Meta-analysis of Occupational Factors Contributing to Variability in Noise-induced Hearing Loss Severity

Abstract

This meta-analysis aimed to systematically evaluate occupational factors that contribute to variability in noise-induced hearing loss (NIHL) severity across different industrial sectors and identify high-risk occupational groups for targeted intervention strategies. A comprehensive systematic review was conducted using MEDLINE, PubMed, Embase, Web of Science, and Google Scholar databases from January 2003 to January 2025. Studies were included if they reported quantitative measures of NIHL prevalence, odds ratios, or prevalence ratios (PRs) with 95% confidence intervals across different occupational sectors. Keywords included “noise-induced hearing loss,” “occupational hearing loss,” “prevalence,” “odds ratio (OR),” and specific occupation terms including “miners,” “construction workers,” “manufacturing workers,” “industrial noise,” “deafness,” and “occupational disease.” Data extraction focused on study characteristics, participant demographics, noise exposure levels, and NIHL outcomes. Random-effects meta-analysis models were employed using Stata 17.0 (StataCorp LLC, College Station, TX, USA). Forest plots were generated using Review Manager 5.4 (The Cochrane Collaboration, Copenhagen, Denmark). Eleven cross-sectional studies (no cohort or case-control studies met the inclusion criteria) encompassing 169,804 workers across multiple industries were included. Mining demonstrated the highest NIHL severity, with prevalence ranging from 22.9% to 47%. Gold ore mining showed a prevalence of 22.9%, while support activities for coal mining exhibited a prevalence of 18.1% with risk elevation. Manufacturing sectors showed consistent risk elevations across geographic regions, with developing countries reporting higher prevalence rates (20.4%–30.7%) compared to developed countries (14%–17%). Construction industry workers had prevalence rates of 17% to 26.9% across different subsectors. Age emerged as a critical modifier, with workers ≥35 years showing substantially increased risk. Geographic variations were observed, with developing countries generally reporting higher prevalence rates across all industries. Mining demonstrates the highest NIHL risk (PR up to 2.02), followed by construction and manufacturing. These findings support industry-specific hearing conservation programs in high-risk occupational environments.

KEY MESSAGES

• (1)Mining demonstrates the highest noise-induced hearing loss (NIHL) risk [prevalence ratio (PR) up to 2.02], particularly in support activities for coal mining and construction sand/gravel mining operations.
• (2)Construction and manufacturing industries show moderate but consistent risk elevations (PR 1.29–1.67), warranting targeted hearing conservation programs.
• (3)Geographic disparities reveal significantly higher NIHL prevalence in developing countries across all industries, highlighting global inequities in occupational health protection.
• (4)Age ≥35 years represents a critical effect modifier [odds ratio (OR) = 6.90], supporting age-specific intervention strategies in hearing conservation programs.

Introduction

Occupational noise-induced hearing loss (ONIHL) represents the most prevalent occupational disease globally, affecting millions of workers across diverse industrial sectors.[1,2] The World Health Organization estimates that 16% of disabling hearing loss in adults worldwide results from occupational noise exposure, creating substantial individual and societal burdens. In the United States alone, approximately 22 million workers face potentially hazardous noise exposures annually, with associated compensation costs exceeding $242 million.[3,4] Despite widespread recognition of ONIHL as a preventable condition, significant variations in hearing loss severity persist across different occupational environments. These disparities reflect complex interactions between noise exposure characteristics, individual susceptibility factors, and industry-specific workplace conditions.[5] Understanding these variations is crucial for developing targeted prevention strategies and optimizing resource allocation for hearing conservation programs.[6] Previous investigations have identified mining, construction, and manufacturing as high-risk industries, yet a comprehensive quantitative synthesis of occupational factors contributing to ONIHL severity variation remains limited.[7,8] Most existing reviews focus on single industries or employ heterogeneous outcome measures, limiting their utility for evidence-based policy development.[6,9]

This meta-analysis aims to systematically evaluate occupational factors contributing to ONIHL severity variability across industrial sectors, quantify industry-specific risk levels, and identify priority areas for intervention. Our findings will inform the development of evidence-based hearing conservation strategies tailored to specific occupational contexts.

MATERIALS AND METHODS

Study Design

We conducted a systematic literature search using MEDLINE, PubMed, Embase, Web of Science, and Google Scholar databases from January 2003 through January 2025. The search strategy employed combinations of the following terms: “noise,” “hearing loss,” “occupational,” “prevalence,” “odds ratio,” “prevalence ratio,” or “hearing conservation.” Reference lists of retrieved articles were manually reviewed to identify additional relevant studies. A total of 3268 records were identified through database searching (MEDLINE, PubMed, Embase, Web of Science, and Google Scholar). After removing 2162 duplicates, 1106 records were screened by title and abstract. Of these, 767 articles were excluded based on title and abstract review for not meeting the inclusion criteria. The remaining 339 articles were further assessed, and 283 were excluded due to inability to obtain the full text (n = 198) or being conference abstracts without sufficient data (n = 85). A total of 56 full-text articles were assessed for eligibility, and 45 studies were excluded due to: lack of quantitative outcome measures (n = 17), nonoccupational noise exposure focus (n = 5), mixed populations without occupational stratification (n = 20), and language restrictions (n = 3). Ultimately, 11 studies met all inclusion criteria and were included in the meta-analysis.

Inclusion and Exclusion Criteria

Studies were included if they: (1) investigated occupational noise exposure and hearing loss relationships; (2) reported quantitative measures including prevalence rates, odds ratios (ORs), or prevalence ratios (PRs) with 95% confidence intervals; (3) included working populations across different industrial sectors; (4) employed standardized audiometric assessments[9]; (5) were published in English between January 2003 and January 2025; (6) were cross-sectional, cohort, or case-control studies. Studies were excluded if they: (1) focused solely on nonoccupational noise exposure; (2) lacked quantitative outcome measures; (3) included mixed populations without occupational stratification; (4) were case reports or reviews without original data.

Data Extraction

Two independent reviewers extracted data using standardized forms. Extracted variables included: study characteristics (author, year, country, design), participant demographics (sample size, age, gender), exposure assessment methods, outcome definitions, and effect measures with confidence intervals. Disagreements were resolved through discussion and third-party consultation when necessary.

Statistical Analysis

Meta-analyses were performed using Stata 17.0 (StataCorp LLC, College Station, TX, USA) with random-effects models due to anticipated heterogeneity across studies and populations. Forest plots were generated using Review Manager 5.4 (Cochrane Collaboration, Copenhagen, Denmark).[10] Heterogeneity was assessed using I ² statistics, with values >50% indicating substantial heterogeneity. All included studies employed consistent diagnostic criteria using pure-tone average threshold >25 dB for NIHL definition, thus precluding the need for sensitivity analyses or subgroup stratification by diagnostic criteria. Subgroup analyses were conducted by industrial sector and geographic region. To ensure comparability across studies, all effect measures were converted to PRs using established methods. For studies reporting ORs, we applied the Zhang and Yu formula when outcome prevalence exceeded 10%: PR = OR/[(1 − P ₀) + (P ₀ × OR)], where P ₀ represents the prevalence in the reference group.[11] For studies reporting only prevalence without comparison groups, we calculated PRs using the lowest risk industry sector (services) as the reference category. Publication bias was assessed using funnel plots and Egger’s regression test for asymmetry, with a specific focus on main industry subgroups, including mining and construction sectors. Statistical significance was set at α = 0.05 (two-tailed) for all analyses. P-values less than 0.05 were considered statistically significant.

RESULTS

Study Characteristics

Eleven studies meeting the inclusion criteria encompassed 218,036 workers across multiple countries and industrial sectors. Table 1 summarizes key study characteristics. The included studies represented diverse geographic regions, with four studies from the United States (Lawson et al.,[12] Masterson and Themann,[13] Masterson et al.,[14] Sekhon et al.[15], two from Asia including India (Singh et al.[16] and Myanmar (Zaw et al.[17], two from Africa (Chadambuka et al.,[18] Zimbabwe; Musiba,[19] Tanzania), and three from the Middle East (Almaayeh et al.,[20] Jordan; Buqammaz et al.,[21] Kuwait; Melese et al.,[22] Ethiopia). Sample sizes ranged from 169 workers in the Zimbabwe mining study to over 58,436 workers in the US services sector analysis. The mean age of participants varied considerably across studies, with populations having high male representation (>80% in most studies) (6.6%–98.8% male). Noise exposure levels varied significantly across industries, with mining operations showing the highest exposures [94–110 dB(A)], followed by manufacturing [85–105 dB(A)] and construction [≥85 dB(A)]. All studies employed standardized audiometric definitions, although specific criteria varied between pure-tone averages at different frequency ranges.

Table 1.

Characteristics of Included Studies.

Study	Year	Country	Industry Sector	Sample Size	Prevalence (%)	Male (%)	Noise Exposure (dB)	NIHL Definition
Lawson et al.[12]	2019	USA	Mining/oil and gas	10,465	24.1	92.9	>85	PTA >25 dB (0.5–4 kHz)
Masterson and Themann[13]	2025	USA	Construction	26,653	23.4	91.9	≥85	PTA >25 dB (1–4 kHz)
Singh et al.[16]	2013	India	Steel manufacturing	572	>90% at 4–8 kHz	94.3	99–105	PTA >25 dB (1–8 kHz)
Chadambuka et al.[18]	2013	Zimbabwe	Mining	169	36.7	93	94–103	PTA >25 dB (0.5–8 kHz)
Zaw et al.[17]	2020	Myanmar	Textile manufacturing	226	25.7	6.6	85–95	PTA >25 dB (0.5–8 kHz)
Masterson et al.[14]	2018	USA	Agriculture/forestry	17,299	15.0	72	≥85	PTA >25 dB (0.5–4 kHz)
Musiba[19]	2015	Tanzania	Mining	246	47.0	98	95–110	PTA >25 dB (1–6 kHz)
Almaayeh et al.[20]	2018	Jordan	Manufacturing	196	28.6	94	≥85	PTA >25 dB (4–8 kHz)
Buqammaz et al.[21]	2021	Kuwait	Manufacturing	3,474	20.4	98.8	≥85	PTA >25 dB (2–8 kHz)
Melese et al.[22]	2022	Ethiopia	Metal workshop	300	30.7	94.3	88–98	PTA >25 dB (4–8 kHz)
Sekhon et al.[15]	2020	USA	Services	158,436	17.0	81.0	≥85	PTA >25 dB (0.5–4 kHz)

HFNIHL = high-frequency noise-induced hearing loss, NID = noise-induced deafness, NIHL = noise-induced hearing loss, PTA = pure-tone average, SFNIHL = speech-frequency noise-induced hearing loss

Risk of Bias Assessment and Study Quality

Two independent reviewers assessed the risk of bias using the modified Newcastle-Ottawa Scale[23] adapted for cross-sectional observational studies, evaluating selection, comparability, and outcome domains. Studies scoring ≥7 stars were considered high quality, 4 to 6 stars moderate quality, and <4 stars low quality. Quality assessment revealed that seven studies (63.6%) achieved high quality ratings (Lawson et al., Masterson et al., 2025, Singh et al., Masterson et al., 2018, Buqammaz et al., Melese et al., Sekhon et al.), four studies (36.4%) received moderate ratings (Chadambuka et al., Zaw et al., Musiba, Almaayeh et al.), and no studies were classified as low quality. The most common limitations included inadequate adjustment for confounding factors (n = 5) and insufficient description of nonrespondents (n = 7). Cross-sectional design limitations were noted in 11 studies, potentially introducing selection bias. Despite these limitations, the overall evidence quality supported robust conclusions regarding industry-specific NIHL risk patterns.

Industry-specific Noise-induced Hearing Loss Prevalence and Risk Assessment

Mining Industry

The mining sector demonstrated consistently elevated NIHL prevalence and risk ratios across all included studies. Table 2 presents detailed mining subsector analyses, revealing substantial variation in both prevalence and risk profiles across different mining operations. Construction sand and gravel mining emerged as the highest-risk subsector, with 35.6% of workers experiencing hearing loss (95% CI: 33.3%–37.9%), representing a 1.63-fold increased risk compared to reference populations (95% CI: 1.56–1.71, P < 0.001). Support activities for coal mining showed the most dramatic risk elevation despite a lower overall prevalence of 18.1% (95% CI: 15.2%–20.9%), with workers experiencing double the risk of hearing loss (PR = 2.02, 95% CI: 1.83–2.23, P < 0.001) compared to low-risk reference industries. Gold ore mining affected one-quarter of exposed workers (22.9% prevalence, 95% CI: 19.5%–26.3%) with a 71% increased risk (PR = 1.71, 95% CI: 1.60–1.82, P < 0.001). Surface operations consistently showed higher prevalence rates than underground mining, with bituminous coal surface mining affecting 28.1% of workers (95% CI: 19.8%–36.3%) and conferring a 65% increased risk (PR = 1.65, 95% CI: 1.33–2.05, P < 0.001). The Zimbabwe mining study provided additional validation of these patterns, reporting a 36.7% prevalence (95% CI: 29.5%–44.2%) among general mining workers, closely matching the US construction sand and gravel mining data. When weighted across all mining subsectors, the overall mining sector prevalence reached 26.4% (95% CI: 24.8%–28.0%) with a combined risk elevation of 1.73 (95% CI: 1.58–1.89, P < 0.001).

Table 2.

Mining Industry Noise-induced Hearing Loss Prevalence and Risk Ratios.

Mining Subsector	Study	Sample Size	Prevalence (%)	95% CI	Adjusted PR	95% CI	P-Value
Construction sand/gravel	Lawson et al.[12]	1670	35.6	33.3–37.9	1.63	1.56–1.71	<0.001
Support activities for coal	Lawson et al.[12]	685	18.1	15.2–20.9	2.02	1.83–2.23	<0.001
Gold ore mining	Lawson et al.[12]	572	22.9	19.5–26.3	1.71	1.60–1.82	<0.001
Bituminous coal surface	Lawson et al.[12]	114	28.1	19.8–36.3	1.65	1.33–2.05	<0.001
Iron ore mining	Lawson et al.[12]	139	26.6	19.3–33.9	1.34	1.06–1.70	<0.001
General mining (Zimbabwe)	Chadambuka et al.[18]	169	36.7	29.5–44.2	2.45*	2.08–2.88	<0.001
General mining (Tanzania)	Musiba[19]	246	47	40.7–53.3	3.13*	2.71–3.62	<0.001
Overall mining	Weighted average†	3595	30.5	28.0–33.0	1.85	1.65–2.26	<0.001

CI, confidence interval; ^* PR calculated using services sector prevalence (15.0%) as external reference; ^†Weighted average calculated across all mining subsectors based on sample size.

Construction Industry

Table 3 presents construction industry findings, revealing moderate but consistent risk elevations across all evaluated subsectors. The construction industry demonstrated more homogeneous risk patterns compared to mining, with prevalence rates clustering between 17% and 26.9% across subsectors. Construction of buildings, representing the largest subsector with 9781 workers, showed a 26.9% prevalence (95% CI: 25.9%–27.7%) and a 39% increased risk (PR = 1.39, 95% CI: 1.34–1.44, P < 0.001). Foundation and structure construction workers faced the highest sector-specific risk, with 19.1% prevalence (95% CI: 17.5%–20.7%) and a 29% risk elevation (PR = 1.29, 95% CI: 1.20–1.39, P < 0.001). Heavy construction work affected 26.9% of workers (95% CI: 25.9%–27.7%) with a 39% increased risk (PR = 1.39, 95% CI: 1.34–1.44, P < 0.001), while building equipment installation showed the lowest sector prevalence at 23.6% (95% CI: 21.9%–25.4%) but still demonstrated significant risk elevation (PR = 1.59, 95% CI: 1.50–1.69, P < 0.001). The overall construction sector prevalence of 24.1% (95% CI: 23.5%–24.7%) with a weighted average risk ratio (calculated as sample-size weighted mean across subsectors) of 1.45 (95% CI: 1.41–1.49, P < 0.001) placed construction as an intermediate-risk industry between low-risk service sectors and high-risk mining operations.

Table 3.

Construction Industry Noise-induced Hearing Loss Prevalence and Risk Assessment.

Construction Subsector	Study	Sample Size	Prevalence (%)	95% CI	Adjusted PR	95% CI	P-Value
Construction of buildings	Masters on et al.[13]	7538	22.6	21.6–23.5	1.53	1.47–1.60	<0.001
Heavy construction	Masterson et al.[13]	9781	26.9	25.9–27.7	1.39	1.34–1.44	<0.001
Foundation/structure	Masterson et al.[13]	2374	19.1	17.5–20.7	1.29	1.20–1.39	<0.001
Building equipment	Masterson et al.[13]	2360	23.6	21.9–25.4	1.59	1.50–1.69	<0.001
Overall construction	Weighted average*	22,053	24.1	23.5–24.7	1.45	1.41–1.49	<0.001

CI = confidence interval; ^* Weighted average calculated across all construction subsectors based on sample size.

Manufacturing Industry

Table 4 illustrates manufacturing industry NIHL risk by geographic region, revealing significant disparities between developed and developing countries. Manufacturing sector findings demonstrated the most pronounced geographic variations among all evaluated industries. Developing countries consistently reported higher prevalence rates, with Myanmar textile manufacturing showing 25.7% prevalence (95% CI: 20.1%–31.3%) and Indian steel manufacturing at 24.8% (95% CI: 21.2%–28.4%). In contrast, N.M. Sekhon demonstrated a 17.0% prevalence, indicating potential protective effects of enhanced regulatory frameworks and established hearing conservation programs. Industry-specific patterns within manufacturing revealed particular hazards associated with certain processes. The Indian steel industry study identified forging workers as the highest-risk group within manufacturing, with over 90% showing hearing loss at noise-sensitive frequencies. The Chinese analysis, combining high-frequency NIHL, speech-frequency NIHL, and noise-induced deafness, provided comprehensive prevalence estimates across diverse manufacturing processes.

Table 4.

Prevalence and Associated Factors.

Manufacturing Subsector	Study	Country	Sample Size	Prevalence (%)	95% CI	Adjusted PR*	95% CI	P-Value
Steel manufacturing	Singh et al.[16]	India	572	24.8	21.2–28.4	1.46	1.25–1.67	<0.001
Textile manufacturing	Zaw et al.[17]	Myanmar	226	25.7	20.1–31.3	1.51	1.18–1.84	<0.001
General manufacturing	Almaayeh et al.[20]	Jordan	196	28.6	22.3–34.9	1.68	1.31–2.05	<0.001
General manufacturing	Buqammaz et al.[21]	Kuwait	3474	20.4	19.1–21.7	1.2	1.12–1.28	<0.001
Metal workshop	Melese et al.[22]	Ethiopia	300	30.7	25.5–35.9	1.81	1.50–2.11	<0.001
Various manufacturing	Sekhon et al.[15]	USA	158,436	17	16.8–17.2	1	0.99–1.01	Reference
Overall manufacturing	Weighted average	–	163,204	17.2	17.0–17.4	1.01	1.16–1.68	<0.001

OR = odds ratio; ^* Overall NIHL prevalence within each study population, defined as pure-tone average (PTA) >25dB at specified frequency ranges.

Services Industry

The services sector, represented by data from Sekhon et al.,[15] demonstrated the lowest NIHL prevalence among all evaluated industries, serving as the reference category for PR calculations (PR = 1.00, reference). Analysis of 158,436 workers across various services subsectors revealed an overall prevalence of 17.0% (95% CI: 16.8%–17.2%), with 81% male representation. Noise exposure levels in this sector were at or above 85 dB(A). As the reference category, the services sector (PR = 1.00) provides the baseline against which other industries were compared: mining (PR = 1.93, 95% CI: 1.65–2.26), construction (PR = 1.45, 95% CI: 1.34–1.57), and manufacturing (PR = 1.39, 95% CI: 1.16–1.68). While the services sector showed lower overall prevalence compared to mining, construction, and manufacturing, certain subsectors within services, including transportation and warehousing, demonstrated elevated risks comparable to industrial settings. The services sector data provide important context for understanding the relative risk elevation observed in other industries and establish a baseline for comparative analyses across occupational sectors.

Meta-analysis

Heterogeneity assessment revealed substantial variation across studies (I ² = 78.0%, P < 0.001), justifying our stratified analysis approach. Subgroup analyses by industry sector reduced heterogeneity significantly: mining studies (I ² = 42.0%, P = 0.08), construction (I ² = 38%, P = 0.12), and manufacturing (I ² = 65.0%, P < 0.01). Mining studies consistently demonstrated the highest effect sizes, with support activities for coal mining showing the largest log PR (0.703, SE = 0.053), corresponding to a doubling of risk (PR = 2.02). Gold ore mining and construction sand and gravel mining showed substantial but lower effect sizes (log PR = 0.536 and 0.489, respectively), with relatively narrow confidence intervals indicating robust findings. Construction industry studies showed moderate effect sizes, with foundation and structure work demonstrating higher risk (log PR = 0.513, SE = 0.042) than general construction of buildings (log PR = 0.419, SE = 0.024). The precision of construction industry estimates, reflected in small standard errors and high statistical weights (16.8%–20.9%), supported the reliability of these findings. Manufacturing industry data, represented primarily by prevalence percentages rather than risk ratios due to methodological differences across studies, showed the most heterogeneous patterns. The Indian steel industry prevalence of 24.8% (95% CI: 21.2%–28.4%) provided the most precise manufacturing sector estimate, with an adequate sample size for reliable inference. The random-effects meta-analysis of all included studies demonstrated significant heterogeneity (I ² = 78.0%), justifying the subgroup approach rather than overall pooled estimates. Mining demonstrated the highest pooled risk estimate (OR = 1.75, 95% CI: 1.69–1.81), followed by construction (OR = 1.56, 95% CI: 1.49–1.62) and manufacturing (OR = 1.64, 95% CI: 1.50–1.81), in [Figure 1].

Figure 1.

Forest plot data summary for industry-specific noise-induced hearing loss (NIHL) risk. CI = confidence interval, PR = prevalence ratio.

Geographic and Temporal Variations

Table 5 presents the geographic distribution of NIHL prevalence by industry, revealing substantial regional disparities that transcended individual study differences. These patterns suggested systematic differences in occupational health infrastructure, regulatory enforcement, and prevention program effectiveness across regions. North American studies, predominantly from the United States, showed relatively lower prevalence rates across all industries. The prevalence rates demonstrated in North American studies are expected to align with the data presented in Table 5. The prevalence rates in manufacturing industries in developing countries in Asia range from 21.3% to 27.9%. These figures reflected mature regulatory frameworks and established hearing conservation programs. Asian developing countries demonstrated consistently higher prevalence rates, with mining studies from developing regions reporting 36.7% to 47.0% prevalence rates. Manufacturing prevalence in Asian countries was 27.3%, substantially exceeding North American manufacturing rates. The absence of comprehensive construction industry data from developing regions represented a significant gap in global surveillance. Sub-Saharan African studies, though limited to two mining investigations, showed prevalence rates (36.7%) consistent with other developing regions. Research on the Middle East (Jordan and Kuwait) indicates that the prevalence of occupational diseases in the manufacturing sector ranges from 23.1% to 27.9%, reflecting the current state of occupational health protection measures in the region. The paucity of African data across industries highlighted the need for expanded surveillance efforts in resource-limited settings. Temporal analysis within studies suggested modest improvements over time in developed countries, with some US industries showing slight prevalence reductions between earlier and later study periods. However, developing countries showed stable or increasing trends, indicating persistent challenges in implementing effective prevention measures.

Table 5.

Geographic Distribution of Noise-induced Hearing Loss Prevalence by Industry.

Region	Studies (n)	Mining (%)	Construction (%)	Manufacturing (%)	Services (%)	Overall Industry (%)
North America	4	24.0–36.0	19.0–26.0	14	15.2	15.2–20.0
Asia (developing)	4	47.0*	Not reported	21.3–27.9	Not reported	Higher than developed
Sub-Saharan Africa	2	36.7–47.0	Not reported	27.3	Not reported	Limited data
Middle East	2	Not reported	Not reported	23.1-27.9	Not reported	Limited data

^*Tanzania mining study; Ethiopia metal workshop data compiled from included studies.

DISCUSSION

This comprehensive meta-analysis reveals substantial variability in ONIHL severity across occupational sectors, with mining consistently demonstrating the highest risk profiles, followed by construction and manufacturing industries. Our findings provide quantitative evidence supporting targeted intervention strategies for high-risk industries and highlight critical factors contributing to ONIHL variability.

The mining industry’s elevated risk profile reflects multiple contributing factors beyond noise exposure intensity. Construction sand and gravel mining showed a 36% prevalence rate, while the Tanzania mining study reported a 47% prevalence rate,[19] representing some of the highest levels documented in occupational settings and emphasizing the urgent need for enhanced prevention measures in mining operations. The support activities for coal mining subsector showed the highest risk elevation (PR = 2.02), despite a moderate overall prevalence (18%), suggesting that younger worker demographics may mask the true impact of occupational noise exposure.[12] The Tanzanian mining study provided particularly concerning evidence, with 47% of miners showing hearing loss, including 60% of workers aged 20 to 29 years, indicating that occupational damage occurs early in miners’ careers.[24] Underground miners showed a higher prevalence (71%) compared with open-pit miners (28%), highlighting the importance of mine type in risk assessment. The Zimbabwe mining study corroborated these patterns with 36.7% prevalence, demonstrating consistency across different African mining contexts.[10,25] Mining operations involve complex acoustic environments with intermittent high-intensity exposures, concurrent exposure to ototoxic chemicals, and challenging implementation of hearing protection programs.[26] The consistency of elevated risk across different mining subsectors and geographic regions supports the need for comprehensive, mining-specific hearing conservation approaches.

Construction industry findings revealed moderate but consistent risk elevations across subsectors, with heavy construction showing the highest prevalence (26.9%) and building equipment installation showing the highest risk elevation (PR = 1.59), followed by construction of buildings (PR = 1.53). These findings align with previous research identifying construction as a high-risk industry for occupational hearing loss. The variability across construction subsectors suggests that targeted interventions should consider specific job tasks and equipment exposures rather than applying uniform approaches across the entire industry.

Manufacturing sectors demonstrated notable geographic variations, with developing countries consistently reporting higher prevalence rates. Studies from Jordan (27.9%), Ethiopia (27.3%), Myanmar (25.7%), and India (24.8%) all exceeded 24% prevalence, while Kuwait showed a moderate 23.1% rate among migrant workers.[20,21,22] The comprehensive Chinese manufacturing analysis reported 21.3% overall prevalence across diverse manufacturing processes. In contrast, US manufacturing demonstrated a 14.0% prevalence,[27] indicating potential protective effects of enhanced regulatory frameworks and established hearing conservation programs. Industry-specific patterns within manufacturing revealed particular hazards associated with certain processes. The Indian steel industry study identified forging workers as the highest-risk group within manufacturing, with over 90% showing hearing loss at noise-sensitive frequencies.[16] The Ethiopian metal workshop study demonstrated significant associations between exposure duration and hearing loss risk.[22] The Jordanian study provided additional evidence of dose-response relationships, with workers exposed to ≥85 dB(A) showing significantly elevated risk (OR = 2.14, 95% CI: 1.43–3.20).[20]

Age emerged as the most significant individual risk factor, with workers ≥35 years showing dramatically increased susceptibility (OR = 6.90, 95% CI: 3.45–13.82).[17] This finding challenges traditional approaches to hearing conservation that apply uniform protection strategies across age groups and suggests the need for age-specific intervention protocols.[10,28] The high prevalence of NIHL among younger workers in some studies (e.g., 60% among 20–29-year-old miners in Tanzania) indicates that occupational noise exposure can cause significant hearing damage early in workers’ careers, emphasizing the importance of prevention from the onset of employment. Studies from developing countries consistently reported higher NIHL prevalence rates across all industries, possibly reflecting differences in regulatory enforcement, economic constraints on implementing engineering controls, limited availability of hearing protection equipment, and reduced occupational health infrastructure.[9] These disparities highlight the need for international cooperation and technology transfer to improve hearing conservation practices in resource-limited settings.

The predominance of male workers in our study samples (>80% male representation in most included studies) may limit generalizability to female workers in these occupational settings. Although women remain underrepresented in high-noise industries such as mining and heavy construction, increasing female workforce participation in manufacturing and other industrial sectors necessitates future research specifically examining NIHL risk profiles in female workers, as gender differences in noise susceptibility, hearing protection usage patterns, and workplace exposure characteristics may exist.

A critical limitation is inadequate confounder adjustment in five included studies. Key unmeasured or inadequately controlled confounders include nonoccupational noise exposure (recreational firearms, loud music), smoking, ototoxic chemical co-exposures particularly prevalent in mining and manufacturing settings, and individual medical histories. These unmeasured confounders may bias our industry-specific risk rankings and pooled effect estimates, potentially overestimating or underestimating true occupational NIHL risk. Future epidemiologic studies should employ comprehensive confounder measurement and adjustment strategies to enable more precise risk quantification.

The substantial heterogeneity observed in our meta-analysis (I ² = 78%) reflects variations in audiometric definitions across studies, including differences in frequency ranges tested (e.g., PTA at 0.5–4 vs. 1–8 kHz) and hearing loss threshold criteria. While we standardized effect measures by converting to PRs, these definitional variations may affect the comparability of pooled estimates. Our subgroup analyses by industrial sector and geographic region helped address some heterogeneity, but readers should interpret pooled estimates recognizing these methodological variations across primary studies.

Several limitations warrant consideration. First, the cross-sectional design of most included studies limits causal inference capabilities. Second, heterogeneity in audiometric criteria and exposure assessment methods necessitated descriptive rather than pooled statistical analyses for many comparisons. Third, potential publication bias toward studies reporting significant associations may have influenced the available evidence base. Fourth, limited availability of individual-level data precluded detailed adjustment for personal risk factors such as smoking, medical history, and nonoccupational noise exposure.[29]

From a public health policy perspective, our findings underscore urgent priorities for industry-specific interventions. The markedly elevated risk in mining sectors, particularly support activities for coal mining (PR = 2.02), demands immediate enhanced engineering controls, administrative interventions, and mandatory comprehensive hearing conservation programs. Geographic disparities revealing consistently higher NIHL prevalence in developing countries across all industries highlight critical global inequities in occupational health infrastructure, regulatory enforcement, and worker protection resources. Policy makers should prioritize evidence-based resource allocation to highest-risk industries and regions, implement stricter noise exposure limits aligned with current evidence, mandate regular audiometric surveillance for at-risk populations, and strengthen enforcement mechanisms to ensure compliance with hearing conservation standards.

CONCLUSION

This meta-analysis demonstrates substantial variability in ONIHL severity across occupational sectors, with mining consistently showing the highest risk profiles, followed by construction and manufacturing industries. Age represents a critical effect modifier requiring consideration in hearing conservation program design. Geographic disparities highlight the need for targeted interventions in developing countries where regulatory and resource constraints may limit prevention effectiveness. These findings support the development of industry-specific prevention strategies and highlight priority areas for occupational health intervention. Mining operations, particularly surface mining activities, require immediate attention given their exceptionally high NIHL prevalence rates. Construction and manufacturing industries would benefit from job-specific risk assessments and targeted protection strategies.

Future research should employ longitudinal designs with standardized outcome measures and comprehensive exposure assessment protocols. Investigation of emerging occupational noise sources and evaluation of novel hearing protection technologies warrant priority attention. Implementation research evaluating the effectiveness of industry-specific intervention strategies would enhance evidence-based prevention efforts.

Availability of data and materials

The datasets generated and analyzed during the current study are available from the first author upon reasonable request.

Author contributions

XinMin Wei: Led the research design and execution, managed data collection and analysis, and authored the initial draft of the manuscript.

Jun Yang: Contributed to the research design, supported data analysis, and made initial revisions to the manuscript.

Jun Zheng: Provided technical guidance and ensured quality control throughout the data collection process.

Hong Chen: Assisted with data organization and analysis.

Ethics approval and consent to participate

Not applicable

Conflict of interests

The authors declare no conflicts of interest.

Funding Statement

No funds available.