A Meta-Analytic Approach to Identification of Clinical Biomarkers of Hearing Loss in Cognitive Impairment Publisher Pubmed



Moghadam AJ ; Sadr H ; Asadi A ; Mehrabi S
Source: Neurological Sciences Published:2026

Abstract

Background: Cognitive impairment (CI) and Hearing Loss (HL) are common in aging populations. Although longitudinal studies identify HL as a risk factor for CI, the reverse association remains unclear. This meta-analysis quantified the cross-sectional prevalence of HL in individuals with established CI and examined associated clinical and biological markers. Methods: A systematic search of EBSCOhost, Embase, PubMed/MEDLINE, Scopus, and Web of Science (Core Collection) was conducted for studies with a case-control or cohort design. Risk of bias was assessed using ROBINS-E. Meta-analyses were performed in RevMan 8.8 using fixed-effects models and statistical significance was set at P < 0.05. Results: Twenty-four studies were included, comprising 6,091 patients with CI, 8,575 individuals with normal cognition, 1,584 CI/HL patients, and 4,507 CI/NH patients. Patients with CI had a markedly higher likelihood of hearing loss (OR = 2.72, 95% CI: 2.39, 3.10, P < 0.00001). Pure tone audiometry (PTA) thresholds were significantly elevated in CI patients compared to those with normal cognition (MD = 5.97 dB, P < 0.00001). The likelihood of smoking was significantly higher in CI/HL patients (OR = 1.22, P = 0.02). APOE ε4 carrier genotype was significantly associated with CI/HL (OR = 1.15, P = 0.03). Cerebrospinal fluid (CSF) biomarkers, including elevated P-tau (MD = 1.96 pg/ml, P = 0.005) and T-tau (MD = 19.40 pg/ml, P < 0.0001), were significantly associated with HL in patients with CI. Conclusion: Hearing loss is substantially more prevalent in individuals with cognitive impairment. Smoking, APOE ε4 genotype, and elevated CSF tau proteins are significant concurrent markers. These findings support routine audiological screening in CI patients and suggest a complex, potentially bidirectional relationship between HL and CI. © Fondazione Societa Italiana di Neurologia 2026.