ABSTRACT

Objective

Schizophrenia is a multifaceted psychiatric disorder that affects about 1% of the world’s population and arises from a combination of genetic, environmental, and neurodevelopmental influences. Recent studies highlight the role of immune system disturbances and neuroinflammation in its development, with tumor necrosis factor-alpha (TNF-α) identified as a pivotal cytokine. This meta-analysis aims to clarify the relationship between the TNF-α rs1800629 genetic variant and the risk of schizophrenia by synthesizing data from published research.

Methods

Two independent reviewers systematically searched PubMed, Web of Science, Embase, Cochrane Library, and Chinese National Knowledge Infrastructure for studies published up to January 19, 2024. Odds ratios and 95% confidence intervals were computed using a fixed-effects model, taking into account the absence of significant heterogeneity.

Results

A total of 33 case-control studies were included, encompassing 7,624 individuals with schizophrenia and 8,933 healthy controls from diverse backgrounds (21 studies on Asian populations, 11 on Caucasian, and one on a mixed group) conducted between 2001 and 2020. The pooled analysis did not reveal a significant link between the TNF-α rs1800629 polymorphism and susceptibility to schizophrenia under any genetic model. Further subgroup analyses by ethnicity (Asian, Caucasian), country (China, Poland), genotyping technique, and publication year also yielded no notable associations.

Conclusions

This comprehensive meta-analysis offers strong evidence that the TNF-α rs1800629 variant is not significantly associated with schizophrenia risk, either globally or within specific ethnic groups. These findings indicate that this polymorphism likely does not play a major role in schizophrenia susceptibility, underscoring the importance of future investigations into other TNF-α variants, gene-gene interactions, or alternative inflammatory mechanisms.

INTRODUCTION

Schizophrenia is a chronic and multifaceted psychiatric illness that affects about 1% of people worldwide. It is marked by a combination of positive symptoms (such as hallucinations and delusions), negative symptoms (including social withdrawal and anhedonia), and cognitive deficits, all of which substantially diminish patients’ quality of life and daily functioning^{1, 2}. The underlying causes of schizophrenia are complex, involving genetic predispositions, environmental exposures, and neurodevelopmental disturbances. Recent research increasingly points to the role of immune system dysfunction and neuroinflammatory processes in the development and progression of the disorder^{3, 4}. Among the inflammatory molecules implicated in schizophrenia, tumor necrosis factor-alpha (TNF-α) stands out as a key cytokine. Elevated TNF-α levels have been repeatedly documented in individuals with chronic schizophrenia and have been linked to greater severity of negative symptoms^{5, 6}.

TNF-α acts as a pro-inflammatory cytokine, playing a pivotal role in immune regulation and, more recently, has been recognized for its influence on brain development, synaptic remodeling, and neuron survival^{7, 8}. Evidence suggests that both TNF-α and interleukin-6 are closely associated with the deficit syndrome subtype of schizophrenia, which is characterized by persistent and primary negative symptoms. This observation supports the idea that deficit schizophrenia may involve distinct immune-related pathophysiology⁹. Moreover, higher TNF-α concentrations have been shown to predict the intensity of specific negative symptoms, such as blunted affect and alogia, as well as overall negative symptom scores, indicating a direct contribution of inflammatory processes to clinical manifestations^{10, 11}. The hypothesis that inflammation plays a central role in schizophrenia is further reinforced by findings that immune-targeted therapies could offer new avenues for treating negative symptoms in individuals with this condition¹².

The TNF-α gene, situated on chromosome 6p21.3, harbors several functional variants that modulate cytokine expression and immune response mechanisms^{13, 14}. One notable variant is the rs1800629 single nucleotide polymorphism (SNP), also known as -308G>A, located in the gene’s promoter region. This SNP has been widely investigated across various inflammatory and autoimmune diseases due to its regulatory effect on TNF-α production. Specifically, the substitution of guanine (G) with adenine (A) at position -308 is associated with increased TNF-α expression in individuals carrying the A allele compared to those with the G allele¹⁵. Research indicates that carriers of the A allele exhibit significantly reduced fractional anisotropy in broad brain regions, alongside more pronounced deficits in both immediate and delayed verbal memory recall and recognition, relative to individuals with the GG genotype¹⁶. These observations imply that the A allele, which drives higher TNF-α levels, may be linked to diminished fronto-temporal white matter connectivity and subsequent memory impairments in patients with schizophrenia^{16, 17}.

Previous case-control studies investigating the association between TNF-α rs1800629 polymorphism and schizophrenia susceptibility have yielded inconsistent and sometimes contradictory results, likely due to differences in study populations, sample sizes, ethnic backgrounds, and methodological approaches⁶. While some studies have reported significant associations between specific genotypes and schizophrenia risk, others have found no significant relationship between the polymorphism and disease susceptibility¹⁸. Additionally, research has indicated that this polymorphism may act as a modulator for disease onset age and cognitive function rather than directly influencing susceptibility, as demonstrated in studies examining the related rs1800629 polymorphism, which showed associations with earlier onset age and cognitive deficits but not with overall disease risk¹⁹. The heterogeneity in study findings underscores the need for a comprehensive meta-analytical approach to synthesize existing evidence and provide more definitive conclusions about the role of rs1800629 in schizophrenia susceptibility.

Understanding genetic variants in inflammatory pathways has significant clinical and therapeutic implications. Given the influence of TNF-α polymorphisms on treatment responses in conditions like autoimmune disorders²⁰, determining if the rs1800629 polymorphism is associated with schizophrenia could inform personalized medicine, guide treatment selection, and aid in developing targeted immunomodulatory interventions. This meta-analysis systematically evaluates case-control studies to determine if the TNF-α rs1800629 polymorphism is associated with increased schizophrenia susceptibility. It also assesses the magnitude of associations across populations and study designs, identifies sources of heterogeneity explaining inconsistencies in prior research, and provides recommendations for future genetic association studies in schizophrenia research.

MATERIALS and METHODS

Search Strategy

Since this meta-analysis did not involve the use of personal data or the recruitment of participants, ethical approval was unnecessary, and patient consent was not applicable. We performed an extensive literature search across a variety of electronic databases to identify publications examining the association between the TNF-α rs1800629 polymorphism and schizophrenia risk, with studies limited to those considered up to January 19, 2024. The databases searched included PubMed, EMBASE, Web of Science, Elsevier, Google Scholar, ScienceDirect, SciELO, Europe PMC, ResearchGate, Circumpolar Health Bibliographic Database, Cochrane Library, Current Contents, Linguamatics, Eye Health Organizations Database, WanFang, China Science and Technology Journal Database, VIP, Chinese Biomedical Database, Chinese National Knowledge Infrastructure, Scientific Information Database, PsycINFO, and ClinicalTrials.gov. To refine our search, we used combinations of keywords such as “schizophrenia”, “TNF-α”, “rs1800629”, “polymorphism”, “genetic susceptibility” and “association study”, as well as related terms including “gene”, “polymorphism”, “DNA sequence”, “single-nucleotide polymorphism”, “SNPs”, “genotype”, “frequency”, “mutation”, “mutant”, “allele”, “variation”, “variant” and “genetic predisposition”. Additionally, we manually reviewed the reference lists of all relevant articles to ensure no pertinent studies were missed. There was no limitation regarding language or publication year; non-English articles were translated when necessary. Our review prioritized human studies published in English or Chinese, and in cases where multiple articles covered the same subjects, the most recent or those with larger sample sizes were selected.

Inclusion and Exclusion Criteria

We established clear criteria for study selection. The inclusion criteria were: (1) Only case-control or cohort studies that investigated the relationship between the TNF-α rs1800629 polymorphism and schizophrenia risk; (2) schizophrenia diagnosis had to be based on recognized clinical standards; (3) studies needed to provide genotype frequencies for both cases and controls to enable calculation of odds ratios (ORs) and 95% confidence intervals (CIs); (4) studies had to provide sufficient demographic information about participants; and (5) only studies published up to January 2024 were considered for relevance. Exclusion criteria were as follows: (1) Reviews, meta-analyses, abstracts, conference proceedings, case reports, letters to the editor, comments, and duplicate publications; (2) studies without a control group or with inappropriate selection criteria; (3) articles with duplicated data from the same author; (4) studies lacking gene frequency data that could not be supplemented; and (5) animal or in vitro research. If multiple publications reported on the same dataset, the study with the largest sample size or the most recent publication was included in the analysis.

Data Extraction

Two independent reviewers assessed the titles, abstracts, and search terms of identified studies to determine eligibility based on the established criteria regarding the TNF-α rs1800629 polymorphism and schizophrenia risk. Any disagreements were resolved through discussion or by consulting a third reviewer, and if needed, the original study authors were contacted for clarification. The screening process began with the evaluation of titles and abstracts to exclude irrelevant studies, followed by a detailed full-text review for final selection. From each eligible study, we extracted the following data: First author’s name, participant ethnicity (categorized as Asian, Caucasian, African, Hispanic, or Mixed), publication year, genotyping method, country of study, total number of schizophrenia cases and controls, genotype frequencies of the TNF-α rs1800629 polymorphism in both groups, Hardy-Weinberg equilibrium (HWE) results, and minor allele frequencies (MAFs) among controls. When research groups published multiple related studies, we included the most recent, or the one with the largest sample size.

Statistical Analysis

To evaluate the association between the TNF-α rs1800629 polymorphism and the risk of developing schizophrenia, ORs with corresponding 95% CIs were calculated. The  statistical significance of the overall effect size was determined using the Z-test, with a p-value below 0.05 considered significant. Five genetic models were analyzed: allele comparison (B vs. A), homozygote comparison (BB vs. AA), heterozygote comparison (BA vs. AA), dominant model (BB+BA vs. AA), and recessive model (BB vs. BA+AA). Here, “A” represented the major allele, while “B” denoted the minor allele. To assess heterogeneity among studies, the Cochran Q-test was applied, with a significance threshold set at p≤0.10. The I² statistic was also used to quantify heterogeneity, with values ranging from 0% (no heterogeneity) to 100% (extreme heterogeneity): 0-25% indicated none, 25-50%, moderate, 50-75%, high, and 75-100%, very high heterogeneity^{22, 23}. Depending on the degree of heterogeneity, either the DerSimonian and Laird random-effects model or the Mantel-Haenszel fixed-effects model was used for pooling effect sizes. HWE in the control groups was checked using the chi-square test, and a p-value less than 0.05 indicated a significant deviation from equilibrium^{24, 25}. Subgroup analyses were performed based on ethnicity and schizophrenia subtype to explore sources of heterogeneity. Sensitivity analyses were conducted by sequentially removing individual studies to test the robustness of the results^{26, 27}. Potential publication bias was assessed through Begg’s and Egger’s tests, along with visual inspection of funnel plots for asymmetry. If bias was detected, the Duval and Tweedie “trim-and-fill” method was employed to adjust the results. All statistical analyses were performed using Comprehensive Meta-Analysis (CMA) software version 2.0 (Biostat, USA). A two-sided p-value less than 0.05 was considered statistically significant.

RESULTS

Characteristics of Selected Studies

A summary of the literature review and study selection process is depicted in Figure 1. The initial search identified 741 articles. After screening titles and abstracts, 312 duplicates and 218 articles related to cell or animal studies, reviews, case reports, and other non-eligible formats were excluded. The full texts of the remaining 211 articles were reviewed in detail, leading to the exclusion of 178 studies based on the set inclusion and exclusion criteria. Ultimately, 33 studies comprising 7,624 schizophrenia cases and 8,933 controls were included in the meta-analysis. Of these, 21 studies focused on Asian populations, 11 on Caucasian populations, and one on a mixed ethnic group. The studies were published between 2001 and 2020 and represented research from various countries, including Italy, Germany, Korea, Singapore, China, Finland, the United States, Canada, Japan, Poland, Pakistan, Saudi Arabia, and Türkiye. With the exception of three studies, the genotype distributions in the control groups conformed to HWE, as outlined in the study protocols. Detailed genotypic frequency data for all included studies are provided in Table 1.

Quantitative Synthesis

Overall Analysis

In the present meta-analysis, no statistically significant relationship was found between the TNF-α rs1800629 polymorphism and the risk of schizophrenia. This conclusion is based on the quantitative synthesis summarized in Table 2 and depicted in the forest plot (Figure 2), which presents results across multiple genetic models (allelic, homozygous, heterozygous, dominant, and recessive). For example, the pooled OR for the allelic comparison (A vs. G) was 1,148 (95% CI: 0.947-1.391, p=0.161), indicating only a slight, nonsignificant increase in risk. Similarly, the homozygous (AA vs. GG: OR=1,332, 95% CI: 0.785-2,260, p=0.289) and heterozygous (AG vs. GG: OR=1,081, 95% CI: 0.905-1,291, p=0.390) comparisons did not reveal significant associations. This lack of significant association was consistent across subgroup analyses stratified by ethnicity (such as Asian, Caucasian, and East Asian populations), country (e.g., China, Poland), HWE status, genotyping technique [polymerase chain reaction - restriction fragment length polymorphism (PCR-RFLP), ankle-brachial index (ABI)], and publication period (before and after 2010). For instance, among Asian populations, the OR for A vs. G was 1,198 (95% CI: 0.892-1,595, p=0.235), and for Caucasian populations, it was 1,043 (95% CI: 0.853-1,275, p=0.689), neither reaching statistical significance. However, a notable exception emerged in the subgroup analysis of studies published after 2010, where the combined AA+AG versus GG genotype model showed a significant association with schizophrenia risk, (OR=1,700, 95% CI: 1,073–2.693, p=0.024). When considering both AA and AG genotypes together,  more recent research  found a statistically significant increase in schizophrenia risk linked to the TNF-α rs1800629 polymorphism.

Heterogeneity Testing and Sensitivity Analysis

The heterogeneity analysis, as indicated by I2 values, reveals substantial variability across the different subgroups and genetic models. Specifically, the overall analysis shows high heterogeneity (I2=74.19% to 86.33%, p≤0.001). Similar trends are observed within ethnicity-based subgroups, with Asians, (I2=89.69% to 80.50%, p≤0.001), Caucasians, (I2=65.40% to 48.83%), and East Asians (I2=89.48% to 81.84%, p≤0.001) all exhibiting considerable heterogeneity. Country-based analyses in countries, such as China (I2=90.62% to 53.49%, p≤0.011) and Poland (I2=97.01% to 83.22%, p=0.001 to 0.009), also demonstrate significant heterogeneity. Subgroups based on HWE and genotyping methods (PCR-RFLP and ABI) also show high levels of heterogeneity. Finally, stratification by publication year (before and after 2010) reveals high heterogeneity in both periods, suggesting that the observed associations are influenced by various confounding factors. These findings suggest that the overall effect estimates should be interpreted with caution, as the true effect may vary across different populations and study designs.

Publication Bias

The evaluation of publication bias related to the TNF-α rs1800629 polymorphism and its association with schizophrenia involved analyzing various genetic models and subgroups using Begg’s and Egger’s tests. The results indicated differing levels of publication bias across models (Figure 3). Specifically, Figure 3 (Begg’s funnel plot) visually represents the publication bias test for the correlation between TNF-α rs1800629 polymorphism and schizophrenia development. For the A vs. G allele comparison, Begg’s test returned a p-value of 0.675, and Egger’s test yielded 0.288, both suggesting no significant bias. Similarly, the AA vs. GG model had p-values of 0.417 (Begg’s) and 0.333 (Egger’s), indicating an absence of substantial bias. The AG vs. GG comparison also showed no evidence of bias, with p-values of 0.232 and 0.148, respectively. However, subgroup analyses based on ethnicity revealed more complex results: in the Asian subgroup, the AG vs. GG model indicated significant publication bias with p-values of 0.037 (Begg’s) and 0.033 (Egger’s), suggesting a potential influence of publication bias for this genetic model within this specific ethnic group. Although no significant bias was found in other comparisons within this group, further investigation is required. For Caucasians and East Asians, results generally showed no significant publication bias, with p-values consistently above 0.05. When analyzing by country, Poland showed potential bias with A vs. G model p-values of 1.000 (Begg’s) and 0.908 (Egger’s), which may indicate a lack of studies showing a significant association in the Polish population for this specific allele comparison. In contrast, analyses for China revealed no publication bias in the assessed models. Examining the HWE subgroup, it was found that no significant publication bias was evident. Similarly, analyses based on genotyping methods (PCR-RFLP and ABI) and publication year (before and after 2010) generally indicated no substantial publication bias.

Minor Allele Frequencies

MAFs ranged from 0.008 to 0.463, demonstrating considerable geographical and ethnic variation. Asian populations exhibited lower MAFs in Japan (0.008) and China (0.071, 0.077) compared to Singapore (0.220) and healthy Chinese parents (0.463). Chinese studies showed diverse MAFs, some exceeding 0.300. Caucasian populations generally had higher MAFs, as seen in Germany (0.169) and Poland (0.171), while Finnish studies reported 0.136 and 0.128. These MAF differences are attributable to nationality, study design, and methodologies, underscoring the importance of population-specific genetic factors in genetic epidemiology. The results indicate a complex interaction of genetic and environmental influences on MAF distribution.

DISCUSSION

The association between the TNF-α rs1800629 polymorphism and schizophrenia susceptibility remains a complex and controversial topic, as evidenced by conflicting findings across various studies. While our CMA, encompassing 33 investigations with 7624 cases and 8933 controls, revealed no significant correlation between this polymorphism and schizophrenia risk across five genetic models, this contrasts with some prior research. For instance, Sacchetti et al.²⁸ (2007) reported a weak association between the AA genotype and schizophrenia susceptibility in Caucasoids, further supported by a replication case-control study indicating an association between the A allele and increased schizophrenia susceptibility, particularly in males, with correlations to a later schizophrenia onset at age. However, these initial observations have not been consistently replicated.

More recent meta-analyses, including those by Qin et al.¹⁸ (2013) and He et al.⁶ (2022), have found no substantial correlation between the TNF-α rs1800629 polymorphism and schizophrenia susceptibility. Qin et al.¹⁶ analysis of 21 studies showed a lack of association among Caucasian and Asian subgroups, as well as between males and females. Similarly, He et al.⁶ pooled analysis of 24 studies found no significant association. Further complicating the picture, studies have explored the influence of specific populations or other factors. Alfimova et al. found that childhood adversity influences the relationship between schizophrenia development and the TNF-α promoter polymorphism rs1800629, while Kang et al.¹⁶ discovered that the A-allele at TNF-α rs1800629 is associated with reduced white matter connectivity in the fronto-temporal region in Korean patients; While Kang et al.¹⁶ discovered that the A-allele at TNF-α rs1800629 is associated with reduced white matter connectivity in the fronto-temporal region in Korean patients²⁶. Conversely, Aytec et al.²⁹ (2022) found no significant difference in the prevalence of TNF-α rs1800629 between Turkish individuals with schizophrenia and healthy controls, and Lang et al.¹⁹ (2020) found no significant relationship between rs1800629 and schizophrenia or suicide. These discrepancies highlight the need for further research to clarify the role of this polymorphism in schizophrenia susceptibility, considering potential influences from ethnicity, environmental exposures, and interactions with other genes.

Clinical Implication

The meta-analysis suggests that the TNF-α rs1800629 polymorphism, on its own, is unlikely to be a strong predictor of schizophrenia risk across diverse populations. Clinically, this implies that routine screening for this specific polymorphism in individuals to assess their risk of developing schizophrenia is not currently warranted. However, clinicians should be aware of the potential for gene-environment interactions and the influence of ethnicity on genetic associations. Further research exploring these factors may refine risk prediction models in the future. It is important to consider other established risk factors and diagnostic criteria when assessing individuals for schizophrenia.

Study Limitations

The study entailed a thorough analysis of the network database. However, there are limitations to this meta-analysis. The primary constraint is that most studies focused on Asian and Caucasian populations, making it challenging to evaluate the impact of TNF-α rs1800629 polymorphism on other groups. This limitation could impact the true association between the polymorphism and schizophrenia. Due to limited data, the relationship between TNF-α rs1800629 polymorphism and the clinical features of schizophrenia could not be fully explored. The analysis was unadjusted; however, an analysis that considered factors like gender, family history of schizophrenia, pregnancy complications, and exposure to toxins or viruses could have been more beneficial. Genetic and environmental interactions were not examined due to insufficient original data. Therefore, further validation with a larger, diverse sample is necessary to confirm the study’s findings. Future research should also account for the potential influence of other genetic polymorphisms and lifestyle factors on schizophrenia development. Additionally, exploring the role of epigenetic modifications in conjunction with TNF-α rs1800629 polymorphism could offer a more comprehensive understanding of the underlying mechanisms involved in schizophrenia susceptibility. Collaborative efforts among researchers from diverse ethnic backgrounds and regions could help overcome current limitations and provide a more nuanced perspective on the relationship between TNF-α rs1800629 polymorphism and schizophrenia. Ultimately, a multidisciplinary approach encompassing genetics, epigenetics, environmental factors, and clinical characteristics is crucial for advancing our understanding of the complex etiology of schizophrenia.

CONCLUSION

In summary, our comprehensive analysis does not support a consistent link between the TNF-α rs1800629 polymorphism and increased susceptibility to schizophrenia. However, these findings should be interpreted with caution due to considerable heterogeneity among studies and the limited representation of ethnic groups, as the current meta-analysis primarily included Asian and Caucasian populations. To better understand the potential involvement of TNF-α rs1800629 in schizophrenia, future research should incorporate larger and more ethnically diverse cohorts. Additionally, exploring gene-gene and gene-environment interactions will be essential for a more complete understanding of the genetic and environmental factors that contribute to schizophrenia risk.