Genetic Landscape and Carrier Screening of Deafness-Associated Genes in a Prenatal Cohort from Fujian, China

Abstract

Purpose

This study aimed to investigate the carrier frequencies of pathogenic variants in the GJB2, SLC26A4, GJB3, and MT-RNR1 genes among a Chinese pregnant population and their male partners in southern China.

Methods

In this retrospective cohort, 1402 pregnant women underwent carrier screening for these four deafness-related genes using targeted next-generation sequencing, and those not willing to participate the study were excluded. Spouses received follow-up testing and counseling were also included. Prenatal diagnosis and postnatal genetic testing with hearing screening were performed to assess newborn outcomes.

Results

Among the 1192 pregnant women included in the final analysis, 611 were identified as carriers of pathogenic or likely pathogenic variants in GJB2, SLC26A4, GJB3, or MT-RNR1. GJB2 showed the highest carrier frequency (23.6%), with c.109G>A and c.235delC being the most frequently detected variants. Additionally, 349 spouses underwent the testing, with similar predominance of GJB2. Cascade screening identified 41 at-risk couples, including three high-risk couples undergoing prenatal diagnosis. Among 39 newborns, most were heterozygous carriers of deafness-associated variants; three infants homozygous for GJB2 c.109G>A variant exhibited mild hearing loss but maintained normal communication ability. These findings support the utility of carrier screening and genetic counseling in assessing hereditary hearing loss risk.

Conclusions

This study demonstrated a substantial carrier frequency of deafness-associated variants among pregnant women and their partners in southern China. Sequential partner screening following identification of maternal carriers enabled the detection of couples at reproductive risk, facilitating genetic counseling and prenatal diagnosis for high-risk pregnancies.

Introduction

According to the latest data of World Health Organization, over 500 million individuals experience hearing loss (HL) during their lifetime, with at least 430 million requiring treatment.¹ The affected individuals are reported to suffer from auditory implications, along with negative psychological and mental impacts.^2,3 The etiologies of HL encompass genetic and/or non-genetic factors,^4,5 with approximately 60% of them having hereditary hearing loss (HHL).⁶

In clinical practice, HHL is characterized by substantial heterogeneity in both genotype and phenotype.⁷ According to literature, 70% of newborns with HHL are asymptomatic, while the remaining newborns are often accompanied by additional abnormalities.⁸ Advances in high-throughput sequencing technologies has revolutionized the identification of the molecular bases of HHL,⁹ and large-scale parallel sequencing has been established as an effective conventional diagnostic and investigative method for HHL.¹⁰ For instance, Liu et al characterized variants in GJB2, GJB3, SLC26A4 and MT-RNR1 in patients from Northwest China,¹¹ whereas, Wu et al evaluated the contribution of monogenic inheritance to HL in a large Chinese cohort and further delineated the spectrum of HL-associated genes.¹² In addition, Du et al investigated the mutation spectra of GJB2, mitochondrial 12S rRNA, and SLC26A4 genes among Han Chinese, Hui people, and Uyghur ethnicities in sensorineural HL patients in Northwest China.¹³ These studies have demonstrated substantial geographic and ethnic heterogeneity in the genetic architecture of HL. However, data from southern China remain comparatively limited, particularly regarding the combined spectrum of major HL-associated variants and their subsequent clinical follow-up.

Such regional variation may directly affect the interpretation of carrier screening and reproductive risk assessment. Population-specific carrier frequencies are also clinically relevant for reproductive counseling. When a pregnant woman is identified as carrying a clinically relevant variant associated with autosomal recessive HL, targeted testing of her partner may help identify couples at increased reproductive risk and guide subsequent prenatal diagnosis. Previous studies have also emphasized that prenatal genetic testing should be accompanied by appropriate counseling, as reproductive decisions are influenced by both medical and psychosocial factors.^14,15 In the present study, we aimed to characterize the spectrum and frequencies of variants in GJB2, GJB3, SLC26A4, and mitochondrial 12S rRNA (MT-RNR1) in a retrospective cohort from southern China. In addition, we also evaluated partner screening, prenatal diagnosis, and available newborn hearing outcomes among genetically at-risk pregnancies.

Methods

Study Population

In this retrospective cohort study, 1402 pregnant women who underwent deafness gene carrier screening at Fujian Maternity and Child Health Hospital between January 2023 and August 2024 were initially assessed for eligibility. Using the HearingCare kit purchased from BGI (Shenzhen, China), the screening panel included 127 loci across four HL-associated genes (GJB2, SLC26A4, GJB3, MT-RNR1), which was designed based on a curated database of pathogenic variants associated with hearing loss in the Chinese population. Pregnant women not willing to participate the study were excluded. Before screening, all women were well informed about hearing health and the genetic screening process. Following this, they provided written informed consent. This study was approved by the Ethics Review Committee of Fujian Maternity and Child Health Hospital (approval No.: 2023KY021). All aspects of the investigation were conducted in accordance with the ethical guidelines of the Declaration of Helsinki.

Next-Generation Sequencing (NGS) Analysis

We detected 127 variant sites of the 4 genes (GJB2, SLC26A4, GJB3, MT-RNR1) using the deafness gene detection kit (Combinatorial Probe-Anchor Synthesis, BGI Biotech, Shenzhen, China). Twenty common variant hotspots were selected for the detection: GJB2 gene (c.35delG, c.176del16, c.235delC, and c.299_300delAT), GJB3 gene (c.538C>T and c.547G>A), SLC26A4 gene (c.281C>T, c.589G>A, c.2168A>G, c.2162C>T, c.1174A>T, c.1226G>A, c.1229C>T, 1975G>C, c.2027T>A, 1VS15+5G>A and IVS7-2A>G), as well as mitochondrial 12SrRNA gene (m.1494C>T, m.1095T>C and m.1555A>G).

For the NGS, 2 mL of venous blood was extracted into EDTA anticoagulant tubes, and genomic DNA was extracted using a whole blood nucleic acid extraction kit (BGI Biotech, Shenzhen, China). Fragmented DNA underwent end-repair, A-tailing, and adapter ligation following standard Illumina library preparation protocols. Hybridization was conducted at 47°C for 16–20 hours, followed by capture washing and post-capture PCR amplification. Enriched libraries were quantified by Qubit fluorometer and qPCR using the KAPA Library Quantification Kit (BGI Biotech, Shenzhen, China). Qualified libraries were converted into DNA nanoballs (DNBs) through rolling circle amplification and sequenced on the MGISEQ-200 platform (BGI, Shenzhen, China) using the combinatorial Probe Anchor Synthesis (cPAS) chemistry.

Trimmomatic (v0.39) was used for the processing of raw sequencing data by removing adapter sequences and low-quality reads. Clean reads were aligned to the human reference genome GRCh37/hg19 using BWA-MEM (v0.7.17). Duplicate reads were marked using Picard tools (v2.23.0) and local realignment around indels was performed with GATK (v4.1.9.0). Single nucleotide variants and small insertions/deletions were identified using GATK HaplotypeCaller and Samtools mpileup. Variant annotation was performed with Ensembl Variant Effect Predictor against multiple databases including OMIM, MGI, Gene Ontology, and HGNC. Population frequencies were queried against gnomAD, ExAC, 1000 Genomes, dbSNP, and an in-house BGI database. Variants with minor allele frequency >0.5% in any population database were filtered out. Candidate pathogenic variants were classified according to the American College of Medical Genetics and Genomics (ACMG-AMP) guidelines.

Genetic Counseling

High-risk cases are defined as couples where both partners carry mutations graded as pathogenic, or where both mutations reside on the same gene (excluding the 109 site). Conversely, instances where at least one or both partners carry the 109 site mutation are classified as at risk. Follow up and recall were performed to pregnant women who are carriers of deafness gene variant sites. For these individuals, we recommend their spouses to undergo deafness gene testing, provide genetic counseling, and reproductive guidance. Those declined NSG screening were excluded.

Statistical Analysis

The statistical analysis of genetic variations associated with deafness was conducted using SPSS software (version 25.0). As this study was primarily descriptive, categorical variables were summarized as frequencies and percentages (%). The frequency of each deafness-associated variant was calculated as the number of individuals carrying the corresponding variant divided by the total number of participants included in the analysis. For the nuclear genes GJB2, GJB3, and SLC26A4, allele frequencies were calculated based on the number of variant alleles among the total number of alleles (2N). For the mitochondrial gene MT-RNR1, carrier frequency was calculated as the number of individuals carrying the corresponding variant divided by the total number of tested females × 100%, because a diploid denominator is not applicable to mitochondrial DNA.

Result

Cohort Characteristics

Among the 1402 women, 1192 were finally included in this study after excluding women not willing to participate the study due to personal reasons (Figure S1), with 611 (51.25%) carrying at least one variant in GJB2, SLC26A4, GJB3 and MT-RNR1 (Table 1). Specifically, the detection rates of GJB2, SLC26A4, GJB3 and mtDNA mutations in pregnant women were 23.6%, 0.74%, 1.76% and 2.43%, respectively (Table 1). Among the spouses of the 611 pregnant women carrying HL-related variants, 262 opted not to undergo genetic testing after genetic counseling, whereas 349 completed testing and were included in the final analysis.

Table 1.

Summary of GJB2, GJB3, SLC26A4 and MT-RNR1 Variants in Pregnant Women (N=1192)

Gene	Mode of Inheritance	Heterozygote	Homozygote	Compound Heterozygote	VAF
GJB2	AR	508 (21.2%)	24 (2.0%)	9 (0.4%)	565 (23.6%)
GJB3	AR/AD	16 (0.70%)	N/A	1 (0.04%)	17 (0.74%)
SLC26A4	AR	42 (1.76%)	N/A	N/A	42 (1.76%)
MT-RNR1	MI	1 (0.08%)	28 (2.35%)	N/A	29 (2.43%)

Notes: For the nuclear genes GJB2, GJB3, and SLC26A4, allele frequencies were calculated based on the number of variant alleles among the total number of alleles (2N). For the mitochondrial gene MT-RNR1, carrier frequency was calculated as the number of individuals carrying the corresponding variant divided by the total number of tested females × 100%, because a diploid denominator is not applicable to mitochondrial DNA.

Abbreviations: AR, autosomal recessive; AD, autosomal dominant; MI, maternal inheritance; VAF, variant allele frequency; N/A, not available.

Variant Spectrum in Women

For hotspot GJB2 variants, 63 c.235delC variants were identified in pregnant women, including 56 heterozygotes and 7 compound heterozygotes. Additionally, we identified 466 women with the c.109G>A variant, including 436 heterozygotes, and 23 homozygotes, and 7 compound heterozygotes (Table 2). These variants, along with the c.299_300delAT, c.139G>T, c.176_191del16, c.1692dupA, c.269T>C, c.35dupG, and c.35dupG, yielded a total variant frequency of 23.6% for GJB2.

Table 2.

The Carrier Frequency of GJB2, GJB3, SLC26A4 and MT-RNR1 Variant in the Pregnant Women (N=1192)

Gene	Variant	Inheritance	Pathogenicity	Heterozygote	Homozygote	Compound Heterozygote	VAF
GJB2
	c.235delC	AR	P	56 (2.35%)	N/A	7 (0.29%)	63 (2.64%)
	c.299_300delAT	AR	P	7 (0.29%)	N/A	N/A	7 (0.29%)
	c.109G>A	AR	P	436 (18.29%)	23 (1.93%)	7 (0.29%)	489 (20.51%)
	c.139G>T	AR	P	1 (0.04%)	N/A	N/A	1 (0.04%)
	c.176_191del16	AR	P	2 (0.08%)	N/A	1 (0.04%)	3 (0.12%)
	c.1692dupA	AR	P	1 (0.04%)	N/A	N/A	1 (0.04%)
	c.269T>C	AR	P	1 (0.04%)	N/A	N/A	1 (0.04%)
	c.35dupG	AR	P	3 (0.13%)	N/A	N/A	1 (0.04%)
	c.35G>T	AR	P	1 (0.04%)	N/A	N/A	1 (0.04%)
GJB3
	c.538C>T	AD/AR	LP	8 (0.33%)	N/A	1 (0.04%)	9 (0.37%)
	c.547G>A	AD/AR	LP	8 (0.33%)	N/A	N/A	8 (0.33%)
SLC26A4
	IVS7-2A>G	AR	P	33 (1.38%)	N/A	N/A	33 (1.38%)
	c.2168A>G	AR	P	2 (0.08%)	N/A	N/A	2 (0.08%)
	c.1229C>T	AR	P	3 (0.13%)	N/A	N/A	3 (0.13%)
	c.1174A>T	AR	P	1 (0.04%)	N/A	N/A	1 (0.04%)
	c.589G>A	AR	P	1 (0.04%)	N/A	N/A	1 (0.04%)
	c.1975G>C	AR	P	1 (0.04%)	N/A	N/A	1 (0.04%)
	c.1692dupA	AR	P	1 (0.04%)	N/A	N/A	1 (0.04%)
MT-RNR1
	m.1555A>G	MI	P	1 (0.08%)	11 (0.92%)	N/A	12 (1.00%)
	m.1494C>T	MI	P	N/A	1 (0.08%)	N/A	1 (0.08%)
	m.1095T>C	MI	P	N/A	16 (1.34%)	N/A	16 (1.34%)

Notes: For the nuclear genes GJB2, GJB3, and SLC26A4, allele frequencies were calculated based on the number of variant alleles among the total number of alleles (2N). For the mitochondrial gene MT-RNR1, carrier frequency was calculated as the number of individuals carrying the corresponding variant divided by the total number of tested females × 100%, because a diploid denominator is not applicable to mitochondrial DNA.

Abbreviations: AR, autosomal recessive; AD, autosomal dominant; MI, maternal inheritance; VAF, variant allele frequency; N/A, not available; P, pathologic; LP, likely pathologic.

SLC26A4 was the second most frequently mutated gene in the pregnant women. In total, 42 pregnant women carried IVS7-2A>G variants. Therefore, IVS7-2A>G was identified as the predominant mutation of SLC26A4 in pregnant women (Table 2). For the GJB3 variant, 17 were carriers of c.538C>T and c.547G>A variants. For MT-RNR1, 29 were carriers of m.1555A>G, m.1494C>T and m.1095T>C. A total of 16 pregnant women carried compound heterozygous variants in GJB2 and GJB3. Of these, 8 individuals carried compound heterozygous variants in GJB2, including c.235delC+c.109G>A combination being the most prevalent (n=7) and c.176_191del16+c.109G>A (n=1). Additionally, 1 individual harbored compound heterozygous variants in GJB3 (Table 2).

Thirteen pregnant women were found to carry variants in two different genes associated with HL. The most prevalent combination was GJB2 c.109G>A/wt with SLC26A4 IVS7-2A>G/wt (n=5). Six mitochondrial variants coexisted with GJB2 (3 variants), and SLC26A4 (3 variants), respectively (Table 3).

Table 3.

Pregnant Women Identified with Dual Gene Variants

GJB2	SLC26A4	GJB3	MT-RNR1	N
c.109G>A/wt	IVS7-2A>G/wt	-	-	5
c.109G>A/c.109G>A	c.1229C>T/wt	-	-	1
c.109G>A/wt	-	c.538C>T/wt	-	1
c.109G>A/wt	-	-	m.1494C>T homo	1
c.109G>A/wt	-	-	m.1095T>C homo	1
c.109G>A/c.109G>A	-	-	m.1555A>G homo	1
c.35dupG/wt	-	-	m.1095T>C homo	1
-	IVS7-2A>G/wt	-	m.1555A>G homo	1
-	c.1229C>T/wt	-	m.1095T>C homo	1

Abbreviation: wt, wild type.

Cascade Screening in Spouses

In this section, we detected the frequency of the variant in the spouses of pregnant women who had been identified with hearing loss variants for genetic testing. The analysis showed that 161 male subjects carried at least one variant allele (Table 4). Specifically, the detection rates of GJB2, GJB3, SLC26A4 and mtDNA variants in tested spouse were 21.78%, 0.72%, 3.44% and 2.0%, respectively (Table 4). The GJB2 gene emerged as the preeminent deafness-associated genetic variant, exhibiting a notable prevalence in our study. Specifically, we identified 126 heterozygous, 11 homozygous and 4 compound heterozygous.

Table 4.

Summary of GJB2, GJB3, SLC26A4 and MT-RNR1 Variant in the Tested Male Partners (N=349)

Gene	Mode of Inheritance	Heterozygote	Homozygote	Compound Heterozygote	VAF
GJB2	AR	126 (18.05%)	11 (3.15%)	4 (0.58%)	152 (21.78%)
GJB3	AR/AD	5 (0.72%)	N/A	N/A	5 (0.72%)
SLC26A4	AR	20 (2.87%)	2 (0.29%)	N/A	24 (3.44%)
MT-RNR1	MI	N/A	7 (2.0%)	N/A	7 (2.0%)

Notes: For the nuclear genes GJB2, GJB3, and SLC26A4, allele frequencies were calculated based on the number of variant alleles among the total number of alleles (2N). For the mitochondrial gene MT-RNR1, carrier frequency was calculated as the number of individuals carrying the corresponding variant divided by the total number of tested females × 100%, because a diploid denominator is not applicable to mitochondrial DNA.

Abbreviations: AR, autosomal recessive; MI, maternal inheritance; AD, autosomal dominant; N/A, not available; VAF, variant allele frequency.

Notably, the c.109G>A variant was the most prevalent, encompassing 130 patients, further classified into 106 heterozygotes, 11 homozygotes and 2 compound heterozygotes. The c.235delC variant was detected in 19 male individuals, including 17 heterozygotes and 2 compound heterozygotes (Table 5). For SLC26A4, IVS7-2A>G variants were found in 22 male individuals, including 20 with heterozygous variants, and 2 with homozygous variants. In the GJB3 gene, we identified 5 patients with heterozygous variants: 4 harbored c.538C>T and 1 harbored c.547G>A. Furthermore, 7 patients were found to carry variants in the MT-RNR1, including 6 with the homoplasmic m.1555A>G variant and 1 with homoplasmic m.1095T>C (Table 5).

Table 5.

The Carrier Frequency of GJB2, GJB3, SLC26A4 and MT-RNR1 Variant in the Tested Male Partners (N=349)

Gene	Variant	Inheritance	Pathogenicity	Heterozygote	Homozygote	Compound Heterozygote	VAF
GJB2
	c.235delC	AR	P	17 (2.44%)	N/A	2 (0.29%)	19 (2.73%)
	c.299_300delAT	AR	P	2 (0.29%)	N/A	N/A	2 (0.29%)
	c.109G>A	AR	P	106 (15.19%)	11 (3.15%)	2 (0.29%)	130 (18.63%)
	c.34G>T	AR	LP	1 (0.14%)	N/A	N/A	1 (0.14%)
GJB3
	c.538C>T	AD/AR	LP	4 (0.57%)	N/A	N/A	4 (0.57%)
	c.547G>A	AD/AR	LP	1 (0.14%)	N/A	N/A	1 (0.14%)
SLC26A4
	IVS7-2A>G	AR	P	20 (2.87%)	2 (0.57%)	N/A	24 (3.44%)
MT-RNR1
	m.1555A>G	MI	P	N/A	6 (1.72%)	N/A	6 (1.72%)
	m.1095T>C	MI	P	N/A	1 (0.29%)	N/A	1 (0.29%)

Notes: For the nuclear genes GJB2, GJB3, and SLC26A4, allele frequencies were calculated based on the number of variant alleles among the total number of alleles (2N). For the mitochondrial gene MT-RNR1, carrier frequency was calculated as the number of individuals carrying the corresponding variant divided by the total number of tested females × 100%, because a diploid denominator is not applicable to mitochondrial DNA.

Abbreviations: AR, autosomal recessive; AD, autosomal dominant; MI, maternal inheritance; VAF, variant allele frequency; N/A, not available; P, pathologic; LP, likely pathologic.

Fourteen patients carried variants in two HL-related genes, involving combinations of variants in genes such as GJB2, SLC26A4, GJB3 and MT-RNR1. The most prevalent combination was GJB2 c.109G>A/wt with SLC26A4 IVS7-2A>G/wt (n=4), followed by GJB2 c.235delC/wt with SLC26A4 IVS7-2A>G/wt (n=3) and GJB2 c.235delC/wt with SLC26A4 c.109G>A/wt (n=2). Other combinations included GJB2 c.109G>A/wt with homozygous SLC26A4 IVS7-2A>G (n=1); homozygous GJB2 c.109G>A with SLC26A4 IVS7-2A>G/wt (n=1); GJB2 c.109G>A/wt with GJB3 c.538C>T/wt (n=1); and GJB2 c.109G>A/wt with homoplasmic MT-RNR1 m.1095T>C (n=1) (Table 6).

Table 6.

Tested Male Partners Identified with Dual Gene Variants

GJB2	SLC26A4	GJB3	MT-RNR1	N
c.235delC/wt	IVS7-2A>G/wt	--	-	3
c.235delC/wt	c.109G>A/wt			2
c.109G>A/wt	IVS7-2A>G/wt	-	-	4
c.109G>A/wt	IVS7-2A>G/IVS7-2A 2A>G	-	-	1
c.109G>A/c.109G>A	IVS7-2A>G/wt	-	-	1
c.109G>A/wt	-	c.538C>T/wt	-	1
c.109G>A/wt	-	-	m.1095T>C homo	1

Abbreviation: wt, wild type.

At-Risk Couples and Prenatal Diagnosis

Among the 349 spouses participated in the testing, 161 were identified as carriers of at least one variant. Among these 161 couples, 41 couples were found to carry variants in the same gene. Specifically, 31 couples were both heterozygous for the GJB2 c.109G>A variant; 4 couples consisted of husbands carrying GJB2 c.235delC and wives carrying c.109G>A; 3 couples showed the reverse combination (husbands with c.109G>A and wives with c.235delC); 1 couple had a husband carrying c.235delC and a wife carrying c.299_300delAT; 1 couple had a wife carrying c.109G>A and a husband carrying c.34G>T; and 1 couple were both heterozygous for the SLC26A4 IVS7-2A>G variant.

NGS-based variant screening was first performed to identify at-risk couples. Prenatal diagnosis through amniocentesis was subsequently offered only to couples with a high risk of having affected offspring, particularly when both partners carried pathogenic variants in the same gene. Ultimately, two couples were identified as carrying compound heterozygous variants in the same gene, and one couple carried the same pathogenic variant in the same gene. After obtaining informed consent, amniotic fluid samples were collected from these three couples for genetic analysis of hearing loss–related variants. The fetal genotyping results revealed GJB2 c.109G>A/wt, GJB2 c.235delC/wt, and wild type, respectively (Table 7). The couple in which both partners carried the SLC26A4 IVS7-2A>G variant declined amniocentesis and chose to proceed with postnatal genetic testing instead, stating that they would accept cochlear implantation if necessary.

Table 7.

Genetic Profiles of 3 Couples Undergoing Prenatal Diagnosis for HHL

Case	Fetus	Father	Mother
1	GJB2 c.109G>A/wt	GJB2 c.235delC>A/wt	GJB2 c.109G>A/wt
2	GJB2 c.235delC>A/wt	GJB2 c.235delC>A/wt	GJB2 c.235delC>A/wt
3	GJB2 wt/wt	GJB2 c.235delC>A/wt	GJB2 c.235delC>A/wt

Abbreviation: wt, wild type.

Newborn Outcomes

In total, 38 pregnancies resulted in live births, including one twin pregnancy (male and female twins) and 37 singleton pregnancies, yielding a total of 39 children. All newborns underwent genetic testing for hearing loss–related genes after birth. Among the 39 children, 3 were homozygous for the GJB2 c.109G>A variant, 26 were heterozygous for c.109G>A, 2 were heterozygous for c.235delC, 1 was heterozygous for SLC26A4 IVS7-2A>G, and 7 were negative for the tested variants. Among the three children homozygous for c.109G>A, one passed the newborn hearing screening at birth, while the other two did not. After 6 months of follow-up, their hearing thresholds were within the range of 25–40 dB HL, indicating mild hearing loss; however, they were able to communicate normally.

Discussion

In the present study, we conducted preconception or prenatal variant screening for deafness-associated genes (GJB2, GJB3, SLC26A4, and MT-RNR1) in a cohort of 1192 pregnant Chinese women and 611 spouses of women identified as variant carriers. The variant frequencies of GJB2, GJB3, SLC26A4, and MT-RNR1 variants among the pregnant women were 23.6%, 0.74%, 1.76%, and 2.43%, respectively. Within GJB2, the most common variant was c.109G>A, followed by c.235delC. For SLC26A4, the predominant variant was IVS7-2A>G. The main variants identified in GJB3 and MT-RNR1 were c.547G>A and m.1555G>A, respectively. We employed a cascade screening strategy in which pregnant women were screened first, and their spouses were subsequently tested if the women were identified as carriers. This approach led to the identification of three high-risk couples, supporting its practicality in a clinical setting. In addition, the results highlight the importance of preconception or prenatal variant screening for HHL in evaluating reproductive risk and supporting informed reproductive decision-making.

The GJB2 c.109G>A variant is one of the most common HL-associated variants in East Asian populations and has long been debated because of its variable expressivity and incomplete penetrance. The variant exhibits a relatively high carrier rate and allele frequency within the general Chinese population, with significant regional variation.¹⁶ In our study, the carrier frequency of c.109G>A was comparable to that reported in previous Chinese screening studies, although slightly lower than that reported in some northern populations, further supporting regional genetic heterogeneity. Although c.109G>A was initially considered a benign polymorphism, it is now recognized as a disease-associated variant, particularly in homozygous or compound heterozygous states.¹⁷ However, the presence of this variant does not necessarily predict clinically significant hearing impairment, as many homozygous individuals exhibit normal or mild hearing phenotypes.¹⁸ In our cohort, c.109G>A variants were detected in 20.51% of pregnant women and 18.63% of males, with homozygotes accounting for 1.93% and 3.15%, respectively, most of whom showed no obvious hearing impairment. Therefore, identification of c.109G>A through prenatal carrier screening should be interpreted cautiously, as a positive screening result indicates increased genetic risk rather than a definitive prediction of hearing impairment. Appropriate nondirective genetic counseling is essential to communicate the uncertainty associated with incomplete penetrance and to support informed reproductive decision-making. The c.235delC variant is another major pathogenic GJB2 variant in East Asian populations, including Chinese, Japanese, and Korean populations.^17,19,20 In our study, it was the second most frequent variant, with carrier frequencies of 2.64% in pregnant women and 2.73% in males. Importantly, therapeutic prospects for GJB2-related hearing loss are rapidly emerging: combined AAV-mediated delivery of GJB2 with a supporting-cell-specific promoter has been shown to restore hearing in DFNB1 mouse models, with efficacy and safety further demonstrated in large-animal cochleae.²¹ These advances highlight the value of identifying GJB2 carriers during preconception and prenatal screening, as early detection may enable timely intervention as molecular therapies move toward the clinic.

Variants in the SLC26A4 gene are the second most common cause of autosomal recessive nonsyndromic hearing loss in the Chinese population. In our cohort, the carrier frequency of SLC26A4 was 1.76% in pregnant women and 3.44% in the male individuals. The most common variants identified were IVS7-2A>G, followed by c.2168A>G and c.1229C>T in pregnant women, which was generally consistent with the previous reports from Chinese populations showing that IVS7-2A>G is a predominant SLC26A4 variant.^22,23 The clinical relevance of SLC26A4 variants is closely related to their role in inner ear development and function. SLC26A4 encodes pendrin, which is expressed in the inner ear, thyroid, and kidneys. Pathogenic variants in SLC26A4 are mainly associated with enlarged vestibular aqueduct syndrome and Pendred syndrome, with clinical manifestations ranging from congenital or progressive hearing loss to vestibular dysfunction and thyroid abnormalities.²⁴ Since hearing impairment associated with SLC26A4 variants may show variable onset and progression, identification of carriers through prenatal screening should be interpreted together with genetic counseling and appropriate clinical follow-up rather than as a direct prediction of disease severity. For couples in whom both partners carry clinically relevant SLC26A4 variants, additional genetic counseling and prenatal diagnostic options may help clarify the potential risk for offspring and support informed reproductive decision-making. Future studies incorporating longitudinal audiological assessments and detailed genotype–phenotype analyses are needed to better define the clinical implications of SLC26A4 variants identified through population-based screening.

GJB3 encodes connexin 31 and has been implicated in both autosomal dominant and autosomal recessive forms of hearing loss.²⁵ GJB3 c.538C>T has been reported as the predominant variant in Chinese populations.²⁶ In our cohort, the carrier frequency of GJB3 variants was 0.74% in pregnant women and 0.72% in male individuals, which was comparable to previous reports from Chinese populations.^27,28 Consistent with previous studies, c.538C>T was the most frequently observed GJB3 variant, followed by c.547G>A.²⁹ Compared with GJB2 and SLC26A4, pathogenic GJB3 variants are less frequently detected in Chinese populations, and their contribution to hereditary hearing loss remains relatively limited.³⁰ However, the clinical interpretation of GJB3 variants may be complex because both dominant and recessive inheritance patterns have been reported, and some variants may exhibit variable expressivity. Therefore, identification of GJB3 variants through population-based screening should be interpreted in conjunction with inheritance patterns, family history, and clinical findings and appropriate genetic counseling. Further studies incorporating comprehensive audiological phenotyping and functional analyses are needed to clarify the genotype–phenotype relationship of GJB3 variants.

In our study, the predominant MT-RNR1 variants were m.1555A>G and m.1494C>T in both pregnant women and males, which was consistent with previous reports from Chinese and other East Asian populations.³¹ This finding further supports the conserved distribution pattern of major MT-RNR1 variants across East Asian populations. These variants are well recognized as susceptibility factors for aminoglycoside-induced HL, and exposure to aminoglycoside antibiotics may result in irreversible auditory damage in genetically susceptible individuals.³² Unlike nuclear gene-associated HL, MT-RNR1-related hearing loss follows a maternal inheritance pattern,³² and therefore genetic findings may have implications for multiple generations within a family. However, identification of MT-RNR1 variants through carrier screening does not necessarily indicate the presence of hearing impairment, but rather an increased susceptibility to drug-induced auditory damage. Therefore, the primary clinical value of MT-RNR1 screening lies in risk prevention through medication guidance rather than prediction of congenital hearing loss. Appropriate genetic counseling and avoidance of unnecessary aminoglycoside exposure are essential for individuals carrying MT-RNR1 variants.

Beyond reproductive risk assessment, the therapeutic landscape for HHL has advanced considerably, which further strengthens the clinical value of carrier screening. AAV-mediated Otof gene replacement therapy has achieved marked hearing restoration in patients with DFNB9 deafness, providing the first clinical validation of gene therapy for hereditary hearing loss.³³ Moreover, precise DNA base editing has restored near-normal auditory function in adult DFNB9 mouse models, with a surprisingly wide therapeutic time window extending to adulthood.³⁴ Complementing gene-replacement and gene-editing strategies, promoting hair cell regeneration by enhancing supporting cell plasticity, such as AAV-mediated delivery of Sparcl1, offers an additional promising avenue for the treatment of sensorineural hearing loss.³⁵ Collectively, these therapeutic advances underscore that identifying at-risk individuals and couples through preconception and prenatal screening not only informs reproductive decision-making but may also facilitate timely access to emerging molecular interventions.

This study has several limitations. First, this was a single-center retrospective study, and therefore the findings may not be fully generalizable to other populations or regions. Second, the genetic screening panel was limited to four HL-associated genes and 127 known variant sites. Given the substantial genetic heterogeneity of hereditary hearing loss, other important genes and rare or novel pathogenic variants may have been missed. Future studies incorporating broader gene panels or exome/genome sequencing may improve variant detection. Third, partner testing was performed only among spouses of variant-positive pregnant women and was dependent on voluntary participation. Therefore, the male participants in this study do not represent an unselected male population, and the observed frequencies among male partners should not be interpreted as population-level carrier frequencies. Fourth, clinical information, including detailed audiological phenotypes, family history, and consanguinity, was incomplete, limiting comprehensive genotype–phenotype interpretation. In addition, the limited number of prenatal diagnostic cases and available newborn follow-up data restricted the assessment of clinical outcomes and long-term hearing trajectories. Finally, as this study was primarily descriptive, statistical analyses focused on frequencies and percentages without confidence interval estimation or formal between-group comparisons. Future prospective multicenter studies with larger cohorts, standardized genetic counseling, comprehensive phenotyping, and longitudinal audiological follow-up are warranted to further evaluate the clinical significance of variants identified through prenatal carrier screening.

Conclusions

In summary, this study characterizes the carrier profiles of HL-associated variants among pregnant women and their spouses, including single nucleotide variants, compound heterozygous variants, and digenic variations. We found that 28.53% of pregnant women and 27.94% of their male partners carried at least one common deafness-related variant. Among these, GJB2 c.109G>A and c.235delC were the most frequently detected. These findings enrich the regional epidemiological data on hereditary hearing loss and further support the clinical applicability of cascade screening strategies. However, given the limitations of the retrospective single-center design, restricted gene panel, incomplete partner participation, and limited clinical follow-up, the clinical utility of such screening strategies requires further validation through larger multicenter prospective studies incorporating broader genetic testing, standardized genetic counseling, and long-term audiological follow-up.

Funding Statement

This work was supported by Fujian Provincial Natural Science Foundation of China (2023Y9366), Fujian Maternity and Child Health Hospital Natural Science Foundation (FY YCXY 24-03), Startup Fund for scientific research, Fujian Medical University (2023QH1189), and Joint Funds for the innovation of science and Technology, Fujian province (2021Y9168), Fujian Provincial Medical Peak Development Program (FY26030X07014).

Data Sharing Statement

The datasets used and analysed during the current study are available from the corresponding author on reasonable request.

Ethics Approval and Informed Consent

This study has been conducted in accordance with the guidelines of the Declaration of Helsinki and has been approved by the institutional ethics committee of Fujian Provincial Maternity and Children’s Hospital (approval number: 2023KY021). Informed consent was obtained from all individual participants included in the study.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors report no conflicts of interest in this work.