Association between hearing impairment and cardiometabolic multimorbidity among middle-aged and older adults in China

Abstract

This study examined the association between hearing impairment (HI) and the risk of cardiometabolic multimorbidity (CMM) in middle-aged and older adults, providing evidence to support early warning and screening of high-risk CMM populations. This study analyzed data from the China Health and Retirement Longitudinal Study (CHARLS) spanning 2011 to 2018. The analysis included 9,035 participants aged 45 years and older. The association between HI and the risk of CMM was investigated using multivariable Cox proportional hazards regression, Kaplan-Meier survival curves, and subgroup analyses. The results were as follows: Over a median follow-up of 7 years, 382 incident CMM cases were documented. In the fully adjusted Cox proportional hazards model (Model 3), those with mild HI had a 32% increased risk of CMM (Hazard Ratio [HR] = 1.32, 95% Confidence Interval [CI]: 1.06–1.68), while those with severe HI had a 54% increased risk (HR = 1.54, 95% CI: 1.14–2.07). Subgroup analyses revealed that males with severe HI had an increased risk of CMM (HR = 1.72, 95% CI: 1.09–2.27), while females with severe HI also showed an elevated risk (HR = 1.55, 95% CI: 1.07–2.26). Among individuals aged ≥ 60 years, severe HI was associated with a 70% increased risk of CMM (HR = 1.70, 95% CI: 1.15–2.52). Our findings suggest an association between HI and the risk of CMM in middle-aged and older adults, with the association being more pronounced in males with severe HI conditions and individuals aged 60 years or older.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-45165-1.

Introduction

Cardiometabolic multimorbidity (CMM) refers to the coexistence of two or more cardiometabolic diseases (CMD), such as diabetes, cardiovascular disease (CVD), and stroke^1,2. With the accelerating aging of the global population, the prevalence of CMM has risen significantly, establishing it as a major public health issue that threatens the health of middle-aged and older adults^3,4. Compared to individuals with a single CMD, those with CMM face substantially higher risks of all-cause mortality, recurrent cardiovascular events, and healthcare resource utilization, imposing a considerable burden on healthcare systems^5,6. Therefore, identifying modifiable risk factors for CMM and developing early intervention strategies are of significant clinical importance.

Hearing impairment (HI) is a highly prevalent sensory disorder in older adults and represents the third leading cause of disability globally^7,8, and its incidence increases with age. Approximately 1.57 billion people globally experienced some degree of hearing loss (HL) in 2019, representing 19.3% of the world’s population⁹. Among adults aged 60 years and older, the prevalence of disabling HL surpasses 25%¹⁰. In recent years, the association between HI and systemic chronic disorders has garnered increasing attention, particularly in the field of cardiometabolic health. Previous studies have indicated that individuals with HI have a 20% increased risk of developing CVD compared to those with normal hearing¹¹, the underlying mechanisms may be associated with pathophysiological pathways such as cochlear microvascular ischemia, chronic inflammation, and oxidative stress¹². Furthermore, a cross-sectional study revealed a significantly higher risk of CMM among individuals with dual impairment of sensory vision and hearing (odds ratio [OR] = 1.862, 95% confidence interval [CI]: 1.387–2.500)¹³. However, the independent association between HI and CMM as well as their potential severity-response relationship remains unclear. Although cross-sectional studies have explored the association between HI and CMM¹⁴, large-sample, long-term longitudinal follow-up data in Asian elderly populations remain relatively scarce, and the association between the two has not been fully elucidated. HI may lead to the occurrence of CMM through the synergistic effects of multiple pathophysiological pathways, including behavioral patterns, vascular pathology, inflammatory stress, and psychoneuroendocrine responses. Therefore, HI is speculated to be an independent risk factor for CMM. This study employs nationally representative data from the China Health and Retirement Longitudinal Study (CHARLS) to systematically examine the ‌longitudinal association between HI and CMM. We hypothesized that HI is associated with a higher risk of developing CMM, aiming to establish a scientific foundation for the early prevention and screening of high-risk populations.

Materials and methods

Data and sample sources

The data for this study were derived from the CHARLS. CHARLS is a nationally representative prospective cohort study of middle-aged and older adults (≥ 45 years) in China. Its sampling framework spans 28 provinces, 150 county-level units, and 450 village-level units, ensuring strong regional and population representativeness¹⁵. The study encompasses comprehensive assessments of sociodemographic characteristics (e.g., age, sex, education level), health status and functioning (e.g., history of chronic diseases, hearing function, and activities of daily living), and socioeconomic status (e.g., income, health insurance coverage). CHARLS data collection began with the baseline survey in 2011, and follow-up surveys were conducted in 2013, 2015, 2018, and 2020. The CHARLS was approved by the Biomedical Ethics Committee of Peking University (IRB00001052-11015). All participants provided written informed consent. The datasets can be accessed through the official website (http://charls.pku.edu.cn/en/).

This study employed data from the CHARLS spanning from 2011 to 2018 for analysis. The inclusion criteria were as follows: (1) age ≥ 45 years; (2) completion of hearing function assessment and cardiometabolic-related evaluations at baseline survey; and (3) availability of complete clinical data. The exclusion criteria were as follows: (1) age < 45 years (n = 649); (2) missing data on key covariate information (e.g., sex, marital status, educational level, smoking history and alcohol use history, n = 3952); (3) incomplete data on chronic disease history (including hypertension, diabetes, heart disease, stroke, and chronic kidney disease [CKD], n = 217); (4) pre-existing CMM at baseline survey (n = 400); and (5) death or loss to follow-up during the study period (n = 3445). Ultimately, a total of 9,035 participants were included in the analysis. The detailed screening process is illustrated in Fig. 1.

Fig. 1.

Flowchart of the participants screening process.

Assessment of hearing status and CMM

This study assessed hearing status using a single self-reported question from the CHARLS survey¹⁶, a pragmatic approach with recognized limitations in large-scale epidemiological studies. This self-reported question has been validated and widely used among the Chinese population¹³, with evidence showing good agreement between self-reported hearing status and objective hearing assessments of HI¹⁷. The assessment of hearing status includes the following questions: “Is your hearing excellent, very good, good, fair, or poor (with using hearing aid if they use)?”¹⁸. Hearing status was classified into three groups based on established diagnostic criteria‌^7,19: normal hearing (self-reported “excellent,” “very good,” or “good”), mild HI (self-reported “fair”), and severe HI (self-reported “poor”). All categories include individuals with or without hearing aids.

The presence of CMD was assessed through self-reported physician diagnosis of any of the following conditions: (1) diabetes or hyperglycemia; (2) stroke; (3) CVD (including coronary heart disease, angina, congestive heart failure, or other cardiac disorders)²⁰. CMM was defined as the concurrent presence of two or more CMD during follow-up. The diagnosis of CMM was established after the second CMD has been clinically confirmed²¹. The onset period of CMM is defined as the interval between the confirmation of CMM and the first interview.

Assessment of covariates

The covariates included in this study encompassed demographic characteristics (including age, sex, educational level, marital status), lifestyle factors (including smoking history, alcohol use history), body mass index (BMI), medical history (including hypertension, CKD, dyslipidemia), physical activity (PA), cardiometabolic baselines (including systolic blood pressure [SBP], diastolic blood pressure [DBP], fasting blood glucose [FBG], triglycerides [TG], total cholesterol [TC], low-density lipoprotein cholesterol [LDL-C] and high-density lipoprotein cholesterol [HDL-C]), and single pre-existing CMD (including diabetes, stroke and CVD). Educational level was reclassified from the original 12 categories into four levels: illiterate, primary school, junior high school, and high school or above. Marital status was categorized into married/cohabiting and other (including divorced, widowed, and unmarried). Smoking and alcohol use histories were recorded as binary variables (yes/no). BMI was calculated as weight (kg) divided by the square of height (m). Medical history of hypertension, CKD, and dyslipidemia were recorded as binary variables (yes/no). Hypertension was defined as self-reported physician diagnoses, or physical examination revealing DBP ≥ 90mmHg or SBP ≥ 140mmHg. Meeting any of these criteria is considered to have hypertension. CKD diagnosis was established through self-reported physician diagnoses, which adhered to professional guidelines²². Dyslipidemia was diagnosed by combining self-reported physician diagnoses with clinical thresholds: elevated TG (≥ 150 mg/dL), TC (≥ 200 mg/dL), or LDL-C (≥ 130 mg/dL), or reduced HDL-C (< 40 mg/dL in men/<50 mg/dL in women). PA is generally divided into three levels: vigorous activities, moderate activities, and mild activities. The level of PA is determined based on participants’ self-reported weekly physical activity: participants engaging in moderate to high intensity physical activity at least once per week are classified as vigorous; those engaging in low intensity physical activity at least once per week are classified as moderate; and those who reported no physical activity are classified as mild. The presence of CMD was assessed through self-reported physician diagnosis²⁰. All covariate data were obtained from standardized questionnaires and physical examination records to ensure data consistency and completeness.

Statistical analysis

Normally distributed continuous variables were expressed as mean±standard deviation (SD) and compared using the t-test. Non-normally distributed continuous variables were expressed as median and interquartile range and compared using the Mann-Whitney U test. Categorical variables were presented as frequencies and percentages and compared between groups using Fisher’s exact test or Pearson’s chi-square test, as appropriate.

The Kaplan-Meier survival curves were used to describe the cumulative incidence risk of CMM for different hearing status groups, and differences between groups were compared using the log-rank test. Cox proportional hazards regression model was used to calculate hazard ratio (HR) and 95% CI for the association between hearing status and CMM. Cox proportional hazards regression model was confirmed by the Schoenfeld residual test. The following three models were constructed: Model 1 was unadjusted; Model 2 was adjusted for age and sex; and Model 3 was further adjusted for smoking history, alcohol use history, educational level, marital status, BMI, hypertension, dyslipidemia, CKD, PA, SBP, DBP, FBG, TG, TC, LDL-C, HDL-C, diabetes, stroke and CVD based on Model 2.

To verify the robustness of the results, subgroup analyses were stratified by age (45–60 years/≥60 years) and sex (male/female) to explore the heterogeneity of the association between HI and CMM across different populations. The significance level (α) for subgroup tests was adjusted by using the Bonferroni correction method. Conduct sensitivity analyses (such as excluding early cases) to verify the robustness of the results. All tests were two-sided, and a P-value < 0.05 was considered statistically significant. Data analysis was performed using Zstats software (version 1.0; http://www.zstats.net) and R (version 4.3.3).

Results

Baseline characteristics of study participants

A total of 9,035 participants aged 45 years or older were included in this study, including 3,998 participants (44.25%) in the normal hearing group, 3,802 participants (42.08%) in the mild HI group, and 1,235 participants (13.70%) in the severe HI group. Statistically significant differences were observed in the three groups in terms of age, sex, BMI, alcohol use history, smoking history, marital status, educational level, SBP, and the prevalence of hypertension, stroke, and CKD (all P < 0.05).

Clinically meaningful differences were also evident across groups. The normal hearing group was younger (57.12 ± 8.42 years), had a lower prevalence of hypertension (35.64%) and CKD (4.65%), and a higher education level (12.91% with high school or above) at baseline. In contrast, the severe HI group was on average 5.26 years older (62.38 ± 9.37 years), had a 7.52% points higher prevalence of hypertension (43.16%), a 2.31% points higher prevalence of CVD, and a lower educational level (40.16% were illiterate). The baseline characteristics of study participants are depicted in Table 1.

Table 1.

Baseline characteristics of study participants.

Characteristics	Overall N = 9035	Normal hearing N = 3998	Mild HI N = 3802	Severe HI N = 1235	P-value
Age (years, mean ± SD)	58.57 ± 8.78	57.12 ± 8.42	58.86 ± 8.55	62.38 ± 9.37	< 0.001
Sex (N, %)					0.042
Male	4167 (46.12)	1903 (47.60)	1707 (44.90)	557 (45.10)
Female	4868 (53.88)	2095 (52.40)	2095 (55.10)	678 (54.90)
BMI (N, %)					< 0.001
<24 kg/m²	6245 (69.12)	2708 (67.73)	2624 (69.02)	913 (73.93)
≥ 24 kg/m²	2790 (30.88)	1290 (32.27)	1178 (30.98)	322 (26.07)
Alcohol use history (N, %)					0.002
No	6049 (66.95)	2610 (65.28)	2568 (67.54)	871 (70.53)
Yes	2986 (33.05)	1388 (34.72)	1234 (32.46)	364 (29.47)
Smoking history (N, %)					0.024
No	6253 (69.21)	2708 (67.73)	2670 (70.23)	875 (70.85)
Yes	2782 (30.79)	1290 (32.27)	1132 (29.77)	364 (29.47)
Marital status (N, %)					< 0.001
Other	947 (10.48)	385 (9.63)	380 (9.99)	182 (14.74)
Married/cohabiting	8088 (89.52)	3613 (90.37)	3422 (90.01)	1053 (85.26)
Educational level (N, %)					< 0.001
Illiterate	2535 (28.06)	1014 (25.36)	1025 (26.96)	496 (40.16)
Primary school	3743 (41.43)	1546 (38.67)	1681 (44.21)	516 (41.78)
Junior high school	1850 (20.48)	922 (23.06)	773 (20.33)	155 (12.55)
High school or above	907 (10.04)	516 (12.91)	323 (8.50)	68 (5.51)
CKD (N, %)					< 0.001
No	8496 (94.03)	3812 (95.35)	3554 (93.48)	1130 (91.50)
Yes	539 (5.97)	186 (4.65)	248 (6.52)	105 (8.50)
PA (N, %)					0.292
Vigorous activities	815 (9.02)	373 (9.33)	320 (8.42)	122 (9.88)
Moderate activities	2669 (29.54)	1150 (28.76)	1154 (30.35)	365 (29.55)
Mild activities	5551 (61.44)	2475 (61.91)	2328 (61.23)	748 (60.57)
Hypertension (N, %)					< 0.001
No	5601 (61.99)	2573 (64.36)	2326 (61.18)	702 (56.84)
Yes	3434 (38.01)	1425 (35.64)	1476 (38.82)	533 (43.16)
SBP (mmHg, mean ± SD)	129.56 ± 21.00	128.85 ± 20.53	129.65 ± 20.77	131.62 ± 22.99	< 0.001
DBP (mmHg, mean ± SD)	75.58 ± 11.99	75.81 ± 11.88	75.54 ± 12.05	74.94 ± 12.16	0.086
Dyslipidemia (N, %)					0.228
No	8299 (91.85)	3694 (92.40)	3479 (91.50)	1126 (91.17)
Yes	736 (8.15)	304 (7.60)	323 (8.50)	109 (8.83)
TG (mg/dL, mean ± SD)	132.17 ± 102.58	133.18 ± 99.80	131.73 ± 104.45	130.28 ± 105.66	0.645
TC (mg/dL, mean ± SD)	193.80 ± 38.68	194.00 ± 39.30	194.31 ± 38.44	191.57 ± 37.32	0.088
LDL-C (mg/dL, mean ± SD)	116.68 ± 34.97	116.64 ± 35.72	117.30 ± 34.57	114.94 ± 33.73	0.119
HDL-C (mg/dL, mean ± SD)	51.37 ± 15.13	51.00 ± 15.00	51.71 ± 15.32	51.50 ± 14.92	0.106
FBG (mg/dL, mean ± SD)	108.76 ± 33.65	108.52 ± 33.48	108.57 ± 32.70	110.16 ± 36.93	0.293
Diabetes (N, %)					0.296
No	8133 (90.02)	3585 (89.67)	3444 (90.58)	1104 (89.39)
Yes	902 (9.98)	413 (10.33)	358 (9.42)	131 (10.61)
Stroke (N, %)					0.013
No	8938 (98.93)	3966 (99.20)	3759 (98.87)	1213 (98.22)
Yes	97 (1.07)	32 (0.80)	43 (1.13)	22 (1.78)
CVD (N, %)					< 0.001
No	8245 (91.26)	3740 (93.55)	3439 (90.45)	1066 (86.32)
Yes	790 (8.74)	258 (6.45)	363 (9.55)	169 (13.68)

Association between HI and the risk of CMM

During the 7-year follow-up period, 382 participants developed CMM, with a cumulative incidence rate of 4.23%. Cox proportional hazards regression model revealed a significant positive association between the degree of HI and the risk of CMM, which persisted after accounting for potential confounders (Table 2). In the unadjusted model (Model 1), compared with the normal hearing group, both the mild HI group (HR = 1.33, 95% CI: 1.06–1.66) and the severe HI group (HR = 1.62, 95% CI: 1.22–2.17) exhibited a significantly elevated risk of CMM (all P < 0.05). After adjustment for age and sex (Model 2), the association for severe HI remained statistically significant (HR = 1.52, 95% CI: 1.13–2.03), whereas mild HI was associated with a marginally significant increased CMM risk (HR = 1.29, 95% CI: 1.03–1.61). In the fully adjusted model (Model 3), the independent association between HI and CMM risk remained significant for both severity strata. Specifically, compared with the normal hearing group, both the mild HI group (HR = 1.32, 95% CI: 1.06–1.68) and the severe HI group (HR = 1.54, 95% CI: 1.14–2.07) exhibited a significantly elevated risk of CMM (all P < 0.05).

Table 2.

Cox proportional hazards regression between HI and CMM.

Hearing status	Model 1 (Unadjusted)		Model 2 (Adjusted for age and sex)		Model 3 (Fully adjusted)
	HR (95% CI)	P	HR (95% CI)	P	HR (95% CI)	P
Normal hearing	1 (Reference)		1 (Reference)		1 (Reference)
Mild HI	1.33 (1.06–1.66)	0.014	1.29 (1.03–1.61)	0.026	1.32 (1.06–1.68)	0.016
Severe HI	1.62 (1.22–2.17)	0.001	1.52 (1.13–2.03)	0.005	1.54 (1.14–2.07)	0.005

Model 1: Unadjusted for covariates; Model 2: Adjusted for age and sex; Model 3: Adjusted for all the covariates.

The proportional hazards assumption of the Cox regression model was tested using Schoenfeld residual test (Table S1), which confirmed that the proportional hazards assumption was met (P > 0.05). Excluding study subjects diagnosed with CMM within 2 years prior to follow-up and refitting the Cox model yielded highly consistent HRs and 95% CIs for key covariates compared with the main analysis, verifying the stability of the results (Table S2). Notably, the minimal attenuation of risk estimates across sequential models (Model 1 to Model 3) underscores the robustness of the association between HI and CMM, which persists independently of the measured socioeconomic and health-related confounders included in our analysis.

The Kaplan-Meier survival curves (Fig. 2) intuitively illustrate the significant differences in the occurrence rate of CMM among groups with different hearing statuses, with the severe HI group exhibited the highest cumulative incidence rate (Log-rank P = 0.002). Notably, the curves overlap significantly during the first approximately 42 months of follow-up, and the divergence in cumulative incidence rates between groups becomes more apparent over time, with this divergence primarily manifesting after 42 months. This early “lag phase” suggests that the association between HI and elevated CMM risk may not be immediate, but rather may develop gradually with prolonged exposure, potentially implying a cumulative threshold effect of hearing impairment on the risk of CMM.

Fig. 2.

Kaplan-Meier curves for the risk of CMM in different hearing status.

Subgroup analysis and interaction tests

To explore the heterogeneity of the association between HI and CMM across different populations, subgroup analyses were stratified by age (45–60 years/≥60 years) and sex (male/female). Females had a higher risk of developing CMM under mild HI conditions (HR = 1.36, 95% CI: 1.02–1.8) compared to males, while males had a higher risk of developing CMM under severe HI conditions than females (HR = 1.55, 95% CI: 1.07–2.26). No significant interaction was observed between sex and the severity of HI (P for interaction = 0.816). Participants aged ≥ 60 years exhibited a significantly higher risk of both mild HI (HR = 1.7, 95% CI: 1.15–2.52) and severe HI (HR = 1.41, 95% CI: 1.0–1.99) compared with those aged 45–60 years. No significant interaction was observed between age and the severity of HI (P for interaction = 0.72). Generally, the risk of CMM tends to increase with advancing age and the severity of HI. (Table 3)

Table 3.

Subgroup analysis of CMM risk by hearing status and interaction tests.

Variable	Normal hearing HR (95% CI)	Mild HI HR (95% CI)	Severe HI HR (95% CI)	P for interaction
Sex				0.816
Male	1 (reference)	1.25 (0.87–1.8)	1.72 (1.09–2.27)
Female	1 (reference)	1.36 (1.02–1.8)	1.55 (1.07–2.26)
Age				0.72
45–60 years	1 (reference)	1.24 (0.92–1.67)	1.36 (0.85–2.17)
≥ 60 years	1 (reference)	1.41 (1.0–1.99)	1.7 (1.15–2.52)

Discussion

Using a nationally representative sample from the CHARLS database, this study included 9,035 participants to investigate the association between HI and the risk of CMM in middle-aged and older adults. The results demonstrated that HI was associated with an elevated risk of CMM, and this association was more pronounced in males with severe HI conditions and individuals aged 60 years or older. These findings provide new insights into the identification of risk factors for CMM and the early screening of high-risk populations.

Although many between-group differences in baseline characteristics were statistically significant, these should be interpreted cautiously given the large sample size, which may confer statistical significance even to small differences that are not necessarily clinically meaningful. For instance, the approximately 5-year age gap between participants with normal hearing and those with severe HI is clinically meaningful, as advanced age is a well-established risk factor for both HL and CMM^9,23. Similarly, the higher prevalence of hypertension and CVD in the severe HI group, albeit modest in absolute magnitude, is consistent with existing evidence linking hearing impairment to adverse cardiovascular risk profiles and supports the use of hearing status as a potential indicator of multisystem vulnerability in older adults^12,13.

This national cohort study first identified HI is associated with an increased risk of CMM. In the fully adjusted model, individuals with mild and severe HI showed a 32% and 54% elevated risk of CMM, respectively. This indicated that HI may be an early warning signal for the development of CMM. This finding aligns with previous studies suggesting an association between HL and CVD¹². A previous meta-analysis demonstrated that HL significantly increases the risk of cardiovascular mortality by 28% (HR = 1.28, 95% CI: 1.10–1.50)²⁴. A dose-response relationship was also identified, showing that each 30-dB increment in audiometric thresholds doubles the hazard for all-cause mortality (HR = 2.05, 95% CI: 1.45–2.90)²⁴. The results confirm the observed dose-response relationship between HL severity and risk elevation, where the hazard increases progressively with higher audiometric thresholds in a dose-dependent relationship. Kim et al. confirmed a significant positive correlation between the degree of HL and the risk of CVD in occupational noise-exposed populations²⁵. A Korean longitudinal study further substantiated this association within specific populations. For instance, among individuals with diabetes and comorbid hearing impairment, the risk of myocardial infarction rose by 11.7% and that of stroke by 13.4%²⁶. Previous studies have shown that the severity of hearing impairment in older adults worsens over time and is often irreversible without intervention²⁷. Existing hearing rehabilitation treatments (including medication, hearing aids, surgery and cochlear implants) may reduce the severity of hearing impairment²⁸.

Notably, the Kaplan-Meier curves revealed a “lag phase” in the association between HI and CMM, with curves overlapping significantly during the first approximately 42 months of follow-up, and divergence in cumulative incidence becoming apparent thereafter. This lag phase suggests a cumulative threshold effect. Pathophysiological mechanisms linking HI to CMM, such as chronic inflammation, vascular dysfunction, social isolation and low physical activity, may take time to accumulate and present as clinically overt cardiometabolic events. Our findings underscore the value of long-term follow-up when exploring associations between sensory impairment and chronic diseases, as short-term observation may underestimate the actual risk.

The association between HI and increased risk of CMM, its potential mechanisms may involve the synergistic effects of multiple pathophysiological pathways, including behavioral patterns, vascular pathology, inflammatory stress, and psychoneuroendocrine responses. Firstly, changes in behavioral patterns represent important driving factors. ‌Specifically, communication barriers caused by moderate to severe hearing loss can lead to ‌social isolation‌, which in turn may discourage participation in physical activities²⁹, and the decrease in physical activities can lead to the development of metabolic disorders, including obesity and insulin resistance^30,31. A systematic review demonstrated that both eccentric and traditional concentric resistance training significantly improve functional capacity in middle-aged and older adults³². Specifically, the evidence provided by the systematic review indicates that eccentric resistance training is highly effective for functional performance in older adults³². In addition, targeted walking protocols may offer practical benefits: intermittent walking training has been shown to improve thyroid function and cardiometabolic risk factors in postmenopausal women³³, a subgroup that our analysis identified as being at heightened CMM risk. These findings indicate that exercise interventions for individuals with HI might lower the risk of CMM. Secondly, CVD and HI are directly related through a common vascular pathological basis. The high metabolic characteristics of microcirculation in the inner ear make stria vascularis highly sensitive to ischemia and hypoxia³⁰. Vascular dysfunction represents a core pathological mechanism in CMM, which can simultaneously compromise cochlear perfusion and disrupt cardiovascular and cerebral blood supply. Furthermore, after adjusting for factors including smoking status and hypertension, the association remained significant, suggesting that pathological mechanisms independent of established vascular risk factors may be involved. Thirdly, chronic inflammation and oxidative stress play pivotal roles in this process. The elevation of inflammatory markers such as interleukin-6 and C-reactive protein is associated not only with HL, but also serves as a key trigger for the development of CMM^23,34,35. HL may trigger systemic inflammation through persistent sensory damage or by reducing anti-inflammatory factors due to social withdrawal. This establishes a vicious cycle of “inflammation-metabolic disorders-vascular injury”, which exacerbates hearing deterioration and increases cardiometabolic risk. Finally, psychological and neuroendocrine disorders exert a cumulative effect. Negative emotions such as depression and anxiety caused by HI, promote abnormal vasoconstriction and hemorheological alterations through neuroendocrine pathways involving imbalance of the 5-hydroxytryptamine and norepinephrine, ultimately inducing tissue ischemia and hypoxia and increasing the risk of cardiovascular events³⁶. The overlapping risk factors, including advanced age and unhealthy diet, further amplify these effects through multisystem interactions.

This study further explored the differences in the association between HI and CMM risk using stratified analyses by sex and age. Results from the sex subgroup analysis revealed the differences in the risk of CMM between males and females under different hearing statuses. Males had a higher risk of developing CMM under severe HI conditions than females (HR = 1.55, 95% CI: 1.07–2.26). This may be related to significant sex-based differences in the structure and function of the auditory system, neurotransmitter systems, and brain plasticity³⁷. These differences resulted in more pronounced negative impacts among male groups with more severe degrees of HI^38,39. Age-stratified analysis revealed a 70% increased risk of CMM (HR = 1.7, 95% CI:1.15–2.52) in individuals aged ≥ 60 years with severe HI compared to those with normal hearing, while no significant association was observed among participants aged 45–60 years. This result may be associated with the age-dependent increase in CMM risk. With advancing age, the decline in metabolic function, accumulation of chronic inflammation, and reduction in physiological reserves across multiple systems collectively contribute to an elevated risk of CMM development.

As the first longitudinal cohort study in China to examine the association between HI and the incidence of CMM, this research addresses an important gap in epidemiological evidence and offers a theoretical foundation for identifying CMM risk factors and screening high-risk populations. Furthermore, this study draws on a nationally representative sample from the CHARLS database, which includes middle-aged and older adults across diverse regions, urban and rural settings, and socioeconomic strata in China. The results demonstrate strong extrapolation and effectively circumvent the regional selection bias inherent in single-center studies. This study employed a multifactorial Cox proportional hazards regression model to adjust for potential confounders, including demographic characteristics (including age, sex, educational level, and marital status), lifestyle factors (including smoking history, alcohol use history, PA), BMI, medical history (including hypertension, dyslipidemia, and CKD), cardiometabolic baselines (including SBP, DBP, TG, TC, LDL-C, HDL-C, FBG), and single pre-existing CMD (including diabetes, stroke and CVD). Stratified analyses by sex and age were further conducted to assess the stability of the associations, which strengthens the internal validity of the findings.

However, this study has several limitations. The definition of hearing status in this study was based on self-reported questionnaires, which may introduce classification bias. Notably, self-reported data are prone to subjective perception and underreporting of mild HL, potentially diluting effect sizes and biasing estimates toward the null. Reliance on a single self-reported question to assess HI severity further limits the granularity of our findings. Notwithstanding, the CHARLS self-reported HI measure has been validated in Chinese older adult populations¹³, showing good agreement with objective audiometry and supporting its utility as a reliable proxy for population-level research¹⁷. This approach aligns with the practical constraints of large-scale surveys, where pure-tone audiometry is often infeasible due to cost and operational limitations, especially in rural settings. Additionally, the CHARLS dataset lacks data on specific HL etiologies (e.g., occupational noise exposure, ototoxic medication use), precluding exploration of etiological heterogeneity in the association between different types of HI and the risk of CMM. Future studies should integrate objective hearing tests (e.g., pure-tone audiometry) with self-reported measures to improve assessment precision and clarify these relationships. Although this study adjusted for multiple covariates, residual confounding from unmeasured factors (e.g. dietary patterns) may persist, potentially influencing the observed associations. Furthermore, the classification of PA was based on a simple self-reported frequency (“at least once per week”), which is relatively crude. This simplified classification may not fully capture the frequency, duration, or intensity of PA accurately and could introduce residual confounding in our estimates. Future intervention studies are needed to address these limitations and validate the findings.

In summary, this study provides epidemiological evidence linking HI to CMM. However, further ‌large-scale‌, multicenter prospective studies are still needed, combined with etiological classification and ‌exploration of molecular mechanisms in HI, to clarify the longitudinal association ‌between HI and CMM‌ and identify potential intervention targets. Furthermore, to overcome the limitations of self-reporting in the current study, future research designs should integrate advanced technologies. For example, wearable sensors can objectively monitor physical activity and biomechanics⁴⁰, with artificial intelligence-driven analytics to better phenotype multimorbidity clusters in patients with sensory impairments^40,41. This approach would address self-reporting limitations and provide deeper insights into the characteristics and progression of CMM⁴².

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

CMM

Cardiometabolic multimorbidity

CMD

Cardiometabolic disease

CVD

Cardiovascular disease

HI

Hearing impairment

HI

Hearing loss

OR

Odds Ratio

CI

Confidence interval

CHARLS

China Health and Retirement Longitudinal Study

BMI

Body mass index

CKD

Chronic kidney disease

PA

Physical Activity

SBP

Systolic blood pressure

DBP

Diastolic blood pressure

FBG

Fasting blood glucose

TG

Triglycerides

TC

Total cholesterol

LDL-C

Low-density lipoprotein cholesterol

HDL-C

High-density lipoprotein cholesterol

SD

Standard deviation

HR

Hazard ratio

RR

Relative risk

Author contributions

X-D Z collected and analyzed the data. Q-C Z drafted the manuscript. B D, Z-J P, and Y-Z S assessed the methodological quality and data integrity of the included studies. Y-P T provided critical feedback and revisions throughout the manuscript’s development. All authors reviewed and approved the final manuscript.

Funding

No funding was received for this study.

Data availability

The data for this study originated from a publicly available database.

Declarations

Competing interests

The authors declare no competing interests.