BACKGROUND: We conducted a nationwide population-based case-control study to analyse potential predisposing factors for hearing loss (HL) that present during the fetal, perinatal, and postnatal periods in prematurely born children.
METHODS: This study enrolled 21,576 children born at < 37 weeks of gestation; 3,596 with HL and 17,980 with normal hearing born between 2002 and 2015, matched for sex, age at diagnosis, and enrollment time. Data were abstracted from the concatenation of three nationwide databases for overall risk factors till the diagnosis of HL.
RESULTS: Maternal HL, maternal diabetes, particularly type 1 diabetes mellitus, and at or before 32 weeks of gestation were the major obstetric risk factors for HL. Prematurely born children who were born via cesarean section and received a combination of antenatal steroids and magnesium sulfate exhibited a significantly reduced risk of developing HL. Ear malformation was a critical predictor for HL. The major postnatal risk factors included seizure and ototoxic drugs use. Premature infants diagnosed with more than 1 diagnosis of bronchopulmonary dysplasia, necrotizing enterocolitis, and intracerebral hemorrhage were at an increased risk of developing HL. Congenital CMV infection and recurrent acute otitis were also independent postnatal factors for HL in prematurely born children.
CONCLUSIONS: To reduce the incidence of childhood HL in prematurely born children, aggressive management of premature birth-related consequences and treatable causes and longitudinal audiological follow-up with early detection and adequate intervention are crucial.

Studies on risk factors for childhood hearing loss (HL) are usually based on questionnaires or small sample sizes. We conducted a nationwide population-based case-control study to comprehensively analyze the maternal, perinatal, and postnatal risk factors for HL in full-term children.
We retrieved data from three nationwide databases related to maternal characteristics, perinatal comorbidities, and postnatal characteristics and adverse events. We used 1:5 propensity score matching to include 12,873 full-term children with HL and 64,365 age-, sex-, and enrolled year-matched controls. Conditional logistic regression was used to evaluate the risk factors for HL.
Among the various maternal factors, maternal HL (adjusted odds ratio [aOR]: 8.09, 95% confidence interval [95% CI]: 7.16-9.16) and type 1 diabetes (aOR: 3.79, 95% CI: 1.98-7.24) had the highest odds of childhood hearing impairment. The major perinatal risk factors for childhood hearing impairment included ear malformations (aOR: 58.78, 95% CI: 37.5-92.0) and chromosomal anomalies (aOR: 6.70, 95% CI: 5.25-8.55), and the major postnatal risk factors included meningitis (aOR: 2.08, 95% CI: 1.18-3.67) and seizure (aOR: 3.71, 95% CI: 2.88-4.77). Other factors included acute otitis media, postnatal ototoxic drug use, and congenital infections.
Many risk factors for childhood HL identified in our study are preventable, such as congenital infection, meningitis, ototoxic drug use, and some maternal comorbidities. Accordingly, more effort is required to prevent and control the severity of maternal comorbidities during pregnancy, initiate genetic diagnostic evaluation for high-risk children, and aggressive screening for neonatal infections.