Introduction

Hearing loss has become the fourth leading cause of disability worldwide [1], affecting approximately one to two out of every 1000 children and significantly impacting their normal development [2]. The commonly used hearing screening methods in clinical practice are the Otoacoustic Emission (OAE) and Auditory Brainstem Response (ABR) tests. Guidelines recommend a two-step screening program for healthy and low-risk newborns, with the ABR test performed if the OAE test is not passed. However, for individuals with auditory neuropathy spectrum disorders (ANSDs), both the OAE and ABR should be used to avoid missed diagnoses [3]. The updated 2019 JCIH guidelines recommend conducting a comprehensive audiological evaluation between hospital discharge and 9 months of age when risk factors for delayed-onset or progressive hearing loss are present [4]. The recommended hearing screening plan for neonatal intensive care unit (NICU) infants and infants in the well-baby nursery (WBN) also differs. Infants admitted to the NICU face a 10–20 times greater risk of permanent hearing loss due to underlying health conditions. Furthermore, 50% of cases involve genetic factors and are not related to other risk factors [5]. Therefore, hearing screening is necessary for both infants in the NICU and healthy infants without related risk factors.

A timely diagnosis of hearing loss in children is crucial, as studies have confirmed that hearing loss is typically diagnosed between 24 and 30 months of age. A delayed diagnosis significantly impacts normal growth and brain development in infants [3]. The severity of hearing loss in children is directly proportional to its negative effects on cognitive, language, and literacy skills [6, 7]. Additionally, communication difficulties in childhood can lead to psychological symptoms such as anger, loneliness, and burnout [8].

Studies have demonstrated that appropriate intervention measures during the first 6 months of life are essential for mitigating the adverse effects of hearing loss [9]. Numerous studies have consistently shown that infants with risk factors for hearing loss are more likely to experience impairment, highlighting the importance of identifying these risk factors and implementing standardized hearing screening programs for early detection and intervention.

This study aimed to systematically review the recent literature on the risk factors for infant hearing loss, conduct a meta-analysis to identify the main risk factors, and provide reliable, evidence-based medical evidence for the prevention and treatment of infant hearing loss.

Methods

The study was conducted according to the 2020 Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines.

Inclusion and exclusion criteria

Inclusion criteria: (i) case‒control studies, cross-sectional studies, or cohort studies; (ii) studies reporting odds ratios (ORs) or relative risks (RRs) with 95% confidence intervals (CIs); and (iii) studies in which the research subjects were infants diagnosed with hearing loss by the ABR, OAE, automatic auditory brainstem response (A-ABR), brainstem auditory evoked response (BAER) and auditory event-related potential (AERP) hearing tests.

Exclusion criteria: (i) repeated publications; (ii) reviews, case reports, lectures and conference abstracts; (iii) studies of nonhuman subjects; and (iv) studies with incomplete information on ORs or lacking sufficient information to calculate OR values.

Search strategy and selection criteria

Searches were conducted in various databases, including the China National Knowledge Infrastructure, China Science and Technology Journal (VIP), Wanfang, Chinese Biology Medicine Disc, PubMed, Web of Science, Scopus, Cochrane Library, SinoMed, Embase, and Clinical Trial Registry databases in China and the USA. The search spanned from inception to November 2023 and involved the use of a combination of subject and free word retrieval methods. The English subject terms used were determined based on PubMed’s MeSH thesaurus. The search terms used included ‘Infant’, ‘Infants’, ‘Infant, Newborn”, ‘Newborn Infant’, ‘Hearing Loss’, ‘Hypoacusis’, ‘Hypoacuses’, ‘Hearing Impairment’, ‘Transitory Deafness’, ‘cohort studies’ and ‘relative risk’. Additionally, a manual search of the reference lists of the included studies was performed. The detailed search strategy for PubMed is shown as an example; details are provided in the supplementary online material (Box 1).

Study selection and data extraction

Two researchers independently conducted the literature screening, and any discrepancies were resolved through discussion or consultation with a third party. The following data were extracted from the eligible studies: (i) basic information about the study (e.g. first author, publication date, research country, and research type), (ii) baseline characteristics such as sample size and age, and (iii) risk factors and specific data on infant hearing loss. Endnote X9 was used for managing and screening the literature. The abstracts and full texts were further reviewed to determine eligibility.

Quality assessment

Two investigators independently assessed bias in the included studies and cross-verified the results. Any disagreements were resolved through discussion with a third party until consensus was reached. The quality of the case‒control and cohort studies was evaluated using the Newcastle–Ottawa Scale (NOS), while cross-sectional studies were assessed for bias based on criteria recommended by the Agency for Healthcare Research and Quality (AHRQ).

Data analysis

RevMan 5.4 software was used for meta-analysis, with ORs/RRs and 95% CIs as the effect indices. Heterogeneity among the included studies was assessed using the χ2 test (α = 0.1) and I2 statistic. A fixed-effects model was employed if heterogeneity was acceptable (P > 0.10 and I2 ≤ 50%); otherwise, a random-effects model was used. The significance level for the meta-analysis was set at α = 0.05. Furthermore, the influence of individual studies on the overall results was evaluated by conducting a sensitivity analysis, whereby studies were eliminated one by one. In addition, funnel plots were drawn for outcome indicators with data from ≥ 6 studies to observe whether publication bias existed. Additionally, the quality of evidence for each risk factor was assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach to evaluate the quality of evidence for each risk factor [10].

Results

Study identification

A total of 6008 relevant studies were obtained during the preliminary examination, and 18 studies were ultimately identified after screening, including 1,110,943 participants. The literature screening process and results are shown in Fig. 1.

Fig. 1
figure 1

Search flowchart

Full size image

Characteristics of the included studies

All the basic characteristics of the included studies are shown in Table 1. The risk of bias assessment results of the included case‒control studies, cohort studies, and cross-sectional studies is shown in Supplementary document 1.

Table 1 Characteristics of included studies
Full size table

Meta-analysis results

The findings of the comprehensive meta-analysis are presented in Table 2. The studies were categorized into perinatal factors, perinatal or postnatal factors, and other factors. Among all factors, eight were perinatal factors; all of them showed statistical significance. A family history of hearing loss [11,12,13,14,15,16,17,18,19] (OR = 2.20, P < 0.001) and the use of ototoxic medications [11,12,13,14,15,16, 20,21,22,23] (OR = 2.75, P < 0.001) were risk factors for hearing loss. Admission to neonatal intensive care unit (NICU) for > 5 days [14, 16, 23] (OR = 2.08, P < 0.001) and hyperbilirubinemia [11,12,13,14,15, 20, 21, 24,25,26] (OR = 2.17, P = 0.009) were strongly associated with hearing loss. Factors such as intrauterine infection [11, 13, 18, 22, 26] (OR = 6.07, P < 0.001), asphyxia [12, 13, 27] (OR = 1.76, P = 0.009), craniofacial anomalies [11, 15, 16, 21, 24, 27] (OR = 6.43, P < 0.001), and congenital malformations and syndromes [14, 19, 28] (OR = 5.01, P < 0.001) strongly increased the risk of hearing loss. Because of the high heterogeneity, this study divided participants with ototoxic medication use into two subgroups and investigated the risk factors for hearing loss in infants in the Asian group and the non-Asian group; the results were statistically significant. Two perinatal or postnatal factors, sepsis or meningitis [12,13,14, 18,19,20, 25](OR = 2.99, P = 0.005) and intracranial hemorrhages [11, 13, 25] (OR = 2.67, P < 0.001), which were closely related to the occurrence of infant hearing loss, were included in this meta-analysis. Among the other factors, SGA [12, 18, 19, 25] (OR = 1.71, P = 0.05), premature birth [11, 13, 26, 28] (OR = 1.95, P = 0.20 > 0.05) and male sex [15, 20, 21] (OR=1.04, P = 0.77 > 0.05) had no statistical significance. Mechanical ventilation [11, 14, 21, 25] (OR = 1.71, P < 0.001) and mechanical ventilation duration > 5 days [12, 13, 18, 22, 27] (OR = 2.10, P = 0.03) were significant risk factors for infant hearing loss. LBW [11, 19, 23, 28] (OR = 1.78, P = 0.001) and VLBW [12, 13, 15, 26] (OR = 3.47, P = 0.003) were also associated with hearing loss.

Table 2 Meta-analysis results of risk factors for infants hearing loss
Full size table

GRADE assessment

Table 3 shows all the results found in the GRADE evaluation. Four factors with moderate-certainty evidence were identified, while the seven factors had low-certainty evidence. In addition, the certainty of evidence for six factors was extremely low.

Table 3 GRADE assessment scores
Full size table

Publication bias

The results revealed a symmetrical distribution of research sites, indicating the absence of publication bias (Fig. 2).

Fig. 2
figure 2

Funnel chart. a Family history of hearing loss; b abnormal factors; c ototoxic medications; d sepsis or meningitis; e hyperbilirubinemia; f mechanical ventilation

Full size image

Discussion

This study is the first to identify 14 risk factors for infant hearing loss based on meta-analysis and hierarchical evidence assessment.

The findings of this study provide moderate evidence that low birth weight, craniofacial anomalies, congenital malformations, and intracranial hemorrhages are significant risk factors for hearing loss in infants. Craniofacial anomalies may increase infants’ risk of developing hearing loss, consistent with the findings of previous research [29]. According to the JCIH statement, craniofacial anomalies and more than 400 syndromes and genetic disorders associated with atypical hearing thresholds are classified as risk factors for perinatal hearing loss [4]. Therefore, early hearing screening for infants with congenital malformations, especially those involving craniofacial anomalies, may be needed. The present study showed a significantly increased risk of hearing loss in infants who were admitted to the NICU for more than 5 days. With respect to LBW, the lower an infant’s weight is, the greater their risk of hearing loss. Newborns residing in the NICU often have complex congenital diseases and poor physical conditions, resulting in an incidence of hearing loss ranging from 2 to 5% [30]. The immature development of various organs in LBW infants, especially in VLBW infants in the NICU, coupled with potential malnutrition, increases susceptibility to auditory nerve cell damage and subsequent hearing impairment due to prolonged exposure to sound sources such as ventilator alarms and vital sign monitors [31]. It can be inferred from these findings that a history of NICU hospitalization, LBW, and mechanical ventilation factors have synergistic effects on infant hearing loss and lead to a greater incidence of hearing loss in the NICU than in the WBN. Therefore, it is necessary to strengthen hearing screening for infants with a history of NICU hospitalization and implement appropriate measures to prevent and control this risk factor. In addition, some VLBW infants may not pass their first OAE test due to the problem of middle ear effusion; as the effusion subsides a few weeks after birth, a large proportion of these infants will pass the subsequent ABR test [32]. Therefore, for hearing assessment in VLBW infants, careful examination and the performance of the ABR by a professional audiologist is needed. There is a strong correlation between intracranial hemorrhage and hearing loss; a large amount of intracranial hemorrhage will lead to serious brain damage, resulting in auditory and various system dysfunction. This study also confirmed that the following risk factors had low or extremely low evidence levels. Our study showed that asphyxia is a risk factor for hearing loss in infants, which is similar to the results of the present study [20]. When asphyxia combined with a history of NICU residence increases the risk of hearing loss in 2-year-old infants [33].Thus, the essence of asphyxia is hypoxia, which can have some effect on inner and outer hair cells, mainly outer hair cells [34]. The damage to cochlear cells caused by severe hypoxia is irreversible, but there is currently no clear threshold of hypoxia available to define the critical point of hearing risk [32]. Therefore, it is necessary to follow up on the hearing of this high-risk population to take preventive measures as soon as possible. The administration of ototoxic drugs is associated with a substantial increase in susceptibility to hearing impairment among infants. In particular, the vestibular or cochlear toxicity of aminoglycoside drugs results in irreversible hearing loss [35]. Moreover, the A155G mutation carried on the 12s rRNA gene in mitochondria, and the simultaneous use of ethylene propionic acid can increase the ototoxicity of aminoglycosides [29]. Subgroup analysis of this factor revealed that the non-Asian group exhibited a high degree of heterogeneity, comprising only developing countries. Conversely, the Asian group encompassed not only developing nations but also three developed countries. This observed heterogeneity is tentatively associated with the limited prevalence of domestic hearing screening in developing countries, insufficient funding and a shortage of testing professionals. Septicemia or meningitis, an infectious disease in infants, is one of the risk factors. Sensorineural hearing loss is the most common serious adverse effect of bacterial meningitis [36]. Swedish guidelines recommend that all patients with meningitis undergo otoscopy and be followed up with audiometry [37]. An association between factors such as male sex and hearing loss was not found in this study. The low level of evidence might be due to the collection of data from different countries with a potential admission and detection bias. This study found no significant association between SGA and infant hearing loss, contradicting previous studies [12, 25]. The variation in the proportion of non-LBW infants classified as SGA across different studies and the heterogeneity between studies may explain this discrepancy, suggesting the need for larger sample sizes and rigorous clinical designs to clarify their relationship. Our study also did not observe a significant correlation between infant hearing loss and preterm birth, which may be caused by the improvement of perinatal care conditions and the overall decrease of complications in preterm infants [38].This study also revealed strong correlations between family history of hearing impairment and intrauterine infection and hearing loss in infants. While 60–70% of deafness cases are caused by genetic factors [39], such as the GJB2, GJB3 and SLC26A4 genes, mutations in the GJB2 gene are the most common [40]. In contrast to the results of Karaca et al. [41], we found that hyperbilirubinemia is a significant risk factor for hearing loss, possibly due to variations in hearing screening methods. Elevated bilirubin levels in the blood can damage the auditory nerve and central nervous system. At this time, auditory brainstem response (ABR) tests, which assess the complete function of the outer ear to the lower brainstem pathway, have a greater detection rate for hearing loss than otoacoustic emission (OAE) tests. There has been increasing evidence that the auditory nervous system is the most sensitive nervous system to bilirubin toxicity [42]. Infants with severe jaundice are at increased risk for auditory nerve disorders [43]. Without timely intervention, these children may face problems related to abnormal language development [44].

The limitations of this study are as follows: (1) The exclusion of grey literature in the analyzed studies may introduce publication bias. (2) Several influencing factors, such as racial differences in Africa and Latin America, were not included due to the limited sample size. It is worth noting that 80% of hearing-impaired children worldwide come from low- and middle-income countries, which further reduces confidence in assessing certain risk factors. (3) Inconsistencies between subgroup results and overall findings suggest potential instability in the results of these studies. Future research should involve larger sample sizes from multiple centers to clarify the risk of hearing loss.

Conclusion

The study revealed 14 risk factors that are strongly linked to infant hearing loss, with moderate evidence for four of these risk factors. Health care professionals need to perform premarital counseling, provide medical screening and fertility guidance and perform TORCH screening for pregnant women. Raising awareness and educating the public on the importance of new-born hearing screening are crucial for identifying infants with hearing loss and intervening as soon as possible. Future large-scale, multicenter studies are needed to investigate the combined impact of multiple risk factors on infant hearing loss and to translate these factors into risk-based scoring systems through prospective research.