatients and outpatients, while two studies did not mention if individuals were receiving inpatient or outpatient care. The psychopathological state of individuals was characterised using the PANSS in 22 studies, the Brief Psychiatric Rating Scale (BPRS) [48] in six studies, the SANS was used in three studies, the Scale for assessment of Positive Symptoms (SAPS) [49] in two studies, the Negative Symptom Rating Scale (NSRS) [50] in one study and the CDSS in one study.

Meta-analysis of HMGB1

Individuals diagnosed with schizophrenia had significantly higher peripheral blood concentrations of HMGB1 compared to healthy controls. The observed standardised mean differences between the groups ranged from 0.2653 to 1.5697 with all the estimates having positive directionality (Figure 3). The estimated standardised mean difference between case and control groups determined from the random effects model was g=0.9197, 95% CI (0.5181 to 1.3169) an effect size that differed significantly from zero (z=4.5030, p<0.001). Heterogeneity was measured using: Tau²=0.0692 (SE=0.1252), I²= 55.83% and Q=4.6331.656, p=0.098. Analysis of studentised residuals uncovered that all studies had a value smaller than ±2.3940. Publication bias was assessed and reported as: failsafe, n=51.00, p<0.001; Begg and Mazumdar rank correlation= -0.33, p>0.999; Egger’s regression=-0.50, p=0.537; trim and fill number of studies=0.00.

Whole group meta-analysis of S100B

Similarly, S100B concentrations were significantly higher in peripheral blood of individuals diagnosed with schizophrenia compared to healthy controls. The observed standardised mean differences between the groups ranged from -0.5027 to 2.5247, with only three of the twenty-eight studies not reporting a positive effect (Figure 4). The estimated standardised mean difference between case and control groups determined from the random effects model was g=0.8695, 95% CI (0.5749 to 1.1641) with this effect size differing significantly from zero (z=5.7852, p<0.001). Heterogeneity was measured using: Tau²=0.5580 (SE=0.172), I²=91.79% and Q= 339.1195, p<0.001. Analysis of studentised residuals uncovered that all studies had a value smaller that ±3.1237. Publication bias was assessed and reported as: fail-safe, n=3,638.00, p<0.001; Begg and Mazumdar rank correlation=0.05, p= 0.710; Egger’s regression=-0.03, p=0.978; trim and fill number of studies=5.00.

Subgroup meta-analysis of S100B

Subgroup analysis was performed to assess interstudy heterogeneity and determine the influence of potentially confounding factors on the standardised mean difference between cases and controls, through stratification of participants. The sample (serum or plasma) used to quantitatively measure S100B, the assay used to measure S100B, the medication status of patients, and patient care status (inpatient or outpatient) were the subgroups used, with the standardised mean difference determined from the random effects model. The mean effect size was reported as hedges g with 95% CIs, and interstudy heterogeneity reported as I². Subgroup analysis is summarised in Table 1.

DISCUSSION

To our knowledge, this is the first meta-analysis to assess peripheral HMGB1 and the third such meta-analysis to compare peripheral S100B concentrations in adults diagnosed with schizophrenia and healthy controls [51,52]. This meta-analysis detected significantly elevated peripheral S100B and HMGB1 concentrations in adults diagnosed with schizophrenia compared to healthy controls. An overrepresentation of individuals diagnosed with schizophrenia using inpatient services (Figure 2) suggests that this study is more representative of individuals in an acute phase of illness compared to the whole population of individuals diagnosed with schizophrenia.

In agreement with previous meta-analyses, S100B concentrations are significantly elevated in the peripheral blood of adults diagnosed with schizophrenia compared with healthy controls [51,52]. Novel to this meta-analysis was differential S100B concentrations detected in the peripheral blood of medicated and unmedicated subgroups of individuals diagnosed with schizophrenia. The subgroup of unmedicated individuals diagnosed with schizophrenia had the largest standardised mean difference between case and control populations, while medicated individuals diagnosed with schizophrenia did not report a statistically significant standardised mean difference between case and control groups. This suggests that antipsychotic medication may influence the immune system and partially elicit therapeutic effects through modulation of immune related pathways. At present, a distinct lack of clinical research examining the influence of antipsychotic medication on peripheral S100B concentrations exists. An in-vitro study has provided evidence that both typical and atypical antipsychotic medication can reduce S100B release from neuronal cell lines [53]. Further research is needed to understand if this effect translates to clinical populations. Studies on drug naïve individuals have reported significantly reduced IL-1β and IL-6 levels in the months following treatment with antipsychotic medication [54,55]. A consensus remains to be reached with longitudinal studies reporting that levels of pro-inflammatory cytokines return to their drug naïve baseline or even higher concentrations over time [56-59].

The triangulation of available evidence suggests that elevated peripheral S100B and HMGB1 levels are characteristic of schizophrenia. Physiological concentrations of S100B and HMGB1 are critical for homeostasis. However, sustained supraphysiological expression can induce altered cellular functioning, cellular damage and death, which is hypothesised to contribute to the symptomology of schizophrenia [6,60,61]. These DAMPs are also known to engage with pattern recognition receptors and therefore induce further inflammation [16]. Future longitudinal research exploring their potential causative or consequential roles in schizophrenia is required.

High levels of statistical heterogeneity were a major limiting factor of this study and were also noted in previous meta-analyses [51]. Understanding the sources of interstudy heterogeneity is of paramount importance as the validity of meta-analyses is dependent on this. Subgroup analysis was employed to assess potential contributors to the whole group heterogeneity but failed to reduce heterogeneity despite segregating participants based on potential effect modifiers and confounds including medication status, assay used, sample source and patient treatment status. It is possible that other potentially confounding factors that were not controlled for in this study including body mass index levels [62], diet [63], smoking habits [64], and exercise levels [65] may have contributed to the observed heterogeneity.

Recent criticisms have highlighted the broad and heterogeneous grouping of individuals into diagnostic categories that encompass a vast number of overlapping symptoms, while failing to account for symptom severity [66,67]. Symptomology is hypothesised to more likely reflect neurobiological alterations as opposed to classical diagnostic groupings, indicating that heterogeneity introduced by clinical diagnosis is a barrier to understanding the neurobiological mechanisms underpinning dominant symptomology of neuropsychiatric disorders [68]. Efforts are underway to uncover and integrate neurobiological markers of neuropsychiatric disorders with more traditional diagnostic approaches [69]. Finally, the absence of an independent assessor or second person verification step may increase the risk of missing potentially relevant information.

The disparity between health of individuals diagnosed with schizophrenia and that of the general population has widened in recent decades, despite the advent of first and second generation antipsychotics and continuing research [70,71]. Understanding the role of current diagnostic groupings in contributing to clinical heterogeneity, and its validity as a conceptual model for which to study human thoughts, behaviours, and emotions should be a research priority as the gap between the health of individuals diagnosed with schizophrenia and the general population widens.

Supplementary Materials

The online-only Data Supplement is available with this article at https://doi.org/10.30773/pi.2022.0173.

Supplementary Table 1.

Systematic review of S100B and HMGB1 levels in patients with schizophrenia and healthy controls

pi-2022-0173-suppl1.pdf

Notes

Availability of Data and Material

The datasets generated or analyzed during the study are available from the corresponding author on reasonable request.

Conflicts of Interest

The authors have no potential conflicts of interest to disclose.

Author Contributions

Conceptualization: all authors. Data curation: Michael Mackey, Declan P. McKernan. Formal analysis: Michael Mackey, Declan P. McKernan. Investigation: all authors. Project administration: Laurena Holleran, Gary Donohoe, Declan P. McKernan. Software: Michael Mackey, Declan P. McKernan. Supervision: Declan P. McKernan. Validation: Gary Donohoe, Declan P. McKernan. Visualization: Michael Mackey, Declan P. McKernan. Writing—original draft: Michael Mackey, Declan P. McKernan. Writing— review & editing: Laurena Holleran, Gary Donohoe, Declan P. McKernan.

Funding Statement

None

Figure 1.

PRISMA flow diagram illustrating study selection process for the systematic review and meta-analysis in accordance with PRISMA guidelines[72]. PRISMA, Preferred Reporting Items for Systematic reviews and Meta-Analyses.

Figure 2.

Pie charts illustrating characteristics of individuals included in the S100B systematic review and meta-analysis. A: Case and control. B: Assay. Pie charts A and B are representative of the whole population of individuals. C: Medication status. D: Patient setting. Pie charts C and D are exclusively representative of individuals diagnosed with schizophrenia. Unmedicated individuals described in pie chart C represents individuals who were not taking antipsychotic medication for at least one week prior to study commencement. ELISA, enzyme linked immunosorbent assay.

Figure 3.

HMGB1 meta-analysis. This forest plot illustrates the whole group meta-analysis analysing peripheral HMGB1 concentrations in schizophrenia, with hedges g effect size and 95% CIs. CI, confidence interval.

Figure 4.

S100B meta-analysis. This Forest plot illustrates the whole group meta-analysis analysing peripheral S100B concentrations in schizophrenia, with hedges g effect size and 95% confidence intervals. CI, confidence interval.

Table 1.

Summary of subgroup analysis

Cohort		Sample (N)	Control (N)	Sample	Effect size [CI]	I²	Fail-safe N
All individuals		1,544	1,248	Plasma and serum	0.87 [0.57, 1.16]	91.8%	3,638.00 (p<0.001)
Sample
	Serum measurement	1,385	1,093	Serum	0.87 [0.56, 1.17]	91.2%	2,566.00 (p<0.001)
	Plasma measurement	159	155	Plasma	0.89 [-0.11, 1.88]	93.5%	89.00 (p<0.001)
Assay
	ELISA	1,144	902	Plasma and serum	0.91 [0.44, 1.39]	95.6%	1,449.00 (p<0.001)
	Immunoluminometric	400	346	Plasma and serum	0.82 [0.49, 1.14]	75.5%	3,638.00 (p<0.001)
Medication status
	Medicated individuals	343	214	Plasma and serum	0.73 [0.01, 1.46]	92.9%	158.00 (p<0.001)
	Non-medicated individuals	466	412	Plasma and serum	1.08 [0.68, 1.48]	86.2%	714.00 (p<0.001)
	Medicated and unmedicated individuals	735	622	Plasma and serum	0.79 [0.25, 1.33]	92.3%	422.00 (p<0.001)
Patient setting
	Inpatient	970	781	Plasma and serum	0.99 [0.64, 1.34]	90.4%	2,458.00 (p<0.001)
	Outpatient	174	112	Plasma and serum	0.21 [-0.42, 0.84]	83.3%	2.00 (p<0.001)
	Mixed cohorts including outpatients and inpatients	60	55	Serum	N/A	N/A	N/A
	Not stated	340	300	Serum	N/A	N/A	N/A

This table depicts the subgroup meta-analysis effect sizes based on sample source, assay conducted, medication status and patient setting.

ELISA, enzyme linked immunosorbent assay; CI, confidence interval; N/A, not applicable

REFERENCES

1. McCutcheon RA, Reis Marques T, Howes OD. Schizophrenia-an overview. JAMA Psychiatry 2020;77:201-210.

2. World Health Organization. ICD-10: international statistical classification of diseases and related health problems: 10th revision (2nd ed). Geneva: World Health Organization; 2004.

3. American Psychiatric Association. Diagnostic and statistical manual of mental disorders (5th ed). Arlington: American Psychiatric Association; 2013.

4. Cannon TD, Chung Y, He G, Sun D, Jacobson A, van Erp TG, et al. Progressive reduction in cortical thickness as psychosis develops: a multisite longitudinal neuroimaging study of youth at elevated clinical risk. Biol Psychiatry 2015;77:147-157.

5. Weinhard L, di Bartolomei G, Bolasco G, Machado P, Schieber NL, Neniskyte U, et al. Microglia remodel synapses by presynaptic trogocytosis and spine head filopodia induction. Nat Commun 2018;9:1228

6. Khandaker GM, Cousins L, Deakin J, Lennox BR, Yolken R, Jones PB. Inflammation and immunity in schizophrenia: implications for pathophysiology and treatment. Lancet Psychiatry 2015;2:258-270.

7. van Kesteren CF, Gremmels H, de Witte LD, Hol EM, Van Gool AR, Falkai PG, et al. Immune involvement in the pathogenesis of schizophrenia: a meta-analysis on postmortem brain studies. Transl Psychiatry 2017;7:e1075

8. Dennison CA, Legge SE, Pardiñas AF, Walters JTR. Genome-wide association studies in schizophrenia: recent advances, challenges and future perspective. Schizophr Res 2020;217:4-12.

9. O’Connell G, Lawrie SM, McIntosh AM, Hall J. Schizophrenia risk genes: implications for future drug development and discovery. Biochem Pharmacol 2011;81:1367-1373.

10. Barker V, Gumley A, Schwannauer M, Lawrie SM. An integrated biopsychosocial model of childhood maltreatment and psychosis. Br J Psychiatry 2015;206:177-180.

11. Engel GL. The clinical application of the biopsychosocial model. J Med Philos 1981;6:101-124.

12. Miller BJ, Buckley P, Seabolt W, Mellor A, Kirkpatrick B. Meta-analysis of cytokine alterations in schizophrenia: clinical status and antipsychotic effects. Biol Psychiatry 2011;70:663-671.

13. Canetta S, Sourander A, Surcel HM, Hinkka-Yli-Salomäki S, Leiviskä J, Kellendonk C, et al. Elevated maternal C-reactive protein and increased risk of schizophrenia in a national birth cohort. Am J Psychiatry 2014;171:960-968.

14. Eaton WW, Byrne M, Ewald H, Mors O, Chen CY, Agerbo E, et al. Association of schizophrenia and autoimmune diseases: linkage of Danish national registers. Am J Psychiatry 2006;163:521-528.

15. Land W. Allograft injury mediated by reactive oxygen species: from conserved proteins of Drosophila to acute and chronic rejection of human transplants. Part II: role of reactive oxygen species in the induction of the heat shock response as a regulator of innate. Transplant Rev 2003;17:31-44.

16. Zindel J, Kubes P. DAMPs, PAMPs, and LAMPs in immunity and sterile inflammation. Annu Rev Pathol 2020;15:493-518.

17. Cochran AJ, Lu HF, Li PX, Saxton R, Wen DR. S-100 protein remains a practical marker for melanocytic and other tumours. Melanoma Res 1993;3:325-330.

18. Oris C, Pereira B, Durif J, Simon-Pimmel J, Castellani C, Manzano S, et al. The biomarker S100B and mild traumatic brain injury: a meta-analysis. Pediatrics 2018;141:e20180037

19. Rickmann M, Wolff JR. S100 protein expression in subpopulations of neurons of rat brain. Neuroscience 1995;67:977-991.

20. Michetti F, Rende M, Lauriola L, Stolfi V, Calogero G, Cocchia D. Immunohistochemical localization of S-100-containing cells in non-nervous structures of the human eye. Cell Biol Int Rep 1986;10:765-773.

21. Li D, Li K, Chen G, Xia J, Yang T, Cai P, et al. S100B suppresses the differentiation of C3H/10T1/2 murine embryonic mesenchymal cells into osteoblasts. Mol Med Rep 2016;14:3878-3886.

22. Riuzzi F, Sorci G, Donato R. S100B protein regulates myoblast proliferation and differentiation by activating FGFR1 in a bFGF-dependent manner. J Cell Sci 2011;124(Pt 14):2389-2400.

23. Xiong Z, O’Hanlon D, Becker LE, Roder J, MacDonald JF, Marks A. Enhanced calcium transients in glial cells in neonatal cerebellar cultures derived from S100B null mice. Exp Cell Res 2000;257:281-289.

24. Gógl G, Alexa A, Kiss B, Katona G, Kovács M, Bodor A, et al. Structural basis of ribosomal S6 kinase 1 (RSK1) inhibition by S100B protein: modulation of the extracellular signal-regulated kinase (ERK) signaling cascade in a calcium-dependent way. J Biol Chem 2016;291:11-27.

25. Donato R. Calcium-independent, pH-regulated effects of S-100 proteins on assembly-disassembly of brain microtubule protein in vitro. J Biol Chem 1988;263:106-110.

26. Bianchi R, Kastrisianaki E, Giambanco I, Donato R. S100B protein stimulates microglia migration via RAGE-dependent up-regulation of chemokine expression and release. J Biol Chem 2011;286:7214-7226.

27. Businaro R, Leone S, Fabrizi C, Sorci G, Donato R, Lauro GM, et al. S100B protects LAN-5 neuroblastoma cells against Abeta amyloid-induced neurotoxicity via RAGE engagement at low doses but increases Abeta amyloid neurotoxicity at high doses. J Neurosci Res 2006;83:897-906.

28. Andersson U, Yang H, Harris H. Extracellular HMGB1 as a therapeutic target in inflammatory diseases. Expert Opin Ther Targets 2018;22:263-277.

29. Youn JH, Shin JS. Nucleocytoplasmic shuttling of HMGB1 is regulated by phosphorylation that redirects it toward secretion. J Immunol 2006;177:7889-7897.

30. Bonaldi T, Talamo F, Scaffidi P, Ferrera D, Porto A, Bachi A, et al. Monocytic cells hyperacetylate chromatin protein HMGB1 to redirect it towards secretion. EMBO J 2003;22:5551-5560.

31. Sha Y, Zmijewski J, Xu Z, Abraham E. HMGB1 develops enhanced proinflammatory activity by binding to cytokines. J Immunol 2008;180:2531-2537.

32. Yang M, Yang X, Wang S, Xu L, Ke S, Ding X, et al. HMGB1-induced endothelial cell pyroptosis is involved in systemic inflammatory response syndrome following radiofrequency ablation of hepatic hemangiomas. Am J Transl Res 2019;11:7555-7567.

33. Brezniceanu ML, Völp K, Bösser S, Solbach C, Lichter P, Joos S, et al. HMGB1 inhibits cell death in yeast and mammalian cells and is abundantly expressed in human breast carcinoma. FASEB J 2003;17:1295-1297.

34. Cuijpers P. Meta-analyses in mental health research. A practical guide. Amsterdam: VU University Amsterdam; 2016.

35. Laird NM, Ware JH. Random-effects models for longitudinal data. Biometrics 1982;38:963-974.

36. Hedges LV, Olkin I. Statistical methods for meta-analysis. San Diego: Academic Press; 1985.

37. Cochran WG. The combination of estimates from different experiments. Biometrics 1954;10:101-129.

38. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003;327:557-560.

39. Orwin RG. A fail-safe N for effect size in meta-analysis. J Educ Stat 1983;8:157-159.

40. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics 1994;50:1088-1101.

41. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629-634.

42. Duval S, Tweedie R. A nonparametric “trim and fill” method of accounting for publication bias in meta-analysis. J Am Stat Assoc 2000;95:89-98.

43. Şahin M, Aybek E. Jamovi: an easy to use statistical software for the social scientists. Int J Assess Tool Educ 2019;6:670-692.

44. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71

45. Kozłowska E, Brzezińska-Błaszczyk E, Agier J, Wysokiński A, Żelechowska P. Alarmins (IL-33, sST2, HMGB1, and S100B) as potential biomarkers for schizophrenia. J Psychiatr Res 2021;138:380-387.

46. Mousa RF, Al-Hakeim HK, Alhaideri A, Maes M. Chronic fatigue syndrome and fibromyalgia-like symptoms are an integral component of the phenome of schizophrenia: neuro-immune and opioid system correlates. Metab Brain Dis 2021;36:169-183.

47. Yilmaz N, Yelboga Z, Yilmaz Y, Demirpence O. High mobility group box-1 levels in schizophrenia: potential biomarker of remission phase. J Med Biochem 2021;40:295-301.

48. Overall JE, Gorham DR. The Brief Psychiatric Rating Scale (BPRS): recent developments in ascertainment and scaling. Psychopharmacol Bull 1988;24:97-99.

49. Andreasen NC. Scale for the assessment of positive symptoms (SAPS). Iowa: University of Iowa; 1984.

50. Iager AC, Kirch DG, Wyatt RJ. A negative symptom rating scale. Psychiatry Res 1985;16:27-36.

51. Aleksovska K, Leoncini E, Bonassi S, Cesario A, Boccia S, Frustaci A. Systematic review and meta-analysis of circulating S100B blood levels in schizophrenia. PLoS One 2014;9:e106342

52. Schroeter ML, Abdul-Khaliq H, Krebs M, Diefenbacher A, Blasig IE. Neuron-specific enolase is unaltered whereas S100B is elevated in serum of patients with schizophrenia--original research and meta-analysis. Psychiatry Res 2009;167:66-72.

53. Steiner J, Schroeter ML, Schiltz K, Bernstein HG, Müller UJ, Richter-Landsberg C, et al. Haloperidol and clozapine decrease S100B release from glial cells. Neuroscience 2010;167:1025-1031.

54. Capuzzi E, Bartoli F, Crocamo C, Clerici M, Carrà G. Acute variations of cytokine levels after antipsychotic treatment in drug-naïve subjects with a first-episode psychosis: a meta-analysis. Neurosci Biobehav Rev 2017;77:122-128.

55. Song XQ, Lv LX, Li WQ, Hao YH, Zhao JP. The interaction of nuclear factor-kappa B and cytokines is associated with schizophrenia. Biol Psychiatry 2009;65:481-488.

56. Romeo B, Brunet-Lecomte M, Martelli C, Benyamina A. Kinetics of cytokine levels during antipsychotic treatment in schizophrenia: a meta-analysis. Int J Neuropsychopharmacol 2018;21:828-836.

57. Song X, Fan X, Li X, Zhang W, Gao J, Zhao J, et al. Changes in pro-inflammatory cytokines and body weight during 6-month risperidone treatment in drug naïve, first-episode schizophrenia. Psychopharmacology (Berl) 2014;231:319-325.

58. Tourjman V, Kouassi É, Koué MÈ, Rocchetti M, Fortin-Fournier S, Fusar-Poli P, et al. Antipsychotics’ effects on blood levels of cytokines in schizophrenia: a meta-analysis. Schizophr Res 2013;151:43-47.

59. Zhang XY, Zhou DF, Cao LY, Zhang PY, Wu GY, Shen YC. Changes in serum interleukin-2, -6, and -8 levels before and during treatment with risperidone and haloperidol: relationship to outcome in schizophrenia. J Clin Psychiatry 2004;65:940-947.

60. Kang R, Chen R, Zhang Q, Hou W, Wu S, Cao L, et al. HMGB1 in health and disease. Mol Aspects Med 2014;40:1-116.

61. Rothermundt M, Peters M, Prehn JH, Arolt V. S100B in brain damage and neurodegeneration. Microsc Res Tech 2003;60:614-632.

62. Tzoulaki I, Jarvelin MR, Hartikainen AL, Leinonen M, Pouta A, Paldanius M, et al. Size at birth, weight gain over the life course, and lowgrade inflammation in young adulthood: northern Finland 1966 Birth Cohort study. Eur Heart J 2008;29:1049-1056.

63. Nettleton JA, Steffen LM, Mayer-Davis EJ, Jenny NS, Jiang R, Herrington DM, et al. Dietary patterns are associated with biochemical markers of inflammation and endothelial activation in the Multi-Ethnic Study of Atherosclerosis (MESA). Am J Clin Nutr 2006;83:1369-1379.

64. Lee J, Taneja V, Vassallo R. Cigarette smoking and inflammation: cellular and molecular mechanisms. J Dent Res 2012;91:142-149.

65. Starkie R, Ostrowski SR, Jauffred S, Febbraio M, Pedersen BK. Exercise and IL-6 infusion inhibit endotoxin-induced TNF-alpha production in humans. FASEB J 2003;17:884-886.

66. Allsopp K, Read J, Corcoran R, Kinderman P. Heterogeneity in psychiatric diagnostic classification. Psychiatry Res 2019;279:15-22.

67. Wardenaar KJ, de Jonge P. Diagnostic heterogeneity in psychiatry: towards an empirical solution. BMC Med 2013;11:201

68. Khandaker GM, Stochl J, Zammit S, Lewis G, Dantzer R, Jones PB. Association between circulating levels of C-reactive protein and positive and negative symptoms of psychosis in adolescents in a general population birth cohort. J Psychiatr Res 2021;143:534-542.

69. Cuthbert BN. The RDoC framework: facilitating transition from ICD/DSM to dimensional approaches that integrate neuroscience and psychopathology. World Psychiatry 2014;13:28-35.

70. Saha S, Chant D, McGrath J. A systematic review of mortality in schizophrenia: is the differential mortality gap worsening over time? Arch Gen Psychiatry 2007;64:1123-1131.

71. Murray CJ, Lopez AD. Global mortality, disability, and the contribution of risk factors: Global Burden of Disease Study. Lancet 1997;349:1436-1442.

72. Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ 2009;339:b2535

73. Chen S, Tian L, Chen N, Xiu M, Wang Z, Yang G, et al. Cognitive dysfunction correlates with elevated serum S100B concentration in drugfree acutely relapsed patients with schizophrenia. Psychiatry Res 2017;247:6-11.

74. Dai N, Jie H, Duan Y, Xiong P, Xu X, Chen P, et al. Different serum protein factor levels in first-episode drug-naive patients with schizophrenia characterized by positive and negative symptoms. Psychiatry Clin Neurosci 2020;74:472-479.

75. Hendouei N, Hosseini SH, Panahi A, Khazaeipour Z, Barari F, Sahebnasagh A, et al. Negative correlation between serum S100B and leptin levels in schizophrenic patients during treatment with clozapine and risperidone: preliminary evidence. Iran J Pharm Res 2016;15:323-330.

76. Hong W, Zhao M, Li H, Peng F, Wang F, Li N, et al. Higher plasma S100B concentrations in schizophrenia patients, and dependently associated with inflammatory markers. Sci Rep 2016;6:27584

77. Lara DR, Gama CS, Belmonte-de-Abreu P, Portela LV, Gonçalves CA, Fonseca M, et al. Increased serum S100B protein in schizophrenia: a study in medication-free patients. J Psychiatr Res 2001;35:11-14.

78. Ling SH, Tang YL, Jiang F, Wiste A, Guo SS, Weng YZ, et al. Plasma S-100B protein in Chinese patients with schizophrenia: comparison with healthy controls and effect of antipsychotics treatment. J Psychiatr Res 2007;41:36-42.

79. Milleit B, Smesny S, Rothermundt M, Preul C, Schroeter ML, von Eiff C, et al. Serum S100B protein is specifically related to white matter changes in schizophrenia. Front Cell Neurosci 2016;10:33

80. Morera-Fumero AL, Díaz-Mesa E, Abreu-Gonzalez P, Fernandez-Lopez L, Cejas-Mendez MDR. Day/night changes in serum S100B protein concentrations in acute paranoid schizophrenia. Prog Neuropsychopharmacol Biol Psychiatry 2017;75:207-212.

81. Qi LY, Xiu MH, Chen DC, Wang F, Kosten TA, Kosten TR, et al. Increased serum S100B levels in chronic schizophrenic patients on long-term clozapine or typical antipsychotics. Neurosci Lett 2009;462:113-117.

82. Rothermundt M, Missler U, Arolt V, Peters M, Leadbeater J, Wiesmann M, et al. Increased S100B blood levels in unmedicated and treated schizophrenic patients are correlated with negative symptomatology. Mol Psychiatry 2001;6:445-449.

83. Rothermundt M, Ponath G, Glaser T, Hetzel G, Arolt V. S100B serum levels and long-term improvement of negative symptoms in patients with schizophrenia. Neuropsychopharmacology 2004;29:1004-1011.

84. Rothermundt M, Falkai P, Ponath G, Abel S, Bürkle H, Diedrich M, et al. Glial cell dysfunction in schizophrenia indicated by increased S100B in the CSF. Mol Psychiatry 2004;9:897-899.

85. Rothermundt M, Ohrmann P, Abel S, Siegmund A, Pedersen A, Ponath G, et al. Glial cell activation in a subgroup of patients with schizophrenia indicated by increased S100B serum concentrations and elevated myo-inositol. Prog Neuropsychopharmacol Biol Psychiatry 2007;31:361-364.

86. Kim HR, Lee MK, Park DB. Increased serum S100B protein in chronic schizophrenic patients in Korea. Clin Chem Lab Med 2007;45:1561-1563.

87. Šakić M, Karlović D, Vidrih B, Peitl V, Crnković D, Vrkić N. Increased calcium-independent lipoprotein phospholipase A2 but not protein S100 in patients with schizophrenia. Psychiatr Danub 2016;28:45-50.

88. Sarandol A, Kirli S, Akkaya C, Altin A, Demirci M, Sarandol E. Oxidative-antioxidative systems and their relation with serum S100 B levels in patients with schizophrenia: effects of short term antipsychotic treatment. Prog Neuropsychopharmacol Biol Psychiatry 2007;31:1164-1169.

89. Schmitt A, Bertsch T, Henning U, Tost H, Klimke A, Henn FA, et al. Increased serum S100B in elderly, chronic schizophrenic patients: negative correlation with deficit symptoms. Schizophr Res 2005;80:305-313.

90. Schroeter ML, Abdul-Khaliq H, Frühauf S, Höhne R, Schick G, Diefenbacher A, et al. Serum S100B is increased during early treatment with antipsychotics and in deficit schizophrenia. Schizophr Res 2003;62:231-236.

91. Steiner J, Bielau H, Bernstein HG, Bogerts B, Wunderlich MT. Increased cerebrospinal fluid and serum levels of S100B in first-onset schizophrenia are not related to a degenerative release of glial fibrillar acidic protein, myelin basic protein and neurone-specific enolase from glia or neurones. J Neurol Neurosurg Psychiatry 2006;77:1284-1287.

92. Steiner J, Myint AM, Schiltz K, Westphal S, Bernstein HG, Walter M, et al. S100B serum levels in schizophrenia are presumably related to visceral obesity and insulin resistance. Cardiovasc Psychiatry Neurol 2010;2010:480707

93. Uzbay T, Goktalay G, Kayir H, Eker SS, Sarandol A, Oral S, et al. Increased plasma agmatine levels in patients with schizophrenia. J Psychiatr Res 2013;47:1054-1060.

94. Wiesmann M, Wandinger KP, Missler U, Eckhoff D, Rothermundt M, Arolt V, et al. Elevated plasma levels of S-100b protein in schizophrenic patients. Biol Psychiatry 1999;45:1508-1511.

95. Xiong P, Zeng Y, Wu Q, Han Huang DX, Zainal H, Xu X, et al. Combining serum protein concentrations to diagnose schizophrenia: a preliminary exploration. J Clin Psychiatry 2014;75:e794-801.

96. Zhang XY, Xiu MH, Song C, Chen DC, Wu GY, Haile CN, et al. Increased serum S100B in never-medicated and medicated schizophrenic patients. J Psychiatr Res 2010;44:1236-1240.

97. Gattaz WF, Lara DR, Elkis H, Portela LV, Gonçalves CA, Tort AB, et al. Decreased S100-beta protein in schizophrenia: preliminary evidence. Schizophr Res 2000;43:91-95.

98. O’Connell K, Thakore J, Dev KK. Levels of S100B are raised in female patients with schizophrenia. BMC Psychiatry 2013;13:146