Related Experiment Video
Updated: Jul 9, 2026
An Automated System for Sound Localization Testing in Hearing-Impaired Listeners
Published on: March 13, 2026
Hearing Loss in Older Adults: Consistent Determinants Across Two Community-Based Cohorts in Southern China
Dian Zhu1, Xutong Zhong1, Ruiqiang Li1
1Department of Epidemiology, School of Public Health, Sun Yat-sen University, Guangzhou, Guangdong, China, sysu.edu.cn.
Background:
Hearing loss (HL) is common in older adults and is associated with substantial functional decline, yet community-based evidence on its determinants remains limited in China, particularly across different methods of hearing assessment.
Objectives:
To investigate associations and predictors of HL among older adults in southern China using audiometric and self-reported assessments, and to compare patterns across two community-based cohorts.
Methods:
Data were analyzed from 2664 adults aged ≥ 60 years in Shenzhen and 30,518 adults aged ≥ 50 years from the Guangzhou Biobank Cohort Study (GBCS). Moderate-to-severe HL was defined as a pure-tone average (PTA) ≥ 35 dB hearing level in the better-hearing ear, calculated from air-conduction thresholds at 500-8000 Hz. HL was assessed using pure-tone audiometry in Shenzhen and a validated self-reported measure in GBCS. Multivariable logistic regression estimated adjusted odds ratios (aORs) with 95% confidence intervals (CIs). Extreme gradient boosting with SHAP values was used to assess predictor importance.
Results:
Older age, male sex, and lower household income were consistently associated with higher odds of HL in both cohorts. In Shenzhen, metabolic disease (aOR = 1.28, 95% CI: 1.05-1.57) and otitis media (aOR = 2.65, 95% CI: 1.63-4.33) were positively associated with HL, whereas thyroid disease showed an inverse association. In GBCS, alcohol consumption (aOR = 1.28, 95% CI: 1.15-1.43), arthritis (aOR = 1.37, 95% CI: 1.23-1.52), and stroke (aOR = 1.64, 95% CI: 1.06-2.45) were positively associated, while overweight status and nonmanual occupation were inversely associated. Machine-learning analyses consistently identified age, sex, education, income, and chronic diseases as key predictors.
Conclusions:
HL in older adults shows both shared and cohort-specific associations across assessment methods, highlighting sociodemographic and health-related disparities. Targeted community-based screening and prevention strategies are warranted.
