Between May and December 2020, we conducted a retrospective cohort study and a cross-sectional study based on noncoal mines and assessed the qualitative and quantitative risks of silicosis. The retrospective cohort study, carried out on 3 noncoal mines and 1 public hospital, was designed to establish formulas for predicting the silicosis cumulative hazard ratio (HR; H) and incidence (I) in noncoal mines. The cross-sectional study, carried out on 155 noncoal mines, was designed to investigate recent levels of silica dust exposure in Chinese noncoal mines. Due to the uneven geographical distribution of mineral resources, nonprobability convenience-based sampling was used to select the 155 noncoal mines from 10 provinces in northern, western, southern, eastern, and central China. The inclusion criteria for silica dust–exposed noncoal mines were as follows: (1) nonferrous metal, ferrous metal, or nonmetal (except fuels) mines; (2) 10%≤free silica content≤100%; (3) complete basic information about noncoal mines; and (4) no extremely abnormal determination result(s). The term “exposure” used in this study was defined as the concentration (mg/m³) of airborne occupational dust that was measured in workers’ upper anterior chest area or breathing zone (about 1 foot from the mouth or nose), and the term “job” was defined as a collection of types of work requiring similar skills, responsibilities, training, and process of production [33].

In this retrospective cohort study, monitoring data, collected from 1940 to 2020, included 2 parts: (1) silica dust–exposed workers’ jobs and occupational health records and (2) the free silica content and silica dust concentrations the workers were exposed to in their jobs. Occupational health records included initial (first) and end (last) dates of silica dust exposure, chest X-ray examination results, and the date of initial diagnosis. Silicosis was diagnosed according to each worker’s chest X-ray examination, occupational exposure history, clinical manifestation, and workplace environment measurement. All workers with silicosis were followed up until the date of initial diagnosis. Monitoring data of the silica dust concentration were used to create a job exposure matrix (JEM). Since the 1950s, the Chinese government has enforced systematic dust-sampling regulations, which require mines to measure dust concentrations monthly in dusty work areas and measure the silica content in bulk dust, if necessary [34]. The respirable dust–sampling regulations were implemented in 1992. Therefore, before 1992, the respirable dust concentration was estimated based on the contemporaneous monitoring of the total dust concentration of the same or a similar job. The formula of respirable dust concentration estimation was 0.30 × total dust concentration of the same or similar job + 0.08, where the coefficient of determination (R²)=0.94. Both total dust and respirable dust concentrations were measured as 8-hour time-weighted average concentrations (C_TWA, mg/m³) using gravimetric samplers and microbalance:

,where c is the dust concentration (mg/m³), m₂ is the postsample filter weight (mg), m₁ is the presample filter weight (mg), V is the average sample flow rate (L/minute), t is the sampling duration (minutes), and T is the working hours under exposure to different concentrations of dust. The free silica content (%) was determined using the pyrophosphate method:

where m₄ is the crucible and free silica weight (g), m₃ is the crucible weight (g), and m is the bulk dust weight (g).

Regarding survey data for this cross-sectional study, the basic information about noncoal mines included the mining method, mine category, product type, production scale [35], number of workers, and prevalence of silica dust exposure (PDE). We carried out 3-day continuous sampling determination once a day, and 3 samples were taken each time from the same sampled site: 1 air dust sample for total dust concentration measurement, 1 air dust sample for respirable dust concentration measurement, and 1 bulk dust sample for free silica content determination. Overall, 9 dust samples were collected for each job during the 3-day sampling period. Air dust samples at the mobile and fixed work sites, such as the inspector and the crusher, were separately collected via personal sampling and area sampling. Personal sampling dust samples at the mobile work site were collected from workers’ upper anterior chest area using air sampling pumps at an air flow rate of 1-5 L/minute. Area sampling dust samples at the fixed work site were collected from workers’ breathing zone (near the collar) using air sampling pumps at an air flow rate of 5-80 L/minute. In addition, to make sure the highest and lowest dust concentrations were included, the sample time of area sampling per day was ≥45 minutes, which was divided into 3 intermittent sample times (≥15 minutes each time). All on-site measurements were performed on days with no rain, and calibration of the air sampler was performed before and after each use. Dust concentrations were gravimetrically analyzed using an XP105DR Microbalance (METTLER TOLEDO Inc). The free silica content in the dustfall sample (bulk dust) was determined using the pyrophosphate method. Total silica dust concentrations were calculated by multiplying total dust concentrations by the free silica content, and respirable silica dust concentrations were calculated by multiplying respirable dust concentrations by the free silica content.

Statistical analysis consisted of 3 parts. First, we established prediction formulas for the silicosis cumulative HR (H) and incidence (I). Kaplan-Meier survival analysis and multivariable Cox regression analysis were conducted to establish the formulas based on cumulative exposure to total silica dust (D_t) and cumulative exposure to respirable silica dust (D_r). The male gender was coded as 1, and the female gender worker was coded as 0.

Second, we determined Chinese noncoal mines’ characteristics and silica dust exposure levels. To estimate the overlimit degree of silica dust exposure, we calculated the percentages of dust sample (concentrations) exceeding China’s occupational exposure limit (OEL) or exceeding the exposure limit recommended by the American Conference for Governmental Industrial Hygienists (ACGIH; recommended exposure limit [REL] of respirable silica dust=0.025 mg/m³).

Finally, we conducted qualitative and quantitative silicosis risk assessments of Chinese noncoal miners. The International Mining and Metals Commission's risk-rating table (ICMM) and the occupational hazard risk index (INDEX) were used in qualitative assessment [36-39], and prediction formulas of the silicosis cumulative HR (H) and incidence (I) were used in quantitative risk assessment (Figure 1). There are 5 kinds of qualitative risk levels for both the ICMM and INDEX: tolerable risk/no hazard, potential risk/mild hazard, high risk/moderate hazard, very high risk/severe hazard, and intolerable risk/extreme hazard. To assess the comprehensive qualitative risk of noncoal mines, each characteristic’s weighted score (S_weighted) was calculated based on the percentage of dust samples in the raw 5-level risks, separately coded as 1, 2, 3, 4, and 5:

The comprehensive qualitative risk level was classified as very low risk (0≤S_weighted≤1), low risk (1.01≤S_weighted≤2.00), medium risk (2.01≤S_weighted≤3.00), high risk (3.01≤S_weighted≤4.00), and very high risk (4.01≤S_weighted≤5.00). In addition, if S_weighted of both the ICMM and INDEX was ≥3.01 or S_weighted of either was ≥4.01, the comprehensive risk level was identified as high. Continuous data of dust exposure and silicosis risk were expressed as the median (IQR).

The OEL for dust exposure depends on the free silica content: if 10%≤free silica content≤50%, OELs for total dust and respirable dust are 1 and 0.7 mg/m³, respectively; if 50%<free silica content≤80%, OELs are 0.7 and 0.3 mg/m³, respectively; and if free silica content>80%, OELs are 0.5 and 0.2 mg/m³, respectively. All statistical analyses were conducted using SAS version 9.4 (SAS Institute).

The Medical Ethical Review Committee of the National Institute for Occupational Health and Poison Control, Chinese Center for Disease Control and Prevention, approved this study (approval number 201811), with the requirement for informed consent being waived. In this study, there was no monetary or nonmonetary compensation provided to participants. No identification of individual participants was shown in any figures or multimedia appendices. Access to worker- and mine-identifiable data was restricted to authorized individuals and was only accessible on 2 designated computers.

In this retrospective cohort study, a total of 2551 silica dust–exposed workers were selected from 3 noncoal mines, including 1 large and 2 medium-size mines. Monitoring data showed that the C_TWA of total dust ranged from 0.06 to 41.94 mg/m³, of respirable dust ranged from 0.05 to 12.54 mg/m³, and of free silica content ranged from 16.8% to 17.3%. Of the 2551 workers, 1512 (59.3%) were still working in the enterprises, while the other 1039 (40.7%) had left, retired, or died. In addition, their follow-up time ranged from 1 year to 68 years, and the total follow-up time was up to 57,480 person-years. By December 2020, 247 (16.3%) workers were diagnosed with silicosis, 221 (89.5%) of whom had stage I silicosis at the initial diagnosis (see Table 1 and Table S1 in Multimedia Appendix 1).

Kaplan-Meier survival analysis showed that there were significant differences in the silicosis probability between silica dust–exposed male and female miners (log-rank test χ²₁=7.52, P=.01), as shown in Figure 2, but there was no difference among different mine categories, the production scale, mining methods, the initial silica dust exposure age, the initial silica dust exposure year, or smoking history.

Regarding total silica dust exposure, multivariable Cox regression analysis showed that gender (hazard ratio [HR]=4.96, β=1.60, P=.02) and cumulative exposure to total silica dust (HR=1.07, β=0.07, P<.001) were independent predictors of silicosis morbidity in noncoal miners. Over the next 10, 20, and 30 years, prediction formulas of the silicosis cumulative HR (H) caused by total silica exposure would be as follows:

H₁₀ = 0.09 × exp(1.60[gender – 0.95] + 0.07[D_t – 8.01])

H₂₀ = 0.27 × exp(1.60[gender – 0.95] + 0.07[D_t – 8.01])

H₃₀ = 0.50 × exp(1.60[gender – 0.95] + 0.07[D_t – 8.01])

The prediction formulas of the silicosis cumulative incidence (I) would be as follows:

I₁₀ = 1 – 0.92^{exp(1.60[gender – 0.95] + 0.07[Dt – 8.01])}

I₂₀ = 1 – 0.76^{exp(1.60[gender – 0.95] + 0.07[Dt – 8.01])}

I₃₀ = 1 – 0.61^{exp(1.60[gender – 0.95] + 0.07[Dt – 8.01])}

Regarding respirable silica dust exposure, gender (HR=4.96, β=1.58, P=.03) and cumulative exposure to respirable silica dust (HR=1.27, β=0.24, P<.001) were also independent predictors of silicosis morbidity in noncoal miners. Over the next 10, 20, and 30 years, the prediction formulas of H caused by respirable silica dust exposure would be as follows:

H₁₀ = 0.09 × exp(1.58[gender – 0.95] + 0.24[D_r – 2.64])

H₂₀ = 0.27 × exp(1.58[gender – 0.95] + 0.24[D_r – 2.64])

H₃₀ = 0.50 × exp(1.58[gender – 0.95] + 0.24[D_r – 2.64])

The prediction formulas of I would be as follows:

I₁₀ = 1 – 0.92^{exp(1.58[gender – 0.95] + 0.24[Dr – 2.64])}

I₂₀ = 1 – 0.76^{exp(1.58[gender – 0.95] + 0.24[Dr – 2.64])}

I₃₀ = 1 – 0.61^{exp(1.58[gender – 0.95] + 0.24[Dr – 2.64])}

In this cross-sectional study, of the 155 noncoal mines, 22 (14.2%) nonmetal mines and 6 (3.9%) nonferrous metal mines were excluded due to no silica dust exposure (free silica content<10%), and 1 (0.6%) nonferrous metal mine was excluded due to the extremely abnormal concentration of total dust (196.16 mg/m³). Ultimately, a total of 126 (81.3%) noncoal mines, with 29,835 miners and 4623 dust samples, were included in this cross-sectional study. Of the 29,835 miners, 13,037 (43.7%) were exposed to silica dust, of whom 12,952 (99.0%) were male. The 126 noncoal mines included were selected from 10 Chinese provinces: (1) 37 (29.4%) nonferrous metal mines from Liaoning, Inner Mongolia, Qinghai, Hubei, Hunan, Sichuan, and Guangdong Provinces; (2) 19 (15.1%) ferrous metal mines from Liaoning, Inner Mongolia, Qinghai, Sichuan, Jiangsu, Hubei, and Shandong Provinces; and (3) 70 (55.6%) nonmetal mines from Liaoning, Inner Mongolia, Qinghai, Gansu, Hubei, Sichuan, Hunan, Jiangsu, and Shandong Provinces. The 4623 dust samples included 1500 (32.4%) dust samples for total dust concentration measurement, 1557 (33.7%) dust samples for respirable dust concentration measurement, and the remaining 1566 (33.9%) dust samples for free silica content determination (see Figures 3-6 and Table S2 in Multimedia Appendix 2).

The overall median PDE, free silica content, total dust concentration, total silica dust concentration, respirable dust concentration, and respirable silica dust concentration were 61.6%, 27.6%, 1.30 mg/m³, 0.39 mg/m³, 0.58 mg/m³, and 0.15 mg/m³, respectively. In addition, 963 (64.2%) of total dust samples (concentrations) and 688 (44.2%) of respirable dust samples (concentrations) exceeded China’s OELs, while 1551 (99.6%) of respirable dust samples exceeded the ACGIH REL. Nonmetal mines had the highest median PDE (51.9%), total silica dust concentration (0.48 mg/m³), and percentages of total dust samples and respirable dust samples exceeding China’s OELs. Nonferrous metal mines had the highest median total dust concentration (1.61 mg/m³) and respirable silica dust concentration (0.17 mg/m³). By product type, the highest median concentrations of total dust, total silica dust, respirable dust, and respirable silica dust occurred separately in sand for building (2.85 mg/m³), silica rock (1.25 mg/m³), zeolite (1.50 mg/m³), and silica rock (0.72 mg/m³). Small mines had the highest median PDE (64.8%), total dust concentration (1.40 mg/m³), total silica dust concentration (0.45 mg/m³), respirable silica dust concentration (0.18 mg/m³), and (n=621, 71.4%) of total dust samples exceeding China’s OELs. By mining method, open-pit mines had a higher median PDE (66.7%), total silica dust concentration (0.43 mg/m³), and percentage (n=597, 65.3%) of total dust samples exceeding China’s OELs; underground mines had higher total dust (1.50 mg/m³) and respirable dust (0.66 mg/m³) concentrations (see Tables 2 and 3 and Table S3 in Multimedia Appendix 3). By mine category and job, winch control workers of nonmetal mines had the highest total dust concentration (3.13 mg/m³) and ratios of total dust concentration (384%) and respirable dust concentration (345.9%) to China’s OELs. Blasters in ferrous metal mines faced the highest respirable dust concentration (2.35 mg/m³). Signal workers in both ferrous and nonferrous metal mines had the lowest dust exposure levels (see Figure S1 in Multimedia Appendix 4 and Figure S2 in Multimedia Appendix 5).

Given that 99.0% of silica dust–exposed miners were male, when predicting the silicosis cumulative HR (H) and incidence (I) over the next 10, 20 and 30 years, we assumed that the gender was male (gender=1). Under total noncoal mine silica dust exposure, the comprehensive risk of silicosis was high (S_weighted of ICMM and INDEX assessments was 3.41 and 3.05, respectively), and the cumulative incidence were predicted to be 6.8%, 25.1%, and 49.9% over the next 10, 20, and 30 years, respectively. Under respirable silica dust exposure, the comprehensive risk of silicosis was medium regardless of the mine category, production scale, or mining method, and the cumulative incidence were predicted to be 6.8%, 27.7%, and 57.4% over the next 10, 20, and 30 years, respectively. Nonmetal miners, nonferrous metal miners, small-noncoal-mine workers, and open-pit noncoal miners had a higher risk and cumulative incidence of silicosis. In addition, packing workers and crushers had a higher risk, while signal workers and strokers had a lower risk (see Tables 4-7, Table S4 in Multimedia Appendix 6, and Table S5 in Multimedia Appendix 7).

By using nonprobability convenience-based sampling, 155 noncoal mines were selected from 10 provinces in northern, western, southern, eastern, and central China, and 126 noncoal mines were included in this study. We found that Chinese noncoal miners still suffer high-level silica dust exposure and medium-level silicosis risk, especially miners in nonmetal, nonferrous metal, small, and open-pit noncoal mines. In addition, the silicosis cumulative incidence caused by total silica dust exposure is predicted to be 6.8%, 25.1%, and 49.9% over the next 10, 20, and 30 years, respectively, and the cumulative incidences caused by respirable silica dust exposure is predicted to be 6.8%, 27.7%, and 57.4%, respectively. These findings are of great reference value for improving occupational health in China, especially for implementing the Occupational Health Protection Action (2019-2030).

Silica is the key component of dust in a mine working environment. Silicosis, a preventable lung disease, has been a serious public health issue worldwide for many decades. The recent outbreaks of silicosis have made governments and scientists frustrated and confused [1,14,40], and high incidences of silicosis in Asia, Africa, and South America are particularly concerning [40]. Occupational health plays an important role in China’s broader health strategy. Assessment of occupational health risk is deemed 1 of the tasks of the health administrative department. Therefore, it is vital to conduct surveillance of the risk of silicosis in Chinese noncoal mines. Many previous medical studies have focused on the effect of noncoal mine silica dust exposure on silicosis morbidity and mortality; however, the visibility of silicosis risk has never been complete or consistent. This study established prediction formulas of quantitative silicosis risk, distinguishing gender and silica dust types. Kaplan-Meier survival analysis showed that male noncoal miners have a higher silicosis probability than female noncoal miners. The finding is consistent with previously published results that men have a 77% higher incidence of silicosis than women [41]. However, we did not find significant difference among different mine categories or smoking histories. This result might possibly be explained by the healthy worker survivor effect. In this cross-sectional study, most silica dust–exposed miners were male. This is consistent with the reality in China. Therefore, when predicting the silicosis cumulative HR and incidence over the next 10, 20, and 30 years, we assumed that the gender was male.

A previous study based on 29 Chinese metal mines and pottery factories indicated that the mean respirable silica dust concentration of the mines fell to less than 0.1 mg/m³ after 1970 and that the concentrations of pottery factories ranged from 0.12 to 0.15 mg/m³ after 1990 [22]. The updated study, based on 1 iron mine, also showed that the mean respirable silica dust concentration gradually decreased to below 0.01 mg/m³ in 2012 [42]. However, survey data of our cross-sectional study showed that 64.2% of total dust samples and 44.2% of respirable dust samples exceeded China’s OELs and that 99.6% of respirable dust samples exceeded the ACGIH REL. In addition, workers in nonmetal, nonferrous metal, small noncoal mines, and open-pit noncoal mines suffered high silica dust exposure levels. This was consistent with results of qualitative and quantitative silicosis risk assessments. The most probable reason for the different findings is the differences in mining technology levels, which would normally be related to the regional economic development in China. Moreover, our survey data were collected from northern, western, southern, eastern, and central China, including nonferrous metal, ferrous metal, and nonmetal mines, not just central and southern China.

It should be noted that the latest WHO/ILO joint estimates indicated that the pooled prevalence of occupational exposure to silica dust in mining was 0.75 and that the pooled level of silica dust was 0.04 mg/m³ [42]. The quality of the prevalence evidence is moderate. More worryingly, the quality of the level evidence is low. Our study showed that Chinese noncoal mines have lower prevalence (61.6%) and higher levels of occupational exposure to silica dust. Furthermore, the overall total dust concentration was 1.30 mg/m³, and the respirable dust concentration was 0.58 mg/m³. It is foreseeable that these findings would be high-quality evidence for both prevalence and level estimation of exposure to silica dust in noncoal mines.

Although this study provides novel findings on the risk of silicosis in Chinese noncoal miners, several limitations need to be considered regarding these findings. The estimation formula of the historical respirable concentration (before 1992) was rough without distinguishing the mining method, mine category, or production scale. Furthermore, only 1 antimony mine and 1 limestone mine were included, limiting the representativeness of these 2 types of mines. Finally, compared to the United States, the accuracy of Chinese sampling procedures might be lower.

An interesting result was observed in the qualitative risk assessment. The qualitative risks of silicosis caused by total silica dust exposure and respirable silica dust exposure were high and medium, respectively. This result demonstrates that both total silica dust and respirable silica dust data are vital for occupational health risk assessment, rather than respirable silica dust data alone. Quantitative risk assessment showed that the silicosis cumulative incidence caused by total silica dust exposure and respirable silica dust exposure would be up to 49.9% and 57.4% over the next 30 years. These findings suggest that silica dust exposure and silicosis risk in Chinese noncoal mines remain a problem. As China has yet to completely fulfill the National Occupational Disease Prevention and Control Plan (2021-2025), these nationally representative data of noncoal mine silica dust exposure would provide valuable evidence for improving miners’ working environments and reducing the risk of silicosis. Effective control measures for reducing noncoal mine silica dust levels, especially primary prevention strategies, are critical to achieve the goal of improving occupational health in China.