A decision tree-based prediction model for hearing loss in middle-aged and older adults: Secondary analysis of the Korea National Health and Nutrition Examination Survey

PURPOSE: This study aimed to determine high-risk groups and develop a predictive model of hearing loss among adults in South Korea. METHODS: We analyzed 1,873 participants aged 40 years and older using the Korean National Health and Nutrition Examination Survey (KNHANES). Complex sample analysis and decision tree analysis were conducted. RESULTS: Among the participants, 764 (36.0%) had hearing loss, with 54.5% aged 65 or older. The decision tree analysis identified age as the most significant predictor, revealing six high-risk groups: five in older adults and one in middle-aged adults. The characteristics of high-risk groups varied by age group. In high-risk groups among older adults, sarcopenia, self-reported health status, and economic activity were key predictors, whereas in middle-aged groups, education level and lifestyle factors such as self-reported oral health played a major role. The highest-risk group was older adults with sarcopenia, with 81.3% experiencing hearing loss. CONCLUSION: It is essential to implement age-specific strategies to effectively prevent and manage hearing loss. For older adults, strengthening support systems to address sarcopenia, enhance perceived health status, and reduce social isolation is crucial. For middle-aged adults, prevention should focus on health management, lifestyle modifications, and improved accessibility to healthcare.