Plasma exosomal microRNA profiles associated with agitation symptoms in schizophrenia.

Plasma exosomal microRNA profiles associated with agitation symptoms in schizophrenia

Background

Inflammatory dysregulation and synaptic dysfunction have been implicated in the pathophysiology of agitation in schizophrenia (SZ), but the underlying molecular mechanisms remain unclear. Plasma exosomal microRNAs (miRNAs) are stable circulating molecules that may reflect disease-related molecular alterations associated with schizophrenia. This study aimed to investigate plasma exosomal miRNA expression profiles and peripheral inflammatory factors in schizophrenia patients with agitation, and to identify potential biomarkers for agitation symptoms.

Methods

In this study, plasma exosomal miRNA expression profiles and peripheral inflammatory factors were compared between schizophrenia patients with agitation assessed within 24 h of hospital admission using PANSS-EC (SZA group, n = 35) and those without agitation (SZN group, n = 29). Differentially expressed plasma exosomal miRNAs were identified and subjected to GO and KEGG enrichment analyses. Peripheral inflammatory factors, including cytokine levels, neutrophil counts, and neutrophil-to-albumin ratio (NAR), were also measured. Partial correlation analyses (adjusting for age, sex, BMI, chlorpromazine equivalent, and diazepam equivalent) with Bonferroni correction were performed. A binary logistic regression (LR) model with forward stepwise selection (fixed covariates: age, sex, BMI, chlorpromazine equivalent, and diazepam equivalent) was constructed to identify discriminating factors, followed by receiver operating characteristic (ROC) curve analysis and five‑fold cross‑validation.

Results

A total of 18 differentially expressed plasma exosomal miRNAs (15 upregulated, 3 downregulated) were identified in the SZA group. Target genes of upregulated miRNAs were enriched in MAPK and Ras signaling pathways, as well as cholinergic synapse and efferocytosis. Conversely, target genes of the downregulated miRNAs were specifically enriched in glutamatergic synapse. NAR showed a trend toward elevation in the SZA group (uncorrected p = 0.004, Bonferroni‑adjusted p = 0.088), while no significant differences in cytokine levels were observed. Partial correlation analysis revealed that in the SZA group, IL‑12p70 was positively correlated with miR‑3620‑5p (r = 0.893, adjusted p < 0.001) and miR‑17‑3p (r = 0.768, adjusted p = 0.028). Multivariable logistic regression (including fixed covariates: age, sex, BMI, chlorpromazine equivalent, diazepam equivalent) retained NAR (OR = 3.778, p = 0.006) and miR‑543 (OR = 0.162, p = 0.026). The combined model (NAR + miR‑543 + fixed covariates) achieved an AUC of 0.812, outperforming single‑marker models (NAR AUC = 0.752, miR‑543 AUC = 0.715). Five‑fold cross‑validation yielded a mean AUC of 0.795 (95% CI 0.675–0.916), with sensitivity 80.1 ± 16.5%, specificity 79.5 ± 26.1%, and accuracy 77.8 ± 8.9%.

Conclusion

The findings suggest that schizophrenia patients with agitation exhibit distinct plasma exosomal miRNA expression profiles and peripheral inflammatory alterations compared to those without agitation. These findings also imply potential involvement of inflammation-associated signaling and synaptic pathways in the biological mechanisms underlying agitation in schizophrenia. Furthermore, the combination of NAR and miR-543 may serve as a potential biomarker for agitation assessment in schizophrenia, offering new insights into the biological mechanisms underlying this clinically challenging symptom.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12888-026-08459-2.

Introduction

Schizophrenia is a complex psychiatric disorder characterized by a broad range of clinical features, including positive symptoms, negative symptoms, cognitive deficits, agitation and affective symptoms [1]. Agitation is one of the vital symptoms of schizophrenia (SZ), defined as a state of motor restlessness accompanied by mental tension, characterized by restlessness, irritability, anxiety, and excessive movement or speech without a specific purpose [2]. As a major cause of hospitalization in patients with schizophrenia [3], agitation occurs with an incidence ranging from 15% to 53% [4]. Clinically, agitation is considered as a continuum that extends from inner tension and restlessness to further aggressive or violent behavior, with no clear boundaries between these states [5]. Such uncontrolled behavior poses serious risks to patients’ families, healthcare providers, and society [6]. In addition, agitation is also considered to be associated with more severe positive symptoms [7] and poorer prognosis [8]. Therefore, it is essential to further clarify the biological mechanisms underlying agitation in schizophrenia.

Synaptic dysfunction is a key contributor in schizophrenia. Synaptic abnormalities, including altered expression of synaptic proteins (e.g., postsynaptic density protein 95 (PSD95), synaptophysin), reduced dendritic spine density, and impaired synaptic plasticity, have been consistently observed in postmortem brain tissues of schizophrenia patients [9, 10]. These alterations are particularly pronounced in brain regions implicated in emotion and behavior regulation, such as the prefrontal cortex and hippocampus [11]. Notably, glutamatergic dysfunction [12] and NMDA receptor abnormalities in GABAergic synapses [13] have been documented in these regions in schizophrenia. Despite synaptic dysfunction being well established in schizophrenia, studies directly examining synaptic alterations in schizophrenia patients with agitation symptoms remain limited. Inflammatory dysregulation is another important pathophysiological mechanism underlying agitation in schizophrenia. Clinical studies indicate that patients with schizophrenia exhibiting agitation symptoms demonstrate elevated levels of multiple pro-inflammatory factors in their peripheral blood. A study involving 119 schizophrenia patients found that the high-agitation group had higher levels of the pro-inflammatory cytokine interleukin-6 (IL-6) than the low-agitation group. Markers associated with M1-type macrophages and indicators of the immune-inflammatory response system also showed an upward trend [14]. Elevated tumor necrosis factor-α levels (TNF-α) [15] and C-reactive protein (CRP) [16] levels have also been reported in schizophrenia patients with agitation. Although direct evidence in patients with agitation remains limited, central neuroinflammation has been associated with overall symptom severity in schizophrenia. Neuropathological and neuroimaging studies have revealed activated microglia and regional gray matter volume loss in the prefrontal cortex [17]. In a pro-inflammatory environment, activated microglia can promote excessive synaptic pruning, particularly of glutamatergic synapses, resulting in excitation/inhibition imbalance [18]. Through microglia-mediated synaptic remodeling, neuroinflammation may further disrupt glutamatergic neurotransmission and neural circuit function, thereby providing a mechanistic link between inflammatory dysregulation and synaptic abnormalities. Such alterations are particularly evident in the prefrontal [19, 20] and cingulate cortex [21], brain regions critical for emotional regulation and impulse control, and may partially underlie the pathophysiological mechanisms of agitation in schizophrenia. Collectively, while synaptic dysfunction and inflammatory dysregulation are well-established features of schizophrenia, their specific contributions to agitation remain incompletely understood and warrant further investigation.

Given the difficulty in obtaining brain tissue and cerebrospinal fluid samples, it is challenging to investigate structural and functional abnormalities in synapses. Due to the presence of the blood-brain barrier (BBB), peripheral and central inflammatory levels are not perfectly correlated. Indeed, only weak correlations have been observed for IL-6 and TNF-α [22]. A review of patients with schizophrenia spectrum disorders revealed a dissociation between central and peripheral immune markers. Specifically, individuals exhibiting signs of central immune dysfunction, such as positive oligoclonal bands and elevated albumin quotient (QAlb) levels in cerebrospinal fluid (CSF), did not show corresponding elevations in peripheral blood inflammatory markers like CRP or neutrophil counts [23]. Moreover, peripheral cytokine concentrations are influenced by multiple biological and pre-analytical factors, including antipsychotic treatment, circadian variation, fasting status, sample collection, and processing conditions, which may further affect their reproducibility and interpretation [24, 25]. These findings suggested the limitations of using peripheral blood markers to reflect central immune status. Therefore, there is a need for biomarkers that can capture disease-related molecular alterations beyond conventional peripheral inflammatory markers.

Exosomes, as carriers discovered in recent years, have double-layered lipid vesicles with a diameter of approximately 150 nm that can cross the BBB and be detected from peripheral blood [26]. Exosomes are widely present in bodily fluids, including blood [27], urine [28], and cerebrospinal fluid [29], and can transport and deliver substances such as proteins, lipids, and RNA [30]. Unlike free circulating miRNAs, which are vulnerable to pre-analytical factors such as temperature and storage duration [31, 32], exosomes package their cargo within a protective lipid bilayer, conferring greater stability and resistance to degradation in the peripheral circulation [33]. Consistent with these advantages, accumulating evidence has demonstrated that peripheral blood exosomes contain disease-related molecular cargo associated with schizophrenia. For instance, inflammatory-related miRNAs from schizophrenia patients, such as the miR-15/107 family and miR-103, have been reported to exhibit increased expression in brain tissue [34, 35], while similar dysregulation has also been observed in plasma exosomes [36]. These findings suggest that plasma exosomal miRNAs may capture disease-related molecular alterations associated with schizophrenia and therefore represent promising peripheral biomarkers. Furthermore, evidence suggests that exosomes also carry synaptic-related proteins, such as microtubule-associated protein 1B (MAP1B) [37] and dysbindin-1, which is encoded by the schizophrenia susceptibility gene Dystrobrevin Binding Protein 1(DTNBP1) [38]. To sum up, plasma exosomes carry molecular cargo associated with inflammatory and synaptic processes and may serve as peripheral biomarkers for investigating the biological mechanisms underlying agitation in schizophrenia.

Accordingly, using plasma exosomal miRNAs as biomarkers, this study aimed to investigate the expression profiles of plasma exosomal miRNAs in schizophrenia patients with and without agitation symptoms and identify predictive exosomal miRNAs associated with agitation. We hypothesized that agitation is accompanied by distinct alterations in exosomal miRNAs related to inflammatory and synaptic processes. By exploring these post-transcriptional molecular changes, this study sought to provide insights into the biological mechanisms underlying agitation in schizophrenia and to identify potential biomarkers for agitation.

Materials and methods

Participants

Sixty-four SZ patients were recruited from Nanjing Brain Hospital (NBH) between 2025 and 2026.The inclusion criteria were as follows: (i) age between 18 and 55 years old; (ii) meeting the Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5) criteria for schizophrenia, as confirmed by two experienced psychiatrists; (iii) elementary school education and above; (iv) without other psychiatric disorders; (v) without recent infection, brain injury, metabolic diseases or autoimmune diseases; (vi) without current or past use of tobacco, alcohol, or other psychoactive substances. The antipsychotic medications taken by schizophrenia patients upon admission and before blood sampling and PANSS assessment were converted to chlorpromazine equivalent [39], and any benzodiazepines administered were converted to diazepam equivalent [40]. All participants provided written informed consent, and the study was approved by the Ethics Committee of Nanjing Brain Hospital. A summary of demographic and clinical characteristics for all the participants was shown in Table 1.

Positive and Negative Syndrome Scale (PANSS) assessments were administered within 24 h of admission. The agitation symptoms were evaluated by the PANSS excitatory component (PANSS-EC), which included excitement (P4), hostility (P7), tension (G4), uncooperativeness (G8), and poor impulse control (G14). The PANSS-EC total score ≥ 14, with one or more items scoring ≥ 4, indicates the presence of agitation [15, 41, 42]. According to the criteria, the SZ patients in this study were divided into non-agitated (SZN, n = 29) and agitated (SZA, n = 35) subgroups.

Plasma collection

Within 24 h of admission, fasting blood samples were drawn from the anterior elbow vein. All blood samples were collected using EDTA-K2 anticoagulant tubes or gel-separator tubes with clotting agents, depending on the testing purpose, and processed within 4 h at 4 °C.

The concentration of plasma cytokines, including 10 selected pro-inflammatory cytokines (IL-2, IL-5, IL-6, IL-8, IL-1β, IL-12p70, IL-17, IFN-α, IFN-γ, and TNF-α) and 2 anti-inflammatory cytokines (IL-4 and IL-10), was quantified using the enzyme-linked immunosorbent assay (ELISA) method. Fasting venous blood samples were collected from all subjects within 24 h of admission and transferred to EDTA anticoagulant tubes. The samples were then centrifuged (1000 g/min, 4 °C, 10 min) to obtain the supernatant, which was stored at − 80 °C for further analysis. For each cytokine, a 96-well ELISA plate was set up with blank wells, standard wells, and sample wells, as well as duplicate wells. The optical density of the samples was determined using a standard curve based on the measurement of standard samples. All samples were analyzed by two experienced technicians who were blinded to the patients’ identities and clinical information. The within-run and between-run coefficients of variation for all analyses were found to be less than 10%.

Neutrophil absolute count, lymphocyte absolute count, and monocyte absolute count were determined using flow cytometry. Platelet count was measured by impedance method. The neutrophil to lymphocyte ratio (NLR) was calculated by dividing the neutrophil absolute count by the lymphocyte absolute count. The neutrophil to monocyte ratio (NMR), platelet to lymphocyte ratio (PLR), and monocyte to lymphocyte ratio (MLR) were calculated similarly. Albumin concentration was measured using the bromocresol green method. The neutrophil to albumin ratio (NAR) was calculated by dividing the neutrophil count by the albumin concentration.

Isolation of plasma exosomes

Blood samples were processed using a two-step centrifugation protocol. First, samples were centrifuged at 1900 g for 10 min, followed by collection of the supernatant plasma. The supernatant was then centrifuged at 3000 g for 15 min. After the second centrifugation, the resulting supernatant was collected and stored at − 80 °C until further processing for sequencing.

Plasma exosomes were isolated by ultracentrifugation. Frozen plasma samples were thawed at 37 °C and centrifuged at 2,000 × g for 30 min and subsequently at 10,000 × g for 45 min at 4 °C to remove cellular debris and larger vesicles. The resulting supernatant was filtered through a 0.45 μm membrane (Millipore, R6BA09493) and ultracentrifuged at 100,000 × g for 70 min at 4 °C. The exosome pellet was resuspended in 10 mL of pre-cooled 1× PBS (Sangon Biotech, E607008), washed once by ultracentrifugation under the same conditions, and finally resuspended in 100 µL of pre-cooled 1× PBS. The purified exosome suspension was subsequently characterized by transmission electron microscopy (TEM) and nanoparticle tracking analysis (NTA), whereas the remaining suspension was stored at − 80 °C.

Characterization of exosomes

Exosome characterization was performed using transmission electron microscopy (TEM) and nanoparticle tracking analysis (NTA). For TEM analysis, purified exosome suspensions were mounted onto copper grids, negatively stained with uranyl acetate (Henan Xinyue, GZ02625-5), air-dried, and examined using a transmission electron microscope (Hitachi HT-7700) operated at 80 kV. For NTA analysis, cryopreserved exosome samples were thawed, diluted appropriately with 1× PBS, and analyzed using a ZetaVIEW nanoparticle tracking analyzer (Particle Metrix).

RNA isolation

Total RNA was extracted from plasma exosomes using TRIzol® LS Reagent (10296028, Thermo Fisher Scientific) according to the manufacturer’s instructions. Thawed exosome suspensions were lysed with TRIzol LS reagent and supplemented with miRNeasy Serum/Plasma Spike-In Control (219610, QIAGEN). Following chloroform extraction and centrifugation for phase separation, RNA was precipitated with isopropanol (YC-SJ06035, Sinopharm Group) and glycogen (R0551, Thermo Fisher Scientific), washed twice with pre-cooled 75% ethanol, and resuspended in RNase-free water (10977015, Thermo Fisher Scientific). RNA quality was assessed before library preparation, and the remaining RNA was stored at − 80 °C until further analysis.

Library preparation

Small RNA libraries were constructed using the NEBNext® Multiplex Small RNA Library Prep Set for Illumina (NEB, UK, #E7560S) according to the manufacturer’s instructions. Total RNA (100 ng–1 µg) extracted from plasma-derived exosomes was used as input material. Briefly, sequential 3′ and 5′ adaptor ligation, reverse transcription, PCR amplification, purification, and gel-based size selection were performed to generate miRNA libraries (140–160 bp). The purified libraries were resuspended in TE Buffer (NEB, UK) for subsequent quality control and sequencing.

Library quality control and RNA sequencing

Library size distribution and concentration were evaluated using an Agilent 2100 Bioanalyzer. Libraries that met the quality control criteria (expected peak size of approximately 147 bp) were pooled at equimolar concentrations before sequencing on an Illumina NovaSeq 6000 platform using a paired-end 150 bp (PE150) strategy.

Differential expression analysis

The quality of the raw sequencing data, including sequencing accuracy and base composition at each position, was assessed using FastQC (v0.11.8). Next, fastp (v0.19.5) was used to remove adapter sequences and low-quality reads, producing clean reads. Quality control of miRNA sequences was performed using blastn (v2.9.0). The filtered sequences were aligned to the reference genome using bowtie (v1.2.3), and miRNA expression levels (counts) were determined with miRDeep2 (v0.1.3) based on the alignment data. Differential expression analysis was carried out using the R package DESeq2 (v1.10.1), with the selection criteria of |log₂ fold change (FC)| ≥ 1 and P Adjust (after Benjamini-Hochberg correction) < 0.05. MiRanda (v3.3) and RNAhybrid were used to predict target genes for differentially expressed miRNAs, and the intersection of the results from both algorithms was retained for subsequent enrichment analysis. Enrichment analysis of target genes was performed using clusterProfiler (v2.4.2) on Gene Ontology (GO) categories, including biological process (BP), molecular function (MF), and cellular component (CC), as well as Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. Fisher’s exact test was used for enrichment analysis with a threshold of FDR < 0.05. (Benjamini‑Hochberg correction)

Statistical analysis

Descriptive statistics

Data distribution normality was tested using the Shapiro-Wilk test. Normally distributed continuous variables are presented as mean ± standard deviation (SD) and compared between two groups using independent samples t-test. Non-normally distributed data are presented as median with (interquartile range, IQR) and compared with Mann–Whitney U tests. Categorical variables are presented as frequencies (%) and compared using chi-square tests. A two-tailed p < 0.05 was considered statistically significant.

Multiple comparison corrections

For differentially expressed plasma exosomal miRNAs, the Benjamini-Hochberg method was applied to control the false discovery rate (FDR). MiRNAs with adjusted p-values < 0.05 were considered significantly differentially expressed. These significantly differentially expressed miRNAs were subjected to GO functional enrichment and KEGG pathway enrichment analyses, and only pathways with Benjamini-Hochberg adjusted p values < 0.05 were discussed. For demographic and clinical data and peripheral blood inflammatory factors, the Bonferroni correction was applied for multiple comparisons. A corrected p-value < 0.05 was considered statistically significant.

Partial correlation analysis

Partial correlation analysis was performed to assess the associations between 22 peripheral blood inflammatory factors and 18 differentially expressed plasma exosomal miRNAs, with age, sex, body mass index (BMI), diazepam equivalent, and chlorpromazine equivalent as covariates. The p values from the correlation analyses were adjusted for multiple comparisons using the Bonferroni correction. Only correlations with a Bonferroni-corrected p-value < 0.05 were discussed.

Binary logistic regression for discriminant model

To identify key factors discriminating agitated from non-agitated schizophrenia patients, binary logistic regression was performed. The dependent variable was group status. The independent variables included: (1) fixed covariates (Model 1): age, sex (female as reference), BMI, diazepam equivalent, and chlorpromazine equivalent; (2) candidate variables: 22 peripheral blood immune‑inflammatory markers and 18 plasma exosomal miRNAs. Due to large differences in measurement scales, all candidate variables were Z‑score standardized before analysis.

Forward stepwise selection based on the likelihood ratio test was used for variable selection, with an entry criterion of p < 0.05. The fixed covariates were forced into the model. After the stepwise procedure, three candidate variables (PLR, NAR and miR-543) met the entry criterion and were retained in the model. Among them, two variables (NAR and miR-543) remained statistically significant (p < 0.05) in the final model. Since PLR was not statistically significant in the final model, the receiver operating characteristic analyses (ROC) included only NAR and miR-543, together with the fixed covariates. The area under the curve (AUC) with 95% confidence interval (CI) was calculated. All tests were two‑tailed, and a p‑value < 0.05 was considered statistically significant. SPSS version 27.0 was used for all analyses.

Cross‑validation

To assess the generalization ability of the final model, five‑fold cross‑validation was performed on the dataset of 64 samples. The dataset was randomly divided into 5 equal folds. In each iteration, one fold served as the validation set and the remaining four folds as the training set. For the training set, a logistic regression model was built with the two identified markers (NAR and miR-543) together with the fixed covariates (age, sex, BMI, diazepam equivalent, chlorpromazine equivalent) as independent variables, and the group status as the dependent variable. The predicted probability for each sample in the validation set was saved. This process was repeated five times, with each fold used exactly once as the validation set. For each validation set, the AUC was calculated based on the predicted probabilities and the actual group labels. The optimal cutoff was determined by maximizing the Youden index. Sensitivity, specificity, accuracy, and Youden index were then computed. The average AUC across the five folds was reported as the arithmetic mean.

Result

Demographic and clinical characteristics

Demographic and clinical characteristics are shown in Table 1. A total of 64 schizophrenia patients were included in this study, comprising 29 patients without agitation symptoms (SZN) and 35 patients with agitation symptoms (SZA). There were no significant differences in age, gender, BMI, duration of illness, age of onset or chlorpromazine equivalent, or diazepam equivalent between the two groups (all p > 0.05). Analysis revealed significant differences in several clinical scales: the SZA group had significantly higher PANSS-EC scores (t = 12.527, P Adjust < 0.001), PANSS total scores (t = 4.630, P Adjust < 0.001), PANSS positive subscale scores (t = 4.742, P Adjust < 0.001), and PANSS general psychopathology subscale scores (t = 3.304, P Adjust = 0.024) compared to the SZN group. No significant difference was found in the PANSS negative subscale scores between the groups (t = -0.95, P Adjust = 1).

Table 1Demographic and clinical data of SZN and SZA

SZN(n = 29) SZA(n = 35) t/X2/Z P value P Adjust

Age(year) 29.10(8.74) 29.63(8.28) 0.25 0.81b 1

Male (%) 31.0 45.7 1.436 0.23 c 1

BMI (kg/m2) 22.51(4.23) 23.36(4.77) 0.75 0.458 b 1

Duration of illness(year) 7.03(5.84) 8.67(6.73) 1.03 0.31 b 1

Age of Onset(year) 22.17(7.55) 21.00(6.78) 0.65 0.52 b 1

Chlorpromazine Equivalent 340.56(188.74) 344.82(187.69) 0.089 0.93 b 1

Diazepam Equivalent a 0.00(5.00) 4.00(5.00) 0.271 0.787 d 1

PANSS total score 89.45(10.47) 103.20(12.84) 4.63 < 0.001***b < 0.001***

PANSS-EC 8.93(2.70) 17.43(2.71) 12.53 < 0.001***b < 0.001***

PANSS positive score 20.62(5.07) 25.86(3.75) 4.74 < 0.001***b < 0.001***

PANSS Negative Score 26.69(5.53) 25.46(4.89) -0.95 0.35b 1

PANSS General Score 38.07(4.94) 42.97(6.60) 3.30 0.002***b 0.024*

Data are presented as mean (SD) or median (IQR). N is sample size

a non-normal distribution data

b Independent samples T-test

c Chi-square test

d Mann-Whitney U test

* significant at P < 0.05.** significant at P < 0.005.***significant at P < 0.001. Using the Bonferroni correction to adjust P value. (P value multiplied by 12, capped at1.000)

* significant at P Adjust < 0.05. ** significant at P Adjust < 0.005. ***significant at P Adjust < 0.001. Abbreviation: SZN: non-agitated schizophrenia group, SZA: agitated schizophrenia group, BMI: body mass index, PANSS: positive and negative syndrome scale, PANSS-EC: Positive and Negative Syndrome Scale-Excited Component

Isolation and characterization of plasma exosomes

Exosomes were isolated from the plasma of SZA and SZN groups via ultracentrifugation. The isolated exosomes were then characterized by Transmission Electron Microscope (TEM) and Nanoparticle Tracking Analysis (NTA), as presented in Fig. 1. The isolated exosomes exhibited a typical cup-shaped morphology with a visible bilayer membrane under TEM (Fig. 1A). NTA revealed that the exosomes size were approximately 100 nm in both SZN and SZA group (Fig. 1B).

Differential expression of miRNAs in SZA group

Plasma exosomal microRNA sequencing was performed on all participants (SZN = 29, SZA = 35). Following quality control and normalization, microRNAs with low expression levels were filtered out. MiRNA-seq found 1635 exosomal miRNAs (Supplementary Table 1). Differentially expressed miRNAs were defined as |Log2FoldChange| ≥1 and an adjusted p-value < 0.05 using the Benjamini-Hochberg correction. A total of 18 differentially expressed miRNAs were identified, including 15 upregulated and 3 downregulated miRNAs (Table 2).

Table 2Differentially expressed microRNAs in SZA group

Differentially expressed miRNAs Log2 Fold Change P Adjust Differentially expressed miRNAs Log2 Fold Change P Adjust

hsa-miR-3198 4.85 0.0047 hsa-miR-874-3p 1.73 0.0181

hsa-miR-10400-5p 4.03 0.0416 hsa-miR-200b-3p 1.71 0.0153

hsa-miR-6515-5p 3.96 4.0148 × 10− 6 hsa-miR-4508 1.58 0.0324

hsa-miR-6741-5p 3.60 1.1541 × 10− 5 hsa-miR-17-3p 1.57 0.0495

hsa-miR-6734-5p 3.47 1.4547 × 10− 6 hsa-miR-2110 1.52 0.0047

hsa-miR-3620-5p 3.13 0.0181 hsa-miR-1292-5p 1.46 0.0153

hsa-miR-671-5p 3.07 0.0047 hsa-miR-4488 -4.34 8.7932 × 10− 10

hsa-miR-296-3p 2.72 0.0153 hsa-miR-9-5p -2.85 0.0002

hsa-miR-4665-5p 2.08 0.0194 hsa-miR-543 -1.14 0.0495

Significant differential expressed miRNAs were identified in the SZA group (n = 35) compared to SZN (n = 29) group. The miRNAs listed all have an adjusted P value < 0.05 (using Benjamini-Hochberg correction)

Abbreviation: SZN: non-agitated schizophrenia group, SZA: agitated schizophrenia group, miRNAs: microRNAs

Target gene prediction and functional enrichment analysis of differentially expressed miRNAs

The predicted target genes of the 18 differentially expressed miRNAs were subjected to GO and KEGG pathway enrichment analyses (both after Benjamini-Hochberg correction). Figure 2 shows the most enriched GO and KEGG elements. The results of the other pathway enrichment analyses can be found in Supplementary Material 2.

For the upregulated group, biological processes were enriched in modulation of chemical synaptic transmission, regulation of trans-synaptic signaling, synapse organization, axonogenesis and gland development. For the cellular component, these target genes were predominantly localized to neuronal cell body, vacuolar membrane, synaptic membrane, cell leading edge and endosome (recycling and early). Molecular functions were significantly enriched in phospholipid binding, GTPase regulator activity, nucleoside-triphosphatase regulator activity, actin binding and phosphatidylinositol binding.

In contrast, for the downregulated group, GO analysis focused on synapse and exocytosis. Biological processes were significantly associated with modulation of chemical synaptic transmission, regulation of transsynaptic signaling, synapse organization and exocytosis. Cellular components were chiefly localized to synapse, including the glutamatergic synapse, neuron to neuron synapse, asymmetric synapse, postsynaptic density, and postsynaptic specialization.

KEGG pathway analysis further elucidated their functional divergence. Target genes of the upregulated group were significantly involved in the MAPK signaling pathway, Ras signaling pathway, proteoglycans in cancer and efferocytosis. Conversely, target genes of the downregulated group were only enriched in glutamatergic synapse.

Peripheral immune inflammatory characteristics in the SZA group

Peripheral blood inflammatory factors are presented in Table 3. A total of 64 schizophrenia patients were included in the analysis, comprising 29 patients without agitation symptoms (SZN) and 35 patients with agitation symptoms (SZA).

In uncorrected analysis, the SZA group showed a trend toward higher NAR levels (Z = 2.81, p = 0.004) compared with the SZN group, but this difference did not pass Bonferroni correction (adjusted p = 0.088, Fig. 3), suggesting only a statistical trend. No significant differences were found between the two groups in any other peripheral blood markers (all adjusted p > 0.05), including neutrophil counts, interleukins (IL-2, IL-4, IL-5, IL-6, IL-8, IL-10, IL-17, IL-1β, IL-12p70), interferons (IFN-α, IFN-γ), tumor necrosis factor‑α (TNF-α), lymphocytes, monocytes, platelets, albumin, NLR, NMR, PLR, and MLR.

Table 3Comparison of peripheral blood inflammatory factors between SZA and SZN group

SZN(n = 29) SZA(n = 35) t/Z P value P Adjust

IL-2(pg/mL) 1.99(0.52) 1.79(0.43) -1.47 0.149a 1

IL-4(pg/mL) 1.86(0.17) 1.82(0.17) -0.86 0.396a 1

IL-5(pg/mL) 1.96(0.44) 1.89(0.30) -0.64 0.524a 1

IL-6(pg/mL) 3.50(2.66) 4.26(2.68) 0.96 0.344a 1

IL-8(pg/mL) 8.14(11.33) 7.21(8.69) -0.32 0.754a 1

IL-10(pg/mL) 1.45(0.38) 1.35(0.30) -1.00 0.323a 1

IL-17(pg/mL) 2.81(1.45) 2.40(0.59) 0.474 0.635b 1

IL-1β(pg/mL) 3.11(1.37) 2.91(1.45) -0.46 0.649a 1

IL-12p70(pg/mL) 1.69(0.19) 1.81(0.62) 0.80 0.429a 1

IFN-α(pg/mL) 1.53(0.40) 1.48(0.32) -0.43 0.673a 1

IFN-γ(pg/mL) 3.93(6.03) 2.90(1.96) -0.81 0.422a 1

TNF-α(pg/mL) 2.37(0.31) 2.46(0.72) 0.51 0.609a 1

Neutrophils(109/L) 3.49(0.95) 4.67(1.84) 2.778 0.005b 0.110

Lymphocytes(109/L) 1.86(0.57) 2.10(0.79) 1.36 0.178a 1

Monocytes(109/L) 0.47(0.16) 0.50(0.15) 0.71 0.483a 1

Platelets(109/L) 235.55(59.91) 251.37(66.18) 0.99 0.324a 1

Albumin(g/L) 40.80(3.34) 40.63(2.62) -0.24 0.815a 1

NLR 2.22(1.79) 2.88(3.06) 1.02 0.31a 1

NMR 7.77(2.34) 10.08(5.48) 1.578 0.115b 1

PLR 137.00(48.71) 138.96(75.01) 0.843 0.399b 1

NAR 0.09(0.02) 0.12(0.05) 2.81 0.004b 0.088

MLR 0.28(0.14) 0.27(0.16) -0.19 0.848a 1

Data are presented as means (standard deviation). N is sample size

a Independent samples T-test

b Mann-Whitney U test

P Adjust: Using the Bonferroni correction to adjust P value (P value multiplied by 22, capped at1.000)

Abbreviation: SZN: non-agitated schizophrenia group; SZA: agitated schizophrenia group; IL: interleukin, IFN: interferon; TNF: tumor necrosis factor, NLR: neutrophil to lymphocyte ratio, NMR: neutrophil to monocyte ratio, PLR: platelet to lymphocyte ratio, NAR: neutrophil to albumin ratio, MLR: monocyte to lymphocyte ratio

Correlation of peripheral inflammatory factors and differentially expressed miRNAs

Correlation of peripheral inflammatory factors and differentially expressed miRNAs was presented in Fig. 4. After Bonferroni‑corrected partial correlation analyses adjusting for age, sex, BMI, chlorpromazine equivalent, and diazepam equivalent, no significant correlations between 18 differentially expressed miRNAs and peripheral inflammatory factors were observed in SZN group. However, in SZA group, IL‑12p70 showed strong positive correlations with miR‑3620‑5p (Fig. 4A, r = 0.893, p Adjust < 0.001) and miR-17-3p (Fig. 4B, r = 0.768, p Adjust = 0.028). No other peripheral inflammatory factors exhibited a significant correlation with any differentially expressed miRNAs in SZA group. The heatmap also included correlations among inflammatory factors or among differentially expressed miRNAs themselves, which are not described here as they are beyond the focus of this study (see Supplementary Fig. 1).

Predictive value of plasma exosomal miRNAs and peripheral inflammatory markers

A two-step approach was used to build the multivariable logistic regression model for agitation in patients with schizophrenia. First, forward stepwise logistic regression was performed among 18 differentially expressed miRNAs and 22 peripheral inflammatory factors, together with fixed covariates (age, sex, BMI, diazepam equivalent, and chlorpromazine equivalent), yielding PLR, NAR, and miR-543 as candidate variables. These candidates were then entered into the final multivariable logistic regression model together with fixed variables (Table 4). All continuous biomarkers were standardized to Z-scores.

All variance inflation factor (VIF) was below 1.5, indicating no serious multicollinearity. The model demonstrated good calibration (Hosmer–Lemeshow χ² = 4.433, df = 8, p = 0.816) and explained 42.5% of the variance in agitation status (Nagelkerke R² = 0.425). Among the biomarkers, NAR was significantly associated with increased odds of agitation (OR per SD increase = 3.778, 95% CI 1.459–9.781, Wald χ² = 7.501, p = 0.006). In contrast, miR-543 was significantly associated with reduced odds of agitation (OR per SD increase = 0.162, 95% CI: 0.032–0.806, Wald χ² = 4.943, p = 0.026). None of the other variables, including age, sex, medication doses, BMI, or PLR, reached statistical significance (all p > 0.05). These findings indicate that higher NAR levels and lower plasma exosomal miR-543 levels were independently associated with agitation in this sample.

We further built logistic regression models incorporating the fixed variables (age, sex, BMI, diazepam equivalent, chlorpromazine equivalent) together with each of the three biomarkers individually or in combination to discriminate between the SZA and SZN groups. The combined model including NAR and miR‑543 together with the fixed variables yielded an AUC of 0.812, while each single biomarker showed AUC of 0.752 for NAR and 0.715 for miR‑543 (Fig. 5).

Table 4Logistic regression models for schizophrenia patients with agitation

Variable B (SE) Wald χ² OR (95% CI) p value VIF

Fixed variables

 Age(years) 0.028(0.04) 0.488 1.028(0.951,1.112) 0.485 1.202

 Male 1.194(0.721) 2.742 3.300(0.803,13.557) 0.098 1.312

 Chlorpromazine Equivalent 0.000(0.002) 0.031 1.000(0.997,1.004) 0.861 1.254

 Diazepam Equivalent 0.125(0.113) 1.225 1.133(0.908,1.412) 0.268 1.387

 BMI, kg/m² -0.107(0.08) 1.801 0.899(0.769,1.050) 0.180 1.355

Z-score standardized biomarkers

 PLR (per SD) -0.522(0.424) 1.513 0.594(0.258,1.363) 0.219 1.340

 NAR (per SD) 1.329(0.424) 7.501 3.778(1.459,9.781) 0.006 1.358

 miR-543 (per SD) -1.822(0.82) 4.943 0.162(0.032,0.806) 0.026 1.117

Model summary

Nagelkerke R² 0.425

Hosmer–Lemeshow χ², p value (df = 8) 4.433,0.816

The model was adjusted for age, sex, BMI, diazepam equivalent, and chlorpromazine equivalent. All continuous biomarkers were standardized to Z‑scores (per SD increase)

Model summary: Nagelkerke R² = 0.425; Hosmer–Lemeshow test χ² (df = 8) = 4.433, p = 0.816

Abbreviation: SE: standard errors, OR: odds ratio, CI: confidence interval, VIF: variance inflation factor, PLR: platelet to lymphocyte ratio, NAR: neutrophil to albumin ratio, BMI: body mass index

Validation of the performance of the NAR and miR-543 model

We examined variations in the performance of the two biomarkers and performed five-fold cross-validation, in which we iterated the following process five times: a logistic model was built using 80% of the participants, and its performance was assessed using the remaining 20%. The average AUC was 0.795 (95% CI 0.675–0.916). The sensitivity, specificity, and accuracy were 80.1 ± 16.5% (mean ± SD), 79.5 ± 26.1%, and 77.8 ± 8.9%, respectively, and the Youden index was 0.596 ± 0.117.

Discussion

To our knowledge, this study is the first to investigate the plasma exosomal miRNA expression profiles and peripheral inflammatory factors between schizophrenia patients with agitation and patients without agitation. The key findings include that: (1) 18 differentially expressed miRNAs (15 upregulated and 3 downregulated) were identified in schizophrenia patients with agitation, with target genes enriched in pathways related to synaptic regulation and inflammation-associated signaling; (2) patients with agitation showed a trend toward elevated NAR levels, while no significant differences in other inflammatory factors were observed between patients with agitation and patients without agitation; (3) ROC analysis confirmed that the combination of NAR and plasma exosomal miR-543 showed moderate discriminative performance for agitated symptoms in schizophrenia. The diagnostic efficiency still remained stable after five-fold cross-validation. These findings provide preliminary evidence regarding plasma exosomal miRNA profiles in schizophrenia patients with agitation, and together with the higher peripheral NAR, suggest a potential multi-level biological pattern. This may offer new insights into the post‑transcriptional regulation and inflammation‑related pathways underlying agitation symptoms.

Pathway enrichment analysis of predicted target genes from the 18 differentially expressed miRNAs revealed distinct functional patterns that may contribute to the neurobiology of agitation in schizophrenia. Because pathway enrichment analysis is exploratory in nature, we focused our interpretation on pathways most closely related to inflammatory dysregulation and synaptic dysfunction, which were the primary biological processes underlying our study hypothesis. Targets of upregulated miRNAs showed GO enrichment primarily in intracellular signaling cascades and cytoskeletal remodeling, with KEGG pathway enrichment prominently in the MAPK signaling pathway and Ras signaling pathway. Although direct evidence in schizophrenia patients with agitation remains limited, previous studies have implicated the Ras-MAPK pathway in aggressive behavior across different species [43]. This interpretation is also consistent with reports that Ras-MAPK pathway alterations are associated with irritability-related phenotypes in RASopathies [44], despite not directly addressing agitation in schizophrenia. The Ras-MAPK signaling pathway is involved in various biological processes including synaptic development, dendritic spine remodeling, neuronal survival, and intracellular signal transduction. Studies using cellular and animal models suggest that the Ras-MAPK pathway is involved in regulating synaptic structure [45]. Postmortem brain tissue analyses of patients with schizophrenia also indicated that dysregulation of Ras-MAPK signaling pathway can lead to synaptic structural abnormalities and reduced neural plasticity [46, 47].Combined with the GO analysis results from our study, the upregulated miRNA targets were significantly associated with synaptic structural component, including the synaptic membrane, distal axon, and neuronal cell body, as well as molecular functions such as actin binding and phospholipid binding. This finding may suggest alterations in synaptic regulation associated with agitation symptoms in schizophrenia. In addition to synaptic structure, the Ras-MAPK pathway may also be involved in inflammatory processes. The Ras-MAPK signaling pathway can regulate the expression of proinflammatory factors (IL-1β, IL-6, TNF-α), and has been implicated in neuroinflammatory processes within the central nervous system [48]. This subsequently activates microglia, causing dysfunction in neural circuits in the amygdala. As the core brain region for emotional regulation, the dysfunction of neural circuits leads to impaired control over emotions and behavior, increasing the risk of aggressive behavior [49]. Besides the Ras-MAPK signaling pathway, upregulated miRNAs were also enriched in cholinergic synapses. Consistent with prior findings, positron emission tomography (PET) and single photon emission computed tomography (SPECT) scans of SZ patients have revealed that lower expression of β2 nicotinic acetylcholine receptors (nAChRs) in the striatum correlated with more severe agitation symptoms [50]. Meanwhile, cholinergic synapse agonists, such as nicotine, effectively reduce agitation scores when used via nicotine patches to acute-phase schizophrenia patients. Upon activation by nicotine, cholinergic synapses negatively regulate striatal dopamine release via M1/M4 muscarinic acetylcholine receptor (mAChR), thereby relieving agitation symptoms [51]. These findings suggest that cholinergic signaling may be involved in the modulation of agitation. Furthermore, the enrichment in efferocytosis-related pathways suggests that impaired clearance of apoptotic material may sustain inflammatory signaling while simultaneously intersecting with cytoskeletal reorganization. Pathway enrichment analysis of the predicted target genes of downregulated miRNAs also showed notable dysregulation in the glutamatergic synaptic pathway. Similarly, autopsy and imaging studies revealed dysregulation of glutamatergic synapses across multiple brain regions in schizophrenia patients. This dysregulation included increased glutamatergic synaptic density in the striatum and decreased density in the prefrontal cortex, leading to impaired control over emotions and behavior [12, 52]. Moreover, evidence indicated that inflammatory signaling pathways, like Ras-MAPK and efferocytosis signaling pathways, can modulate synaptic plasticity and neurotransmitter systems, suggesting a potential link between immune activation and synaptic dysfunction [53]. These findings suggest that alterations in synaptic regulation and inflammation-associated signaling may be associated with agitation symptoms in schizophrenia. It should be noted that since this study performed pathway analysis based on plasma exosomal miRNAs, the results cannot yet clearly reflect the specific brain region localization of these alterations within the central nervous system.

The mildly elevated NAR in the SZA group may indicate subtle alterations in peripheral immune status. However, this finding should be interpreted cautiously because this difference did not remain significant after strict multiple-comparison correction. Previous studies have reported elevated NAR in patients with schizophrenia across different disease stages, suggesting that this marker may reflect relatively stable changes in peripheral immune-related processes [54]. In contrast, circulating cytokines are more dynamic inflammatory mediators and can be influenced by disease stage, antipsychotic treatment, circadian variation, and other biological factors [24, 25]. Although several pro-inflammatory cytokines, including IL-1β, IL-6, and IFN-γ, have been reported to increase during the acute phase of schizophrenia [55, 56], our sample mainly consisted of patients with chronic illness and a history of relapse (mean disease duration: 8.67 years), and antipsychotic exposure may further attenuate peripheral cytokine alterations [55, 57]. Therefore, the absence of significant cytokine differences between the SZA and SZN groups does not necessarily exclude subtle immune-related alterations. Nevertheless, larger longitudinal studies are required to further clarify the relationship between peripheral immune markers and agitation symptoms in schizophrenia.

Correlation analysis indicated that the relationship between peripheral inflammatory activity and differentially expressed plasma exosomal miRNAs was inconsistent across the groups. After Bonferroni correction and adjustment for gender, age, BMI, chlorpromazine equivalent and diazepam equivalent, no significant associations were observed between the 18 differentially expressed miRNAs and peripheral inflammatory factors in the SZN group. In contrast, in the SZA group, IL-12p70 showed a significant positive correlation with miR-3620-5p and miR-17-3p, suggesting a possible association between plasma exosomal miRNA alterations and the IL-12-related immune axis in the SZA group. As a key driver of Th1 immune polarization, IL-12p70 promotes IFN-γ secretion and enhances cytotoxic immune responses [58, 59]. The correlation between IL-12p70 and miR-17-3p is particularly noteworthy. MiR-17-3p originates from the miR-17-92 gene cluster, which is known to regulate cell proliferation and inflammation regulation [60, 61]. The correlation between miR-17-3p and IL-12p70 may reflect either a shared upstream inflammatory response or a preliminary regulatory relationship that warrants further experimental validation. At present, the functional role of miR-17-3p has been demonstrated only in cell and mouse studies, and there are no schizophrenia animal models available. Further experiments are needed to investigate the regulatory relationship between miR-17-3p and IL-12p70. The function of miR-3620-5p remains unclear. But the correlation coefficient between miR-3620-5p and IL-12p70 is as high as 0.893, suggesting that it may be a candidate molecule worthy of further study in agitation-related biology. Nevertheless, further research is needed to identify its target genes and functional pathways. It is worth noting that, apart from IL-12p70, no other inflammatory factors showed significant correlations with the differentially expressed miRNAs in the SZA group, which suggests a potentially selective association. The heatmaps also visualize the correlations among inflammatory factors and among miRNAs, which may provide clues for future mechanistic studies. Future studies involving larger sample sizes of schizophrenia patients with agitation are needed, incorporating longitudinal follow-up and in vitro functional experiments, to verify whether the miR-3620-5p/miR-17-3p–IL-12p70 axis is involved in the pathogenesis of agitation.

Using a forward stepwise logistic regression approach, we first screened 18 differentially expressed miRNAs and 22 peripheral inflammatory factors, and PLR, NAR, and miR‑543 were retained as the selected variables. These three biomarkers were then entered into a multivariable logistic regression model together with fixed variables (age, sex, BMI, diazepam equivalent, and chlorpromazine equivalent). In this adjusted model, NAR was significantly associated with increased odds of agitation, while miR‑543 was associated with reduced odds. However, PLR and all fixed variables did not reach statistical significance. The model demonstrated good calibration and explained 42.5% of the variance in agitation status. The combined model incorporating NAR, and miR‑543 together with fixed variables yielded an AUC of 0.812, showing improved discriminative performance compared with the single-marker models (NAR AUC = 0.752, miR‑543 AUC = 0.715). Five-fold cross-validation suggested that the predictive performance of the model remained relatively stable. The average AUC of 0.795 (95% CI 0.675–0.916) indicated that the model still retained moderate predictive performance. MiR-543 has been experimentally shown to regulate the inflammation, which can suppress the expression of Kruppel-like factor 6(KLF6), blocking the transcriptional expression of its downstream proinflammatory factors (TNF-α, IL-1β, IL-6, etc.) [62]. As discussed above, elevated NAR levels indicate the presence of a dysregulation in inflammation [63]. Downregulation of miR-543 together with the trend toward higher NAR may reflect subtle immune dysregulation in the SZA group. In summary, through a systematic forward stepwise screening and multivariable modeling strategy, plasma exosomal miR‑543 and peripheral inflammatory factor NAR emerge as potentially useful and easily accessible peripheral biomarkers associated with agitation in schizophrenia.

Limitations

This study had some limitations. First, the sample size was relatively small. Although we additionally used alternative statistical methods (the Bonferroni correction and the Benjamini–Hochberg correction) to obtain reliable and accurate results. A retrospective power analysis showed that among the miRNAs identified through correction, the miRNA with the smallest absolute log₂ fold change had a Hedges’ g of 0.357. Given the current sample size and a two-sided α = 0.05, the statistical power to detect this effect was only 29.5%. Sensitivity analysis further showed that an effect size of d = 0.715 is required to achieve 80% power. Therefore, although this study has sufficient statistical power to detect moderate-to-large miRNAs differences, it lacks sufficient power to detect more subtle but potentially biologically significant changes. Secondly, although five-fold cross-validation was performed, there is still a lack of an independent external validation cohort. Future studies with larger sample sizes and independent external validation are needed to confirm robustness and clinical value of plasma exosomal miRNAs. Third, although exosomes are capable of crossing the BBB, plasma exosomal miRNAs still cannot fully reflect the central inflammatory activity or identify the specific brain regions involved. Future research should explore the analysis of exosomes derived from specific cell sources, such as neuron-derived or glial cell-derived exosomes. Fourth, although patients with tobacco or alcohol use were excluded and statistical adjustment was performed for BMI, chlorpromazine equivalent, and diazepam equivalent, residual confounding related to medication exposure cannot be completely ruled out. Because patients with agitation frequently experience poor insight, altered consciousness, or reduced compliance during acute episodes, reliable assessment of pre-admission medication history was difficult. Therefore, future studies involving first-episode drug-free patients are warranted. Fifth, although antipsychotic medications were statistically analyzed using chlorpromazine equivalent for both groups, the long-term effects of antipsychotic medication on plasma exosomal miRNAs and peripheral inflammatory factors cannot be ruled out. Sixth, this study focused exclusively on acute agitation assessed within 24 h of hospital admission. Due to the cross‑sectional design, we cannot determine whether the observed molecular alterations are specific to the acute agitated state or represent a persistent trait. Longitudinal follow‑up studies are needed to assess whether these changes normalize after symptom resolution or remain stable over time. Finally, agitation in schizophrenia frequently co-occurs with more severe positive symptoms, making it difficult to fully disentangle specific molecular alterations from those associated with broader psychopathological severity. Therefore, the identified miRNAs may reflect overlapping biological mechanisms related to heightened psychotic symptom burden rather than a completely independent molecular signature specific to agitation symptoms.

Conclusion

In summary, the findings indicate that schizophrenia patients with agitation symptoms exhibit alterations in plasma exosomal miRNA expression profiles, providing novel insights into the potential involvement of synaptic dysfunction and inflammation-associated signaling in the pathophysiology of agitation. Compared with the SZN group, the SZA group exhibited 18 differentially expressed plasma exosomal miRNAs. Functional enrichment analyses indicated that the predicted target genes of upregulated miRNAs were mainly enriched in the Ras-MAPK signaling pathway, efferocytosis, and cholinergic synapse-related pathways, whereas the predicted target genes of downregulated miRNAs were primarily enriched in glutamatergic synapse-related pathways. From the perspective of post-transcriptional gene regulation, these alterations may suggest that schizophrenia patients with agitation have changes in synaptic dysregulation and inflammation-associated signaling. Concurrently, the observed trend toward increased NAR further supported this interpretation, although these findings were preliminary. Furthermore, a diagnostic model combining plasma exosomal miR-543 and NAR showed moderate and stable differentiation efficacy for agitation symptoms in SZ patients, with a comprehensive AUC of 0.812 and with a 5-fold cross-validation mean AUC of 0.795, suggesting the potential of plasma exosomal microRNAs as candidate non-invasive biomarkers for agitation assessment. Future studies should include larger independent cohorts for external validation and should clarify the regulatory mechanisms of differentially expressed miRNAs through cell and animal experiments, in order to explore the pathophysiological processes underlying agitated symptoms in schizophrenia.

Supplementary Information

Below is the link to the electronic supplementary material.

References

1.Wallwork RS Fortgang R Hashimoto R Weinberger DR Dickinson D Searching for a consensus five-factor model of the Positive and Negative Syndrome Scale for schizophrenia Schizophr Res 2012 137 1–3 246 50 10.1016/j.schres.2012.01.031 22356801 PMC3351536 Wallwork RS, Fortgang R, Hashimoto R, Weinberger DR, Dickinson D. Searching for a consensus five-factor model of the Positive and Negative Syndrome Scale for schizophrenia. Schizophr Res. 2012;137(1–3):246–50.10.1016/j.schres.2012.01.031 PMC3351536 22356801

2.Citrome L Agitation in schizophrenia: origins and evidence-based treatment Curr Opin Psychiatry 2021 34 3 216 21 10.1097/YCO.0000000000000685 33395103 Citrome L. Agitation in schizophrenia: origins and evidence-based treatment. Curr Opin Psychiatry. 2021;34(3):216–21.10.1097/YCO.0000000000000685 33395103

3.Piechniczek-Buczek J Psychiatric emergencies in the elderly population Emerg Med Clin North Am 2006 24 2 467 90 viii10.1016/j.emc.2006.01.008 16584967 Piechniczek-Buczek J. Psychiatric emergencies in the elderly population. Emerg Med Clin North Am. 2006;24(2):467–90. viii.10.1016/j.emc.2006.01.008 16584967

4.Zhou JS Zhong BL Xiang YT Chen Q Cao XL Correll CU Ungvari GS Chiu HF Lai KY Wang XP Prevalence of aggression in hospitalized patients with schizophrenia in China: A meta-analysis Asia Pac Psychiatry 2016 8 1 60 9 10.1111/appy.12209 26346165 Zhou JS, Zhong BL, Xiang YT, Chen Q, Cao XL, Correll CU, Ungvari GS, Chiu HF, Lai KY, Wang XP. Prevalence of aggression in hospitalized patients with schizophrenia in China: A meta-analysis. Asia Pac Psychiatry. 2016;8(1):60–9.10.1111/appy.12209 26346165

5.Bak M Weltens I Bervoets C De Fruyt J Samochowiec J Fiorillo A Sampogna G Bienkowski P Preuss WU Misiak B The pharmacological management of agitated and aggressive behaviour: A systematic review and meta-analysis Eur Psychiatry 2019 57 78 100 10.1016/j.eurpsy.2019.01.014 30721802 Bak M, Weltens I, Bervoets C, De Fruyt J, Samochowiec J, Fiorillo A, Sampogna G, Bienkowski P, Preuss WU, Misiak B, et al. The pharmacological management of agitated and aggressive behaviour: A systematic review and meta-analysis. Eur Psychiatry. 2019;57:78–100.10.1016/j.eurpsy.2019.01.014 30721802

6.Darmedru C Demily C Franck N Cognitive remediation and social cognitive training for violence in schizophrenia: a systematic review Psychiatry Res 2017 251 266 74 10.1016/j.psychres.2016.12.062 28219026 Darmedru C, Demily C, Franck N. Cognitive remediation and social cognitive training for violence in schizophrenia: a systematic review. Psychiatry Res. 2017;251:266–74.10.1016/j.psychres.2016.12.062 28219026

7.Sun Y Jiang W Yu H Zhang J Zhou Y Yin F Su H Jia Y Construction and verification of aggressive behavior risk prediction model in stable patients with schizophrenia BMC Psychiatry 2023 23 1 800 10.1186/s12888-023-05296-5 37919744 PMC10621096 Sun Y, Jiang W, Yu H, Zhang J, Zhou Y, Yin F, Su H, Jia Y. Construction and verification of aggressive behavior risk prediction model in stable patients with schizophrenia. BMC Psychiatry. 2023;23(1):800.10.1186/s12888-023-05296-5 PMC10621096 37919744

8.Sagayadevan V Satghare P Jeyagurunathan A Koh YS Shafie S Chang S Samari E Subramaniam M Mediating effect of symptom severity on the relationship between aggression, impulsivity and quality of life outcomes among patients with schizophrenia and related psychoses Front Psychiatry 2023 14 1154083 10.3389/fpsyt.2023.1154083 37810606 PMC10556254 Sagayadevan V, Satghare P, Jeyagurunathan A, Koh YS, Shafie S, Chang S, Samari E, Subramaniam M. Mediating effect of symptom severity on the relationship between aggression, impulsivity and quality of life outcomes among patients with schizophrenia and related psychoses. Front Psychiatry. 2023;14:1154083.10.3389/fpsyt.2023.1154083 PMC10556254 37810606

9.de Bartolomeis A, Vellucci L, De Simone G, Mazza B, Barone A, Ciccarelli M. Dysregulated signaling at postsynaptic density: a systematic review and translational appraisal for the pathophysiology, clinics, and antipsychotics’ treatment of schizophrenia. Cells. 2023;12(4).10.3390/cells12040574 PMC9954794 36831241

10.Osimo EF Beck K Reis Marques T Howes OD Synaptic loss in schizophrenia: a meta-analysis and systematic review of synaptic protein and mRNA measures Mol Psychiatry 2019 24 4 549 61 10.1038/s41380-018-0041-5 29511299 PMC6004314 Osimo EF, Beck K, Reis Marques T, Howes OD. Synaptic loss in schizophrenia: a meta-analysis and systematic review of synaptic protein and mRNA measures. Mol Psychiatry. 2019;24(4):549–61.10.1038/s41380-018-0041-5 PMC6004314 29511299

11.Li W Lv L Luo XJ In vivo study sheds new light on the dendritic spine pathology hypothesis of schizophrenia Mol Psychiatry 2022 27 4 1866 8 10.1038/s41380-022-01449-2 35079121 Li W, Lv L, Luo XJ. In vivo study sheds new light on the dendritic spine pathology hypothesis of schizophrenia. Mol Psychiatry. 2022;27(4):1866–8.10.1038/s41380-022-01449-2 35079121

12.Kruse AO Bustillo JR Glutamatergic dysfunction in Schizophrenia Transl Psychiatry 2022 12 1 500 10.1038/s41398-022-02253-w 36463316 PMC9719533 Kruse AO, Bustillo JR. Glutamatergic dysfunction in Schizophrenia. Transl Psychiatry. 2022;12(1):500.10.1038/s41398-022-02253-w PMC9719533 36463316

13.Huang Y Jiang H Zheng Q Fok AHK Li X Lau CG Lai CSW Environmental enrichment or selective activation of parvalbumin-expressing interneurons ameliorates synaptic and behavioral deficits in animal models with schizophrenia-like behaviors during adolescence Mol Psychiatry 2021 26 6 2533 52 10.1038/s41380-020-01005-w 33473150 Huang Y, Jiang H, Zheng Q, Fok AHK, Li X, Lau CG, Lai CSW. Environmental enrichment or selective activation of parvalbumin-expressing interneurons ameliorates synaptic and behavioral deficits in animal models with schizophrenia-like behaviors during adolescence. Mol Psychiatry. 2021;26(6):2533–52.10.1038/s41380-020-01005-w 33473150

14.Dong Y Tang Y Li Y Cao P Xu G Zhu R Li R Sui Y Role of peripheral cytokines and orbitofrontal cortex subregion structure in schizophrenia agitation Sci Rep 2025 15 1 14125 10.1038/s41598-025-99033-5 40269239 PMC12019167 Dong Y, Tang Y, Li Y, Cao P, Xu G, Zhu R, Li R, Sui Y. Role of peripheral cytokines and orbitofrontal cortex subregion structure in schizophrenia agitation. Sci Rep. 2025;15(1):14125.10.1038/s41598-025-99033-5 PMC12019167 40269239

15.Wang X Chen W Gou M Li W Li N Tong J Zhou Y Xie T Yu T Feng W Relationship between plasma TNF-α levels and agitation symptoms in first episode patients with schizophrenia BMC Psychiatry 2024 24 1 480 10.1186/s12888-024-05796-y 38956509 PMC11218378 Wang X, Chen W, Gou M, Li W, Li N, Tong J, Zhou Y, Xie T, Yu T, Feng W, et al. Relationship between plasma TNF-α levels and agitation symptoms in first episode patients with schizophrenia. BMC Psychiatry. 2024;24(1):480.10.1186/s12888-024-05796-y PMC11218378 38956509

16.Barzilay R Lobel T Krivoy A Shlosberg D Weizman A Katz N Elevated C-reactive protein levels in schizophrenia inpatients is associated with aggressive behavior Eur Psychiatry 2016 31 8 12 10.1016/j.eurpsy.2015.09.461 26657596 Barzilay R, Lobel T, Krivoy A, Shlosberg D, Weizman A, Katz N. Elevated C-reactive protein levels in schizophrenia inpatients is associated with aggressive behavior. Eur Psychiatry. 2016;31:8–12.10.1016/j.eurpsy.2015.09.461 26657596

17.Müller N Inflammation in Schizophrenia: Pathogenetic Aspects and Therapeutic Considerations Schizophr Bull 2018 44 5 973 82 10.1093/schbul/sby024 29648618 PMC6101562 Müller N. Inflammation in Schizophrenia: Pathogenetic Aspects and Therapeutic Considerations. Schizophr Bull. 2018;44(5):973–82.10.1093/schbul/sby024 PMC6101562 29648618

18.Sellgren CM Gracias J Watmuff B Biag JD Thanos JM Whittredge PB Fu T Worringer K Brown HE Wang J Increased synapse elimination by microglia in schizophrenia patient-derived models of synaptic pruning Nat Neurosci 2019 22 3 374 85 10.1038/s41593-018-0334-7 30718903 PMC6410571 Sellgren CM, Gracias J, Watmuff B, Biag JD, Thanos JM, Whittredge PB, Fu T, Worringer K, Brown HE, Wang J, et al. Increased synapse elimination by microglia in schizophrenia patient-derived models of synaptic pruning. Nat Neurosci. 2019;22(3):374–85.10.1038/s41593-018-0334-7 PMC6410571 30718903

19.Chen X He L Zhang Q Luo Y Cao Y Zhu H Chen W Chemogenetic Restoration of the Excitation-Inhibition Imbalance in the Medial Prefrontal Cortex Ameliorates Schizophrenia-Like Sensory Gating Deficits Acta Physiol (Oxf) 2026 242 2 e70151 10.1111/apha.70151 41489040 Chen X, He L, Zhang Q, Luo Y, Cao Y, Zhu H, Chen W. Chemogenetic Restoration of the Excitation-Inhibition Imbalance in the Medial Prefrontal Cortex Ameliorates Schizophrenia-Like Sensory Gating Deficits. Acta Physiol (Oxf). 2026;242(2):e70151.10.1111/apha.70151 41489040

20.Lányi O Koleszár B Schulze Wenning A Balogh D Engh MA Horváth AA Fehérvari P Hegyi P Molnár Z Unoka Z Excitation/inhibition imbalance in schizophrenia: a meta-analysis of inhibitory and excitatory TMS-EMG paradigms Schizophrenia (Heidelb) 2024 10 1 56 10.1038/s41537-024-00476-y 38879590 PMC11180212 Lányi O, Koleszár B, Schulze Wenning A, Balogh D, Engh MA, Horváth AA, Fehérvari P, Hegyi P, Molnár Z, Unoka Z, et al. Excitation/inhibition imbalance in schizophrenia: a meta-analysis of inhibitory and excitatory TMS-EMG paradigms. Schizophrenia (Heidelb). 2024;10(1):56.10.1038/s41537-024-00476-y PMC11180212 38879590

21.Shukla DK Wijtenburg SA Chen H Chiappelli JJ Kochunov P Hong LE Rowland LM Anterior Cingulate Glutamate and GABA Associations on Functional Connectivity in Schizophrenia Schizophr Bull 2019 45 3 647 58 10.1093/schbul/sby075 29912445 PMC6483591 Shukla DK, Wijtenburg SA, Chen H, Chiappelli JJ, Kochunov P, Hong LE, Rowland LM. Anterior Cingulate Glutamate and GABA Associations on Functional Connectivity in Schizophrenia. Schizophr Bull. 2019;45(3):647–58.10.1093/schbul/sby075 PMC6483591 29912445

22.Gigase FAJ Smith E Collins B Moore K Snijders G Katz D Bergink V Perez-Rodriquez MM De Witte LD The association between inflammatory markers in blood and cerebrospinal fluid: a systematic review and meta-analysis Mol Psychiatry 2023 28 4 1502 15 10.1038/s41380-023-01976-6 37055513 PMC10266485 Gigase FAJ, Smith E, Collins B, Moore K, Snijders G, Katz D, Bergink V, Perez-Rodriquez MM, De Witte LD. The association between inflammatory markers in blood and cerebrospinal fluid: a systematic review and meta-analysis. Mol Psychiatry. 2023;28(4):1502–15.10.1038/s41380-023-01976-6 PMC10266485 37055513

23.Campana M Strauß J Münz S Oviedo-Salcedo T Fernando P Eichhorn P Falkai P Hasan A Wagner E Cerebrospinal Fluid Pathologies in Schizophrenia-Spectrum Disorder-A Retrospective Chart Review Schizophr Bull 2022 48 1 47 55 10.1093/schbul/sbab105 34480476 PMC8781327 Campana M, Strauß J, Münz S, Oviedo-Salcedo T, Fernando P, Eichhorn P, Falkai P, Hasan A, Wagner E. Cerebrospinal Fluid Pathologies in Schizophrenia-Spectrum Disorder-A Retrospective Chart Review. Schizophr Bull. 2022;48(1):47–55.10.1093/schbul/sbab105 PMC8781327 34480476

24.Tourjman V Kouassi É Koué M Rocchetti M Fortin-Fournier S Fusar-Poli P Potvin S Antipsychotics’ effects on blood levels of cytokines in schizophrenia: a meta-analysis Schizophr Res 2013 151 1–3 43 7 10.1016/j.schres.2013.10.011 24200418 Tourjman V, Kouassi É, Koué M, Rocchetti M, Fortin-Fournier S, Fusar-Poli P, Potvin S. Antipsychotics’ effects on blood levels of cytokines in schizophrenia: a meta-analysis. Schizophr Res. 2013;151(1–3):43–7.10.1016/j.schres.2013.10.011 24200418

25.Zhou X Fragala MS McElhaney JE Kuchel GA Conceptual and methodological issues relevant to cytokine and inflammatory marker measurements in clinical research Curr Opin Clin Nutr Metab Care 2010 13 5 541 7 20657280 PMC2955626 10.1097/MCO.0b013e32833cf3bc Zhou X, Fragala MS, McElhaney JE, Kuchel GA. Conceptual and methodological issues relevant to cytokine and inflammatory marker measurements in clinical research. Curr Opin Clin Nutr Metab Care. 2010;13(5):541–7.10.1097/MCO.0b013e32833cf3bc PMC2955626 20657280

26.Stahl PD Raposo G Extracellular Vesicles: Exosomes and Microvesicles, Integrators of Homeostasis Physiol (Bethesda) 2019 34 3 169 77 10.1152/physiol.00045.2018 30968753 Stahl PD, Raposo G. Extracellular Vesicles: Exosomes and Microvesicles, Integrators of Homeostasis. Physiol (Bethesda). 2019;34(3):169–77.10.1152/physiol.00045.2018 30968753

27.Cai H Pang Y Wang Q Qin W Wei C Li Y Li T Li F Wang Q Li Y Proteomic profiling of circulating plasma exosomes reveals novel biomarkers of Alzheimer’s disease Alzheimers Res Ther 2022 14 1 181 10.1186/s13195-022-01133-1 36471423 PMC9720984 Cai H, Pang Y, Wang Q, Qin W, Wei C, Li Y, Li T, Li F, Wang Q, Li Y, et al. Proteomic profiling of circulating plasma exosomes reveals novel biomarkers of Alzheimer’s disease. Alzheimers Res Ther. 2022;14(1):181.10.1186/s13195-022-01133-1 PMC9720984 36471423

28.Tomita Y Suzuki K Yamasaki S Toriumi K Miyashita M Ando S Endo K Yoshikawa A Tabata K Usami S Urinary exosomal microRNAs as predictive biomarkers for persistent psychotic-like experiences Schizophrenia (Heidelb) 2023 9 1 14 10.1038/s41537-023-00340-5 36906656 PMC10008540 Tomita Y, Suzuki K, Yamasaki S, Toriumi K, Miyashita M, Ando S, Endo K, Yoshikawa A, Tabata K, Usami S, et al. Urinary exosomal microRNAs as predictive biomarkers for persistent psychotic-like experiences. Schizophrenia (Heidelb). 2023;9(1):14.10.1038/s41537-023-00340-5 PMC10008540 36906656

29.Xu H Li M Pan Z Zhang Z Gao Z Zhao R Li B Qi Y Qiu W Guo Q miR-3184-3p enriched in cerebrospinal fluid exosomes contributes to progression of glioma and promotes M2-like macrophage polarization Cancer Sci 2022 113 8 2668 80 10.1111/cas.15372 35411604 PMC9357622 Xu H, Li M, Pan Z, Zhang Z, Gao Z, Zhao R, Li B, Qi Y, Qiu W, Guo Q, et al. miR-3184-3p enriched in cerebrospinal fluid exosomes contributes to progression of glioma and promotes M2-like macrophage polarization. Cancer Sci. 2022;113(8):2668–80.10.1111/cas.15372 PMC9357622 35411604

30.Pegtel DM Gould SJ Exosomes Annu Rev Biochem 2019 88 487 514 10.1146/annurev-biochem-013118-111902 31220978 Pegtel DM, Gould SJ. Exosomes. Annu Rev Biochem. 2019;88:487–514.10.1146/annurev-biochem-013118-111902 31220978

31.Kim SH MacIntyre DA Sykes L Arianoglou M Bennett PR Terzidou V Whole Blood Holding Time Prior to Plasma Processing Alters microRNA Expression Profile Front Genet 2021 12 818334 10.3389/fgene.2021.818334 35096023 PMC8795683 Kim SH, MacIntyre DA, Sykes L, Arianoglou M, Bennett PR, Terzidou V. Whole Blood Holding Time Prior to Plasma Processing Alters microRNA Expression Profile. Front Genet. 2021;12:818334.10.3389/fgene.2021.818334 PMC8795683 35096023

32.Glinge C Clauss S Boddum K Jabbari R Jabbari J Risgaard B Tomsits P Hildebrand B Kääb S Wakili R Stability of Circulating Blood-Based MicroRNAs - Pre-Analytic Methodological Considerations PLoS ONE 2017 12 2 e0167969 10.1371/journal.pone.0167969 28151938 PMC5289450 Glinge C, Clauss S, Boddum K, Jabbari R, Jabbari J, Risgaard B, Tomsits P, Hildebrand B, Kääb S, Wakili R, et al. Stability of Circulating Blood-Based MicroRNAs - Pre-Analytic Methodological Considerations. PLoS ONE. 2017;12(2):e0167969.10.1371/journal.pone.0167969 PMC5289450 28151938

33.Cheng L, Sharples RA, Scicluna BJ, Hill AF. Exosomes provide a protective and enriched source of miRNA for biomarker profiling compared to intracellular and cell-free blood. J Extracell Vesicles. 2014;3.10.3402/jev.v3.23743 PMC3968297 24683445

34.Beveridge NJ Gardiner E Carroll AP Tooney PA Cairns MJ Schizophrenia is associated with an increase in cortical microRNA biogenesis Mol Psychiatry 2010 15 12 1176 89 10.1038/mp.2009.84 19721432 PMC2990188 Beveridge NJ, Gardiner E, Carroll AP, Tooney PA, Cairns MJ. Schizophrenia is associated with an increase in cortical microRNA biogenesis. Mol Psychiatry. 2010;15(12):1176–89.10.1038/mp.2009.84 PMC2990188 19721432

35.Banigan MG Kao PF Kozubek JA Winslow AR Medina J Costa J Schmitt A Schneider A Cabral H Cagsal-Getkin O Differential expression of exosomal microRNAs in prefrontal cortices of schizophrenia and bipolar disorder patients PLoS ONE 2013 8 1 e48814 10.1371/journal.pone.0048814 23382797 PMC3559697 Banigan MG, Kao PF, Kozubek JA, Winslow AR, Medina J, Costa J, Schmitt A, Schneider A, Cabral H, Cagsal-Getkin O, et al. Differential expression of exosomal microRNAs in prefrontal cortices of schizophrenia and bipolar disorder patients. PLoS ONE. 2013;8(1):e48814.10.1371/journal.pone.0048814 PMC3559697 23382797

36.Du X Lv J Feng J Li X Gao Y Wang X Zhao W Ren Z Zhang R Cao X Plasma exosomes lncRNA-miRNA-mRNA network construction and its diagnostic efficacy identification in first-episode schizophrenia BMC Psychiatry 2023 23 1 611 10.1186/s12888-023-05052-9 37605121 PMC10441745 Du X, Lv J, Feng J, Li X, Gao Y, Wang X, Zhao W, Ren Z, Zhang R, Cao X, et al. Plasma exosomes lncRNA-miRNA-mRNA network construction and its diagnostic efficacy identification in first-episode schizophrenia. BMC Psychiatry. 2023;23(1):611.10.1186/s12888-023-05052-9 PMC10441745 37605121

37.Barnett MM Reay WR Geaghan MP Kiltschewskij DJ Green MJ Weidenhofer J Glatt SJ Cairns MJ miRNA cargo in circulating vesicles from neurons is altered in individuals with schizophrenia and associated with severe disease Sci Adv 2023 9 48 eadi4386 10.1126/sciadv.adi4386 38019909 PMC10686555 Barnett MM, Reay WR, Geaghan MP, Kiltschewskij DJ, Green MJ, Weidenhofer J, Glatt SJ, Cairns MJ. miRNA cargo in circulating vesicles from neurons is altered in individuals with schizophrenia and associated with severe disease. Sci Adv. 2023;9(48):eadi4386.10.1126/sciadv.adi4386 PMC10686555 38019909

38.Zhu CY Shen Y Xu Q Propagation of dysbindin-1B aggregates: exosome-mediated transmission of neurotoxic deposits Neuroscience 2015 291 301 16 10.1016/j.neuroscience.2015.02.016 25704251 Zhu CY, Shen Y, Xu Q. Propagation of dysbindin-1B aggregates: exosome-mediated transmission of neurotoxic deposits. Neuroscience. 2015;291:301–16.10.1016/j.neuroscience.2015.02.016 25704251

39.Leucht S Samara M Heres S Davis JM Dose Equivalents for Antipsychotic Drugs: The DDD Method Schizophr Bull 2016 42 Suppl 1Suppl 1 S90 94 10.1093/schbul/sbv167 27460622 PMC4960429 Leucht S, Samara M, Heres S, Davis JM. Dose Equivalents for Antipsychotic Drugs: The DDD Method. Schizophr Bull. 2016;42(Suppl 1Suppl 1):S90–94.10.1093/schbul/sbv167 PMC4960429 27460622

40.Brunner E Chen CA Klein T Maust D Mazer-Amirshahi M Mecca M Najera D Ogbonna C Rajneesh KF Roll E Joint Clinical Practice Guideline on Benzodiazepine Tapering: Considerations When Risks Outweigh Benefits J Gen Intern Med 2025 40 12 2814 59 10.1007/s11606-025-09499-2 40526204 PMC12463801 Brunner E, Chen CA, Klein T, Maust D, Mazer-Amirshahi M, Mecca M, Najera D, Ogbonna C, Rajneesh KF, Roll E, et al. Joint Clinical Practice Guideline on Benzodiazepine Tapering: Considerations When Risks Outweigh Benefits. J Gen Intern Med. 2025;40(12):2814–59.10.1007/s11606-025-09499-2 PMC12463801 40526204

41.Montoya A Valladares A Lizán L San L Escobar R Paz S Validation of the Excited Component of the Positive and Negative Syndrome Scale (PANSS-EC) in a naturalistic sample of 278 patients with acute psychosis and agitation in a psychiatric emergency room Health Qual Life Outcomes 2011 9 18 10.1186/1477-7525-9-18 21447155 PMC3078838 Montoya A, Valladares A, Lizán L, San L, Escobar R, Paz S. Validation of the Excited Component of the Positive and Negative Syndrome Scale (PANSS-EC) in a naturalistic sample of 278 patients with acute psychosis and agitation in a psychiatric emergency room. Health Qual Life Outcomes. 2011;9:18.10.1186/1477-7525-9-18 PMC3078838 21447155

42.Liang Q Li Y Zhou C Zhang R Lu S Shen X Jiang F Xie S The relationship between cognitive function and cortical thickness in first-episode drug-naive schizophrenia patients with agitation Front Psychiatry 2025 16 1576215 10.3389/fpsyt.2025.1576215 40343104 PMC12058731 Liang Q, Li Y, Zhou C, Zhang R, Lu S, Shen X, Jiang F, Xie S. The relationship between cognitive function and cortical thickness in first-episode drug-naive schizophrenia patients with agitation. Front Psychiatry. 2025;16:1576215.10.3389/fpsyt.2025.1576215 PMC12058731 40343104

43.Zhang-James Y Fernàndez-Castillo N Hess JL Malki K Glatt SJ Cormand B Faraone SV An integrated analysis of genes and functional pathways for aggression in human and rodent models Mol Psychiatry 2019 24 11 1655 67 10.1038/s41380-018-0068-7 29858598 PMC6274606 Zhang-James Y, Fernàndez-Castillo N, Hess JL, Malki K, Glatt SJ, Cormand B, Faraone SV. An integrated analysis of genes and functional pathways for aggression in human and rodent models. Mol Psychiatry. 2019;24(11):1655–67.10.1038/s41380-018-0068-7 PMC6274606 29858598

44.Serur Y Fuhrmann N Russo O Green T Irritability in children with RASopathies, insights into emotional dysregulation and social impairment Eur Child Adolesc Psychiatry 2025 34 11 3497 507 10.1007/s00787-025-02767-w 40459753 PMC12335977 Serur Y, Fuhrmann N, Russo O, Green T. Irritability in children with RASopathies, insights into emotional dysregulation and social impairment. Eur Child Adolesc Psychiatry. 2025;34(11):3497–507.10.1007/s00787-025-02767-w PMC12335977 40459753

45.Koh YH Ruiz-Canada C Gorczyca M Budnik V The Ras1-mitogen-activated protein kinase signal transduction pathway regulates synaptic plasticity through fasciclin II-mediated cell adhesion J Neurosci 2002 22 7 2496 504 10.1523/JNEUROSCI.22-07-02496.2002 11923414 PMC6758305 Koh YH, Ruiz-Canada C, Gorczyca M, Budnik V. The Ras1-mitogen-activated protein kinase signal transduction pathway regulates synaptic plasticity through fasciclin II-mediated cell adhesion. J Neurosci. 2002;22(7):2496–504.10.1523/JNEUROSCI.22-07-02496.2002 PMC6758305 11923414

46.Trubetskoy V Pardiñas AF Qi T Panagiotaropoulou G Awasthi S Bigdeli TB Bryois J Chen CY Dennison CA Hall LS Mapping genomic loci implicates genes and synaptic biology in schizophrenia Nature 2022 604 7906 502 8 10.1038/s41586-022-04434-5 35396580 PMC9392466 Trubetskoy V, Pardiñas AF, Qi T, Panagiotaropoulou G, Awasthi S, Bigdeli TB, Bryois J, Chen CY, Dennison CA, Hall LS, et al. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nature. 2022;604(7906):502–8.10.1038/s41586-022-04434-5 PMC9392466 35396580

47.Funk AJ McCullumsmith RE Haroutunian V Meador-Woodruff JH Abnormal activity of the MAPK- and cAMP-associated signaling pathways in frontal cortical areas in postmortem brain in schizophrenia Neuropsychopharmacology 2012 37 4 896 905 10.1038/npp.2011.267 22048463 PMC3280643 Funk AJ, McCullumsmith RE, Haroutunian V, Meador-Woodruff JH. Abnormal activity of the MAPK- and cAMP-associated signaling pathways in frontal cortical areas in postmortem brain in schizophrenia. Neuropsychopharmacology. 2012;37(4):896–905.10.1038/npp.2011.267 PMC3280643 22048463

48.Eissa N, Awad MA, Thomas SD, Venkatachalam K, Jayaprakash P, Zhong S, Stark H, Sadek B. Simultaneous antagonism at H3R/D2R/D3R reduces autism-like self-grooming and aggressive behaviors by mitigating MAPK activation in mice. Int J Mol Sci. 2022;24(1).10.3390/ijms24010526 PMC9820264 36613969

49.Albert-Gascó H, Ros-Bernal F, Castillo-Gómez E, Olucha-Bordonau FE. MAP/ERK signaling in developing cognitive and emotional function and its effect on pathological and neurodegenerative processes. Int J Mol Sci. 2020;21(12).10.3390/ijms21124471 PMC7352860 32586047

50.Saint-Georges Z MacDonald J Al-Khalili R Hamati R Solmi M Keshavan MS Tuominen L Guimond S Cholinergic system in schizophrenia: A systematic review and meta-analysis Mol Psychiatry 2025 30 7 3301 15 10.1038/s41380-025-03023-y 40394282 PMC12185319 Saint-Georges Z, MacDonald J, Al-Khalili R, Hamati R, Solmi M, Keshavan MS, Tuominen L, Guimond S. Cholinergic system in schizophrenia: A systematic review and meta-analysis. Mol Psychiatry. 2025;30(7):3301–15.10.1038/s41380-025-03023-y PMC12185319 40394282

51.Allen MH Debanné M Lazignac C Adam E Dickinson LM Damsa C Effect of nicotine replacement therapy on agitation in smokers with schizophrenia: a double-blind, randomized, placebo-controlled study Am J Psychiatry 2011 168 4 395 9 10.1176/appi.ajp.2010.10040569 21245085 Allen MH, Debanné M, Lazignac C, Adam E, Dickinson LM, Damsa C. Effect of nicotine replacement therapy on agitation in smokers with schizophrenia: a double-blind, randomized, placebo-controlled study. Am J Psychiatry. 2011;168(4):395–9.10.1176/appi.ajp.2010.10040569 21245085

52.Mayeli A, Sonnenschein SF, Yushmanov VE, Wilson JD, Blazer A, Foran W, Perica M, Calabro FJ, Luna B, Hetherington HP, et al. Dorsolateral Prefrontal Cortex Glutamate/Gamma-Aminobutyric Acid (GABA) alterations in clinical high risk and first-episode schizophrenia: a preliminary 7-T magnetic resonance spectroscopy imaging study. Int J Mol Sci. 2022;23(24).10.3390/ijms232415846 PMC9781166 36555487

53.Joyce MKP Uchendu S Arnsten AFT Stress and Inflammation Target Dorsolateral Prefrontal Cortex Function: Neural Mechanisms Underlying Weakened Cognitive Control Biol Psychiatry 2025 97 4 359 71 10.1016/j.biopsych.2024.06.016 38944141 PMC11671620 Joyce MKP, Uchendu S, Arnsten AFT. Stress and Inflammation Target Dorsolateral Prefrontal Cortex Function: Neural Mechanisms Underlying Weakened Cognitive Control. Biol Psychiatry. 2025;97(4):359–71.10.1016/j.biopsych.2024.06.016 PMC11671620 38944141

54.Balcioglu YH Kirlioglu SS C-Reactive Protein/Albumin and Neutrophil/Albumin Ratios as Novel Inflammatory Markers in Patients with Schizophrenia Psychiatry Investig 2020 17 9 902 10 10.30773/pi.2020.0185 32894927 PMC7538240 Balcioglu YH, Kirlioglu SS. C-Reactive Protein/Albumin and Neutrophil/Albumin Ratios as Novel Inflammatory Markers in Patients with Schizophrenia. Psychiatry Investig. 2020;17(9):902–10.10.30773/pi.2020.0185 PMC7538240 32894927

55.Lv H, Guo M, Guo C, He K. The interrelationships between cytokines and schizophrenia: a systematic review. Int J Mol Sci. 2024;25(15).10.3390/ijms25158477 PMC11313682 39126046

56.León-Ortiz P Rivera-Chávez LF Torres-Ruíz J Reyes-Madrigal F Carrillo-Vázquez D Moncada-Habib T Cassiano-Quezada F Cadenhead KS Gómez-Martín D, de la Fuente-Sandoval C: Systemic inflammation and cortical neurochemistry in never-medicated first episode-psychosis individuals Brain Behav Immun 2023 111 270 6 10.1016/j.bbi.2023.05.001 37149107 PMC10330452 León-Ortiz P, Rivera-Chávez LF, Torres-Ruíz J, Reyes-Madrigal F, Carrillo-Vázquez D, Moncada-Habib T, Cassiano-Quezada F, Cadenhead KS. Gómez-Martín D, de la Fuente-Sandoval C: Systemic inflammation and cortical neurochemistry in never-medicated first episode-psychosis individuals. Brain Behav Immun. 2023;111:270–6.10.1016/j.bbi.2023.05.001 PMC10330452 37149107

57.Halstead S Siskind D Amft M Wagner E Yakimov V Shih-Jung Liu Z Walder K Warren N Alteration patterns of peripheral concentrations of cytokines and associated inflammatory proteins in acute and chronic stages of schizophrenia: a systematic review and network meta-analysis Lancet Psychiatry 2023 10 4 260 71 10.1016/S2215-0366(23)00025-1 36863384 Halstead S, Siskind D, Amft M, Wagner E, Yakimov V, Shih-Jung Liu Z, Walder K, Warren N. Alteration patterns of peripheral concentrations of cytokines and associated inflammatory proteins in acute and chronic stages of schizophrenia: a systematic review and network meta-analysis. Lancet Psychiatry. 2023;10(4):260–71.10.1016/S2215-0366(23)00025-1 36863384

58.Posseme C Llibre A Charbit B Bondet V Rouilly V Saint-André V Boussier J Bergstedt J Smith N Townsend L Early IFNβ secretion determines variable downstream IL-12p70 responses upon TLR4 activation Cell Rep 2022 39 13 110989 10.1016/j.celrep.2022.110989 35767946 PMC9237956 Posseme C, Llibre A, Charbit B, Bondet V, Rouilly V, Saint-André V, Boussier J, Bergstedt J, Smith N, Townsend L, et al. Early IFNβ secretion determines variable downstream IL-12p70 responses upon TLR4 activation. Cell Rep. 2022;39(13):110989.10.1016/j.celrep.2022.110989 PMC9237956 35767946

59.Jana M Mondal S Jana A Pahan K Interleukin-12 (IL-12), but not IL-23, induces the expression of IL-7 in microglia and macrophages: implications for multiple sclerosis Immunology 2014 141 4 549 63 10.1111/imm.12214 24224652 PMC3956429 Jana M, Mondal S, Jana A, Pahan K. Interleukin-12 (IL-12), but not IL-23, induces the expression of IL-7 in microglia and macrophages: implications for multiple sclerosis. Immunology. 2014;141(4):549–63.10.1111/imm.12214 PMC3956429 24224652

60.Li Q, Zhang J, Liu S, Zhang F, Zhuang J, Chen Y. MicroRNA-17-3p is upregulated in psoriasis and regulates keratinocyte hyperproliferation and pro-inflammatory cytokine secretion by targeting < em>CTR9. Eur J Histochem. 2022;66(1).10.4081/ejh.2022.3275 PMC8764465 35016493

61.Cai Y Zhang Y Chen H Sun XH Zhang P Zhang L Liao MY Zhang F Xia ZY Man RY MicroRNA-17-3p suppresses NF-κB-mediated endothelial inflammation by targeting NIK and IKKβ binding protein Acta Pharmacol Sin 2021 42 12 2046 57 10.1038/s41401-021-00611-w 33623121 PMC8633290 Cai Y, Zhang Y, Chen H, Sun XH, Zhang P, Zhang L, Liao MY, Zhang F, Xia ZY, Man RY, et al. MicroRNA-17-3p suppresses NF-κB-mediated endothelial inflammation by targeting NIK and IKKβ binding protein. Acta Pharmacol Sin. 2021;42(12):2046–57.10.1038/s41401-021-00611-w PMC8633290 33623121

62.Li W, Wang J, Hao W, Yu C. MicroRNA-543-3p down-regulates inflammation and inhibits periodontitis through KLF6. Biosci Rep. 2021;41(5).10.1042/BSR20210138 PMC8144941 33955459

63.Wei Y Gao H Luo Y Feng J Li G Wang T Xu H Yin L Ma J Chen J Systemic inflammation and oxidative stress markers in patients with unipolar and bipolar depression: A large-scale study J Affect Disord 2024 346 154 66 10.1016/j.jad.2023.10.156 37924985 Wei Y, Gao H, Luo Y, Feng J, Li G, Wang T, Xu H, Yin L, Ma J, Chen J. Systemic inflammation and oxidative stress markers in patients with unipolar and bipolar depression: A large-scale study. J Affect Disord. 2024;346:154–66.10.1016/j.jad.2023.10.156 37924985