Hearing Loss and Dementia Prevalence Across Countries: Global Ecological Evidence in Females and Males

Abstract

Background: Hearing loss is a well-established modifiable risk factor for dementia at the individual level, but less is known about whether hearing loss prevalence aligns with dementia prevalence at the country level across diverse socioeconomic and demographic contexts. This study examined sex-stratified country-level associations between hearing loss prevalence and dementia prevalence, while avoiding causal interpretation of ecological associations. Methods: Country-level sex-specific prevalence estimates for hearing loss and dementia prevalence were obtained from the Global Burden of Disease data resources. Analyses were conducted separately for females and males and included descriptive sample summaries, Pearson and Spearman correlations, principal component analysis, partial correlations used as sequential robustness checks, and theory-informed multivariable linear regression. Variables were log10-transformed for the primary analyses. Models adjusted for economic development, urbanisation, the Henneberg Index, and sex-specific life expectancy at age 60. Formal sex comparison was evaluated using coefficient-comparison testing of the hearing-loss effect from comparable adjusted models. Results: The global bivariate association between hearing loss prevalence and dementia prevalence was strong in females (Pearson r = 0.827; Spearman rho = 0.845; p < 0.001) and males (Pearson r = 0.848; Spearman rho = 0.855; p < 0.001). Principal component analyses indicated that hearing loss prevalence clustered with socioeconomic and demographic indicators in both sexes, explaining more than 70% of total variance. In multivariable regression models, adding hearing loss increased explanatory power in females (R² from 0.661 to 0.827) and males (R² from 0.674 to 0.811). A formal coefficient-comparison test did not indicate a statistically meaningful difference between the female and male hearing-loss coefficients in the fully adjusted models (z = 0.82; p = 0.41). Conclusions: Hearing loss prevalence is strongly associated with dementia prevalence at the country level in both females and males after adjustment for macro-level socioeconomic and demographic covariates. The findings support the population-level relevance of hearing health for dementia surveillance and prevention, while indicating broad convergence rather than clear sex-based divergence in the hearing loss-dementia association.

Keywords

hearing loss dementia prevalence sex-stratified analysis ecological analysis global health socioeconomic development

Background

Dementia is a growing global public health challenge, with prevalence rising rapidly as populations age and life expectancy increases worldwide. In the absence of effective disease-modifying therapies, prevention strategies have increasingly focused on identifying modifiable risk factors that may reduce dementia burden at the population level (Norton et al., 2014). Among these, hearing loss has emerged as one of the most prominent and potentially preventable contributors to cognitive decline and dementia (Lin et al., 2011).

At the individual level, extensive evidence from longitudinal cohort studies and meta-analyses demonstrates that hearing loss is associated with increased risk of incident dementia, accelerated cognitive decline, and adverse structural and functional brain changes (Deal et al., 2015; Lin et al., 2011; Loughrey et al., 2018). Proposed mechanisms include increased cognitive load resulting from degraded auditory input, sensory deprivation, social isolation, and shared neuropathological or vascular pathways (Peelle, 2018; Pichora‐Fuller et al., 1995; Tun et al., 2009). On the basis of this evidence, hearing loss has been identified in major dementia prevention frameworks as a leading modifiable risk factor across the life course (WHO, 2019; WHO, 2021).

Despite this growing body of individual-level evidence, most existing studies have been conducted in cohorts of adults, often in high-income settings (Readman et al., 2025; Yu et al., 2024). Considerably less is known about whether the hearing loss-dementia association is expressed at the population level across diverse socioeconomic, demographic, and epidemiological contexts. Country-level ecological analyses cannot infer individual risk, but they provide a complementary perspective by examining whether national hearing loss prevalence aligns with national dementia prevalence after accounting for broader contextual factors (Greenland, 2001; Morgenstern, 1995; Readman et al., 2025).

Sex is relevant to this question because dementia prevalence, longevity, health behaviours, occupational exposures, and access to healthcare differ between females and males (Huque et al., 2023; LaPlume et al., 2022). However, sex-stratified ecological analyses should not be interpreted as evidence of sex differences unless formal effect-modification testing is undertaken (Altman & Bland, 2003; Greenland & Morgenstern, 1989; Knol & VanderWeele, 2012). In this study, sex stratification was used primarily to describe whether country-level patterns are broadly similar or divergent between females and males, with formal comparison added to avoid overinterpreting parallel models.

Interpretation of population-level associations is additionally challenged by the close interrelationship between hearing loss prevalence and broader macro-level conditions (Haile et al., 2021; Scholes et al., 2018). Economic development, urbanisation, increased longevity, and reduced biological selection pressures may be associated with both hearing loss and dementia prevalence (Mollalo et al., 2025; You et al., 2022; You et al., 2025). Consequently, observed correlations may reflect shared structural determinants rather than a distinct association between hearing loss and dementia (Greenland, 2001). Analytical approaches that account for these interdependencies are therefore essential for assessing the independent explanatory contribution of hearing loss within a broader socioeconomic and demographic framework (Morgenstern, 1995; Vatcheva et al., 2016).

Against this background, the present study uses global country-level data to examine sex-stratified associations between hearing loss prevalence and dementia prevalence. The specific aims were to: (1) characterise the country-level association between hearing loss prevalence and dementia prevalence in females and males; (2) assess whether this association persists after adjustment for socioeconomic development, urbanisation, the Henneberg Index, and longevity; and (3) formally assess whether the adjusted hearing-loss coefficient differs between female and male models. The main contribution is the extension of established individual-level evidence to a global ecological framework, rather than the claim of a new individual-level risk pathway or strong sex-specific divergence.

Methods

Data Sources

Sex-specific national prevalence estimates for dementia and hearing loss were obtained from the Global Burden of Disease (GBD) Study produced by the Institute for Health Metrics and Evaluation, University of Washington, USA (IHME, 2024). Dementia prevalence estimates for 2023, including Alzheimer’s disease and other dementias, were generated separately for females and males using standard GBD modelling frameworks that integrate data from cohort studies, health surveys, hospital and outpatient records, administrative claims, and mortality registries, with case definitions based on International Classification of Diseases (ICD) criteria (IHME, 2024; Nichols et al., 2022).

Hearing loss prevalence estimates for 2023 were also derived from the GBD database using established statistical modelling approaches based on population registries, epidemiological surveys, vital registration systems, and hospital records (IHME, 2023). Dementia and hearing loss were expressed as age-inclusive prevalence rates per 100,000 population and were log10-transformed prior to analysis to improve linearity and cross-national comparability.

Macro-structural indicators were sourced from internationally recognised databases. Economic development was measured using GDP per capita (PPP; constant 2020 US dollars) and urbanisation as the proportion of the population living in urban areas, both from the World Bank World Development Indicators (World Bank, 2022); Population ageing was indexed using sex-specific life expectancy at age 60 (e₆₀; 2020) from the World Health Organization (WHO) Global Health Observatory (WHO, 2020). Population-level exposure to relaxed natural selection was assessed using the Henneberg Index (Biological State Index, I_bs) derived from published mortality and fertility estimates (You & Henneberg, 2017, 2026).

Countries were classified using multiple international frameworks to support stratified analyses, including World Bank income groups, United Nations development status, and the six WHO regions. Additional geopolitical groupings (e.g., APEC, OECD, European Union, Arab World, Latin America and the Caribbean, Southern African Development Community, Shanghai Cooperation Organisation, and English-speaking countries) were included to capture broader contextual variation. Depending on data availability, analyses included up to 204 countries.

Study Design and Variable Specification

A sex-stratified ecological, cross-national design was used. Dementia prevalence was specified as the outcome variable, and hearing loss prevalence was specified as the primary explanatory variable. Analyses were conducted separately for females and males to describe within-sex country-level patterns.

Because parallel sex-specific models do not, by themselves, provide a formal test of sex differences, the adjusted hearing loss coefficients were compared using a coefficient-difference test based on the unstandardised coefficients and standard errors from comparable fully adjusted female and male models. A pooled sex-by-hearing interaction model was not used because the dataset comprised sex-specific country-level prevalence estimates with separate female and male dementia outcomes. The coefficient-difference test was therefore used to directly assess whether the adjusted hearing loss–dementia association differed between the sex-specific models.

This specification was guided by life-course considerations, as hearing loss is commonly identified earlier in adulthood, whereas dementia predominantly occurs in later life. Sex stratification was used to determine whether the population-level association was evident in both females and males, rather than to assume a priori that the association differed by sex.

Except for the Henneberg Index, which was obtained from a single published source, the study variables showed strong and stable country-level correlations across 2009–2023, indicating enduring national patterns. This temporal stability suggests that the use of a single reference year was unlikely to materially affect the observed ecological associations.

Data Preparation

Data from multiple international sources were harmonised into a country-level dataset. Prevalence measures and macro-structural indicators (economic development, urbanisation, Henneberg Index, and ageing) were aligned by country and standardised.

Countries with missing data for variables required in a given analysis were excluded using listwise deletion. Analytic sample sizes are therefore reported for each major analysis. All prevalence and macro-structural variables were log10-transformed in the primary correlation, PCA, partial correlation, and regression analyses to reduce skewness, stabilise variance, and improve linearity.

Multicollinearity was assessed separately in female and male regression models using tolerance and variance inflation factor (VIF) statistics. All models met conventional thresholds (tolerance ≥ 0.10; VIF ≤ 10), indicating no problematic multicollinearity.

Statistical Analysis

Exploratory scatterplots of untransformed rates were examined to assess distributions and visual patterns. Polynomial trend lines in Figure 1 were retained as descriptive visual summaries only and were not used for statistical inference. Primary analyses were based on log10-transformed variables and linear modelling. Associations between hearing loss prevalence and dementia prevalence were evaluated using Pearson and Spearman correlations, both globally and within predefined country groupings. Differences in correlation strength across income levels and development status were tested using Fisher’s r-to-z transformation.

Principal component analysis (PCA) was conducted separately for females and males to assess whether hearing loss prevalence clustered with indicators of population ageing and socioeconomic context, including economic development, urbanisation, the Henneberg Index, and ageing (e₆₀). Sampling adequacy was assessed using the Kaiser–Meyer–Olkin (KMO) statistic and Bartlett’s test of sphericity, and components with eigenvalues greater than one were retained.

Multivariable linear regression models were fitted separately for females and males using the enter method. These theory-informed models were treated as the primary adjusted analyses. Standardised regression coefficients (beta), unstandardised coefficients (B), standard errors, adjusted R² values, and F-statistics were reported. A coefficient-difference test was then used to formally evaluate whether the adjusted hearing-loss coefficient differed between the female and male models. Partial correlations were retained as sequential robustness checks and reported in the supplementary material to show how the hearing loss-dementia association changed after progressive covariate adjustment. Stepwise regression analyses were moved to the supplementary material and treated as exploratory sensitivity analyses only.

Except for exploratory scatterplots generated using Microsoft Excel 2016, all variables were log10-transformed prior to analysis. Statistical analyses were conducted using IBM SPSS Statistics, version 31 (IBM Corp., 2023). Findings are interpreted at the population level and do not imply causality or individual-level risk.

Results

Sample Description and Analytic Samples

The global bivariate analyses included up to 204 countries with sex-specific hearing loss and dementia prevalence estimates. Analyses requiring macro-level covariates used smaller complete-case samples because economic development, urbanisation, Henneberg Index, and life expectancy at age 60 were not available for all countries. The fully adjusted partial correlation and multivariable regression models included 174 countries. Table 1 summarises data availability and analytic sample sizes for the major analyses.

Table 1.

Analytic Sample Sizes and Data Availability Across Major Analyses

Analysis	Main variables included	Analytic n	Notes on missing data
Global bivariate hearing loss-dementia correlations	Sex-specific hearing loss prevalence and dementia prevalence	204	Used maximum available prevalence data
Subgroup correlations by income, development, WHO region, and geopolitical grouping	Sex-specific hearing loss and dementia prevalence within country groups	11–155 across subgroup analyses; see Supplemental Table S2	Group size depended on classification membership and available prevalence data; the worldwide correlation used n = 204 and is reported separately.
PCA of hearing loss and macro-level indicators	Hearing loss prevalence, economic development, urbanisation, Henneberg index, and sex-specific ageing	Female model: 174; male model: 174	Listwise deletion used when macro-level indicators were missing
Sequential partial correlations	Hearing loss prevalence, dementia prevalence, economic development, urbanisation, Henneberg index, and sex-specific ageing	189, 181, or 174 depending on adjustment set	Reported in Supplemental Table S3
Fully adjusted multivariable regression	Dementia prevalence, hearing loss prevalence, economic development, urbanisation, Henneberg index, and sex-specific ageing	174	Primary adjusted models
Exploratory stepwise regression	Same candidate variables as regression models	174	Sensitivity analysis only; reported in Supplemental Table S4.

Notes. Analytic n reflects the maximum available countries for each analysis after listwise deletion of missing variables. All primary statistical analyses used log10-transformed variables. PCA = principal component analysis; WHO = World Health Organization.

Descriptive statistics for the key country-level variables are provided in Supplemental Table S1. This table summarises the available sample size, minimum, maximum, mean, and standard deviation for sex-specific hearing loss prevalence, sex-specific dementia prevalence, economic development, urbanisation, the Henneberg Index, and sex-specific life expectancy at age 60.

Sex-Stratified Association Between Hearing Loss and Dementia Prevalence

Figure 1 presents sex-stratified scatterplots of hearing loss prevalence and dementia prevalence across countries using untransformed rates. In both panels, higher hearing loss prevalence was associated with higher dementia prevalence. The overlaid polynomial curves are descriptive visual summaries of the raw-scale data and were not used for primary inference.

The visual pattern is consistent with the correlation and regression analyses conducted on log10-transformed variables. Because non-linearity was not the focus of the primary statistical framework, inferential interpretation is based on the log10-linear correlation, partial correlation, and multivariable regression models reported below and in the supplementary material.

Overall, Figure 1 provides descriptive visual support for a positive country-level association between hearing loss prevalence and dementia prevalence in both females and males, without implying causal inference or formally tested non-linear effects.

Figure 1.

Association between hearing loss prevalence and dementia prevalence across countries by sex. Panel A shows females and Panel B shows males.

Country-level Correlations Between Hearing Loss and Dementia Prevalence

Sex-stratified bivariate correlation analyses demonstrated strong positive country-level associations between hearing loss prevalence and dementia prevalence in both females and males (Table 2). These analyses show that countries with higher hearing loss prevalence generally had higher dementia prevalence, but they should not be interpreted as evidence of individual-level risk or as a formal test of sex differences.

Table 2.

Pearson and Spearman Correlations Among Hearing Loss, Dementia, Socioeconomic, Demographic, and Ageing Indicators (log₁₀-Transformed)

Variable Pair	Pearson r	Spearman ρ	p-value
Female hearing loss – Male hearing loss	0.968	0.969	<0.001
Female dementia – male dementia	0.972	0.971	<0.001
Female ageing – Male ageing	0.916	0.907	<0.001
Female hearing loss – Female dementia	0.827	0.845	<0.001
Male hearing loss – Male dementia	0.848	0.855	<0.001
Female dementia – Economic development	0.767	0.783	<0.001
Male dementia – Economic development	0.777	0.790	<0.001
Economic development – Henneberg index	0.768	0.891	<0.001
Economic development – Female ageing	0.832	0.843	<0.001
Economic development – Male ageing	0.747	0.750	<0.001
Urbanisation – Economic development	0.674	0.708	<0.001
Henneberg index – Male dementia	0.738	0.843	<0.001
Henneberg index – Female ageing	0.741	0.856	<0.001
Henneberg index – Male ageing	0.637	0.756	<0.001

Notes. Hearing loss and dementia prevalence were sourced from the Global Burden of Disease Study (Institute for Health Metrics and Evaluation); economic development (GDP per capita, PPP) and urbanisation from the World Bank; sex-specific ageing indexed by life expectancy at age 60 from the World Health Organization; and the Henneberg Index from published estimates. All variables were log10-transformed. Correlations were statistically significant unless otherwise indicated (p < 0.001; n = 178-204 depending on variable availability).

Female hearing loss prevalence was strongly correlated with female dementia prevalence (Pearson r = 0.827; Spearman rho = 0.845; p < 0.001). Male hearing loss prevalence was also strongly correlated with male dementia prevalence (Pearson r = 0.848; Spearman rho = 0.855; p < 0.001). The similar coefficients indicate that the country-level association was evident in both sexes, while formal sex comparison is addressed in the regression section below.

Across sex-specific comparisons, Pearson and Spearman coefficients were concordant and statistically significant, supporting the robustness of the bivariate association (Table 2).

Sex-stratified correlations across income levels, development status, WHO regions, and selected geopolitical groupings are presented in Supplemental Table S2. At the global level, associations were strong in both females (Pearson r = 0.827; Spearman rho = 0.845) and males (Pearson r = 0.848; Spearman rho = 0.855).

Correlation strength varied primarily by socioeconomic context rather than sex. Associations were modest in low-income countries but strong in lower-middle- and upper-middle-income countries for both sexes, while correlations in high-income countries remained positive but attenuated. When grouped as LMICs versus high-income countries, correlations were significantly stronger in LMICs for both females and males. Similar patterns were observed across WHO regions and geopolitical groupings, with no region demonstrating a strong association in only one sex.

Overall, variation in correlation strength was driven more by income level, development status, and regional context than by sex, reinforcing the interpretation that the hearing loss-dementia association is broadly similar in females and males at the population level.

Principal Component Structure of Hearing Loss and Structural Indicators by Sex

Sex-stratified principal component analyses were conducted to assess whether hearing loss prevalence clustered with socioeconomic, demographic, and longevity-related indicators. Sampling adequacy was high in both models (females: KMO = 0.839; males: KMO = 0.829), and Bartlett’s tests of sphericity were significant (both p < 0.001), confirming suitability for factor analysis.

In both sexes, a single component (eigenvalue > 1) was retained, explaining 70.86% of total variance in females and 70.24% in males. All variables loaded positively and strongly on the component. Among females, the highest loadings were observed for economic development (0.933), ageing (0.903), Henneberg index (0.880), and hearing loss prevalence (0.762). A comparable pattern was evident in males, with strong loadings for affluence (0.939), Henneberg index (0.873), ageing (0.824), and hearing loss prevalence (0.821).

Overall, these results show that hearing loss prevalence is embedded within a shared socioeconomic-demographic-health structure in both females and males, with similar component patterns across sexes (Table 3).

Table 3.

Principal Component Analysis of Hearing Loss, Socioeconomic, Demographic, and Ageing Indicators by Sex

Variable	Female model component 1 loading	Female communality (h²)	Male model component 1 loading	Male communality (h²)
Hearing loss prevalence	0.762	0.580	0.821	0.675
Economic development	0.933	0.871	0.939	0.882
Urbanisation	0.709	0.503	0.716	0.513
Henneberg index	0.880	0.774	0.873	0.763
Ageing e60	0.903	0.815	0.824	0.680
Analytic n	174	—	174	—
Eigenvalue	3.543	—	3.512	—
Variance explained (%)	70.86	—	70.24	—
KMO	0.839	—	0.829	—
Bartlett’s test, χ² (df)	584.91 (10)	—	549.09 (10)	—
Bartlett’s p-value	<0.001	—	<0.001	—

Notes. Hearing loss prevalence and macro-level indicators were log10-transformed. Principal component analysis was conducted using the correlation matrix with listwise deletion. One component was retained in each sex-specific model based on eigenvalue greater than one. No rotation was applied because only one component was extracted. KMO = Kaiser–Meyer–Olkin measure of sampling adequacy. The retained component represents a shared socioeconomic-demographic-health structure with broadly consistent loading patterns across females and males.

Sequential partial correlations are reported in Supplemental Table S3 as robustness checks. These analyses showed that the hearing loss-dementia association remained positive and statistically significant after progressive adjustment for economic development, urbanisation, Henneberg Index, and sex-specific ageing.

Theory-informed Multivariable Regression and Formal Sex Comparison

Theory-informed multivariable linear regression models were estimated separately for females and males to examine predictors of dementia prevalence (Table 4). All models showed strong overall fit and were statistically significant (p < 0.001).

Table 4.

Multivariable Linear Regression Models Predicting Dementia Prevalence by Sex

Predictor	Female base model β (B ± SE)	p	Female + hearing model β (B ± SE)	p	Male base model β (B ± SE)	p	Male + hearing model β (B ± SE)	p
Economic development	0.258 (0.200 ± 0.077)	0.010	0.128 (0.099 ± 0.056)	0.076	0.339 (0.258 ± 0.069)	<0.001	0.053 (0.041 ± 0.057)	0.474
Urbanisation	0.036 (0.071 ± 0.120)	0.556	0.124 (0.245 ± 0.087)	0.006	−0.005 (−0.009 ± 0.116)	0.936	0.072 (0.139 ± 0.089)	0.121
Henneberg index	0.168 (1.522 ± 0.668)	0.024	−0.018 (−0.161 ± 0.497)	0.747	0.343 (3.061 ± 0.621)	<0.001	0.132 (1.181 ± 0.505)	0.020
Ageing	0.421 (2.253 ± 0.451)	<0.001	0.259 (1.383 ± 0.330)	<0.001	0.229 (1.076 ± 0.313)	<0.001	0.206 (0.969 ± 0.240)	<0.001
Hearing loss	—	—	0.567 (1.412 ± 0.111)	<0.001	—	—	0.574 (1.559 ± 0.142)	<0.001
Constant	−2.520 ± 1.157	0.031	−13.502 ± 1.199	<0.001	0.608 ± 0.833	0.467	−13.023 ± 1.392	<0.001
R²	0.661	—	0.827	—	0.674	—	0.811	—
Adjusted R²	0.653	—	0.822	—	0.667	—	0.805	—
F (df)	82.2 (4,169)	<0.001	160.3 (5,168)	<0.001	87.5 (4,169)	<0.001	144.0 (5,168)	<0.001

Notes. Hearing loss and dementia prevalence were obtained from the Global Burden of Disease Study, produced by the Institute for Health Metrics and Evaluation. Economic development was measured using GDP per capita, PPP; urbanisation was obtained from the World Bank; sex-specific ageing was indexed by life expectancy at age 60 from the World Health Organization; and reduced selection opportunity was measured using the Henneberg Index from published estimates. Dementia prevalence rates for females and males were log10-transformed. β denotes standardised regression coefficients, B denotes unstandardised coefficients, and SE denotes standard error. All models were estimated using listwise deletion (n = 174). Base models included socioeconomic and demographic covariates only, whereas “+ hearing” models additionally incorporated sex-specific hearing loss prevalence. A coefficient-difference test comparing the fully adjusted hearing-loss coefficients showed no statistically significant female–male difference (z = 0.82; p = 0.41).

In baseline models excluding hearing loss, socioeconomic and demographic covariates explained similar proportions of variance in dementia prevalence among females (R² = 0.661) and males (R² = 0.674). Ageing, economic development, and the Henneberg Index were retained as relevant macro-level covariates, whereas urbanisation was not consistently independently associated across baseline models.

After inclusion of sex-specific hearing loss prevalence, model explanatory power increased markedly for both females (R² = 0.827) and males (R² = 0.811). Hearing loss was the strongest correlate of dementia prevalence in both fully adjusted models (female: beta = 0.567, B = 1.412, SE = 0.111; male: beta = 0.574, B = 1.559, SE = 0.142; both p < 0.001).

Formal comparison of the unstandardised hearing-loss coefficients did not indicate a statistically significant difference between the female and male models (difference in B = 0.147; z = 0.82; p = 0.41). These results support a strong association in both sexes, while avoiding interpretation of small differences in parallel models as evidence of meaningful sex-based effect modification.

Exploratory stepwise regression sensitivity analyses are reported in Supplemental Table S4 and are not used as the basis for primary inference.

Discussion

This study adds global ecological evidence to the individual-level literature on hearing loss and dementia by examining whether the association is also evident at the country level. Using sex-stratified ecological data and complementary analytical approaches, the findings show that national hearing loss prevalence aligns closely with dementia prevalence in both females and males. This association persisted after adjustment for economic development, urbanisation, the Henneberg Index, and ageing, suggesting that hearing loss captures a population-level dimension of dementia burden not fully explained by the included macro-level covariates. These findings should be interpreted as ecological associations rather than evidence of individual-level risk or causality.

Population-level Relevance Beyond Individual Risk

At the individual level, hearing loss is well established as a risk factor for cognitive decline and dementia, supported by longitudinal cohorts, neuroimaging studies, and intervention trials (Lin et al., 2014; Loughrey et al., 2018; Mueller, 2024; Yu et al., 2024). The present study complements this evidence by showing that countries with higher hearing loss prevalence also tend to experience higher dementia prevalence, even when major macro-level determinants are considered. The ecological design means that these results cannot be translated directly to individual risk, but they do indicate that hearing health may be informative for population-level dementia surveillance and prevention planning.

The persistence of the association after adjustment challenges the interpretation that hearing loss prevalence merely reflects underlying socioeconomic development or population ageing (Liang et al., 2021; Nichols et al., 2022). While both hearing loss and dementia are shaped by demographic transition, the findings suggest that hearing loss prevalence captures an additional dimension of population-level vulnerability not fully explained by the covariates included in this analysis (Nichols et al., 2022; Alzheimer’s Disease International, 2025).

Structural Embedding Within Demographic and Socioeconomic Systems

These PCA and regression findings indicate that hearing loss prevalence is closely patterned within broader socioeconomic and demographic systems. Its alignment with economic development, urbanisation, longevity, and reduced selection opportunity suggests that hearing loss may function not only as a clinical condition, but also as a population-level marker of ageing, survival patterns, occupational and environmental exposures, and health-system capacity for detection and care (Emmett & Francis, 2015; Haile et al., 2021; Stevens et al., 2013; You & Henneberg, 2022; You et al., 2025). This structural patterning may help explain why hearing loss prevalence remains strongly associated with dementia prevalence across countries, even after adjustment for the included macro-level covariates (Krieger, 2024; Livingston et al., 2020; VanderWeele, 2019).

Rather than undermining the relevance of hearing loss, this contextual positioning strengthens its interpretability as a population-level indicator (WHO, 2021; VanderWeele, 2019). Hearing loss prevalence likely reflects cumulative exposure to multiple upstream influences across the life course, including prolonged survival into older age, diminished biological selection, environmental noise, occupational risk environments, and unequal access to hearing care (Basner et al., 2014; Man et al., 2021; Nelson et al., 2005). In this sense, hearing loss functions as an integrative marker linking biological vulnerability with social and environmental context, making it particularly informative for population-health analyses of dementia burden (Diez Roux, 2001; Peelle, 2018).

Sex-specific Patterns: Convergence Rather Than Divergence

Despite well-documented sex differences in dementia prevalence, longevity, and health trajectories, this study found broadly similar country-level hearing loss-dementia associations in females and males (Anstey, Peters et al., 2021; Huque et al., 2023). The sex-stratified models showed strong positive associations in both sexes, and formal comparison of the fully adjusted hearing-loss coefficients did not support a statistically significant female-male difference. Taken together, these findings support broad convergence between females and males rather than clear sex-based divergence in the country-level hearing loss–dementia association (Altman & Bland, 2003; Knol & VanderWeele, 2012).

Small numerical differences between female and male estimates may reflect differences in occupational noise exposure, cardiovascular profiles, healthcare access, or ascertainment patterns (Gong, Harris et al., 2023; Kim et al., 2018; Wang et al., 2021). However, these differences were not the main finding and should not be overinterpreted. The stronger conclusion is that the association was present in both sexes, while heterogeneity across income levels, development status, and regions appeared more prominent than sex-based divergence.

Income and Development Gradients

The attenuation of associations in high-income and highly developed countries warrants careful interpretation. In these settings, widespread access to hearing aids, cochlear implants, and rehabilitative services may mitigate the functional and social consequences of hearing loss, while more comprehensive cognitive screening and diagnostic infrastructure may alter dementia ascertainment patterns (Lin et al., 2023; Mick et al., 2014; Shukla et al., 2020). In contrast, in low- and middle-income settings, untreated hearing impairment may exert a stronger population-level impact through pathways such as social isolation, communication barriers, delayed healthcare access, and reduced cognitive stimulation (Olusanya et al., 2014; Stern, 2012; World Health Organization, 2021).

An alternative explanation is saturation. In highly aged populations where both hearing loss and dementia prevalence are already elevated, between-country variability may be limited, reducing observable correlations (Bloom et al., 2015; Harper, 2014). Importantly, these income- and development-related gradients were similar for females and males, further supporting the interpretation that contextual factors, rather than sex, are the primary drivers of heterogeneity in association strength (Diez Roux, 2001; Krieger, 2024; You, 2025).

Implications for Dementia Prevention and Public Health

The findings have implications for dementia prevention and surveillance. While individual-level research has established hearing loss as a modifiable risk factor for cognitive decline and dementia, the present analysis shows that hearing loss prevalence is also a country-level correlate of dementia prevalence. This supports the inclusion of hearing health indicators in broader population-health approaches to cognitive ageing.

From a public health perspective, earlier identification of hearing impairment, improved access to affordable hearing technologies, and integration of hearing care into primary healthcare may be relevant for dementia prevention strategies, particularly in low- and middle-income settings where untreated hearing loss remains common. However, intervention effects cannot be inferred from the present ecological analysis.

The observed association should not be interpreted as evidence of direct causality. Hearing loss prevalence likely reflects cumulative biological, social, environmental, and structural influences shaping dementia risk across populations. Nevertheless, the consistency of the association suggests that hearing loss may function as a useful population-level indicator within dementia surveillance frameworks.

Effective prevention requires addressing upstream determinants, as hearing loss is embedded within demographic transition, requiring integrated clinical and policy responses.

Strengths and Limitations

This study has several strengths. It draws on a global dataset with wide geographic coverage, uses a sex-stratified ecological framework, and relies on harmonised international estimates. The use of multiple analytical approaches, together with formal coefficient comparison and relocation of stepwise analyses to the supplementary material, strengthens the transparency of the revised analysis while avoiding overreliance on data-driven model selection.

Several limitations should be acknowledged. As an ecological analysis, the findings cannot be interpreted as reflecting individual-level risk or causal relationships. Residual confounding may persist due to unmeasured or imperfectly captured factors such as education, cardiovascular disease burden, healthcare quality, hearing-aid uptake, diagnostic practice, or dementia ascertainment. Prevalence estimates for hearing loss and dementia are modelled estimates and may vary in precision across regions, particularly where primary data are sparse. Finally, the cross-sectional design precludes assessment of temporal sequencing or lagged effects between hearing loss exposure and subsequent dementia outcomes.

Future research integrating longitudinal population-level data, lagged exposure models, and multilevel designs linking individual and contextual data would help clarify underlying pathways and strengthen causal inference.

Conclusion

This study provides evidence that hearing loss prevalence is a strong and consistent country-level correlate of dementia prevalence worldwide in both females and males. Formal comparison did not support clear sex-based divergence in the adjusted hearing loss-dementia association. Overall, the findings position hearing loss as an important population-level indicator within the global dementia landscape, reflecting the combined influence of biological vulnerability and macro-level social and demographic context.

Supplemental Material

Supplemental Material - Hearing Loss and Dementia Prevalence Across Countries: Global Eco-logical Evidence in Females and Males

Supplemental Material for Hearing Loss and Dementia Prevalence Across Countries: Global Eco-logical Evidence in Females and Males by Wenpeng You in Dementia

Footnotes

Acknowledgements

The author acknowledges the international organisations that curate and maintain the publicly available datasets used in this study.

ORCID iD

Wenpeng You

Ethical Considerations

Ethical approval was obtained from the Office of Research Ethics, Compliance and Integrity (ORECI), The University of Adelaide (Approval No. 36289).

Consent to Participate

Consent to participate was not applicable, as this study used publicly available, de-identified data.

Author Contributions

Wenpeng You: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Writing – original draft; Writing – review and editing.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The data analysed in this study are described in the Methods section and were obtained from publicly available databases maintained by international organisations. As all datasets are openly accessible for non-commercial research use, no additional approvals were required, in accordance with the data access policies of the respective data providers.*

GEN AI Use Statement

Generative AI or AI-assisted tools were not used in the creation, modification, or enhancement of any figures, images, artwork, or graphical materials in this manuscript. All visual elements were produced using standard statistical software based on the original study data. Generative AI did not contribute to the study design, analytical procedures, or methodological processes, and no AI-generated artwork, graphical abstracts, or cover images were included in this submission.

Supplemental Material

Supplemental material for this article is available online.

Author Biography

Wenpeng You, PhD, MSc, BSc is a Lecturer in the School of Nursing and Midwifery at Western Sydney University, Australia, and holds appointments with the School of Pharmacy and Biomedical Sciences and Adelaide Nursing School at the University of Adelaide. His research focuses on population health, dementia epidemiology, and the role of structural, demographic, and biological determinants in shaping cognitive ageing and neurological disease burden across populations. He has extensive experience in ecological and quantitative research methods, with particular expertise in global health data analysis, health inequalities, and sex-specific patterns in chronic disease epidemiology. His work aims to inform population-level surveillance, prevention strategies, and policy responses to dementia and related conditions.

References

Altman

D. G.

Bland

J. M.

(2003). Interaction revisited: The difference between two estimates. BMJ, 326(7382), 219. https://doi.org/10.1136/bmj.326.7382.219

Alzheimer's Disease International . (2025). World Alzheimer Report 2015: The Global Impact of Dementia: An analysis of prevalence, incidence, cost and trends. Retrieved 20 April 2024, from. https://www.alzint.org/u/WorldAlzheimerReport2015.pdf?

Anstey

K. J.

Peters

Mortby

M. E.

Kiely

K. M.

Eramudugolla

Cherbuin

Huque

M. H.

Dixon

R. A.

(2021). Association of sex differences in dementia risk factors with sex differences in memory decline in a population-based cohort spanning 20–76 years. Scientific Reports, 11(1), 7710. https://doi.org/10.1038/s41598-021-86397-7

Basner

Babisch

Davis

Brink

Clark

Janssen

Stansfeld

(2014). Auditory and non-auditory effects of noise on health. The Lancet, 383(9925), 1325–1332. https://doi.org/10.1016/S0140-6736(13)61613-X

Bloom

D. E.

Chatterji

Kowal

Lloyd-Sherlock

McKee

Rechel

Rosenberg

Smith

J. P.

(2015). Macroeconomic implications of population ageing and selected policy responses. The Lancet, 385(9968), 649–657. https://doi.org/10.1016/S0140-6736(14)61464-1

Deal

J. A.

Sharrett

A. R.

Albert

M. S.

Coresh

Mosley

T. H.

Knopman

Wruck

L. M.

Lin

F. R.

(2015). Hearing impairment and cognitive decline: A pilot study conducted within the atherosclerosis risk in communities neurocognitive study. American Journal of Epidemiology, 181(9), 680–690. https://doi.org/10.1093/aje/kwu333

Diez Roux

A. V.

(2001). Investigating neighborhood and area effects on health. American Journal of Public Health, 91(11), 1783–1789. https://doi.org/10.2105/ajph.91.11.1783

Emmett

S. D.

Francis

H. W.

(2015). The socioeconomic impact of hearing loss in US adults. Otology & Neurotology, 36(3), 545–550. https://doi.org/10.1097/MAO.0000000000000562

Gong

Harris

Lipnicki

D. M.

Castro‐Costa

Lima‐Costa

M. F.

Diniz

B. S.

Xiao

Lipton

R. B.

Katz

M. J.

Wang

Preux

P. M.

Guerchet

Gbessemehlan

Ritchie

Ancelin

M. L.

Skoog

Najar

Sterner

T. R.

Scarmeas

Cohort Studies of Memory in an International Consortium COSMIC . (2023). Sex differences in dementia risk and risk factors: Individual‐participant data analysis using 21 cohorts across six continents from the COSMIC consortium. Alzheimer's and Dementia, 19(8), 3365–3378. https://doi.org/10.1002/alz.12962

10.

Greenland

(2001). Ecologic versus individual-level sources of bias in ecologic estimates of contextual health effects. International Journal of Epidemiology, 30(6), 1343–1350. https://doi.org/10.1093/ije/30.6.1343

11.

Greenland

Morgenstern

(1989). Ecological bias, confounding, and effect modification. International Journal of Epidemiology, 18(1), 269–274. https://doi.org/10.1093/ije/18.1.269

12.

Haile

L. M.

Kamenov

Briant

P. S.

Orji

A. U.

Steinmetz

J. D.

Abdoli

Abdollahi

Abu-Gharbieh

Afshin

Ahmed

Ahmed Rashid

Akalu

Alahdab

Alanezi

F. M.

Alanzi

T. M.

Al Hamad

Ali

Alipour

Al-Raddadi

R. M.

Chadha

(2021). Hearing loss prevalence and years lived with disability, 1990–2019: Findings from the Global Burden of Disease Study 2019. The Lancet, 397(10278), 996–1009. https://doi.org/10.1016/s0140-6736(21)00516-x

13.

Harper

(2014). Economic and social implications of aging societies. Science, 346(6209), 587–591. https://doi.org/10.1126/science.1254405

14.

Huque

Eramudugolla

Chidiac

Ehrenfeld

Matthews

F. E.

Peters

Anstey

K. J.

(2023). Could country-level factors explain sex differences in dementia incidence and prevalence? A systematic review and meta-analysis. Journal of Alzheimer's Disease, 91(4), 1231–1241. https://doi.org/10.3233/JAD-220724

15.

IHME . (2023). GBD results. Retrieved March 12 2023, from. https://vizhub.healthdata.org/gbd-results/

16.

IHME . (2024). Global burden of disease 2021: Findings from the GBD 2021 study.

17.

Kim

M.-Y.

Kim

Hong

C. H.

Lee

S. Y.

Jung

Y.-S.

(2018). Sex differences in cardiovascular risk factors for dementia. Biomolecules & therapeutics, 26(6), 521–532. https://doi.org/10.4062/biomolther.2018.159

18.

Knol

M. J.

VanderWeele

T. J.

(2012). Recommendations for presenting analyses of effect modification and interaction. International Journal of Epidemiology, 41(2), 514–520. https://doi.org/10.1093/ije/dyr218

19.

Krieger

(2024). Epidemiology and the people's health: Theory and context. Oxford University Press.

20.

LaPlume

A. A.

McKetton

Anderson

N. D.

Troyer

A. K.

(2022). Sex differences and modifiable dementia risk factors synergistically influence memory over the adult lifespan. Alzheimer's & Dementia: Diagnosis, Assessment & Disease Monitoring, 14(1), Article e12301. https://doi.org/10.1002/dad2.12301

21.

Liang

Qian

Gao

(2021). Hearing loss and dementia: A meta-analysis of prospective cohort studies. Frontiers in Aging Neuroscience, 13(13), 695117. https://doi.org/10.3389/fnagi.2021.695117

22.

Lin

Ferrucci

Goh

Doshi

Metter

Davatzikos

Kraut

M. A.

Resnick

S. M.

(2014). Association of hearing impairment with brain volume changes in older adults. NeuroImage, 90(1), 84–92. https://doi.org/10.1016/j.neuroimage.2013.12.059

23.

Lin

F. R.

Metter

E. J.

O’Brien

R. J.

Resnick

S. M.

Zonderman

A. B.

Ferrucci

(2011). Hearing loss and incident dementia. Archives of Neurology, 68(2), 214–220. https://doi.org/10.1001/archneurol.2010.362

24.

Lin

F. R.

Pike

J. R.

Albert

M. S.

Arnold

Burgard

Chisolm

Couper

Deal

J. A.

Goman

A. M.

Glynn

N. W.

Gmelin

Gravens-Mueller

Hayden

K. M.

Huang

A. R.

Knopman

Mitchell

C. M.

Mosley

Pankow

J. S.

Reed

N. S.

ACHIEVE Collaborative Research Group . (2023). Hearing intervention versus health education control to reduce cognitive decline in older adults with hearing loss in the USA (ACHIEVE): A multicentre, randomised controlled trial. The Lancet, 402(10404), 786–797. https://doi.org/10.1016/S0140-6736(23)01406-X

25.

Livingston

Huntley

Sommerlad

Ames

Ballard

Banerjee

Brayne

Burns

Cohen-Mansfield

Cooper

Costafreda

S. G.

Dias

Fox

Gitlin

L. N.

Howard

Kales

H. C.

Kivimäki

Larson

E. B.

Ogunniyi

Mukadam

(2020). Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. The Lancet, 396(10248), 413–446. https://doi.org/10.1016/S0140-6736(20)30367-6

26.

Loughrey

D. G.

Kelly

M. E.

Kelley

G. A.

Brennan

Lawlor

B. A.

(2018). Association of age-related hearing loss with cognitive function, cognitive impairment, and dementia: A systematic review and meta-analysis. JAMA Otolaryngology – Head & Neck Surgery, 144(2), 115–126. https://doi.org/10.1001/jamaoto.2017.2513

27.

Man

Chen

Zhang

Yin

Yang

(2021). Global, regional, and national burden of age-related hearing loss from 1990 to 2019. Aging (Albany NY), 13(24), 25944–25959. https://doi.org/10.18632/aging.203782

28.

Mick

Kawachi

Lin

F. R.

(2014). The association between hearing loss and social isolation in older adults. Otolaryngology–Head and Neck Surgery, 150(3), 378–384. https://doi.org/10.1177/0194599813518021

29.

Mollalo

Kramer

Cutty

Hoseini

(2025). Systematic review and meta-analysis of rural-urban disparities in Alzheimer’s disease dementia prevalence. The Journal of Prevention of Alzheimer’s Disease, 12(9), 100305. https://doi.org/10.1016/j.tjpad.2025.100305

30.

Morgenstern

(1995). Ecologic studies in epidemiology: Concepts, principles, and methods. Annual Review of Public Health, 16(1), 61–81. https://doi.org/10.1146/annurev.pu.16.050195.000425

31.

Mueller

S. G.

(2024). Traumatic brain injury and post-traumatic stress disorder and their influence on development and pattern of Alzheimer’s disease pathology in later life. Journal of Alzheimer's Disease, 98(4), 1427–1441. https://doi.org/10.3233/JAD-231183

32.

Nelson

D. I.

Nelson

R. Y.

Concha‐Barrientos

Fingerhut

(2005). The global burden of occupational noise‐induced hearing loss. American Journal of Industrial Medicine, 48(6), 446–458. https://doi.org/10.1002/ajim.20223

33.

Nichols

Steinmetz

J. D.

Vollset

S. E.

Fukutaki

Chalek

Abd-Allah

Abdoli

Abualhasan

Abu-Gharbieh

Akram

T. T.

Al Hamad

Alahdab

Alanezi

F. M.

Alipour

Almustanyir

Amu

Ansari

Arabloo

Ashraf

Vos

(2022). Estimation of the global prevalence of dementia in 2019 and forecasted prevalence in 2050: An analysis for the Global Burden of Disease Study 2019. The Lancet Public Health, 7, e105–e125. https://doi.org/10.1016/s2468-2667(21)00249-8

34.

Norton

Matthews

F. E.

Barnes

D. E.

Yaffe

Brayne

(2014). Potential for primary prevention of Alzheimer's disease: An analysis of population-based data. The Lancet Neurology, 13(8), 788–794. https://doi.org/10.1016/S1474-4422(14)70136-X

35.

Olusanya

B. O.

Neumann

K. J.

Saunders

J. E.

(2014). The global burden of disabling hearing impairment: A call to action. Bulletin of the World Health Organization, 92(5), 367–373. https://doi.org/10.2471/BLT.13.128728

36.

Peelle

J. E.

(2018). Listening effort: How the cognitive consequences of acoustic challenge are reflected in brain and behavior. Ear and Hearing, 39(2), 204–214. https://doi.org/10.1097/AUD.0000000000000494

37.

Pichora‐Fuller

M. K.

Schneider

B. A.

Daneman

(1995). How young and old adults listen to and remember speech in noise. Journal of the Acoustical Society of America, 97(1), 593–608. https://doi.org/10.1121/1.412282

38.

Readman

M. R.

Littlejohn

Dodd

Rhodes

Wareing

Polden

Plack

C. J.

Giebel

(2025). Hearing loss as a risk factor for dementia: A systematic review and meta-analysis from a global perspective. Aging & Mental Health, 29(10), 1831–1844. https://doi.org/10.1080/13607863.2025.2515180

39.

Scholes

Biddulph

Davis

Mindell

J. S.

(2018). Socioeconomic differences in hearing among middle-aged and older adults: Cross-sectional analyses using the Health Survey for England. BMJ Open, 8(2), e019615. https://doi.org/10.1136/bmjopen-2017-019615

40.

Shukla

Harper

Pedersen

Goman

Suen

J. J.

Price

Applebaum

Hoyer

Lin

F. R.

Reed

N. S.

(2020). Hearing loss, loneliness, and social isolation: A systematic review. Otolaryngology–Head and Neck Surgery, 162(5), 622–633. https://doi.org/10.1177/0194599820910377

41.

Stern

(2012). Cognitive reserve in ageing and Alzheimer's disease. The Lancet Neurology, 11(11), 1006–1012. https://doi.org/10.1016/S1474-4422(12)70191-6

42.

Stevens

Flaxman

Brunskill

Mascarenhas

Mathers

C. D.

Finucane

Global Burden of Disease Hearing Loss Expert Group . (2013). Global and regional hearing impairment prevalence: An analysis of 42 studies in 29 countries. The European Journal of Public Health, 23(1), 146–152. https://doi.org/10.1093/eurpub/ckr176

43.

Tun

P. A.

McCoy

Wingfield

(2009). Aging, hearing acuity, and the attentional costs of effortful listening. Psychology and Aging, 24(3), 761–766. https://doi.org/10.1037/a0014802

44.

VanderWeele

T. J.

(2019). Principles of confounder selection. European Journal of Epidemiology, 34(3), 211–219. https://doi.org/10.1007/s10654-019-00494-6

45.

Vatcheva

K. P.

Lee

McCormick

J. B.

Rahbar

M. H.

(2016). Multicollinearity in regression analyses conducted in epidemiologic studies. Epidemiology (Sunnyvale, Calif.), 6(2), 227. https://doi.org/10.4172/2161-1165.1000227

46.

Wang

Yang

Han

Huang

(2021). Sex differences in noise-induced hearing loss: A cross-sectional study in China. Biology of Sex Differences, 12(1), 24. https://doi.org/10.1186/s13293-021-00369-0

47.

WHO . (2019). Risk reduction of cognitive decline and dementia: WHO guidelines. World Health Organization.

48.

WHO . (2020). Life expectancy at age 60 (years). Retrieved March 12 2022, from. https://www.who.int/data/gho/data/indicators/indicator-details/GHO/life-expectancy-at-age-60-(years)

49.

WHO . (2021). Global status report on the public health response to dementia. from. https://iris.who.int/bitstream/handle/10665/344701/9789240033245-eng.pdf?sequence=1

50.

World Bank . (2022). DataBank. World Development Indicators (Public) . Retrieved May 18 2022, from. https://databank.worldbank.org/home.aspx

51.

World Health Organization . (2021). World report on hearing. World Health Organization.

52.

You

(2025). Birth rate as a determinant of dementia incidence: A comprehensive global analysis. American Journal of Alzheimer’s Disease & Other Dementias®, 40, 01–11. https://doi.org/10.1177/15333175241287677

53.

You

Henneberg

(2017). Cancer incidence increasing globally: The role of relaxed natural selection. Evolutionary Applications, 11(2), 140–152. https://doi.org/10.1111/eva.12523

54.

You

Henneberg

(2022). Large household reduces dementia mortality: A cross-sectional data analysis of 183 populations. PLoS One, 17(3), e0263309. https://doi.org/10.1371/journal.pone.0263309

55.

You

Henneberg

(2026). Medical progress, relaxed natural selection, and adolescent obesity: Implications for global health. Future Science OA, 12(1), 2602337. https://doi.org/10.1080/20565623.2025.2602337

56.

You

Henneberg

(2022). Healthcare services relaxing natural selection may contribute to increase of dementia incidence. Scientific Reports, 12(1), 1–10. https://doi.org/10.1038/s41598-022-12678-4

57.

You

Koo

F. K.

Sevastidis

Chang

R. H. C.

(2025). Reevaluating the role of biological aging in dementia: A retrospective cross-sectional global analysis incorporating confounding factors. Geriatric Nursing, 63, 643–651. https://doi.org/10.1016/j.gerinurse.2025.04.023

58.

R.-C.

Proctor

Soni

Pikett

Livingston

Lewis

Schilder

Bamiou

Mandavia

Omar

Pavlou

Lin

Goman

A. M.

Gonzalez

S. C.

(2024). Adult-onset hearing loss and incident cognitive impairment and dementia–a systematic review and meta-analysis of cohort studies. Ageing Research Reviews, 98, 102346. https://doi.org/10.1016/j.arr.2024.102346

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.46 MB

0.00 MB