Nomogram for predicting cognitive impairment in middle-aged and elderly individuals with self-reported hearing loss: Insights from the longitudinal CHARLS cohort

TL;DR

A new tool predicts cognitive decline in older adults with hearing loss, using factors like education and social activity to help with early prevention.

Contribution

The first predictive model for cognitive impairment tailored to older adults with self-reported hearing loss.

Findings

Higher education, urban living, and social activity reduce cognitive impairment risk in hearing-impaired older adults.

The nomogram shows strong predictive performance with AUCs between 0.728 and 0.768.

Risk scores vary significantly across demographic subgroups, enabling targeted prevention strategies.

Abstract

•A nomogram predicts cognitive decline in self-reported hearing-impaired older adults.•Education, urban living, and social activity are key protective factors.•The model shows good performance (AUC 0.728–0.768; C-index >0.71).•Risk scores differ significantly across demographic subgroups.•The nomogram facilitates early identification and targeted prevention.

A nomogram predicts cognitive decline in self-reported hearing-impaired older adults.

Education, urban living, and social activity are key protective factors.

The model shows good performance (AUC 0.728–0.768; C-index >0.71).

Risk scores differ significantly across demographic subgroups.

The nomogram facilitates early identification and targeted prevention.

To develop and evaluate a predictive model for cognitive impairment, assessed by a brief cognitive test, among middle-aged and older adults with self-reported hearing loss, aiming to facilitate early identification and targeted intervention in high-risk populations.

This study utilized data from the 2011–2018 waves of the China Health and Retirement Longitudinal Study, including 1,093 individuals aged 45 and above with self-reported hearing impairment but normal cognition at baseline. The average follow-up duration was 5.2-years. Cognitive impairment was defined as a total cognitive score at least one standard deviation below the mean for the corresponding age group. Univariate and multivariate Cox proportional hazards models were used to identify independent predictors, and a nomogram was constructed based on significant variables. Model performance was assessed using calibration curves, time-dependent Concordance indices (C-index), and the Area Under the receiver operating Characteristic Curve (AUC). Risk stratification analysis was conducted across various sociodemographic subgroups.

During follow-up, 152 individuals (13.9%) developed cognitive impairment. Higher education level, urban residence, and participation in leisure social activities (e.g., playing mahjong) were independently associated with lower risk of cognitive impairment. The nomogram demonstrated good discriminative performance, with AUCs of 0.728, 0.762, and 0.768, and corresponding time-dependent C-indices of 0.719, 0.761, and 0.767 at 3-, 5-, and 6-year follow-up points, respectively. Calibration plots indicated good agreement between predicted and observed risks. Risk scores varied significantly across subgroups defined by sex, age, education level, and residential location. Kaplan–Meier analyses confirmed the model’s effective risk stratification capability.

This study presents the first predictive model tailored to cognitive risk heterogeneity among older adults with self-reported hearing loss. By incorporating a brief cognitive test, the nomogram demonstrated reliable performance and holds potential for use in risk stratification, population screening, and personalized preventive interventions.

2 ‒ Prospective cohort study based on the longitudinal CHARLS data.

Figures