Prevalence of GJB2 gene mutations in nonsyndromic hearing impairments: A systematic review and meta-analysis Authors: Ruijie Feng, Ernest Mangantig, Wan Shahriman Yushdie Wan Yusoff, Hui Li, Xiaoze Li, Mohd Nazri Abu Published: 2026 ABSTRACT Background:Nonsyndromic hearing impairment (NSHI) is the most common type of inherited hearing loss. Mutations in GJB2, which encodes connexin 26, are a major cause worldwide, but their prevalence and variant distribution vary across populations. To systematically evaluate the global prevalence of GJB2 mutations and 5 hotspot variants (c.35delG, c.176_191del16, c.235delC, c.299_300delAT, and c.109G > A) in NSHI patients, and assess their diagnostic relevance.Methods:A meta-analysis was conducted on studies published between January 2019 and June 2024 retrieved from PubMed, Scopus, Web of Science, and Cochrane Library. Only high-quality studies (Newcastle-Ottawa Scale ≥ 7) with molecular diagnostic data were included. Pooled prevalence, 95% confidence intervals, and heterogeneity (I2) were calculated. Subgroup and meta-regression analyses were performed according to geographic region.Results:Fifty-one studies from 24 countries covering 18,968 patients with NSHI were included. The overall prevalence of GJB2 mutations was 25.9% (95% confidence intervals: 21.7–30.1%; I2 = 96%). The variant-specific rates included c.235delC (15.2%), c.35delG (12.0%), c.109G > A (7.6%), c.299_300delAT (3.1%), and c.176_191del16 (0.7%). The highest prevalence occurred in the Middle East (38.5%), whereas Africa had the lowest (11.1%).Conclusion:GJB2 mutations show marked geographic and ethnic variability in populations with NSHI. These findings support region-specific genetic screening to improve the early diagnosis, counseling, and targeted interventions for hereditary hearing loss. 1. Introduction Hearing loss is a growing global health concern, currently affecting over 430 million people. This number corresponds to more than 5% of the world’s population and is expected to exceed 700 million (approximately 1 in every ten people) by 2050, according to the World Health Organization.[1] The burden of hearing impairment extends beyond communication difficulties, encompassing substantial economic and health care consequences. These include expenditures for hearing aids, cochlear implants, early intervention in newborns, speech and auditory rehabilitation, and long-term care for the affected individuals and their families. Clinically, hearing loss is defined as a hearing threshold of ≥20 dB and can range in severity from mild to profound.[2] Among the various etiological factors, genetic causes are particularly prominent, accounting for at least 50% of congenital and early onset hearing loss cases. Approximately 70% of genetic hearing loss cases are nonsyndromic (NSHI), with autosomal recessive inheritance accounting for nearly 80% of these cases.[3] One of the most extensively studied genes in NSHI is GJB2, which encodes the gap junction protein connexin 26.[4] This protein facilitates intercellular communication by enabling the passage of ions and small molecules across cell membranes.[5] Mutations in GJB2 disrupt gap junction channels, impair cochlear function, and ultimately result in hearing loss.[6,7] Extensive research has investigated the molecular consequences of GJB2 mutations, including their effects on protein function, membrane transport, and channel conductance.[6] However, despite growing evidence, considerable uncertainty persists regarding the diagnostic value of specific GJB2 variants across different populations. Advances in genetic testing technologies, particularly in direct gene sequencing, have significantly improved the precision of GJB2 mutation detection. Nonetheless, variations in mutation prevalence across regions, ethnic groups, and detection methods underscore the need for comprehensive synthesis. Therefore, the present study aimed to systematically review and meta-analyze the existing literature on GJB2 mutations in NSHI patients. We focused on hotspot variants and performed subgroup analyses based on geographic distribution and detection techniques, with emphasis on direct gene sequencing as the most accurate method for mutation identification. This study aimed to elucidate the clinical utility of GJB2 screening for NSHI and to inform future strategies for early genetic diagnosis and targeted intervention. 2. Methods 2.1. Literature search strategy A systematic literature search was performed using 4 electronic databases: PubMed, Scopus, Web of Science, and Cochrane Library, covering studies published from January 2019 to June 2024, a period marked by heightened research activity in this field. The search covered publications published from January 2019 to June 2024, a period marked by increased research activity in the field of genetic hearing loss. The search strategy included the following keywords and MeSH terms: GJB2, deafness gene, hearing loss, autosomal recessive, connexin 26, and Cx26. No language restrictions were applied. The protocol was prospectively registered with PROSPERO (registration number: CRD420250580566). 2.2. Inclusion and exclusion criteria Studies were selected based on the following inclusion criteria: full-text original research articles published between January 2019 and June 2024; reported the carrier frequency of GJB2 mutations associated with nonsyndromic hearing impairment (NSHI), along with valid and extractable statistical data; and employed molecular diagnostic techniques, such as microarray analysis, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry, or direct gene sequencing. The following were excluded: review articles, case reports, editorials, and conference abstracts; basic experimental studies or mechanistic investigations that did not involve human subjects; and duplicate publications or overlapping datasets. 2.3. Quality assessment and data extraction Two independent reviewers evaluated the methodological quality of the included studies using the Newcastle-Ottawa Scale. Only studies with a score ≥7 were considered high-quality and were included in the final analysis. Relevant data were extracted using a standardized form and cross-verified by both reviewers to ensure accuracy and consistency. 2.4. Statistical analysis Pooled carrier frequencies and the corresponding 95% confidence intervals (CIs) were calculated. Heterogeneity among studies was assessed using Cochran Q test (with P <.10 indicating significance) and quantified using the I2 statistic, where values >50% indicated substantial heterogeneity. A random-effects model was applied in cases of significant heterogeneity; otherwise, a fixed effects model was used. Publication bias was assessed using funnel plot inspection and Egger regression test (P < .05, considered significant). Subgroup and meta-regression analyses were conducted based on geographical region, age, severity of hearing impairment, and detection techniques to explore the potential sources of heterogeneity. Sensitivity analyses were performed by sequentially omitting the individual studies to test the robustness of the results. All statistical analyses were conducted using the R software (version 4.4.1). 3. Results 3.1. Characteristics of included studies A total of 1911 potential records were identified through database searches, using predefined keywords. After screening titles, abstracts, and full texts, 51 studies from 24 countries were included in the final meta-analysis, encompassing a total of 18,968 individuals diagnosed with NSHI. The majority of studies have employed sequencing-based methods for mutation detection, except for a few (e.g., Tian et al 2023, Dai et al 2022, Lin et al 2019)[8–10] that utilized microarray technology. Among the analyzed GJB2 variants, c.235delC was the most frequently reported, particularly in East Asian populations such as China, where multiple studies (e.g., Wu et al 2022, Zhou et al 2019, Sun et al 2022)[11–13] have documented high prevalence rates. Other commonly reported mutations included c.35delG, which was predominantly observed in European cohorts (e.g., Hungary, Portugal, and Russia), and c.109G > A, which showed a notable presence in several Chinese and Southeast Asian studies (Table 1). The detailed literature search and study selection process is illustrated in Figure 1. Table 1Characteristics of studies included in the meta-analysis.StudyLocationTest methodTotal NSHI sample(n) GJB2 35delG176–191del16235delC299–300delAT109G > AEmily2024[14]Toronto, CanadaSequencing7328-----Abdelaziz 2024[15]UAESequencing1062420-10-Sansovic 2024[16]Zagreb, CroatiaSequencing306124-----Li 2024[17]Beijing, ChinaSequencing11716--324Kriukelis 2024[18]Queensland, AustraliaSequencing6254511-2112Khawla 2024[19]MoroccoSequencing1453325---3Shaza 2024[20]QatarSequencing39199---1Vu 2023[21]Ho Chi Minh, Viet NamSequencing4912--129Tian 2023[8]Henan, ChinaMicroarray689-148425-Riza 2023[22]RomaniaSequencing124-10--1-Pshennikova 2023[23]Baikal Lake Region, RussiaSequencing1656723-114Aziz 2023[24]North Ossetia–AlaniaSequencing1094025----Peart 2023[25]South Florida, USASequencing1361210--11Margit 2023[26]Szeged, Pécs, and Debrecen, HungarySequencing1397171---2Liang 2023[27]Western Guangdong, ChinaSequencing9739--8315Lee 2023[28]Taiwan, ChinaSequencing350107-----Hu 2023[29]Wuhan, ChinaSequencing9650--16-47Fu 2023[30]Hainan, ChinaSequencing18313001310Jalal 2023[31]Isfahan, IranSequencing764835--2-Wu 2022[11]Beijing, ChinaSequencing588207--17467-Anke 2022[32]Southwest, GermanSequencing751611---1Sun 2022[13]Shandong, ChinaSequencing566412277741046284-Cláudia 2022[33]PortugalSequencing713-----Liu 2022[34]Zhejiang, ChinaSequencing10525--2124Lin 2022[35]Taiwan, ChinaSequencing34---9-25Güleç 2022[36]Istanbul, TurkeySequencing402141100----Dai 2022[9]Chengdu, ChinaMicroarray14271272611622-Bian 2022[37]Northwest, ChinaSequencing17942954891603157Adebolajo 2022[38]Ibadan, NigeriaSequencing602-----Niu 2021[39]Henan, ChinaSequencing13925021691Liu 2021[40]Shaoxing, ChinaMS27--2102-Guan 2021[41]Beijing, ChinaSequencing40410823511821Duan 2021[42]Qinghai, ChinaSequencing4406842401018Dobric 2021[43]Belgrade, SerbiaSequencing92-37----Al-Janabi 2021[44]Najaf, IraqSequencing95-75-53--Downie 2020[45]Victoria, AustraliaSequencing10622132112Zhou 2019[12]Shanxi, ChinaSequencing12012558121323225Zhang 2019[46]Heze, ChinaSequencing314850061915Yang 2019[47]Yunnan, ChinaSequencing2354102326-Yan 2019[48]Wenzhou, ChinaSequencing50611606951129Sun 2019[49]Eastern Shandong, ChinaSequencing42085066933-Mallory 2019[50]Atlanta, USASequencing5665----Lin 2019[10]Guangzhou, ChinaMicroarray1517100688-Li 2019[51]Linyi, ChinaSequencing116350328114Wang 2019[52]Northern Jiangsu, ChinaSequencing11761274616-Kurtulgan 2019[53]Sivas, TurkeySequencing53149----Khan 2019[54]Khyber Pakhtunkhwa, PakistanSequencing40156----Gal 2019[55]Budapest, HungarySequencing2395340---3Samuel 2019[56]GhanaSequencing14011-----Erick 2019[57]Lima, PeruSequencing1334313-1-2Felippe 2019[58]Rio, BrazilSequencing100144---- Characteristics of studies included in the meta-analysis. Flow diagram for the literature search process. 3.2. Related study indexes in the meta-analysis The pooled detection rate for GJB2 mutations across all included studies was estimated at 25.9% (95% CI: 21.7%–30.1%), with substantial heterogeneity among studies (I2 = 96%, P <.001). Among the specific hotspot variants, c.235delC exhibited the highest estimated carrier rate (15.2%), followed by c.35delG (12.0%), both of which demonstrated significant heterogeneity (I2 > 97%). The estimated rates of c.299_300delAT, c.109G > A, and c.176_191del16 were 3.1%, 7.6%, and 0.7%, respectively. Owing to the high heterogeneity (P <.001) observed in these analyses, random-effects models were consistently applied. These findings underscore the notable regional and ethnic differences in GJB2 mutation frequencies across global NSHI populations. The detailed characteristics of the included studies are summarized in Table 2. Table 2Summary of the related study indexes in the meta-analysis.GeneStudies, NSample sizeEvent numberTest of heterogeneityModelEstimated rate (%)95%CI (%)I2 (%)P value GJB2 4517,907391996<.001Random25.921.7–30.135delG3716,85562697<.001Random12.06.5–17.5176–191del161813,93313366<.001Random0.70.4–1.0235delC3016,143235899<.001Random15.210.4–20.0299–300delAT2916,12161193<.001Random3.12.1–4.2109G > A25739730191<.001Random7.61.9–13.2 Summary of the related study indexes in the meta-analysis. 3.3. Mutation specific meta-analysis Figures 2–7 present the pooled prevalence estimates and effect sizes of overall GJB2 mutations and 5 major hotspot variants (c.35delG, c.176_191del16, c.235delC, c.299_300delAT, and c.109G > A) across the included NSHI cohorts. Significant heterogeneity was detected in the pooled analyses (P < .10 or I2 > 65%), leading to the use of random-effects models for calculating pooled odds ratios. The resulting forest plots illustrate the association between each mutation and the risk of NSHI, highlighting substantial genetic variability among populations. These findings reinforce the importance of region- and ethnicity-specific strategies for genetic screening and early intervention. We then analyzed publication bias and conducted a sensitivity analysis. The results are presented in Figures S1 and S2. Forest plot of the effect of GJB2 mutations on NSHI cohorts. Forest plot of the effect of 35delG variant on NSHI cohorts. Forest plot of the effect of 176 to 191del16 variant on NSHI cohorts. Forest plot of the effect of 235delC variant on NSHI cohorts. Forest plot of the effect of 299-300delAT variant on NSHI cohorts. Forest plot of the effect of 109G>A variant on NSHI cohorts. 3.4. Subgroup analysis Subgroup meta-analysis based on geographic region revealed significant differences in the prevalence of GJB2 mutations across 45 studies (Table 3). The overall proportion of GJB2 mutations among NSHI patients was 25.9% (95% CI: 21.7%–30.1%), accompanied by marked heterogeneity (I2 = 96.1%, P < .001). The highest prevalence was observed in the Middle East (38.5%), followed by Europe (30.8%), and East Asia (25.6%), whereas lower prevalence rates were found in Australia (13.4%) and Africa (11.1%). Table 3Subgroup meta-analysis of GJB2 mutations.SubgroupNProportion(95%CI)I2 P a P b GJB2 mutations North America region318.8% (0.6%–37.0%)91.3%<.001.059 South America region223.1% (5.1%–41.0%)91.5%<.001 Europe region730.8% (18.9%–42.7%)96.4%<.001 Australia region213.4% (0.2%–26.7%)91.0%<.001 African region311.1% (0.0%–22.3%)90.8%<.001 The Middle East region638.5% (26.4%–50.7%)87.5%<.001 East Asia region2225.6% (20.4%–30.7%)96.2%<.001 Overall 45 25.9% (21.7%–30.1%) 96.1% <.001 35delG variant North America region27.8% (4.0%–11.5%)0.0%.721<.001 South America region26.7% (1.0%–12.3%)68.6%.074 Europe region723.6% (12.1%–35.2%)94.0%<.001 Australia region26.5% (0.0%–16.8%)90.5%.001 The Middle East region830.2% (14.8%–45.5%)96.3%<.001 East Asia region160.3% (0.0%–0.7%)74.0%<.001 Overall 37 12.0% (6.5%–17.5%) 96.7% <.001 176–191del16 variant Australia region11.9% (0.0%–4.5%)--0.381 East Asia region170.7% (0.4%–0.9%)67.8%<.001 Overall 18 0.7% (0.4%–1.0%) 66.4% <.001 235delC variant South America region10.75% (0.0%–2.2%)--.118 Europe region10.61% (0.0%–1.8%)-- Australia region20.35% (0.0%–0.8%)0.0%.519 The Middle East region228.1% (0.0%–81.9%)99.1%<.001 East Asia region2416.6% (12.4%–20.8%)95.6%<.001 Overall 30 15.2% (10.4%–20.0%) 98.6% <.001 299–300delAT variant .077 North America region10.7% (0.0%–2.2%)-- Europe region20.7% (0.0%–1.6%)0.0%.842 Australia region20.2% (0.0%–0.5%)0.0%.411 African region20.8% (0.0%–3.1%)45.0%.177 East Asia region224.0% (2.8%–5.3%)89.0%<.001 Overall 29 3.1% (2.1%–4.2%) 93.0% <.001 109G > A variant .800 North America region10.7% (0.0%–2.2%)-- South America region11.5% (0.0%–3.6%)-- Europe region41.5% (0.1%–2.5%)0.0%.866 Australia region21.9% (0.1%–2.9%)0.0%.981 African region22.2% (0.1%–4.3%)0.0%.859 East Asia region1512.1% (2.4%–21.8%)94.7%<.001 Overall 25 7.6% (1.9%–13.2%) 91.2% <.001 Pa for heterogeneity within each subgroup.Pb for heterogeneity between subgroups with meta-regression analysis. Subgroup meta-analysis of GJB2 mutations. Pa for heterogeneity within each subgroup. Pb for heterogeneity between subgroups with meta-regression analysis. Similar regional disparities were observed for the hotspot mutations. For instance, c.35delG was highly prevalent in Europe (23.6%) and the Middle East (30.2%) but was almost absent in East Asia (0.3%). In contrast, c.235delC and c.109G > A were predominantly observed in East Asian populations, with estimated prevalence rates of 16.6% and 12.1%, respectively, but rarely reported in other regions. However, when stratified by genetic variations, it was shown that these “geographical differences” largely reflected the differences in the genetic background of the population (such as the high frequency of 35delG in the European population and its rarity in the East Asian population), rather than being the result of independent geographical environmental influences. Furthermore, subgroup analyses were performed according to age, severity of hearing impairment, and detection methodologies. The findings indicated no significant differences across age groups for the overall GJB2 gene (P = .987) (Table S1). However, within the 35delG mutation subgroup, age demonstrated a highly significant effect (P < .001), primarily attributable to an anomalously elevated value (51.1%) reported in a study comprising a heterogeneous sample of newborns, children, and adults. These results imply that the impact of age may differ depending on the specific gene and could be influenced by the demographic composition of individual study samples. Neither the severity of hearing impairment nor the detection techniques yielded significant effects within each gene subgroup (Tables S2 and S3); nonetheless, certain subgroups may present unstable estimates due to the limited number of studies included (n = 1). 4. Discussion This meta-analysis confirmed that GJB2 mutations, particularly 5 hotspot variants (c.35delG, c.176_191del16, c.235delC, c.299_300delAT, and c.109G > A), were significantly associated with NSHI. Clear evidence of regional and ethnic variability has been observed.[59–66] The overall pooled prevalence of GJB2 mutations was 25.9% (95% CI: 21.7%–30.1%), with estimates ranging widely from 3.3% to 63.2% across individual studies, highlighting substantial heterogeneity (I2 = 96%, P < .001). Possible contributors to this heterogeneity include differences in genetic backgrounds, sample sizes, case selection criteria, and detection methods.[67] This genetic variability underscores the need to implement population-specific screening strategies. Among the hotspot variants, c.235delC demonstrated the highest pooled prevalence (15.2%) and is particularly common in East Asian populations (China, Japan, and Korea).[61] A high frequency of GJB2 variants has also been reported in some Middle Eastern cohorts (including Iraqi samples about 55.8%), indicating that parts of the Middle East may also harbor elevated GJB2 burdens, which may reflect local founder events and the well-documented impact of high consanguinity on recessive disorder prevalence in several Arab populations.[44,63,68] In contrast, c.35delG, with an overall prevalence of 12.0%, was predominantly identified in European and Middle Eastern populations, especially in Hungary (51.1%), Iraq (78.9%), and Iran (46.1%), which is rare or absent in East Asia, likely reflecting distinct evolutionary histories and founder effects, a distribution consistent with a well-documented founder effect and the historical spread of this allele across Caucasian/Eurasian populations.[59,60] c.176_191del16 and c.299_300delAT variants have been reported in Chinese cohorts, with pooled prevalence rates of 0.7% and 3.1%, respectively, reflecting their largely population-specific distribution and contribution to the genetic architecture of NSHI in East Asia.[63] c.109G > A exhibited a moderate overall prevalence of 7.6%, with higher frequencies documented in Eastern China (up to 15.8%), but was infrequently reported in other regions. Additionally, several localized Chinese cohorts have reported notably higher allele frequencies (for example, the Xiamen cohort), underscoring intra-China regional variation and the need for area-tailored screening panels.[69] Finally, many pathogenic GJB2 genotypes occur as compound heterozygotes, and genotype class (biallelic truncating vs truncating/non-truncating vs non-truncating) correlates with the degree of hearing loss, indicating that compound states and variant type meaningfully influence clinical severity and age of onset.[70] Subgroup analyses and meta-regression revealed considerable heterogeneity within regions for all major variants (I2 > 65%), indicating substantial variability even among studies conducted in similar geographic settings. For some variants, including c.176_191del16, c.235delC, and c.109G > A, no statistically significant differences between regional subgroups were observed, suggesting that intrapopulation factors or study-level methodological differences may contribute to variability in prevalence estimates. In contrast, for c.35delG, meta-regression analysis identified geographic region as a significant source of heterogeneity (P < .001), reinforcing its strong regional association. Compared with previous reports, this study offers several advantages. It incorporates a larger number of recent studies (N = 45), represents a broader geographic scope, and applies meta-regression to assess the potential sources of between-study and between-region heterogeneity.[71–73] These strengths not only enhance the accuracy of prevalence estimates, but also provide actionable evidence to guide the development of region-specific genetic testing strategies and targeted public health interventions. This provides an important epidemiological basis for the development of targeted gene therapy intervention measures. For instance, in the European population, base editing therapies directed at the c.35delG mutation should be prioritized, whereas in the East Asian population, gene replacement strategies focusing on the c.235delC and c.299_300delAT mutations warrant greater emphasis. Overall, this meta-analysis reinforces the complexity of the genetic landscape underlying NSHI, and highlights the importance of incorporating regional and ethnic mutation patterns into both clinical practice and public health strategies. 5. Limitations This study had several limitations that may influence the interpretation and generalizability of the findings. First, the quality and methodological rigor of the included studies varied, with some studies assessing only selected hotspot mutations using microarrays, MALDI-TOF-MS, or targeted sequencing panels. Moreover, in the original studies, the reporting of detection methods was insufficient and the number of microarray studies was limited. Consequently, the rare GJB2 variants that were not covered in these studies may have been underestimated. Second, owing to incomplete reporting of relevant data in some original publications, it was not feasible to comprehensively analyze the impact of additional factors, such as age at diagnosis, audiometric phenotypes, or population genetic backgrounds, which could potentially influence prevalence estimates and heterogeneity. This study exclusively concentrated on the 5 predominant mutation sites of the GJB2 gene, excluding other less common pathogenic variants. Consequently, this approach may result in an underestimation of the total mutation frequency within the GJB2 gene. Finally, our meta-analysis was limited by the geographic distribution of the available data, with underrepresentation of studies from certain regions, particularly Africa and South America. As a result, these findings may not fully capture the global distribution of GJB2 mutations. 6. Conclusion This comprehensive meta-analysis evaluated the prevalence and distribution of GJB2 gene mutations and 5 key hotspot variants (c.35delG, c.176_191del16, c.235delC, c.299_300delAT, and c.109G > A) among NSHI populations across 51 regions. Despite the methodological heterogeneity among studies, the findings provide valuable insights into the geographic and ethnic patterns of pathogenic GJB2 variants, reinforcing their critical role in hereditary hearing loss. These results support the development of region-specific genetic screening strategies and highlight the need for additional research in underrepresented populations to further characterize mutation profiles and improve early diagnosis, genetic counseling, and preventive public health initiatives. Acknowledgments We thank all investigators who contributed to this study. We also thank all the patients enrolled in this study. Author contributions Conceptualization: Ruijie Feng, Xiaoze Li, Mohd Nazri Abu. Data curation: Ruijie Feng, Wan Shahriman Yushdie Wan Yusoff. Writing – original draft: Ruijie Feng. Formal analysis: Ernest Mangantig, Hui Li. Visualization: Ernest Mangantig, Wan Shahriman Yushdie Wan Yusoff. Writing – review & editing: Ernest Mangantig, Xiaoze Li, Mohd Nazri Abu. Validation: Hui Li. Project administration: Mohd Nazri Abu.