Concurrent physiologic and gene-based genetic newborn hearing screening in a general population

Abstract

Background

Physiologic newborn hearing screening (NBHS) has several weaknesses requiring improvement. Studies have explored variant-based genetic NBHS as a complement to physiologic NBHS. However, this genetic NBHS strategy has fundamental limitations that challenge logistics.

Methods

Given findings that a small set of genes accounts for the majority of genetic etiologies of hearing loss (HL), we designed a gene-based genetic NBHS strategy that screens for genes instead of limited variants, targeting the coding regions of twelve HL-related genes and the aminoglycoside-induced HL risk gene MT-RNR1. Our cohort study was conducted at Nanjing Women and Children’s Healthcare Hospital in China, between March 2022 and July 2023. All participants were offered concurrent physiologic and gene-based genetic NBHS. Infants with positive physiologic or genetic NBHS results were scheduled for audiological assessments.

Results

Of 20,997 participants with eligible physiologic and genetic NBHS results, 164 (0.8%) had positive physiologic or genetic NBHS results. Of 150 who underwent audiological assessments, 77 had confirmed HL. Among them, 32 were identified by genetic NBHS only, 24 by both physiologic and genetic NBHS, and 21 by physiologic NBHS only. Compared with conventional physiologic NBHS protocol (identified 45 HL cases), incorporating gene-based genetic NBHS into physiologic NBHS (additionally identified 32 HL cases) only required an additional 0.4% (87/20,997) of infants to undergo audiological assessments, yet achieved a 71% (32/45) increase in HL identification and provided genetic etiological information for 73% (56/77) of all HL cases. For 21 HL cases identified by physiologic NBHS only, genome sequencing identified genetic findings in three additional cases, showing that up to 77% (59/77) of infants with HL were caused by genetic factors in this general newborn population.

Conclusions

Our new comprehensive NBHS protocol, concurrent physiologic and gene-based genetic NBHS, significantly improved the identification of infants with HL. The results demonstrated that the gene-based genetic NBHS addressed fundamental limitations of the variant-based strategy and was a powerful complement to physiologic NBHS.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40246-026-00912-3.

Introduction

Hearing loss (HL) is a significant global public health issue [1]. Without adequate language exposure, infants and children with HL fall behind their hearing peers in language, cognition, and social development, with lasting effects into adulthood [2]. Therefore, early identification and intervention are essential [2, 3]. The implementation of physiologic newborn hearing screening (NBHS) worldwide has demonstrated its feasibility and benefits, establishing it as the standard of care [3, 4]. However, this global effort, based on best practices and evidence, has also revealed several weaknesses: inclusion of limited etiological information; the potential to miss permanent and/or progressive mild HL; and the inability to identify delayed-onset HL and risk of aminoglycoside-induced HL. Addressing these weaknesses is key to improving the physiologic NBHS protocol [5].

Advances in our understanding of HL genetics and decreasing sequencing costs have highlighted the significant potential of genetic NBHS as a complementary approach [5–7]. Several pilot studies have explored the clinical value of incorporating variant-based genetic NBHS into physiologic NBHS [8–15]. In China, studies with outcome assessments primarily focused on screening 9 to 20 variants in four genes (GJB2, SLC26A4, GJB3, and MT-RNR1) with a test cost of 30 to 40 US dollars [8–14]. These studies demonstrated that the variant-based genetic NBHS could enhance physiologic NBHS by identifying additional infants at risk for aminoglycoside-induced HL. However, the benefits of identifying additional infants with HL missed by physiologic NBHS were limited and variable across different studies, with the additional detection rate ranged from 0 to 13% [8, 9, 11]. Moreover, as the variant-based strategy screened only selected common HL-causing variants, any carrier result it generated was inherently inconclusive, as it could not determine whether a second pathogenic allele was present. To address this limitation and identify more infants with HL, more aggressive clinical recommendations have been adopted for these inconclusive carriers. A large study of 239,636 Chinese newborns screened for 20 variants, reported 11,974 carriers (5.0%) and recommended retesting physiologic NBHS for carriers who passed physiologic NBHS [12]. Another large study of 180,469 Chinese newborns screened for 9 variants and recommended audiological assessments before one year of age for 7481 carriers (4.1%) who passed physiologic NBHS. In comparison, only 1915 newborns (1.1%) in this study were referred for audiological assessments due to their not passing physiologic NBHS [10]. These inconclusive carrier results complicated genetic counseling, may worsen parental anxiety, and increased healthcare costs due to additional physiologic NBHS and audiological assessments [5, 12].

Based on findings that a small set of genes accounts for the majority of genetic etiologies of HL [5–7], we designed a gene-based strategy that screens for genes instead of limited variants, including twelve HL-related genes and the aminoglycoside-induced HL risk gene MT-RNR1, with a test cost of 70 US dollars. This strategy expanded screening to more genes and, by targeting their entire coding regions, enabled the detection of most pathogenic variants and significantly reduced the probability of undetected second pathogenic alleles in carrier findings. We conducted concurrent physiologic and gene-based genetic NBHS in a general newborn population to explore the feasibility and benefits of this comprehensive NBHS protocol.

Materials and methods

Study design

From March 18, 2022, to July 27, 2023, this study was conducted at Nanjing Women and Children’s Healthcare Hospital in Nanjing, China. Newborns receiving concurrent physiologic and genetic NBHS before discharge, with parental informed consent, were included. Those receiving a blood transfusion before the dried blood spot collection were excluded. This study complied with the Declaration of Helsinki and was approved by the Institutional Review Board of Nanjing Women and Children’s Healthcare Hospital (No. 2021KY-071) and BGI Genomics (No. BGI-IRB 24090).

Concurrent physiologic and genetic NBHS

Physiologic NBHS

Physiologic NBHS was performed according to the Technical Specifications for Newborn Hearing Screening issued by the Ministry of Health of China [4]. Briefly, initial physiologic NBHS was performed using distortion-product otoacoustic emissions (DPOAEs) within 3 days after birth. For newborns who did not pass in one or both ears, automated auditory brainstem response (AABR) or DPOAEs was performed around 42 days of age. Those who did not pass the 42-day retest in one or both ears were classified as referred in the two-step physiologic NBHS, termed positive physiologic NBHS results in our study.

Genetic NBHS

Genetic NBHS was conducted using a custom-designed panel that targeted the entire coding regions, hotspot intronic regions containing pathogenic variants, and the mitochondrial gene MT-RNR1. Gene selection was guided by a workgroup of pediatric audiologists, pediatricians, geneticists, and genetic counselors. Twelve HL-related genes were selected based on three criteria: (1) definitive HL associations determined by the ClinGen Hereditary Hearing Loss Gene Curation Expert Panel [16], (2) classification as category A in BabySeq or category 1 in NC NEXUS newborn screening projects [17, 18], and (3) top HL-causing genes accounting for 90% of all genetic diagnoses in BGI in-house data (Supplemental Table 1). Separately, MT-RNR1 was selected due to its association with a high risk of aminoglycoside-induced HL [19].

Dried blood spot collection, genomic DNA extraction, library construction, and sequencing were performed as previously described [20]. Data analysis and variant interpretation followed published criteria [20], and variants with a matching inheritance pattern were manually reviewed based on American College of Medical Genetics and Genomics guidelines [21]. A positive genetic NBHS result was defined as pathogenic or likely pathogenic variants with a matching inheritance pattern in the twelve HL-related genes. Carriers with variants in cis through pedigree analysis were eliminated (Supplemental Methods). An ototoxicity risk result was defined as a homoplasmic or heteroplasmic pathogenic variant in MT-RNR1.

Audiological assessments

Infants with positive physiologic or genetic NBHS results were scheduled for comprehensive audiological assessments, including otoscopic examination, tympanometry, auditory brainstem response (ABR), and behavioral assessment (if older than 6 months). Transient conductive conditions of the external or middle ears (e.g., vernix or other debris in the ear canal or fluid in the middle ear) were not classified as HL in this study. For ABR, HL was defined as a threshold > 30 dB nHL for click. HL grades based on click-ABR were defined as profound (> 90 dB nHL), severe (71–90 dB nHL), moderate (46–70 dB nHL), or mild (31–45 dB nHL) [22]. For behavioral assessment, HL was defined as a threshold > 25 dB HL, with grades classified as profound (> 90 dB HL), severe (71–90 dB HL), moderate (41–70 dB HL), or mild (26–40 dB HL) [23]. For behavioral assessment which based on parents’ observations and answers to behavioral evaluation questions, HL was defined by the pediatric audiologist, and grade and laterality were not assigned. Follow-up audiological assessments were recommended for infants who received a hearing loss result and for infants who received a normal hearing result with a positive genetic NBHS result. Infants who received a normal hearing result with a positive physiologic and negative genetic NBHS result were managed with routine care. Audiological assessment results in our study were based on the latest records in the hospital system as of May 31, 2024.

Statistical analysis

Statistical analysis was performed using R version 4.1.3. Continuous variables with skewed distributions were reported as medians with interquartile ranges (IQRs). Categorical variables were summarized as frequencies and percentages, with χ2 test used for comparative analysis. A two-sided P value < 0.05 was considered statistically significant. 95% CIs were computed using the exact binomial test.

Results

Study population

A total of 28,102 newborns were born in Nanjing Women and Children’s Healthcare Hospital from March 18, 2022, to July 27, 2023. Of these, 21,008 (74.8%) received concurrent physiologic and genetic NBHS. Their characteristics are outlined in Table 1. The median gestational week was 39 weeks (IQR 38–40), the median birth weight was 3320 g (IQR 3070–3590), and 52.7% were male, 47.3% female. Additionally, neonatal hyperbilirubinemia was documented in 249 newborns (1.2%). Due to quality control failures in genetic NBHS, 11 newborns were excluded, resulting in 20,997 (99.9%) having eligible results for both physiologic and genetic NBHS.

Table 1.

Characteristics of 21,008 study subjects

Characteristic	No. (%) (n = 21,008)
Sex
Female	9940 (47.3%)
Male	11 068 (52.7%)
Gestational age
< 37 wk	815 (3.9%)
37–38 wk	7095 (33.8%)
39–40 wk	12 828 (61.1%)
≥ 41 wk	270 (1.3%)
Birth weight
≤ 2499 g	42 (0.2%)
2500–4000 g	19 936 (94.9%)
> 4000 g	1030 (4.9%)
Delivery
Vaginal delivery	11 961 (56.9%)
Cesarean delivery	9047 (43.1%)

Overview of concurrent physiologic and genetic NBHS and follow-up

Of 20,997 subjects with eligible results, 164 (0.8%) had positive physiologic or genetic NBHS results, and received audiological assessment follow-up. Among these, 150 (91%) had eligible audiological assessment results, while 14 (9%) were lost to follow-up by May 31, 2024. In total, 77 infants were confirmed to have HL at a median age of 3 months (IQR 3–3). The overall positive predictive value (PPV) of concurrent physiologic and genetic NBHS was 51% (77/150, 95% CI 43–60%). Of 77 infants with HL, 32 were identified by genetic NBHS only, 24 by both physiologic and genetic NBHS, and 21 by physiologic NBHS only. Therefore, compared with conventional physiologic NBHS protocol, which identified 45 infants with HL (24 plus 21), incorporating genetic NBHS into physiologic NBHS increased the HL detection rate by 71% (32/45, 95% CI 56–84%) (Fig. 1).

Fig. 1.

Study flowchart. NBHS, newborn hearing screening

Genetic etiology and phenotype of infants with HL

Of 32 infants with HL identified by genetic NBHS only, 30 (94%) had biallelic GJB2 genotypes involving c.109G > A homozygous (23 cases) and compound heterozygotes with another pathogenic variant (7 cases), and 2 (6%) had biallelic SLC26A4 genotypes. Mild HL was most prevalent, accounting for 78% (25/32) (Table 2).

Table 2.

Genotype and phenotype of 77 infants with hearing loss

Genotype	Total, no.	Grade, no.a				Laterality, no.	Positive results by variant-based strategy in China, no.b
		Mild	Moderate	Severe	Profound	Bilateral / Unilateral
Identified by genetic NBHS only (n = 32)	32	25	6	1	0	23/9	1
GJB2:c.[109G > A];[109G > A]c	23	18	4	0	0	15/7	0
GJB2:c.[109G > A];[235del]	6	4	1	1	0	5/1	0
GJB2:c.[109G > A];[257C > G]	1	1	0	0	0	1/0	0
SLC26A4:c.[919-2A > G];[1001 + 5G > C]	1	1	0	0	0	1/0	0
SLC26A4:c.[1975G > C];[2168A > G]	1	0	1	0	0	0/1	1
Identified by both physiologic and genetic NBHS (n = 24)	24	4	15	2	3	18/6	8
GJB2:c.[109G > A];[109G > A]	10	3	6	0	1	6/4	0
GJB2:c.[235del];[235del]	4	0	4	0	0	4/0	4
GJB2:c.[109G > A];[235del]	2	1	1	0	0	1/1	0
GJB2:c.[176_191del];[235del]	1	0	1	0	0	1/0	1
GJB2:c.[235del];[299_300del]	1	0	0	0	1	1/0	1
SLC26A4:c.[349del];[916dup]	1	0	1	0	0	1/0	0
SLC26A4:c.[919-2A > G];[754T > C]	1	0	0	1	0	1/0	0
SLC26A4:c.[919-2A > G];[916dup]	1	0	1	0	0	1/0	0
SLC26A4:c.[919-2A > G];[1975G > C]	1	0	1	0	0	0/1	1
SLC26A4:c.[919-2A > G];[919-2A > G]	1	0	0	1	0	1/0	1
CHD7:c.[5405-17G > A];[=]	1	0	0	0	1	1/0	0
Identified by physiologic NBHS only (n = 21)	21	5	7	2	7	9/12	0
Total	77	34	28	5	10	50/27	9

NBHS newborn hearing screening

^aFor infants with different grades of hearing in each ear, the grade of hearing loss was determined based on the ear with the more severe hearing loss

^bThe variant-based strategy screens 20 variants, including c.35del, c.167del, c.176_191del, c.235del, and c.299_300del of GJB2; c.281C > T, c.589G > A, c.919-2A > G, c.1174A > T, c.1226G > A, c.1229C > T, c.1707 + 5G > A, c.1975G > C, c.2027T > A, c.2162C > T, and c.2168A > G of SLC26A4; m.1494C > T and m.1555A > G of MT-RNR1; c.538C > T and c.547G > A of GJB3 [11, 12]; other variant-based strategies screen 9 and 15 variants, which are the subset of 20 variants[10, 13, 14]

^cGrade and laterality were not assigned for one infant in this subgroup, as hearing loss was defined by the pediatric audiologist based on parents’ observations and answers to behavioral evaluation questions

Of all 77 infants with HL, 56 had positive genetic NBHS results. Our genetic NBHS strategy provided genetic etiological information for 73% (56/77, 95% CI 61–82%) of all HL cases (Table 2 and Fig. 1). Based on the genotype information, we compared our gene-based strategy with three commonly used variant-based strategies in China. The results showed that the variant-based strategy identified as a positive result for only 12% (9/77) HL cases (Table 2).

Of 21 infants with HL identified by physiologic NBHS only, all received genome sequencing. Three had a genetic diagnosis, including two with homozygous deletion in STRC and one with biallelic variants in LRTOMT (Table 3). Finally, our data showed that up to 77% (59/77, 95% CI 66–86%) of infants with HL were caused by genetic factors in this general population. GJB2-related HL (62%, 48/77) was the most prevalent, followed by SLC26A4-related HL (9%, 7/77) and STRC-related HL (3%, 2/77). CHD7 and LRTOMT accounted for HL in the remaining two subjects (Tables 2 and 3).

Table 3.

Genome sequencing results of 21 infants with hearing loss identified by physiologic NBHS only

Patient no	Sex	Gradea	Laterality	Other phenotype	Genotype identified by genetic NBHS	Additional genetic findings diagnosed by genome sequencingb
1	Male	Mild	Bilateral	None	SLC26A4:c.[2009T > C];[=]	STRC:homozygous deletion
2	Male	Mild	Bilateral	None	No variant identified	STRC:homozygous deletion
3	Female	Profound	Bilateral	None	GJB2:c.[176_191del];[=]	LRTOMT:c.[566del];[567dup]
4	Male	Mild	Bilateral	None	GJB2:c.[109G > A];[=]	None
5	Female	Mild	Unilateral	Cleft palate	No variant identified	None
6	Female	Mild	Unilateral	Cleft palate	No variant identified	None
7	Male	Moderate	Bilateral	Craniofacial malformation	No variant identified	None
8	Female	Moderate	Unilateral	None	GJB2:c.[109G > A];[=]	None
9	Male	Moderate	Unilateral	None	No variant identified	None
10	Female	Moderate	Unilateral	None	GJB2:c.[176_191del];[=]	None
11	Male	Moderate	Unilateral	Microtia	No variant identified	None
12	Male	Moderate	Unilateral	Microtia	No variant identified	None
13	Male	Moderate	Unilateral	Microtia	No variant identified	None
14	Male	Severe	Bilateral	Microtia	GJB2:c.[109G > A];[=]	None
15	Male	Severe	Unilateral	None	No variant identified	None
16	Female	Profound	Bilateral	None	SLC26A4:c.[754T > C];[=]	None
17	Male	Profound	Bilateral	None	No variant identified	None
18	Male	Profound	Bilateral	Inner ear malformation	GJB2:c.[109G > A];[=]	None
19	Male	Profound	Unilateral	None	GJB2:c.[109G > A];[=]	None
20	Female	Profound	Unilateral	Auditory nerve dysplasia	No variant identified	None
21	Male	Profound	Unilateral	Auditory nerve dysplasia	No variant identified	None

NBHS newborn hearing screening

^aFor infants with different grades of hearing in each ear, the grade of hearing loss was determined based on the ear with the more severe hearing loss

^bThe gene curation and selection, sequencing, data analysis, and variant interpretation of genome sequencing were carried out as previously described [24–26]

Correlation between positive genotypes and auditory outcomes

Studying the correlation between genotypes and HL rate during infancy benefits genetic counseling. Of 20,997 subjects, 113 (5.4‰, 95% CI 4.4–6.5‰) had positive genetic NBHS results, of whom 104 (92%) completed audiological assessments based on the latest records. The overall HL rate for infants with positive genetic NBHS results (PPV of genetic NBHS alone) was 54% (56/104, 95% CI 44–64%). For infants with GJB2 genotypes, the HL rate was 49% (42/86, 95% CI 38–60%) among those c.109G > A homozygous and compound heterozygotes with another pathogenic variant, and 86% (6/7, 95% CI 42–100%) for other non-109G > A genotypes. For infants with SLC26A4 genotypes, the HL rate was 88% (7/8, 95% CI 47–100%) (Table 4). Notably, follow-up data showed hearing status progression in 11 cases with positive genotypes: 6 transitioned from normal hearing to HL, and 5 progressed in laterality or grade (Supplemental Table 2). For 48 cases with positive genotypes who received a normal hearing result, ongoing audiological assessment was essential as they were at high risk for HL.

Table 4.

Auditory outcomes of 113 infants with positive genetic NBHS results^a

Genotype	Total, no.	Lost to follow-up, no.	Hearing loss, no.	Normal hearing, no.b	Hearing loss rate, % (95% CI)
GJB2 homo/CH	102	9	48	45	52 (41–62)
c.109G > A homo/CH	95	9	42	44	49 (38–60)
c.[109G > A];[109G > A]	67	5	33	29	53 (40–66)
c.[109G > A];[235del]	18	2	8	8	50 (25–75)
c.[109G > A];[299_300del]	4	1	0	3	0 (0–71)
c.[109G > A];[257C > G]	2	0	1	1	50 (1–99)
c.[109G > A];[176_191del]	2	1	0	1	0 (0–98)
c.[109G > A];[583A > G]	2	0	0	2	0 (0–84)
Non-109G > A homo/CH	7	0	6	1	86 (42–100)
c.[235del];[235del]	4	0	4	0	100 (40–100)
c.[176_191del];[235del]	1	0	1	0	100 (3–100)
c.[235del];[299_300del]	1	0	1	0	100 (3–100)
c.[176_191del];[583A > G]	1	0	0	1	0 (0–98)
SLC26A4 homo/CH	8	0	7	1	88 (47–100)
c.[919-2A > G];[919-2A > G]	1	0	1	0	100 (3–100)
c.[919-2A > G];[754T > C]	1	0	1	0	100 (3–100)
c.[349del];[916dup]	1	0	1	0	100 (3–100)
c.[919-2A > G];[916dup]	1	0	1	0	100 (3–100)
c.[919-2A > G];[1001 + 5G > C]	1	0	1	0	100 (3–100)
c.[919-2A > G];[1975G > C]	1	0	1	0	100 (3–100)
c.[1975G > C];[2168A > G]	1	0	1	0	100 (3–100)
c.[919-2A > G];[1229C > T]	1	0	0	1	0 (0–98)
USH2A:c.[2802T > G];[9958G > T]	2	0	0	2	0 (0–84)
CHD7:c.[5405-17G > A];[ =]	1	0	1	0	100 (3–100)
Total	113	9	56	48	54 (44–64)

NBHS newborn hearing screening, homo homozygous, CH compound heterozygous

^aHearing status in 8 infants with negative physiologic NBHS results was defined by the pediatric audiologist based on parents’ observations and answers to behavioral evaluation questions; 7 had normal hearing (6 with GJB2:c.[109G > A];[109G > A] and 1 with GJB2:c.[109G > A];[235del]) and 1 had hearing loss with GJB2:c.[109G > A];[109G > A]

^bOf the 48 infants with positive genetic NBHS results who received a normal hearing result, 2 had positive physiologic NBHS results (both were GJB2:c.[109G > A];[109G > A]), and the remaining 46 had negative physiologic NBHS results

Discussion

In this large general cohort, compared with conventional physiologic NBHS protocol, concurrent physiologic and gene-based genetic NBHS only required an additional 0.4% (87/20,997, Fig. 1) of infants to undergo audiological assessments, who passed physiologic NBHS but had positive genetic NBHS results, yet achieved a 71% increase in HL identification and provided genetic etiological information for 73% of all HL cases. Our results demonstrated that gene-based genetic NBHS was a powerful complement to the physiologic NBHS in a general population.

The Birth Defects Monitoring Center of China reported a nationwide HL detection rate of 0.23% (95% CI 0.15–0.26%) by conventional physiologic NBHS [27]. In our study, physiologic NBHS alone achieved a comparable HL detection rate of 0.21% (45/20,997, 95% CI 0.16–0.29%), supporting the reliability of our findings. Incorporating gene-based genetic NBHS into physiologic NBHS increased the detection rate by 71% (32/45). In contrast, previous studies with outcome assessments showed variant-based genetic NBHS increased the detection rate by only 0–13% [8, 9, 11]. Recently, a cohort study incorporated exome sequencing into NBHS for neonatal intensive care unit (NICU) newborns, increasing HL detection by 15.6% [28]. This limited improvement was primarily due to its focus on the high-risk NICU population.

Importantly, our study found that the HL cases identified by genetic NBHS only were primarily mild HL (78%). WHO 2021 data showed a fivefold increase in moderate or higher-grade HL prevalence from the neonatal period (0.2%) to ages 1–4 years (1.0%) [3]. It is unclear whether these additional moderate or higher-grade HL cases were congenital progressive mild or delayed-onset later, as mild HL is often undetected due to OAE and AABR limitations [2, 5]. The Centers for Disease Control and Prevention (CDC) of the United States reported that mild HL accounted for approximately 20% (1115/5547, 95% CI 19–21%) of infants with HL [29]. In our study, mild HL comprised 20% (9/45) of HL cases identified by physiologic NBHS, and increased to 44% (34/77, 95% CI 33–56%) when including all HL cases, significantly exceeding CDC data (P < 0.001). Our data revealed a larger proportion and a more realistic prevalence (1.6‰ 34/20,997, 95% CI 1.1–2.3‰) of mild HL in newborns and infants, previously obscured by physiologic NBHS.

Identifying infants with mild HL is integral to improving the physiologic NBHS [3, 5]. Several longitudinal studies have demonstrated that a significant proportion of infants and young children initially detected with mild HL, whether unilateral or bilateral, developed progressive HL at age [30, 31]. Our follow-up data also showed hearing status progression in 11 cases. Both permanent and progressive mild HL in this important period significantly impacts speech, language, and cognitive development in children [32, 33]. Therefore, early identification, long-term monitoring, and timely intervention for infants with mild HL are essential. In our study, infants with mild HL were enrolled in long-term audiological follow-up and provided family-centered communication intervention supported by pediatric audiologists and speech-language pathologists [34].

The gene-based strategy also presented an advantage in providing genetic etiology. Of 77 infants with HL, 56 (73%) received genetic etiologic information during NBHS. HL has diverse causes, and clinical manifestations of different forms of HL often overlap in infancy. Providing etiologic information for this complex condition during the NBHS period can significantly impact clinical management, refine genetic counseling, shorten the time to arrive at an etiologic diagnosis, and improve health outcomes [5, 7]. In our study, a boy identified with a de novo variant in CHD7 during NBHS initially presented with unilateral HL and cryptorchidism. He was diagnosed with CHARGE syndrome within three months and received medical management. By 10 months, his condition progressed to bilateral HL and growth retardation.

Our results showed that 94% of HL cases identified by genetic NBHS only were associated with c.109G > A in GJB2. This raises the question: would simply adding this variant to the variant-based strategy be a more cost-effective choice? However, the answer is no. The variant-based strategy has fundamental design limitations that challenge logistics. This strategy, which screens limited common variants, yielded numerous inconclusive carrier results with uncertainty about the presence of another pathogenic allele [10, 12]. As a result, approximately an additional 4% of newborns required audiological assessments due to inconclusive carrier results, despite passing physiologic NBHS [10]. Taking the referral rate (0.3–1.4%) for audiological assessments by conventional physiologic NBHS in China and Japan as the baseline [35, 36], these additional 4% audiological assessments could create a huge burden on the healthcare system. Adding c.109G > A, based on our carrier frequency data of 9.4%, would result in approximately 940 additional inconclusive carrier results per 10,000, further exacerbating this issue. In comparison, the gene-based strategy only required an additional 0.4% (87/20,997) of infants to undergo audiological assessments, who passed physiologic NBHS but had positive genetic NBHS results. Therefore, the gene-based genetic NBHS emerges as an optimal approach in China.

The gene-based strategy has strong potential as a globally effective approach. First, HL cohorts from different countries consistently demonstrate that a small set of genes accounts for the majority of genetic etiologies. Data from 2460 HL cases in the United States showed that the 15 most commonly identified genes accounted for up to 73.6% of genetic diagnoses [5]. Ten genes explained 86.6% among 3877 pre-lingual HL cases in Japan [37], and 9 genes explained 96.4% among 1292 pediatric HL cases in Russia [38]. Our 13 genes covered 49.9% of genetic diagnoses in the United States, 76.5% in Japan, and 93.2% in Russia cohorts (Supplemental Table 3). Second, studies on physiologic NBHS sensitivity from different countries, WHO data, and NBHS working groups indicate that physiologic NBHS misses a substantial proportion of mild to moderate HL [2, 3, 5]. Variants associated with this HL group exhibit high allele frequencies in the general population and high diagnostic frequencies in multiethnic HL cohorts [5, 39, 40]. Specifically, common nontruncating variants in GJB2, such as c.109G > A and c.101T > C, are typically associated with progressive mild to moderate HL, regardless of compound heterozygotes with truncating or nontruncating variants [41]. The c.109G > A shows high allele frequencies in East Asian (4.5%) and Ashkenazi Jewish (0.8%), and c.101T > C in European (2.1%) and Ashkenazi Jewish (0.9%) ancestry groups [40]. Recent data suggest that STRC represents the next most common etiology for mild to moderate HL [7]. In multiethnic countries such as the United States, copy number variants in STRC, and GJB2 variants c.109G > A and c.101T > C, rank among the four most commonly identified variants causing HL [5]. We believe the gene-based strategy should involve selecting a set of genes specific to each country, rather than applying the same panel globally.

This study has limitations. First, gene selection for NBHS was constrained by the lack of large-scale genetic etiology data from Chinese pediatric HL cohorts at the time of study design. Although 12 of our 13 genes were featured in recent pediatric HL studies from China [42, 43], gene selection can be optimized. For instance, replacing less relevant genes with STRC could enhance effectiveness, as STRC accounted for 3% (2/59) of our cohort’s genetic diagnoses (Table 3) and significant proportions in the United States (14.3%), Japan (5.9%), and Russia (2.6%) cohorts (Supplemental Table 3) [5, 37, 38]. Second, the cost-effectiveness of this protocol was not evaluated. Third, although our findings are likely generalizable across populations and countries, further studies are needed. Fourth, among 7 infants with normal hearing defined by the pediatric audiologist based on parents’ observations and answers to behavioral evaluation questions, the presence of unilateral HL cannot be completely excluded, as parental observations primarily reflect hearing status in the better hearing ear.

Conclusions

The findings of this cohort study suggested that gene-based genetic NBHS was a powerful complement to physiologic NBHS in a general population. By only requiring an additional 0.4% of infants to undergo audiological assessments, it addressed fundamental limitations of variant-based strategy, yet achieved the highest additional detection rate reported to date (71%) and provided genetic etiological information for 73% of all HL cases. At this point, expanding screening to multiple HL-related genes, provides an interim step toward genome newborn screening.

Abbreviations

AABR

Automated auditory brainstem response

ABR

Auditory brainstem response

CDC

The centers for disease control and prevention

DPOAEs

Distortion-product otoacoustic emissions

HL

Hearing loss

IQRs

Interquartile ranges

NBHS

Newborn hearing screening

PPV

Positive predictive value

Author contributions

Yun Sun: Conceptualization, Investigation, Data curation, Funding acquisition, Writing—Original draft. Jingyu Zhao: Conceptualization, Formal analysis, Visualization, Project administration, Writing—Original draft. Liping Meng: Investigation, Data curation, Methodology, Writing—Review & Editing. Tao Jiang: Investigation, Data curation, Methodology, Writing—Review & Editing. Jiale Xiang: Formal analysis, Methodology, Project administration, Writing—Review & Editing. Xin Wang, Xianwei Guan, Jingjing Zhang, Ye Fan, Jing Xu: Data curation, Writing—Review & Editing. Zibin Lin, Xiangzhong Sun, Hui Huang, Chunna Fan, Jing Yang: Formal analysis, Writing—Review & Editing. Zhiyu Peng: Conceptualization, Supervision, Methodology, Writing—Review & Editing. Zhengfeng Xu: Conceptualization, Investigation, Funding acquisition, Supervision, Writing—Review & Editing.

Funding

This work was supported by the National Key R&D Program of China (2022YFC2703400), Jiangsu Province Capability Improvement Project through Science, Technology and Education Jiangsu Provincial Medical Key Discipline (ZDXK202211), and Nanjing Health Science and Technology Development Special Fund (YKK21159).

Availability of data and materials

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

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committees of Nanjing Women and Children’s Healthcare Hospital (No. 2021KY-071) and BGI (No. BGI-IRB 24090). Written informed consent was obtained from all the participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.