The investigators recruited 48 middle-aged adults (40–65 years) and matched participants one-to-one by age and socioeconomic status. Twenty-four participants had clinically normal hearing (≤25 dB HL) and 24 had untreated, postlingually acquired hearing loss. Among the hearing-impaired group, nine had unilateral loss and 15 had bilateral loss. Audiometry followed standard procedures (0.25–8 kHz) in a sound-treated room; pure tone average (PTA) was calculated from 0.5, 1, 2 and 4 kHz. Degree of hearing loss classification used Goodman’s criteria. The study excluded congenital hearing loss, current amplification users, and those with neurogenic or psychogenic conditions. All participants scored ≥26 on the MoCA and had at least 10 years of formal education, reducing the likelihood of pre-existing cognitive impairment confounding results.
Cognitive evaluation targeted three core domains: memory, attention and executive function. Memory was assessed with a contextual memory test (CMT) plus forward and backward digit span tasks. Attention was measured using a Go/No-Go paradigm with accuracy and reaction time outcomes. Executive function used timed verbal fluency tasks (letter and category). Except for the CMT (PowerPoint presentation), tasks were delivered visually via jsPsych. Practice trials preceded each test. Stimuli and instructions were visually presented to minimize the influence of hearing status on task delivery. The total test battery lasted approximately 40 minutes, with rest breaks or split sessions permitted for fatigue.
Across the battery, the cohort with untreated hearing loss performed worse than the normal-hearing group on every cognitive measure administered. The study team reported statistically significant differences with notably large effect sizes for memory measures (CMT, forward and backward digit span), for the category verbal fluency task, and for Go/No-Go accuracy. Letter fluency showed a moderate effect size. The authors did not report precise p-values or subgroup-specific pairwise statistics in the source text beyond these effect-size characterizations. Subgroup composition (unilateral versus bilateral) was documented, but the source did not present detailed, separate outcome statistics for UHL and BHL beyond inclusion in subgroup analyses.
Investigators pooled NH and HL participants to examine correlations between better-ear PTA and cognitive outcomes. Spearman correlation analyses, adjusted for multiple comparisons, indicated relationships between worse hearing thresholds and poorer cognitive task performance. To summarise shared variance across the cognitive measures, the team performed principal component analysis. A single principal component emerged (eigenvalue 3.18), explaining 40% of total variance and carrying factor loadings from 0.3 to 0.74. Using that PCA-derived composite cognitive score as the dependent variable, multivariate linear regression identified better-ear pure-tone thresholds as an independent predictor of cognitive performance; age and education were not independent predictors in that model, according to the authors.
The study applied nonparametric tests where distributions deviated from normality (FDS, BDS, Go/No-Go). Group comparisons used Mann–Whitney U tests with Holm step-down adjustments for multiple comparisons; subgroup comparisons used Kruskal–Wallis with Dunn post hoc tests and Bonferroni correction. Effect sizes were emphasized alongside p-values, reflecting the exploratory design and modest sample. The authors reported data-processing steps to ensure PCA suitability (KMO = 0.68; Bartlett’s test p < 0.001). Analyses were conducted in JASP and R, and statistical significance was set at p < 0.05.
The authors explicitly acknowledged limitations that temper interpretation. The sample size was modest and obtained via convenience sampling, limiting generalisability. The cross-sectional design prevents causal or temporal inferences; the source did not report longitudinal follow-up data. Duration of hearing loss was documented for participants but was not included in the primary regression analyses. The source also did not provide detailed, separate outcome statistics for unilateral versus bilateral subgroups in the principal analyses. The investigators stated that longitudinal studies are needed to determine temporal relationships between midlife hearing loss and cognitive change, and to clarify whether treatment of hearing loss alters trajectories identified here.