This systematic review demonstrated strict compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) standards[8]. A systematic search was performed across four electronic databases: PubMed, EMBASE, Web of Science, and the Cochrane Library, on September 19, 2023, without constraints on publication date. The formulated search strategy included essential phrases such as MDD, 5-HTR1A, C-1019G polymorphism, and antidepressant response. These terms were carefully selected to comprehensively address the elements of the Patient, Intervention, Comparison, Outcome framework and guarantee a complete compilation of relevant papers for the meta-analysis. Publications in every language were taken into account. A comprehensive manual examination of the bibliographies of pertinent papers was conducted to uncover any further studies.

Inclusion criteria: Studies that specifically investigated the association between the 5-HTR1A gene C-1019G polymorphism and antidepressant response in patients diagnosed with MDD.

Studies that provided clear data on the genotype of the 5-HTR1A C-1019G polymorphism.

Exclusion criteria: Research involving animal models or in vitro experiments.

Non-peer-reviewed articles, conference abstracts, editorials, and literature reviews.

Studies with a primary focus on other psychiatric disorders, even if they included a subset of patients with MDD.

Publications not available in full text or those that required translation from languages other than English, due to resource constraints.

In the meta-analysis, literature screening and data extraction were performed independently by two reviewers, who subsequently cross-verified their results. In cases of disagreement, the reviewers engaged in discussions to achieve consensus, and a third-party reviewer was consulted when necessary. The extracted data included various parameters: (1) Authorship; (2) Publication year; (3) Sample size; (4) Participant age; (5) Ethnicity; (6) Country; (7) Clinical diagnosis; (8) Diagnostic criteria; (9) Treatment efficacy criteria; (10) Medication duration; (11) Types of antidepressants utilized; and (12) The distribution of the 5-HTR1A gene C1019G polymorphism. When the required data were not available in published reports, the original study investigators were contacted via email to request the pertinent unpublished information.

The meta-analysis involved a thorough evaluation of study quality by two independent assessors using the Newcastle-Ottawa Scale (NOS)[9]. This measure, noted for its robustness, assesses research across three essential domains: (1) Selection; (2) Comparability; and (3) Outcome, each contributing to a thorough awareness of potential biases within the investigations. Following this comprehensive evaluation, studies were assigned ratings from 0 to 9 according to their compliance with these criteria. The score interpretation was organized as follows: (1) A score ranging from 0 to 3 denoted low-quality studies; (2) Scores from 4 to 6 indicated moderate quality; and (3) Studies with scores from 7 to 9 were classified as high quality.

The meta-analysis conducted statistical evaluations to assess the heterogeneity among studies. This was accomplished using χ² statistics, with the degree of heterogeneity measured by the I² value. A situation in which the I² value was below 50% and the P value was 0.10 or greater signified an absence of substantial heterogeneity, resulting in the utilization of the fixed-effect model for determining the aggregated effect size. An I² value of 50% or above, or a P value below 0.10, indicates considerable heterogeneity. In these situations, the random-effects model was employed to ascertain the aggregated effect size. An examination of the funnel plot's symmetry was performed to assess potential publishing bias. A symmetrical arrangement of data points around the apex in the funnel plot would indicate a negligible impact of publication bias on the results of the meta-analysis. Additionally, Egger's linear regression test functioned as a quantitative instrument for identifying publication bias. All statistical tests were bilateral, with a P value threshold established at less than 0.05 for statistical significance. The data analysis utilized Stata version 17 (StataCorp, College Station, TX, United States).

During the initial phase of this systematic review and meta-analysis, a thorough search across various electronic databases yielded 1216 articles potentially relevant to the study. A deduplication algorithm was then applied to eliminate duplicate entries, ensuring the uniqueness of each study included. This was followed by a detailed screening of titles and abstracts, guided by well-defined inclusion and exclusion criteria. These criteria encompassed various aspects such as study methodology, demographic characteristics of the participants, clinical outcomes measured, and the research methods' overall quality. After this preliminary screening, 42 articles were shortlisted for a more detailed evaluation. This next stage involved an independent, full-text review by multiple investigators to ensure an impartial and comprehensive assessment of each article. During this in-depth examination, 31 articles were excluded for specific reasons: (1) Review articles (n = 9); (2) Sequentially published works (n = 9); (3) Studies with insufficient data for analysis (n = 8); and (4) Clinical trials without control groups (n = 5). Consequently, 11 articles were identified that satisfactorily met the stringent criteria outlined in the research protocol[10-20]. These selected studies were then included in the final meta-analysis, as illustrated in Figure 1.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Study inclusion flowchart.

The meta-analysis encompasses a broad spectrum of international studies that investigated the efficacy of antidepressant medication in MDD, as diagnosed by the diagnostic and statistical manual of mental disorders, fourth edition criteria. The included studies, published between 2004 and 2022, examined populations in Asia (India, South Korea, Japan, and China) and Europe (United Kingdom, Italy, and Finland), as well as multi-country studies including Caucasian populations from Belgium, Austria, and Israel. Participant ages across studies varied, with an average age ranging from 35.0 years ± 10.3 years to 88.16 years ± 3.80 years. The gender distribution within the sample sizes, where reported, included both male and female subjects, with total sample sizes ranging from 19 individuals to 283 individuals. The studies utilized a variety of scales to measure antidepressant efficacy, including the Hamilton Depression Scale, with both the 17-item and 21-item versions, the montgomery and asberg depression rating scale, the hamilton depression rating scale, and the geriatric depression scale. Treatment durations varied, with some studies conducting assessments over a period as brief as 2 weeks, while others extended up to 12 weeks. A diverse array of antidepressant medications was analyzed across the studies, including Citalopram, Escitalopram, Mirtazapine, Fluoxetine, Paroxetine, Sertraline, and combinations thereof, along with selective serotonin reuptake inhibitors, Selective Serotonin and Norepinephrine Reuptake Inhibitors, Venlafaxine, and Fluvoxamine. This collection of studies provides a comprehensive dataset for evaluating the impact of various antidepressant treatments on MDD, taking into account a range of demographic variables and clinical outcomes. The data will be instrumental in assessing the overall effectiveness of antidepressants in a diverse, international cohort of patients with MDD (Table 1).

The methodological quality of the papers included in the meta-analysis was meticulously assessed using the NOS, a thorough instrument intended to evaluate the quality of non-randomized research, especially within systematic reviews and meta-analyses. The evaluation criteria encompassed three primary categories: (1) Selection of research groups; (2) Comparability of groups; and (3) Determination of either the exposure or result of interest for case-control or cohort studies, respectively. The study revealed a dispersion of NOS ratings that indicates predominantly excellent methodological quality. Two studies received a score of 7, indicating good quality, three studies were rated 8 points, and the bulk, consisting of six research, achieved the highest score of 9 points, denoting very high quality. Significantly, none of the studies employed blinding, and there was no indication of allocation concealment, features that could possibly introduce bias. Nonetheless, it was noted that none of the research exhibited indications of financing biases, implying that financial issues did not affect the study results. Moreover, all trials presented comprehensive outcome data, and there were no signs of early termination bias or baseline imbalances among participants. This thorough assessment resulted in a summary risk of bias and associated ratios, as outlined in Table 2 of the meta-analysis[10-20].

For the allelic model comparing the G allele to the C allele, the odds ratio (OR) was 0.86 with a 95%CI: 0.69-1.12, suggesting no significant association (P = 0.23). The model utilized for heterogeneity was random due to the moderate level of heterogeneity among the studies (I² = 46.46%), with a significant heterogeneity P value (P = 0.02). Both Egger's and Begg's tests showed P values of 0.64 and 0.83, respectively, indicating no significant publication bias. In the dominant model (CG vs CC), a fixed-effects approach yielded an OR of 0.81 (95%CI: 0.71–1.09), with the association remaining non-significant (P = 0.13). The heterogeneity was modest (I² = 15.81%), and not statistically significant (P = 0.26). Egger's and Begg's tests suggested a low probability of publication bias with P values of 0.10 and 0.09, respectively. The recessive model (GG vs CC) was analyzed using a random-effects model due to significant heterogeneity (I² = 40.64%, P = 0.04), and the OR was 0.80 (95%CI: 0.60–1.31). The association was not statistically significant (P = 0.37), and publication bias tests were non-indicative of bias (Egger's P = 0.25, Begg's P = 0.21).

The combined dominant model (CG+GG vs CC) also applied a random-effects model because of substantial heterogeneity (I² = 33.66%, P = 0.04) and reported an OR of 0.80 (95%CI: 0.59–1.09), with a non-significant association (P = 0.23). Tests for publication bias returned P values of 0.10 for Egger's and 0.08 for Begg's, indicating a low likelihood of bias. The locus model (GG vs CG+CC) using a fixed-effects model (due to moderate heterogeneity, I² = 0.00%, P = 0.907) indicated an OR of 1.15 (95%CI: 0.90–1.47) with a non-significant association (P = 0.20, Figure 2). Publication bias was not suggested by the Egger's (P = 0.29) and Begg's tests (P = 0.16). Lastly, the unfitted model (CC+GG vs GC), analyzed with a fixed-effects model due to non-significant heterogeneity (I² = 0.00%, P = 0.52), showed an OR of 0.97 (95%CI: 0.82–1.20) with a non-significant association (P = 0.95). The publication bias tests were inconclusive (Egger's P = 0.64, Begg's P = 0.21) (Table 3).

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Forest plots of antidepressant efficacy associated with serotonin 1A receptor polymorphism gene C-1019G polymorphism.

The meta-analysis did not reveal a significant impact of the 5-HTR1A C-1019G polymorphism on the efficacy of antidepressant treatment across various genetic models. The analyses demonstrate a consistent lack of statistically significant association, with low to moderate heterogeneity and minimal indications of publication bias across studies. This suggests that the C-1019G polymorphism in the 5-HTR1A gene may not be a major determinant of the clinical response to antidepressant medication in the assessed effective treatment group.

Publication bias within the meta-analysis was systematically evaluated to ascertain the robustness and reliability of the pooled results. Funnel plots, which visually depict the correlation between study size and effect estimates, were created for each analysis. The symmetrical arrangement of papers in these plots did not indicate publication bias, as asymmetry in such plots typically signifies this bias (Figure 3). Additionally, Egger's linear regression test was utilized as a statistical method to quantitatively evaluate the likelihood of publication bias among the papers. The test findings were non-significant for all evaluated variables (P > 0.05), hence reinforcing the lack of publication bias in this meta-analysis. These findings bolster the validity of the meta-analytic results, indicating that the estimated effects are not significantly affected by unpublished research or selective reporting.

Open in New Tab   Full Size Figure   Download Figure

Figure 3 Funnel plot analysis for publication bias.