Electroconvulsive therapy alters serum cytokine levels and correlates with symptom improvement in patients with acute schizophrenia

Abstract

BACKGROUND

Neuroinflammation is strongly implicated in the pathophysiology of schizophrenia. Key inflammatory markers including tumor necrosis factor-alpha (TNF-α), which modulates neuronal survival and synaptic transmission; interleukin (IL)-8, a neutrophil chemoattractant involved in synapse modulation; and IL-18, which regulates neuronal plasticity and cognitive processes, show distinct alterations in acute schizophrenia. Electroconvulsive therapy (ECT) demonstrates efficacy in acute schizophrenia, yet its effects on these specific inflammatory pathways remain unclear.

AIM

To investigate the effects of ECT on serum cytokine levels and their association with clinical symptoms in acute schizophrenia.

METHODS

Seventy-seven patients with acute schizophrenia (first-episode or relapsed after 4 weeks medication discontinuation, diagnosed per Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition) receiving antipsychotic treatment and 55 well-matched healthy controls were recruited. Groups were matched for age, sex, smoking status, and body mass index. Serum TNF-α, IL-8, and IL-18 were measured using Luminex technology. Clinical symptoms were assessed using the Positive and Negative Syndrome Scale (PANSS). Both PANSS and cytokines were remeasured after 8-10 ECT treatments at 48-hour intervals.

RESULTS

Compared to controls, patients exhibited significantly higher serum concentrations of TNF-α (t = 5.445, P < 0.001) and IL-8 (t = 9.612, P < 0.001) but lower IL-18 (t = -10.007, P < 0.001). ECT resulted in significant elevation of IL-8 and IL-18 levels (t = -3.188, P = 0.002; t = -4.682, P < 0.001, respectively), while TNF-α showed no significant change (t = -1.830, P = 0.071). Before ECT, the serum TNF-α concentration positively correlated with the PANSS general psychopathology score (r = 0.251, P = 0.028) and that of IL-8 negatively correlated with the PANSS negative symptom score (r = -0.250, P = 0.028). However, after ECT, the serum IL-8 concentration negatively correlated with the PANSS general psychopathology score (r = -0.320, P = 0.005). In ECT responders, the post-ECT serum IL-8 concentration positively correlated with a reduced PANSS positive symptom score (r = 0.414, P = 0.001).

CONCLUSION

ECT may mitigate the clinical symptoms of acute schizophrenia through modulation of inflammatory signaling.

Key Words: Electroconvulsive therapy; Schizophrenia; Tumor necrosis factor-α; Interleukin-8; Interleukin-18; Treatment response

Core Tip: Compared to healthy controls, acute schizophrenia patients exhibited significantly higher serum tumor necrosis factor-alpha (TNF-α) and interleukin (IL)-8 levels but lower IL-18 levels. After electroconvulsive therapy (ECT), serum IL-8 and IL-18 concentrations increased significantly, while TNF-α remained unchanged. Before ECT, TNF-α positively correlated with Positive and Negative Syndrome Scale general psychopathology scores, while IL-8 negatively correlated with negative symptom scores. After ECT, IL-8 negatively correlated with general psychopathology scores. In ECT responders, post-treatment IL-8 levels positively correlated with improvements in positive symptoms. The study suggests ECT may ameliorate clinical symptoms of acute schizophrenia by modulating inflammatory signaling pathways, particularly IL-8 signaling.

INTRODUCTION

Schizophrenia is an etiologically and symptomatically complex disorder characterized by clusters of positive symptoms (such as hallucinations and delusions), negative symptoms (such as affective flattening and social withdrawal), and cognitive impairments, typically following a chronic and relapsing course that significantly impacts patient quality of life and social functioning[1-3]. With a global prevalence of approximately 0.5%-1%, the disorder imposes a substantial burden on patients, families, and society[4]. Current evidence suggests that schizophrenia arises from a complex interplay among genetic susceptibility factors, neurodevelopmental abnormalities, neurotransmitter imbalances, and dysregulated immune-inflammatory responses[5-7]. Acute schizophrenia refers to the phase when symptoms rapidly worsen or relapse, typically manifesting as prominent positive symptoms, thought disturbances, and behavioral dysregulation, and often requiring urgent intervention and hospitalization[8].

Changes in various blood inflammatory markers across different stages of the disorder strongly suggest that dysregulated immune-inflammatory responses are major drivers of schizophrenia pathophysiology[9]. Indeed, numerous studies have reported abnormal cytokine levels in the peripheral blood of acute schizophrenia patients, suggesting a possible state of low-grade inflammation[10,11]. For instance, serum concentrations of interleukin (IL)-8 were markedly elevated in first-episode schizophrenia patients and correlated with negative symptoms[12], while serum concentrations of IL-18, a critical regulator of innate and adaptive immune responses in the central nervous system as well as neurodevelopment and synaptic plasticity[13], were associated with cognitive dysfunction and agitation symptoms[14-16]. In addition, elevated serum concentrations of tumor necrosis factor-alpha (TNF-α), a major pro-inflammatory cytokine involved in modulating neuronal survival, neurogenesis, and synaptic transmission, have been reported consistently in acute schizophrenia patients and correlated with disease severity[17,18]. Elucidating the precise relationships between these inflammatory factors and the clinical manifestations of acute schizophrenia, as well as their dynamic changes during disease progression and treatment, could provide both clues to schizophrenia pathology and biomarkers for prognosis and treatment evaluation (Figure 1).

Figure 1 Interleukin-8 [chemokine (C-X-C motif) ligand 8] inflammatory pathway in schizophrenia. This diagram illustrates the inflammatory mechanism of interleukin (IL)-8 in schizophrenia pathophysiology. Activation factors (lipopolysaccharide, tumor necrosis factor-alpha, IL-1) stimulate macrophages and microglia to produce IL-8 [chemokine (C-X-C motif) ligand 8], which recruits neutrophils to inflammatory sites. These activated neutrophils release inflammatory mediators including reactive oxygen species and proteases, contributing to neuroinflammation and schizophrenia symptoms. This pathway highlights the role of IL-8 as a key chemokine linking innate immune responses to neuropsychiatric manifestations in schizophrenia. CXCL8: Chemokine (C-X-C motif) ligand 8; IL: Interleukin-8; LPS: Lipopolysaccharide; ROS: Reactive oxygen species; TNF-α: Tumor necrosis factor-alpha.

Electroconvulsive therapy (ECT) is a well-established and effective treatment particularly suitable for patients with schizophrenia[19,20]. Symptom suppression following ECT may result from normalized neuroplasticity, neurotransmitter signaling, and immune-inflammatory responses[21]. Szota et al[22] reported decreased IL-10 and IL-17 concentrations following ECT treatment, while Zincir et al[23] observed changes in TNF-α levels after ECT. However, Kutsuna et al[24] found no significant changes in immune complexes among patients with psychiatric disorders following ECT, although three functionally diverse proteins, DENND1C, ADARB1, and perilipin-4, were significantly reduced. Research specifically examining the effects of ECT on IL-8, IL-18, and TNF-α concentrations in acute schizophrenia patients and the relationship between these changes and clinical improvement remains limited and findings inconsistent.

Based on this background, we hypothesized that ECT may influence clinical symptoms in acute schizophrenia patients through modulation of IL-8, IL-18, and TNF-α signaling and that these effects could be reflected by changes in specific symptoms. Therefore, this study aimed to: (1) Compare serum IL-8, IL-18, and TNF-α concentrations between acute schizophrenia patients and well-matched healthy controls; (2) Investigate changes in these inflammatory markers after ECT treatment; (3) Analyze correlations between these inflammatory marker concentrations and clinical symptoms; and (4) Compare inflammatory marker profiles between ECT responders and non-responders.

MATERIALS AND METHODS

Participant recruitment and ethics approval

This case control study was conducted between December 2022 and October 2024. The patient group consisted of acute schizophrenia inpatients from local psychiatric wards, who were all first episode or relapsed due to medication discontinuation for at least 4 weeks prior to admission. The minimum 4-week medication-free period was selected to ensure adequate clearance of drugs and active metabolites, and that symptom exacerbation represents true relapse rather than discontinuation reaction, consistent with previous research conventions[25]. The control group comprised healthy volunteers from the same geographic region matched for age distribution, sex ratio, smoking habits, and educational level. The study protocol was approved by the Ethics Committee of the Fourth People’s Hospital of Lianyungang (No. 2021 LSYYXLL-P03), and written informed consent was obtained from all participants or their legal guardians. All procedures were conducted in accordance with the Declaration of Helsinki.

All patients met the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition diagnostic criteria for schizophrenia, confirmed by two independent psychiatrists. Schizophrenia subtypes were not specifically differentiated as Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition emphasizes dimensional assessment. Inclusion criteria for the patient group were as follows: (1) Han Chinese ethnicity; (2) 18-65 years of age; (3) At least primary school education with capability to complete cognitive function tests; (4) Ability to cooperate with clinical assessments; (5) Previous history of schizophrenia with current relapse or first episode meeting diagnostic criteria (detailed below); and (6) Voluntary participation with informed consent provided by the participant or guardian. Exclusion criteria for the patient group were as follows: (1) Neurological disorders (such as cerebrovascular disease, intracranial space-occupying lesions, epilepsy, spinal cord disease, or neurodevelopmental abnormalities); (2) History of severe traumatic brain injury, cranial surgery, or other significant trauma; (3) Physical conditions that might interfere with ECT (including severe cardiovascular disease, uncontrolled hypertension, active diabetes, or major organ dysfunction); (4) Use of antibiotics within the past 4 weeks, current acute or chronic infections, or inflammatory diseases (such as rheumatoid arthritis and inflammatory bowel disease); (5) Hematological or immunological disorders; history of substance dependence or alcohol abuse; and (6) Concurrent transcranial magnetic stimulation therapy.

Data collection

A standardized general information questionnaire was constructed to collect baseline demographic information (name, sex, age, education level), lifestyle habits (current smoking status) and physical examination parameters (height, weight). Body mass index (BMI) was calculated as weight (kg) divided by height squared (m²). Current smoking status was defined as regular smoking behavior within 1 month prior to enrollment, categorized as smoker or non-smoker[26]. For the patient group, duration of illness, age of onset, and medication status were also recorded. Antipsychotic medication dosages were converted to chlorpromazine equivalents for recording and analysis.

Assessments of clinical symptoms

Clinical assessments were independently conducted by two senior psychiatrists, and inter-rater reliability coefficient was greater than 0.8. Clinical symptoms were assessed using the Positive and Negative Syndrome Scale (PANSS)[27]. This scale consists of 30 items divided into three subscales: (1) Positive symptoms (7 items); (2) Negative symptoms (7 items); and (3) General psychopathology (16 items). Each item is rated on a 7-point scale [1 (absent) to 7 (extreme)], with higher scores indicating more severe symptoms[28]. After 8-10 sessions of ECT, patients were re-evaluated using the PANSS to assess treatment response. Treatment efficacy was determined by the percentage reduction in total PANSS score, with responders defined as those achieving a reduction ≥ 25%, and non-responders as those with a reduction < 25%[29].

Blood sampling and biochemical assays

Fasting venous blood samples were collected from all participants between 07:00 and 09:00 AM on the day following clinical and cognitive assessments. For each subject, 5 mL of blood was drawn from the antecubital vein into ethylenediaminetetraacetic acid anticoagulant tubes anticoagulant tubes. Samples were centrifuged at 3000 rpm for 15 minutes to separate serum, which was then aliquoted into 0.5 mL cryotubes, coded, and stored at -80 °C until analysis. For patients, blood samples were collected both before the first ECT session and after completion of the ECT course. Serum concentrations of IL-8, IL-18, and TNF-α were measured using the Luminex liquid suspension chip detection system and commercial kits (R and D Systems China Co., Ltd.) according to the manufacturer’s protocols[30].

ECT

ECT was administered through bilateral electrodes using the Thymatron system with brief-pulse stimulation (Somatics, LLC, United States). Treatment protocol consisted of 8-10 sessions administered at 48-hour intervals (3 sessions per week). Brief-pulse square-wave stimulation with pulse width of 1.0 millisecond was employed, following standard stimulus threshold titration and dosing procedures[31]. Prior to each session, patients received anesthesia induction with propofol (1.5-2.5 mg/kg) and muscle relaxation with succinylcholine chloride (0.8-1.0 mg/kg), along with atropine 0.5 mg intravenously.

Prior to treatment, all patients received comprehensive safety evaluations, including coagulation tests, electrocardiogram, and cranial imaging to rule out contraindications. Treatments were conducted under standardized conditions in the morning (07:00-09:00) following an 8-hour fasting period. The anesthesia protocol included atropine sulfate premedication, followed by propofol for anesthetic induction and succinylcholine for muscle relaxation. Stimulus parameters were standardized according to clinical protocols. Throughout the procedure, a specialized medical team closely monitored the electroencephalogram, motor seizure activity, and vital signs, and documented seizure duration, quality, and adverse effects. All procedures followed institutional safety protocols, and patients remained under observation in the recovery area until fully conscious with stable vital signs before discharge.

Statistical analysis

Statistical analyses were performed using IBM SPSS 25.0 software, graphing using GraphPad Prism 8, and both size calculations and power analysis using G*Power 3.1[32]. The Kolmogorov-Smirnov test was used to assess the distribution characteristics of all datasets prior to further statistical testing. Continuous variables with normal distribution are presented as mean ± SD. Non-normally distributed datasets (such as IL-8 and IL-18 concentrations) were first natural log-transformed to meet the assumptions for parametric tests. Categorical variables are expressed as n (%). Independent samples t-tests were used to compare continuous variables while χ² tests were employed to compare categorical variables between patient and healthy control groups. Paired t-tests were used to compare pre-treatment and post-treatment measurements in the patient group. To control for increased type I error due to multiple comparisons, Bonferroni correction was applied, setting the significance level at 0.0167 (0.05/3). Pearson correlation coefficients were calculated to assess associations between serum cytokine concentrations and clinical symptom scores. Multiple linear regression analysis was performed to control for potential confounding factors, including age, sex, BMI, smoking status, education level, duration of illness, and antipsychotic medication dosage (chlorpromazine equivalent dose). Unless otherwise specified, all statistical tests were two-sided, with P < 0.05 was considered statistically significant.

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

RESULTS

Sociodemographic and clinical characteristics

Patients and control groups did not differ in mean age (t = -1.796, P = 0.075, sex ratio (χ² = 0.001, P = 0.977), average BMI (t = 0.255, P = 0.799), or smoking rate (χ² = 0.260, P = 0.610); however, average years of education was significantly lower in the patient group (t = -5.694, P < 0.001). Thus, groups were demographically well matched. For patients, the mean age of onset was 25.42 ± 5.62 years, mean duration of illness was 7.0 years (interquartile range: 3.0-15.5 years), and mean chlorpromazine equivalent dose was 663.10 ± 279.63 mg/day, reflecting the inclusion criteria for first-episode or recently relapsed patients (Table 1).

Table 1 Sociodemographic and clinical characteristics of schizophrenia patients and healthy controls, n (%)/mean ± SD.

	Patients (n = 77)	Healthy controls (n = 55)	t/χ²	P value
Age (years)	35.19 ± 10.50	38.69 ± 11.72	-1.796	0.075
Sex (male/female)	38/39	27/28	0.001	0.977
Body mass index (kg/m2)	25.10 ± 4.80	24.90 ± 3.92	0.255	0.799
Education (years)	10.12 ± 2.83	13.22 ± 3.41	-5.694	< 0.001
Smoking status	18 (23.4)	15 (27.3)	0.260	0.610
Age of onset (years)	25.42 ± 5.62	-	-	-
Duration of illness (years)	7.0 (3.0, 15.50)	-	-	-
Equivalent dose of chlorpromazine (mg/day)	663.10 ± 279.63	-	-	-

Cytokine concentrations: Patient-control comparisons and ECT-related changes

IL-8 and IL-18 concentrations were non-normally distributed and so were first natural log-transformed prior to (parametric) statistical testing (Table 2). Compared to healthy controls, serum concentrations of TNF-α and IL-8 were significantly elevated in patients before ECT (t = 5.445, P < 0.001; t = 9.612, P < 0.001, respectively), while the IL-18 concentration was significantly lower (t = -10.007, P < 0.001). After ECT treatment, TNF-α and IL-8 concentrations remained significantly higher than in healthy controls (t = 7.314, P < 0.001; t = 12.871, P < 0.001, respectively), while the IL-18 concentration remained significantly lower (t = -7.983, P < 0.001). However, serum IL-8 and IL-18 concentrations were significantly higher after ECT compared to pre-treatment baseline (t = -3.188, P = 0.002; t = -4.682, P < 0.001, respectively), while changes in that of TNF-α were not statistically significant (P = 0.071; Figure 2).

Figure 2 Serum tumor necrosis factor-alpha, interleukin-8, and interleukin-18 concentrations in healthy controls and acute schizophrenia patients before and after electroconvulsive therapy. A: Tumor necrosis factor-alpha levels; B: Interleukin-8 levels (log-transformed); C: Interleukin-18 levels (log-transformed). This compared among three groups: (1) Before electroconvulsive therapy; (2) After electroconvulsive therapy; and (3) Healthy controls. ECT: Electroconvulsive therapy; IL: Interleukin-8; TNF-α: Tumor necrosis factor-alpha.

Table 2 Statistical summary of serum cytokine concentrations for healthy controls and patients both before and after electroconvulsive therapy, mean ± SD.

	Healthy controls (n = 55)	Patients (n = 77)		P value
		Before ECT	After ECT	Within-group comparison1	Before ECT vs healthy controls2	After ECT vs healthy controls2
Tumor necrosis factor-alpha (pg/mL)	3.77 ± 1.55	5.72 ± 2.31	6.19 ± 2.07	0.071	< 0.001	< 0.001
Log IL-8 (pg/mL)	0.82 ± 0.17	1.26 ± 0.31	1.41 ± 0.31	0.002	< 0.001	< 0.001
Log IL-18 (pg/mL)	2.44 ± 0.21	2.10 ± 0.19	2.17 ± 0.18	< 0.001	< 0.001	< 0.001

¹Paired t-tests.

²Independent samples t-tests.

ECT: Electroconvulsive therapy; IL: Interleukin-8.

Using a reduction rate of ≥ 25% in the PANSS total score as the criterion for clinical improvement, patients were divided into responders and non-responders. Among 77 patients, 65 (84.4%) were classified as responders and 12 (15.6%) as non-responders. Table 3 presents the changes in serum TNF-α, IL-8, and IL-18 concentrations after ECT for these groups. In the responder group (as in the total group), ECT significantly enhanced serum IL-8 (t = -3.555, P = 0.001) and IL-18 (t = -3.947, P < 0.001) concentrations, while the TNF-α concentration was not significantly changed (t = -1.531, P = 0.131). In the non-responder group, there were no post-treatment changes in the TNF-α (t = -1.167, P = 0.268) or IL-8 (t = 0.556, P = 0.589) concentrations. Importantly, the numeric increase in the IL-18 concentration did not reach statistical significance after Bonferroni correction (t = -2.745, P = 0.019, adjusted α = 0.0167; Figure 3).

Figure 3 Serum tumor necrosis factor-alpha, interleukin-8, and interleukin-18 concentrations before after electroconvulsive therapy in responders and non-responders. A: Tumor necrosis factor-alpha levels; B: Interleukin-8 levels (log-transformed); C: Interleukin-18 levels (log-transformed). They were compared between responders and non-responders before (white bars) and after (gray bars) electroconvulsive therapy treatment. ECT: Electroconvulsive therapy; IL: Interleukin-8; TNF-α: Tumor necrosis factor-alpha.

Table 3 Statistical summary of serum cytokine concentrations before and after electroconvulsive therapy in responders and non-responders, mean ± SD.

	Before ECT	After ECT	t value	P value
Responders (n = 65)
TNF-α (pg/mL)	5.77 ± 2.44	6.21 ± 2.06	-1.531	0.131
Log IL-8 (pg/mL)	1.24 ± 0.31	1.43 ± 0.30	-3.555	0.001
Log IL-18 (pg/mL)	2.10 ± 0.19	2.17 ± 0.19	-3.947	< 0.001
Non-responders (n = 12)
TNF-α (pg/mL)	5.47 ± 1.47	6.08 ± 2.24	-1.167	0.268
Log IL-8 (pg/mL)	1.35 ± 0.32	1.30 ± 0.34	0.556	0.589
Log IL-18 (pg/mL)	2.07 ± 0.15	2.17 ± 0.09	-2.745	0.019

Paired t-tests were used for within-group comparisons. ECT: Electroconvulsive therapy; IL: Interleukin-8; TNF-α: Tumor necrosis factor-alpha.

Associations between cytokine concentrations and clinical symptoms

At baseline, the serum TNF-α concentration was positively correlated with PANSS general psychopathology subscale score (r = 0.251, P = 0.028), while the IL-8 concentration was negatively correlated with the PANSS negative symptom subscale score (r = -0.250, P = 0.028). No significant correlations were found between the serum IL-18 concentration and PANSS total score or subscale scores (all P > 0.05). Additionally, there were no significant correlations of serum TNF-α and IL-8 concentrations with the PANSS total score and other subscale scores (all P > 0.05). Furthermore, the correlations between the serum TNF-α concentration and general psychopathology score (β = 1.358, t = 2.244, P = 0.028) and between the IL-8 concentration and negative symptom score remained significant in multiple linear regression analysis controlling for potential confounding factors including age, sex, BMI, smoking status, disease duration, and chlorpromazine equivalent dose (β = 1.358, t = 2.244, P = 0.028; β = -9.529, t = -2.235, P = 0.028, respectively).

After treatment, the serum IL-8 concentration was negatively correlated with PANSS general psychopathology subscale score (r = -0.320, P = 0.005; Figure 4A) but not with any pre-treatment PANSS score. There were no significant correlations of serum TNF-α and IL-18 concentrations with the PANSS total score or subscale scores after treatment (all P > 0.05).

Figure 4 Correlation of the serum interleukin-8 concentration, Positive and Negative Syndrome Scale general psychopathology subscale score and reduction in the Positive and Negative Syndrome Scale positive symptom subscale score. A: Correlation between the serum interleukin-8 concentration and Positive and Negative Syndrome Scale general psychopathology subscale score after electroconvulsive therapy treatment; B: Correlation between the serum interleukin-8 concentration and reduction in the Positive and Negative Syndrome Scale positive symptom subscale score after electroconvulsive therapy in the responder group. ECT: Electroconvulsive therapy; IL: Interleukin-8; PANSS: Positive and Negative Syndrome Scale.

The serum IL-8 concentration remained negatively correlated with the general psychopathology score after treatment in multiple linear regression analysis adjusting for age, sex, BMI, smoking status, education, duration of illness, and chlorpromazine equivalent dose (B = -9.130, t = -2.927, P = 0.005).

In the responder group, the serum IL-8 concentration was positively correlated with the reduction in the PANSS positive symptom subscale score (r = 0.414, P = 0.001; Figure 4B), but not with reductions in the PANSS total score, negative symptom score, or general psychopathology score (all P > 0.05). The serum IL-18 concentration was not significantly correlated with reductions in the PANSS total score or any subscale scores (all P > 0.05).

DISCUSSION

The main findings of this study are as follows: (1) Compared to healthy controls, patients with acute schizophrenia exhibited significantly higher serum TNF-α and IL-8 concentrations, while the serum IL-18 concentration was significantly lower; (2) After ECT treatment, serum IL-8 and IL-18 concentrations increased significantly compared to pre-treatment values, while changes in the TNF-α concentration were not statistically significant; (3) Before ECT, serum the TNF-α concentration positively correlated with the PANSS general psychopathology subscale score, while the IL-8 concentration negatively correlated with the PANSS negative symptom score; (4) After ECT, the serum IL-8 concentration negatively correlated with the PANSS general psychopathology subscale score; (5) In responders, post-ECT serum IL-8 and IL-18 concentrations were significantly higher than at baseline, while the TNF-α concentration showed no significant change; and (6) In responders, the post-ECT serum IL-8 concentration positively correlated with the reduction in the PANSS positive symptom score. To our knowledge, this is the first study to report changes in serum IL-8 and IL-18 concentrations after ECT treatment and significant relationship with clinical symptoms in a Chinese Han population with acute schizophrenia.

The significantly higher serum TNF-α and IL-8 concentrations and lower IL-18 concentration among patients with acute schizophrenia is consistent with the immune-inflammatory hypothesis of schizophrenia. A meta-analysis by Halstead et al[9] reported significantly higher serum TNF-α concentrations in patients with schizophrenia compared to healthy controls, particularly during acute episodes. Similarly, Wang et al[18] reported that plasma TNF-α concentration was significantly elevated in first-episode schizophrenia patients and correlated with psychopathology symptoms. Eaton et al[33] also found elevated IL-8 Levels in the peripheral blood of schizophrenia patients, and King et al[34] suggested that higher IL-8 concentrations in schizophrenia patients represented a response to chronic stress. In contrast, our finding of a significantly lower serum IL-18 concentration in schizophrenia patients is inconsistent with several previous reports. Wang et al[35] reported that α-linolenic acid alleviated systemic and cerebral inflammation through inhibition of IL-18 in a mouse model of schizophrenia, and a Syed et al[15] concluded through systematic review and meta-analysis that serum IL-18 concentrations were generally higher in schizophrenia and first-episode psychosis patients. Guan et al[14] further suggested that elevated IL-18 is associated with cognitive dysfunction in schizophrenia patients. Possible reasons for lower serum IL-18 concentrations in the current study include disease phase (acute first episode or relapse) and genetic background. Nonetheless, these observed differences in serum TNF-α, IL-8, and IL-18 concentrations suggest that inflammatory mechanisms contribute to the emergence of acute schizophrenia symptoms.

Serum IL-8 and IL-18 concentrations were elevated significantly after ECT treatment, suggesting that ECT may exert therapeutic effects through modulation of inflammatory responses. In patients with depression, Dellink et al[36] found that effective ECT influenced cytokine levels. In addition to canonical inflammatory pathways, cytokine signaling pathways are involved in regulating hippocampal plasticity[37], and ECT has been confirmed to enhance hippocampal neuroplasticity. For instance, in addition to regulating innate and adaptive immune responses, IL-18 influences neurodevelopment and synaptic plasticity[13]. Hjell et al[16] reported that IL-18 signaling was associated with excitement symptoms in severe mental disorders. The observed elevation in serum IL-18 concentration after ECT from below normal prior to treatment suggests that this intervention improves specific psychiatric symptoms by normalizing immune-inflammatory state and neuroplasticity.

In contrast to the observed increases in serum IL-8 and IL-18 after ECT, the serum TNF-α concentration was not altered. Zincir et al[23] reported that serum TNF-α concentrations changed after ECT for treatment-resistant depression, but with complex patterns, suggesting dependence on factor like current clinical state. A recent review by Wang and Zhang[21] on immune-inflammatory mechanisms of ECT in schizophrenia also suggested that different inflammatory factors may have distinct response patterns to ECT. Several explanations may account for the relatively stable TNF-α concentrations after ECT[17]. First, TNF-α-regulated inflammatory pathways may be less sensitive to ECT than those regulated by IL-8 and IL-18. Second, a longer observation period may be necessary to detect significant changes in TNF-α. Third, individual differences may result in non-significant changes at the group level. For instance, Kutsuna et al[24] also reported differential effects of ECT on various immune markers. Therefore, we conducted further analysis on ECT response and non-response populations, although discrepancies remain.

Before ECT treatment, the serum TNF-α concentration positively correlated with PANSS general psychopathology symptoms, while that of IL-8 negatively correlated with negative symptoms, supporting specific associations between inflammatory markers and clinical phenotypes in schizophrenia. The positive correlation between serum TNF-α and general psychopathology score is consistent with Yang et al[30], who reported that the peripheral blood TNF-α concentration was associated with PANSS general psychopathology scores in chronic medicated schizophrenia patients. Zhao et al[12] also reported correlations between serum IL-8 and clinical symptoms among first-episode schizophrenia patients, in accord with current findings and supporting the involvement of IL-8 in disease pathophysiological. IL-8 may indirectly influence the severity of negative symptoms by affecting neuroplasticity and synaptic function[38]. After treatment, we found that the serum IL-8 concentration was negatively correlated with the PANSS general psychopathology score across the entire cohort and positively correlated with a reduced PANSS positive symptom score within the responder group. This latter finding supports the utility of IL-8 as a biomarker for ECT treatment response. However, these correlations do not establish causality. Inflammation and psychopathology may have a bidirectional relationship[10]. In accord with the current findings, Szota et al[22] found that ECT influenced cytokine levels in schizophrenia patients and was associated with symptom improvement. In patients with depression, changes in serum IL-8 among ECT responders were correlated with microstructural changes in cerebral white matter[39]. It has also been suggested that IL-8 is involved in regulating reparative neuroplasticity[40], which may be associated with clinical symptom improvement.

This study has several limitations. First, this is a single-center study with a relatively limited sample size, which may affect the generalizability and statistical power of the results. Second, we only measured three inflammatory markers (TNF-α, IL-8, and IL-18), while the inflammatory network comprises numerous interacting cytokines and chemokines. Third, all patients were receiving antipsychotic medication, which may have influenced inflammatory markers independently of ECT despite controlling for this effect in statistical analyses by expressing dose in chlorpromazine equivalents. Fourth, cytokine measurements were limited to baseline and post-ECT timepoints without extended follow-up. This precludes assessment of whether the observed inflammatory changes are sustained long-term. Fifth, potential biases including selection bias (single-center design), measurement bias (peripheral blood markers may not reflect central nervous system inflammation), and confounding bias (concurrent antipsychotic medication) may have influenced the results. Finally, this is an observational study, which precludes determination of a causal relationship between changes in inflammatory markers and clinical improvement. Future research should address these limitations through multicenter studies with larger sample sizes, expanded inflammatory marker panels, longitudinal follow-up, and mechanistic investigations.

CONCLUSION

In conclusion, patients with acute schizophrenia exhibited elevated serum TNF-α and IL-8 concentrations and a lower serum IL-18 concentration compared to health matched controls. After ECT treatment, serum IL-8 and IL-18 concentrations were significantly greater, while the TNF-α concentration was unchanged. Specific correlations were detected between these inflammatory marker concentrations and clinical symptoms, notably including a positive correlation between the serum IL-8 concentration and positive symptom improvement in the responder group following ECT. These findings suggest an association between inflammatory modulation and the therapeutic effects of ECT in acute schizophrenia. Future studies should validate the potential of IL-8 as a treatment response marker and elucidate the underlying mechanisms.

ACKNOWLEDGEMENTS

We would like to thank the participants in the study.