Combined effects of xylene and noise exposure at realistic low levels on hearing impairment: a cross-sectional study in China Authors: Zhihao Shi, Genyuan Li, Zhehua Zhou Published: 2026 ABSTRACT ObjectivesTo investigate the combined effect of low-level xylene and noise exposure on hearing impairment in Chinese manufacturing workers and to provide evidence for occupational hearing protection strategies in China.MethodsA cross-sectional survey was conducted among 899 workers (295 co-exposed, 604 noise-only) in Jiaxing City, China from 2023 to 2025. Workplace noise and xylene concentrations were monitored, and pure-tone audiometry was performed with age-gender correction. Binary logistic regression and stratified regression analysis were used to evaluate risk factors for high-frequency hearing loss (HFHL).ResultsThe HFHL prevalence was 5.76% in the combined exposure group and 2.48% in the noise-only group (P = 0.01). Logistic regression model shows that the combined exposure group had a 5.49-fold higher HFHL risk (P < 0.01). Hypertension, longer exposure duration, higher noise intensity and higher xylene concentration were independent risk factors for HFHL. Stratified analyses confirmed a higher HFHL risk in the combined exposure group across subgroups of age < 40/ ≥ 40 years, male workers, and exposure duration 0–5/6–10 years.ConclusionsCombined xylene and noise exposure exerts a significant synergistic ototoxic effect even when both exposures are within Chinese national occupational exposure limits. Introduction Hearing loss is a major global public health concern, identified by the World Health Organization (WHO) as one of the six leading causes of the global disease burden, with an estimated 2.45 billion people affected worldwide by 2050 (GBD 2019 Hearing Loss Collaborators, 2021). Occupational noise is a predominant cause of acquired hearing loss: 38% of U.S. manufacturing workers are exposed to hazardous noise, and 12% of all manufacturing workers have hearing difficulty (Masterson et al., 2025). Occupational noise-induced deafness ranks as the second most common occupational disease in China (Statistical Bulletin on the Development of Health Care in China in, 2024). A systematic review by Zhou et al. revealed that the prevalence of occupational noise-induced hearing loss among Chinese workers exceeds 20% (Zhou et al., 2020). Beyond noise, organic solvents have been proven to be potential causes of hearing loss. As early as 1986, Bergström and Nyström discovered the effect of organic solvents in causing or promoting hearing loss. A 20-year longitudinal study conducted by them found a higher prevalence of hearing loss in chemical industry workers (23%)—exposed to both noise and organic solvents—than in non-chemical industry workers (5%–8%), despite lower noise levels in the chemical sector (Bergström and Nyström, 1986). In recent decades, emerging research has revealed that numerous organic solvents possess intrinsic ototoxicity (Rybak, 1992; Dillard et al., 2021; Hormozi et al., 2017; Loukzadeh et al., 2014), which could be considered as ototoxicants. Early animal studies and human observational studies have demonstrated that these solvents can damage cochlear hair cells, stria vascularis, and auditory pathways independently of noise (Campo et al., 2001; Campo et al., 1999; Bale et al., 2005; Lataye et al., 2007; Campo and Maguin, 2007). In industrial settings, simultaneous or sequential exposure to noise and organic solvent is common. According to the most recent data, it is estimated that approximately 2.6 million U.S. manufacturing workers are exposed to both hazardous noise and ototoxic organic solvents each year (Masterson et al., 2025). However, findings regarding the combined effects of noise and solvents remain inconsistent. Some studies confirm a synergistic hearing loss risk from combined noise and organic solvent exposure (Morata et al., 1993; Staudt et al., 2019; Rosati and Jamesdaniel, 2020; Unlu et al., 2014; Kim et al., 2005; Chang et al., 2006; Chang et al., 2003; Rabinowitz et al., 2008)., others report no significant difference in noise-induced permanent threshold shifts (NIPTS) compared with noise-only exposure (De Barba et al., 2005). This inconsistency may stem from differences in study populations, exposure assessment methods, solvent types, and definitions of hearing loss. Xylene, a high-priority ototoxicant (Morata and Little, 2002), is a prevalent component in numerous industrial processes, including automobile manufacturing, furniture production and printing. A significant number of extant studies are of an animal-based nature, exploring the mechanism of its ototoxicity (Campo et al., 2001; Campo et al., 1999; Bale et al., 2005; Lataye et al., 2007; Campo and Maguin, 2007). It is imperative to note that the concentration of xylene utilized in these animal studies does not correspond to the actual concentration to which workers are exposed. The enterprises' emphasis on toxic substance protection and the improvement of protective facility design has led to the observation that the actual concentration of xylene to which workers are exposed is often at a low level. The epidemiological studies have focused predominantly on the single exposure of xylene, with a paucity of data on combined exposure to xylene and noise. Furthermore, it is imperative that further studies be conducted among Chinese manufacturing workers, who represent a sizable and highly exposed population. Therefore, this study was designed to investigate the prevalence of hearing loss in workers co-exposed to xylene and noise compared with workers exposed to noise alone in Jiaxing, China. By elucidating the synergistic ototoxic effect under realistic low-dose exposure conditions, this study will provide evidence to support evidence-based occupational health prevention and hearing protection strategies for co-exposed workers. Materials and methods Study design and population A cross-sectional survey was conducted in Jiaxing City, a key industrial hub within China's Yangtze River Delta. The data were collected from 2023 to 2025 through the databases of Jiaxing City, including the Key Occupational Diseases Active Surveillance System and Workplace Occupational Hazard Monitoring System. Job position matching was performed for the same enterprise, and subjects with mismatched posts or missing data were excluded. The inclusion criteria for study subjects were as follows: (1) Exposure limited to noise alone, or noise combined with xylene-dominated organic solvents; (2) ≥ 6 months of continuous employment in the current post. The exclusion criteria were: History of shooting or military service, blast exposure, ear disease, head trauma, hereditary deafness, ototoxic drug use, and other confounding factors associated with hearing loss. A total of 899 workers with complete personal, occupational, and audiometric data were enrolled, including 295 in the combined xylene-noise exposure group, and 604 in the noise-only exposure group. Each participant was informed of the purpose and design of this study and was asked to sign an informed consent form. The study protocol was approved by the ethics committee of the Jiaxing Center for Disease Control and Prevention, China (approval reference number: JXCDC-2026-017-001). Measures Workplace occupational Hazard monitoring Prior to the occupational hazard monitoring, a field investigation was conducted by experienced occupational hygienists to identify workplace hazards, exposed posts, exposure duration and hearing protection device (HPD) use. Occupational noise exposure was defined as a noise intensity ≥75 dBA. Occupational xylene exposure was defined as scenarios where xylene was confirmed as the predominant organic solvent via MSDS reviewing (MSDS, material safety data sheets) and air sampling. Field sampling and monitoring were performed during normal production. In accordance with the National Occupational Health Standards of China: Measurement of Noise in the Workplace (GBZ/T 189.8–2007), noise intensity was measured using a sound level meter (AWA6228, Hangzhou Aihua Instruments Co., Ltd., China). The measurement of Gaussian noise was repeated thrice, and the mean value was adopted as the representative measurement. In contrast, the measurement of non-Gaussian noise involved conducting individual long-term measurements, with the equivalent sound level calculated. The results were all converted to an A-weighted sound pressure level normalized to a nominal 8-h working day (LAeq,8h). In accordance with GBZ 159–2004 Specifications of Air Sampling for Hazardous Substances Monitoring in the Workplace, xylene in the workplace air was collected using an air sampler (EM-500, Shenzhen Guoji Instrument Co., Ltd., China). In accordance with GBZ/T 300.66–2017 Determination of Xylene in Workplace Air, the concentration of xylene was determined using a gas chromatograph (8890, Agilent Technologies, Inc.). Xylene sampling was performed at three distinct time points at the job site within a single working day, and time-weighted average concentrations were computed. All instruments were calibrated by authorized metrological institutions. All the measured results were uploaded to the Workplace Occupational Hazard Monitoring System. Pure-tone audiometry Pure tone audiometry was a component of occupational health checkups for workers. In the course of these medical examinations, information was collected and blood pressure measurements were taken. Measured blood pressure ≥ 140/90 mmHg were classified as hypertensive. The results of these tests were uploaded to the Key Occupational Diseases Active Surveillance System. Pure-tone audiometry was performed in a soundproof booth by otolaryngologists in accordance with GBZ 49–2014 Diagnosis of Occupational Noise-Induced Deafness. Pure-tone air conduction threshold testing was conducted at frequencies of 0.5, 1, 2, 3, 4, 6, and 8 kHz for both the left and right ears of all study subjects, and otological examinations were performed by the physicians to rule out conductive hearing impairment of the ears. All subjects were required to be removed from noise exposure for 14–16 h prior to audiometric testing. The hearing threshold of an individual is primarily determined by age and noise exposure. Subsequent to accounting for the elevation in hearing threshold attributable to normal aging, the residual effect of noise on hearing can be estimated, which is referred to as NIPTS. According to ISO 7029:2000, the Chinese standard GBZ 49–2014 specifies the statistical distribution of hearing thresholds relative to age among the normal hearing Chinese population. The NIPTS was calculated by subtracting the corresponding age- and sex-specific median hearing threshold shift from the measured pure-tone hearing threshold of each subject. High-frequency hearing loss (HFHL) was defined as a mean NIPTS of ≥40 dB HL at 3, 4, and 6 kHz in both ears. Speech-frequency hearing loss (SFHL) was defined as a mean NIPTS of >25 dB HL at 0.5, 1, and 2 kHz in the better ear. Quality control measures Prior to the initiation of the project, investigators received standardized training on noise and xylene measurement and pure-tone audiometry, and a preliminary survey was conducted to identify problems,which would be promptly reported and rectified. Supplementary training was provided as necessary. The collected data were promptly entered into systems, which were equipped with logical verification and integrity checks to flag incomplete or erroneous entries. Two system administrators were assigned to review every record in the systems monthly. In the event that any errors were identified during these reviews, prompt communication and correction were mandated. Furthermore, a 10% sample of the enrolled enterprises was selected on a yearly basis for the purpose of re-measuring noise and xylene levels. Statistical analysis Normally distributed continuous variables were presented as mean ± standard deviation (SD) and compared between groups using the independent-samples t-test. Non-normally distributed continuous variables were expressed as median (P25, P75) and between-group comparisons were conducted via the Mann-Whitney U test. Categorical variables were described as sample size and percentage (n, %); intergroup differences were examined using the chi-square (χ2) test or Fisher's exact test, as appropriate. A subgroup analysis comparing hearing loss prevalence between males and females was conducted. Binary logistic regression analysis was performed to evaluate HFHL risk (dependent variable: HFHL), with the following covariates: noise intensity, xylene concentration, exposure group (combined exposure and noise-only, with the latter serving as the reference category), hypertension status (hypertension and non-hypertension, with the latter serving as the reference category), exposure duration (0–5 years,6–10 years, and > 10 years, with 0–5 years as the reference), smoking years (no smoke, 1–10 years, 11–20 years, and > 20 years, with no smoke designated as the reference category), and HPD use (nonuse, infrequent use, and frequent use, with nonuse serving as the reference category). Stratified analysis divided the study population into subgroups by age (<40 / ≥40 years), gender (male / female), and exposure duration (0–5 / 6–10 / >10 years). Subsequently, a series of logistic regression analyses were conducted within each subgroup. Stratified logistic regression analyses were aimed to evaluate whether the combined exposure effect is consistent across different subgroups and to control for potential effect modification by these variables. P < 0.05 was considered significant. All analyses were conducted using IBM SPSS Statistics, version 25 (IBM Corp., Armonk, NY). Results General characteristics The combined exposure group had a significantly lower mean age (35.23 ± 8.69 vs. 39.02 ± 10.18 years, P < 0.01), higher proportion of workers <40 years old (73.22% vs. 54.30%, P < 0.01) and males (90.51% vs. 83.61%, P = 0.01) than the noise-only group. Neither noise intensity nor xylene concentration in the combined exposure group exceeded the national OELs (OEL, occupational exposure limit). The median xylene concentration in the combined exposure group was 0.35 mg/m3, 0.7% of its OEL in China (50 mg/m3). A total of 32.1% of the noise-only exposure group demonstrated excessive noise levels, which exceeded the national OEL (85 dBA). The combined exposure group had lower mean noise intensity (76.95 ± 1.51 vs. 81.71 ± 4.88 dBA, P < 0.01) but longer median exposure duration [5(2,9) vs. 2(1,7) years, P < 0.01] (Table 1).Table 1Demographic and occupational characteristics of manufacturing workers exposed to noise and/or xylene in Jiaxing City, China, 2023–2025.CharacteristicsCombined exposure group(n = 295)Noise-only exposure group(n = 604)PAge (years), mean ± SD35.23 ± 8.6939.02 ± 10.18t = 5.79, P < 0.01Age (years), n(%)χ2 = 29.68, P < 0.01 <40216(73.22%)328(54.30%) ≥4079(26.78%)276(45.70%)Gender, n(%)χ2 = 7.78, P = 0.01 Female28(9.49%)99(16.39%) Male267(90.51%)505(83.61%)Exposure duration (years), median(P25, P75)5(2,9)2(1,7)P < 0.01+Exposure duration (years), n(%)χ2 = 38.20, P < 0.01 0–5152(51.53%)435(72.02%) 6–1082(27.80%)86(14.24%) >1061(20.68%)83(13.74%)Noise intensity (dBA), mean ± SD76.95 ± 1.5181.71 ± 4.88t = 21.92, P < 0.01Noise intensity (dBA), n(%)P < 0.01a 75-275(93.22%)296(49.01%) 80-20(6.78%)114(18.87%) 85-0(0)194(32.12%)Xylene concentration, median(P25, P75)0.35(0.35, 1.00)0Z = -20.28, P < 0.01Hypertension, n(%)χ2 = 9.59, P < 0.01 No256(86.78%)472(78.15%) Yes39(13.22%)132(21.85%)HPD, n(%)χ2 = 314.00, P < 0.01 Nonuse4(1.36%)45(7.45%) Infrequent use273(92.54%)179(29.64%) Frequent use18(6.10%)380(62.91%)aFisher's exact test; +Mann-Whitney U test; HPD, hearing protection device; SD, standard deviation. Table 1Demographic and occupational characteristics of manufacturing workers exposed to noise and/or xylene in Jiaxing City, China, 2023–2025.CharacteristicsCombined exposure group(n = 295)Noise-only exposure group(n = 604)PAge (years), mean ± SD35.23 ± 8.6939.02 ± 10.18t = 5.79, P < 0.01Age (years), n(%)χ2 = 29.68, P < 0.01 <40216(73.22%)328(54.30%) ≥4079(26.78%)276(45.70%)Gender, n(%)χ2 = 7.78, P = 0.01 Female28(9.49%)99(16.39%) Male267(90.51%)505(83.61%)Exposure duration (years), median(P25, P75)5(2,9)2(1,7)P < 0.01+Exposure duration (years), n(%)χ2 = 38.20, P < 0.01 0–5152(51.53%)435(72.02%) 6–1082(27.80%)86(14.24%) >1061(20.68%)83(13.74%)Noise intensity (dBA), mean ± SD76.95 ± 1.5181.71 ± 4.88t = 21.92, P < 0.01Noise intensity (dBA), n(%)P < 0.01a 75-275(93.22%)296(49.01%) 80-20(6.78%)114(18.87%) 85-0(0)194(32.12%)Xylene concentration, median(P25, P75)0.35(0.35, 1.00)0Z = -20.28, P < 0.01Hypertension, n(%)χ2 = 9.59, P < 0.01 No256(86.78%)472(78.15%) Yes39(13.22%)132(21.85%)HPD, n(%)χ2 = 314.00, P < 0.01 Nonuse4(1.36%)45(7.45%) Infrequent use273(92.54%)179(29.64%) Frequent use18(6.10%)380(62.91%)aFisher's exact test; +Mann-Whitney U test; HPD, hearing protection device; SD, standard deviation. Demographic and occupational characteristics of manufacturing workers exposed to noise and/or xylene in Jiaxing City, China, 2023–2025. Fisher's exact test; +Mann-Whitney U test; HPD, hearing protection device; SD, standard deviation. Hearing loss prevalence As is shown in Table 2, HFHL prevalence in the combined exposure group (5.76%) was significantly higher than the noise-only group (2.48%, χ2 = 6.21, P = 0.01), with no significant SFHL difference between groups (2.71% vs. 2.48%, P = 0.84). Subjects aged ≥40 years had significantly higher prevalence of both HFHL and SFHL compared with those aged <40 years, with statistically significant differences (χ2 = 9.49, P < 0.01; χ2 = 6.54, P = 0.01). Although male subjects had higher prevalence of hearing loss than female subjects, the differences were not statistically significant (χ2 = 0.28, P = 0.60; χ2 = 0.21, P = 0.65). A subgroup analysis comparing hearing loss prevalence between males and females was conducted (Fig. 1). For workers aged ≥40 years, there was also no significant difference of both HFHL and SFHL between males and females (P > 0.05).Table 2Prevalence of high-frequency and speech-frequency hearing loss among manufacturing workers exposed to noise and/or xylene in Jiaxing City, China, 2023–2025.NHigh-frequency hearing lossSpeech-frequency hearing lossnPrevalence (%)χ2PnPrevalence (%)χ2PTotal899323.56//232.56//Exposure groups6.210.010.040.84 Noise-only exposure group604152.48152.48 Combined exposure group295175.7682.71Age (years)9.49<0.016.540.01 <40544112.0281.47 ≥40355215.92154.23Gender0.280.600.210.65 Female12732.3621.57 Male772293.76212.72Exposure duration (years)11.53<0.01/0.31a 0-5587122.04142.39 6–10168105.9574.17 >10144106.9421.39Noise intensity (dBA)2.010.365/0.01a 75-571172.9871.23 80-13453.7396.72 85-194105.1573.61Hypertension5.080.021.020.31 No728212.88212.88 Yes171116.4321.17HPD7.620.02/0.01a Nonuse4948.1612.04 Infrequent use45291.9951.11 Frequent use398194.77174.27aFisher's exact test; HPD, hearing protection device.Fig. 1Prevalence of hearing loss stratified by gender and age among manufacturing workers exposed to noise and/or xylene in Jiaxing City, China, 2023–2025. (a) Prevalence of high-frequency hearing loss; (b) Prevalence of speech-frequency hearing loss. Blue bars represent male participants, and orange bars represent female participants. Data are presented as prevalence (%). Groups were stratified by age: <40 years and ≥ 40 years. Numerical values are labeled above each bar. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)Fig. 1 Table 2Prevalence of high-frequency and speech-frequency hearing loss among manufacturing workers exposed to noise and/or xylene in Jiaxing City, China, 2023–2025.NHigh-frequency hearing lossSpeech-frequency hearing lossnPrevalence (%)χ2PnPrevalence (%)χ2PTotal899323.56//232.56//Exposure groups6.210.010.040.84 Noise-only exposure group604152.48152.48 Combined exposure group295175.7682.71Age (years)9.49<0.016.540.01 <40544112.0281.47 ≥40355215.92154.23Gender0.280.600.210.65 Female12732.3621.57 Male772293.76212.72Exposure duration (years)11.53<0.01/0.31a 0-5587122.04142.39 6–10168105.9574.17 >10144106.9421.39Noise intensity (dBA)2.010.365/0.01a 75-571172.9871.23 80-13453.7396.72 85-194105.1573.61Hypertension5.080.021.020.31 No728212.88212.88 Yes171116.4321.17HPD7.620.02/0.01a Nonuse4948.1612.04 Infrequent use45291.9951.11 Frequent use398194.77174.27aFisher's exact test; HPD, hearing protection device. Prevalence of high-frequency and speech-frequency hearing loss among manufacturing workers exposed to noise and/or xylene in Jiaxing City, China, 2023–2025. Fisher's exact test; HPD, hearing protection device. Prevalence of hearing loss stratified by gender and age among manufacturing workers exposed to noise and/or xylene in Jiaxing City, China, 2023–2025. (a) Prevalence of high-frequency hearing loss; (b) Prevalence of speech-frequency hearing loss. Blue bars represent male participants, and orange bars represent female participants. Data are presented as prevalence (%). Groups were stratified by age: <40 years and ≥ 40 years. Numerical values are labeled above each bar. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) HFHL prevalence differed significantly across exposure duration subgroups (χ2 = 11.53, P < 0.01), with a significant dose-response trend (χ2_(trend) = 10.76, P < 0.01). Hypertensive workers had higher HFHL prevalence (6.43% vs. 2.88%, P = 0.02), and HPD use status was associated with HFHL (P = 0.02). Binary logistic regression analysis After adjusting for hypertension, smoking years, HPD use, exposure duration, noise intensity (LAeq,8h), and xylene concentration, the combined exposure group had a 5.485-fold increased HFHL risk (P < 0.01). Hypertension (OR = 2.44, P = 0.03), exposure duration (OR = 1.70, P = 0.02), noise intensity (OR = 1.23, P < 0.01), and xylene concentration (OR = 1.60, P < 0.01) were independent HFHL risk factors (Table 3).Table 3Binary logistic regression analysis on risk factors for high-frequency hearing loss among manufacturing workers exposed to noise and/or xylene in Jiaxing City, China, 2023–2025.VariablesbSbPORHypertension0.890.420.032.44Smoking years0.400.220.081.49HPD−0.140.470.770.87Exposure duration0.530.230.021.70Noise intensitya0.210.06<0.011.23Xylene concentrationa0.470.11<0.011.60Exposure groups1.700.51<0.015.49Constant−22.684.82<0.01/acontinuous variables; Hypertension: non-hypertension = 0, hypertension = 1; Smoking years: no smoke = 1, 1–10 years = 2, 11–20 years = 3, >20 years = 4; HPD: nonuse = 0, infrequent use = 1, frequent use = 2; Exposure duration: 0–5 years = 1, 6–10 years = 2, >10 years = 3; Exposure groups: the noise-only group = 0, the combined exposure group = 1; Dependent viable: high-frequency hearing loss = 1, no high-frequency hearing loss = 0; HPD, hearing protection device; b, regression coefficient; Sb, standard error for regression coefficient; OR, odds ration. Table 3Binary logistic regression analysis on risk factors for high-frequency hearing loss among manufacturing workers exposed to noise and/or xylene in Jiaxing City, China, 2023–2025.VariablesbSbPORHypertension0.890.420.032.44Smoking years0.400.220.081.49HPD−0.140.470.770.87Exposure duration0.530.230.021.70Noise intensitya0.210.06<0.011.23Xylene concentrationa0.470.11<0.011.60Exposure groups1.700.51<0.015.49Constant−22.684.82<0.01/acontinuous variables; Hypertension: non-hypertension = 0, hypertension = 1; Smoking years: no smoke = 1, 1–10 years = 2, 11–20 years = 3, >20 years = 4; HPD: nonuse = 0, infrequent use = 1, frequent use = 2; Exposure duration: 0–5 years = 1, 6–10 years = 2, >10 years = 3; Exposure groups: the noise-only group = 0, the combined exposure group = 1; Dependent viable: high-frequency hearing loss = 1, no high-frequency hearing loss = 0; HPD, hearing protection device; b, regression coefficient; Sb, standard error for regression coefficient; OR, odds ration. Binary logistic regression analysis on risk factors for high-frequency hearing loss among manufacturing workers exposed to noise and/or xylene in Jiaxing City, China, 2023–2025. continuous variables; Hypertension: non-hypertension = 0, hypertension = 1; Smoking years: no smoke = 1, 1–10 years = 2, 11–20 years = 3, >20 years = 4; HPD: nonuse = 0, infrequent use = 1, frequent use = 2; Exposure duration: 0–5 years = 1, 6–10 years = 2, >10 years = 3; Exposure groups: the noise-only group = 0, the combined exposure group = 1; Dependent viable: high-frequency hearing loss = 1, no high-frequency hearing loss = 0; HPD, hearing protection device; b, regression coefficient; Sb, standard error for regression coefficient; OR, odds ration. The prevalence in the combined exposure group was consistently higher than that in the noise-only exposure group, with the exception of the subgroup with exposure duration >10 years (Fig. 2). Stratified regression confirmed the combined exposure group had an about 7-fold higher HFHL risk in both age < 40 years (OR = 7.00, P = 0.03) and ≥ 40 years (OR = 7.04, P = 0.01) subgroups. Male workers in the combined exposure group had a 4.332-fold higher HFHL risk (P = 0.01), with no valid model for females due to small sample size. For exposure duration, the combined exposure group had significantly higher HFHL risk for 0–5 years (OR = 5.42, P = 0.02) and 6–10 years (OR = 8.29, P = 0.04); no valid model was established for >10 years (Table 4).Fig. 2Prevalence of high-frequency hearing loss (%) stratified by age, gender, and exposure duration among manufacturing workers exposed to noise and/or xylene in Jiaxing City, China, 2023–2025. (a) Stratified by age: <40 years and ≥ 40 years; (b) Stratified by gender: Male and Female; (c) Stratified by exposure duration: 0–5 years, 6–10 years, and > 10 years. Workers were classified into the noise-only exposure group and the combined xylene-noise exposure group. All data are presented as high-frequency hearing loss prevalence (%).Fig. 2Table 4Stratified logistic regression analysis for the association between combined xylene-noise exposure and high-frequency hearing loss risk among manufacturing workers in Jiaxing City, China, 2023–2025.FactorsSubgroupsORPAge<40 years7.000.03≥40 years7.040.01GenderMale4.330.01Femalea/1.00Exposure duration0–5 years5.420.026–10 years8.290.04>10 yearsa/1.00aNo valid logistic regression model was established for females and duration >10 years; noise intensity and xylene concentration are continuous variables; Hypertension: non-hypertension = 0, hypertension = 1; Smoking years: no smoke = 1, 1–10 years = 2, 11–20 years = 3, >20 years = 4; HPD: nonuse = 0, infrequent use = 1, frequent use = 2; Exposure duration: 0–5 years = 1, 6–10 years = 2, >10 years = 3; Exposure groups: the noise-only group = 0, the combined exposure group = 1; Dependent viable: high-frequency hearing loss = 1, no high-frequency hearing loss = 0; OR, odds ration. Prevalence of high-frequency hearing loss (%) stratified by age, gender, and exposure duration among manufacturing workers exposed to noise and/or xylene in Jiaxing City, China, 2023–2025. (a) Stratified by age: <40 years and ≥ 40 years; (b) Stratified by gender: Male and Female; (c) Stratified by exposure duration: 0–5 years, 6–10 years, and > 10 years. Workers were classified into the noise-only exposure group and the combined xylene-noise exposure group. All data are presented as high-frequency hearing loss prevalence (%). Table 4Stratified logistic regression analysis for the association between combined xylene-noise exposure and high-frequency hearing loss risk among manufacturing workers in Jiaxing City, China, 2023–2025.FactorsSubgroupsORPAge<40 years7.000.03≥40 years7.040.01GenderMale4.330.01Femalea/1.00Exposure duration0–5 years5.420.026–10 years8.290.04>10 yearsa/1.00aNo valid logistic regression model was established for females and duration >10 years; noise intensity and xylene concentration are continuous variables; Hypertension: non-hypertension = 0, hypertension = 1; Smoking years: no smoke = 1, 1–10 years = 2, 11–20 years = 3, >20 years = 4; HPD: nonuse = 0, infrequent use = 1, frequent use = 2; Exposure duration: 0–5 years = 1, 6–10 years = 2, >10 years = 3; Exposure groups: the noise-only group = 0, the combined exposure group = 1; Dependent viable: high-frequency hearing loss = 1, no high-frequency hearing loss = 0; OR, odds ration. Stratified logistic regression analysis for the association between combined xylene-noise exposure and high-frequency hearing loss risk among manufacturing workers in Jiaxing City, China, 2023–2025. No valid logistic regression model was established for females and duration >10 years; noise intensity and xylene concentration are continuous variables; Hypertension: non-hypertension = 0, hypertension = 1; Smoking years: no smoke = 1, 1–10 years = 2, 11–20 years = 3, >20 years = 4; HPD: nonuse = 0, infrequent use = 1, frequent use = 2; Exposure duration: 0–5 years = 1, 6–10 years = 2, >10 years = 3; Exposure groups: the noise-only group = 0, the combined exposure group = 1; Dependent viable: high-frequency hearing loss = 1, no high-frequency hearing loss = 0; OR, odds ration. Discussion This cross-sectional study in Jiaxing City, which is located in the manufacturing agglomeration area in the Yangtze River Delta of China, confirmed that even when xylene concentrations were far below the national OELs and noise intensities were generally within the regulatory limits, their combined exposure still significantly increased HFHL risk in workers. The combined exposure group had over twice the HFHL prevalence of the noise-only group, with a 5.49-fold higher risk after confounder adjustment—consistent with international studies on organic solvent-noise synergism (Fuente et al., 2013; Sliwinska-Kowalska et al., 2004; Liu et al., 2024) and providing key epidemiological evidence for low-dose combined occupational hazard risks. The synergistic effect stems from shared damage targets and overlapping pathophysiological processes in the cochlea: cochlear hair cells, the core structure for auditory signal transduction, are vulnerable to both noise and xylene. Noise induces hair cell damage via mechanical vibration and oxidative stress (Natarajan et al., 2023), while xylene impairs the cochlear stria vascularis (disrupting ionic/nutrient balance) and exacerbates oxidative stress (reactive oxygen species accumulation), superimposing on noise-induced damage to accelerate hair cell apoptosis (Campo et al., 2001; Campo et al., 1999). Xylene also alters the membranous structures of outer hair cells (OHCs) (Campo et al., 2001) and impairs signal transmission in the auditory pathway through its toxic effects on the central nervous system (Bale et al., 2005; Lataye et al., 2007; Campo and Maguin, 2007), reducing the tolerance threshold of the auditory system to noise-induced damage—explaining the higher HFHL risk in the combined exposure group despite lower noise intensity. HFHL was the only significant hearing loss endpoint, as the high-frequency region of the cochlear basilar membrane is more sensitive to physical/chemical damage than mid/low-frequency regions (Morata and Little, 2002; Ren et al., 2023; Morata et al., 2002), making HFHL an early screening indicator for this exposed population. The results of stratified analysis in this study further refined the risk characteristics of HFHL among different populations. Previous occupational hearing loss studies have consistently shown that hearing function begins to decline significantly after 40 years of age, and the susceptibility to noise and chemical ototoxicity increases markedly in workers aged ≥40 years (Fan et al., 2024). The subjects were divided into two subgroups based on age: <40 years and ≥ 40 years of age. In the age-stratified analysis, the combined exposure group had an approximately 7-fold higher HFHL risk in both the <40 years and ≥ 40 years subgroups. These results demonstrate the combined effect on hearing loss in the two age subgroups. This may be associated with young workers' low HPD compliance (92.54% infrequent use) observed in this study. In the gender-stratified analysis, male wokers had a 4.33-fold increased HFHL risk in the combined exposure group; no female model was possible due to the male-dominated manufacturing workforce. It's reported that noise-induced hearing loss develops more rapidly during the initial 10 years of exposure, while, particularly in the case of non-Gaussian noise, it develops expeditiously within the first 5 years (Zhang et al., 2021). The participants were divided into three subgroups based on their exposure duration (0–5 years,6–10 years, and > 10 years). Stratified analysis by exposure duration revealed a significantly increased risk of hearing damage in the combined exposure group with 0–5 years and 6–10 years, whereas no statistically significant HFHL difference was observed in the >10 years subgroup. This phenomenon may be associated with healthy worker selection bias (severely affected workers left their posts) or improved HPD use in long-term exposed workers, alongside small sample size limitations. Hypertension was an independent HFHL risk factor (OR = 2.44), which is consistent with the conclusions of previous studies (FYOMM et al., 2022). Hypertension can impair cochlear microcirculation and exacerbate ischemic-hypoxic hair cell damage (Sliwinska-Kowalska et al., 2007), with combined xylene-noise exposure amplifying this effect. Some studies have reported the damage to inner ear microcirculation caused by nicotine, tar and other substances in tobacco (Lustig, 2006). Although smoking years did not reach statistical significance in this study (P = 0.08), further exploration into the relationship between smoking and hearing loss is warranted. Notably, HPD use showed no significant protective effect, due to extremely low compliance (92.54% infrequent use and 1.36% nonuse) in the combined exposure group—reflecting a common issue of underutilized protective equipment in manufacturing. This study has certain limitations: as a cross-sectional design, it cannot confirm causality; small subgroup sample sizes limit statistical power; and unmonitored trace organic solvents (especially benzene series such as toluene) may confound results. Prospective cohort studies with larger samples and comprehensive solvent monitoring are needed to verify the causal relationship and synergistic mechanisms. From the perspectives of public health prevention and control, this study has important practical guiding significance. This study confirms that combined exposure to low-dose xylene and moderate/low-intensity noise already poses significant health hazards, suggesting OELs for co-exposed posts should be appropriately lowered. Manufacturing enterprises should optimize production processes and adopt closed and automated technologies to reduce xylene volatilization and noise emissions. Meanwhile, enterprises should strengthen occupational health monitoring, incorporate high-frequency hearing threshold testing into the routine physical examinations of combined exposure workers, and strengthen occupational health education to improve HPD compliance. Intensive hypertension monitoring and intervention should also be implemented for co-exposed workers to mitigate superimposed damage. Conclusions Combined low-level xylene and noise exposure significantly increases HFHL risk via synergistic ototoxicity, even within national occupational limits. Hypertension, exposure duration, noise intensity, and xylene concentration are independent risk factors, supporting targeted hearing protection and monitoring. Declaration of generative AI and AI-assisted technologies in the manuscript preparation process During the preparation of this work the authors used Doubao (Version 1.8, an intelligent large language model developed by ByteDance) in order to improve language and generate Fig. 1. After using this tool/service, the authors reviewed and edited the content as needed and takes full responsibility for the content of the published article. CRediT authorship contribution statement Zhihao Shi: Writing – original draft, Visualization, Validation, Methodology, Investigation. Genyuan Li: Software, Project administration, Formal analysis, Data curation. Zhehua Zhou: Writing – review & editing, Supervision, Resources, Funding acquisition, Conceptualization. Funding This work was supported by the Jiaxing Municipal Science and Technology Program, in Jiaxing City, China (Grant Numbers: 2024AD10062), and General Projects of the Jiaxing Municipal Health Science and Technology Program (Grant Numbers: JWKJ-26072). Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.