2014 Volume 56 Issue 6 Pages 444-452
Objective: Few studies have systematically investigated the impact of past occupational dust exposure on mental health. We examined whether retired factory workers exposed to any of the 4 dusts of silica, cement, coal and asbestos had more depressive symptoms and anxiety in southern China, which has experienced rapid economic development. Methods: We used data from the Guangzhou Biobank Cohort Study phase 3. Exposures, lifestyle, symptoms and medical history of the participants were assessed with a structured interview. Self-reported intensity and duration of past occupational dust exposure were used to derive cumulative exposure. Outcome measures were assessed by the 15-item Chinese version of the Geriatric Depression Scale (score ≥5) and the single-item on anxiety. Results: The results revealed that 359 workers were exposed to at least one of the 4 dusts and that 1,253 were unexposed (controls). After adjustment of multiple confounders, greater risks of depressive symptoms were associated with high exposure to silica (odds ratio (OR) of 3.12, 95% CI of 1.17–8.31) and asbestos (OR of 6.90; CI of 1.29–36.75). Risks of anxiety were higher in those with low or high exposures to dust (OR of 2.01 and CI of 1.04–3.87 and OR of 2.29 and CI of 1.30–4.03, respectively) and cement (OR of 3.20 and CI of 1.27–8.07 and OR of 2.30 and CI of 1.09–4.87, respectively), and those with high exposure to silica (OR of 5.29, CI of 1.76–15.92). Conclusions: Past occupational exposures to silica, cement, coal and asbestos dusts were associated with adverse mental health outcomes in retired factory workers. The mechanism underlying the relationship between occupational exposures and psychological symptoms in later life needs to be further studied.
(J Occup Health 2014; 56: 444–452)
Silica, cement, coal and asbestos dusts are well- known fibrogenic dusts that have been ubiquitous in many occupational activities for centuries. Dust-related respiratory diseases and disabilities have been most commonly recognized in both developed and developing countries1). Pneumoconiosis is the most serious occupational lung disease in China, not only because of its high morbidity and mortality, but also because of the large number of workers at high risk2). About 12 million workers are occupationally exposed to fibrogenic dusts in China3). For many decades, numerous studies have focused on radiographic abnormalities, physiologic changes and clinical manifestations in workers exposed to fibrogenic dusts1). Studies on Chinese patients from Hong Kong and mainland China have reported that patients with pneumoconiosis had a high prevalence of depressive symptoms4, 5). However, previous data on the mental health of dust exposed workers with pneumoconiosis are limited. In the US, one study found that older construction workers were significantly more susceptible to chronic lung disease such as chronic bronchitis or emphysema, musculoskeletal problems and emotional disorders6). The results indicated that the high rate of chronic lung disease was most likely related to on-the-job dust exposure, while the increased risk of emotional disorders might be due to the dynamics of the construction labor market. Another study found that New York City transit workers exposed to the September 11th disaster, and to the "dust cloud", had a significantly higher risk of depressive symptoms, with odds ratios (ORs) of 2.31 (95% CI: 1.04, 5.15) and 2.48 (95% CI: 1.12, 5.51), respectively7).
Depression and anxiety are the two most common mental health problems seen in general medical settings8). Depression is also present in significant proportions of the elderly in the community, and known risk factors for depression that are common in the elderly include psychosocial deprivation and loss, multiple chronic illnesses, poor health and functional disability9). There is a wide variation in the prevalence of anxiety and depression by place, even after adjusting for sociodemographic characteristics, such as sex, age, race or ethnicity, education, marital status and employment status10). Higher rates (>10%) of current depressive symptoms were found in Alabama, Arkansas, Michigan, Mississippi, Oklahoma, Tennessee, West Virginia and Puerto Rico, whereas lower rates (<6%) were observed in Connecticut, Iowa, Nebraska and North Dakota10). Regional variation was also found in lifetime diagnoses of depression and anxiety: the former ranged from 6.8 to 21.3%, and the latter ranged from 5.4 to 17.2%10). The examination of regional variability is important both to better understand risk factors for depression and anxiety and to target public health efforts and resources for detection and management10). Most research on environmental risk factors including airborne particulate and water- and foodborne toxins and on cultural, social and physical factors has focused on the social environment. Few studies1, 4–7) on environmental risk factors have focused on the occupational environment. Investigations of occupational environmental risk factors could contribute to our understanding of etiology and identification of vulnerable groups.
The purpose of this study was to examine the associations of depressive symptoms and anxiety with past occupational exposure to silica, cement, coal and asbestos dusts utilizing data from a sample of retired Chinese factory workers.
The Guangzhou Biobank Cohort Study (GBCS) is an ongoing collaboration between the Guangzhou No. 12 Hospital and the Universities of Hong Kong and Birmingham. Detailed descriptions of the methods have been reported elsewhere11). Briefly, participants were recruited from “The Guangzhou Health and Happiness Association for the Respectable Elders (GHHARE)”, which has a membership of more than 100,000 older Guangzhou permanent residents and is a community social and welfare association supported by the municipal government. Its membership is open to anyone aged 50 years or older for a monthly nominal fee of 4 Yuan (50 US cents). More than 95% of the selected members were eligible, and all participants gave written informed consent to participate in the GBCS. Recruitment for the cohort study was conducted in three phases (2003–2004, 2005–2006 and 2006–2008), each drawing upon 10% (about 10,000) of the members. The current analysis was limited to data from retired factory workers in the last recruitment phase (phase 3, n=10,088), which included the mental health measures of interest needed for the present analysis. The eligibility criteria were ability to consent, ability to walk, and no current treatments that, if omitted, could result in an immediate life-threatening risk. The study received approval from the Medical Ethics Committee of the Guangzhou Medical Association. The present analysis was based on participants who were factory workers who self-reported they were exposed to silica, cement, coal and asbestos dusts in the job they had held the longest and had mental health measurements, an area that has not been sufficiently investigated. We focused on these 4 dusts because they can cause pneumoconiosis, which is recognized by Chinese law as an occupational disease. Of the 3,759 subjects who self-reported that they had served longest as factory workers, 762 who reported exposure to chemical or physical agents but no dusts were excluded. Of the 1,613 who reported exposure to dusts, 396 reporting exposure to at least one of the 4 dusts examined in this study (97 workers to silica, 176 to cement, 89 to coal and 34 to asbestos) were eligible as the exposed group, while 1,217 workers were excluded because of exposure to other dusts including chemical fibers, man-made fibers and other inorganic dusts. The 1,384 workers who reported no exposure were eligible for the control group. Of the 1,780 (396 + 1,384) workers above, 168 were excluded because of missing data for some variables, and 1,612 with complete information were analyzed. Of the 1,612 workers, 359 reported exposure to at least one of the 4 dusts (89 workers to silica, 155 to cement, 83 to coal and 32 to asbestos), and 1,253 workers reported no exposure (controls). Of the 359 subjects who reported dust exposure, 33% were from construction material manufacturing, 26% were from the construction industry, 20% were from coal mining, and the others were from machinery manufacturing (6.9%), the metallurgical industry (4.7%), the electric power industry (4.7%) and the transportation industry (4.7%); 113 also reported exposure to chemical agents such as organic solvents and metals, and 170 reported exposure to physical agents such as noise and high temperatures. All workers reported that they did not have pneumoconiosis, chronic obstructive pulmonary disease, tuberculosis, asthma, cardiopulmonary disease or any other chest diseases. Figure 1 presents a detailed description of subject selection.
Participant recruitment in phase 3 of the Guangzhou Biobank Cohort Study.
Participants underwent a detailed interview by trained interviewers using a standardized computer-based questionnaire, followed by anthropometric and medical examinations. They were asked if they had been exposed to dust in the job they had held the longest and to specify the intensity (mild, moderate, heavy) and duration (number of years) of exposure. A cumulative exposure index was calculated by grouping subjects into two categories based on both exposure duration and exposure intensity12, 13). Participants were classified as having low exposure if they reported mild exposure for less than 26 years or high cumulative exposure resulting from moderate or heavy exposure for 26 years or more. This classification depended entirely on the individual's perception of exposure intensity without consideration of the relative levels of exposure across occupations12). Occupational histories were reviewed by two occupational physicians with attention to consistency between job types and exposure statements.
Depression and anxietyDepressive symptoms were assessed by the Chinese version of the 15-item Geriatric Depression scale (GDS). The total score of the scale can range from 0 to 15, with higher scores indicating higher levels of depressive symptoms. A total score of ≥5 on the GDS is considered to be indicative of depression with a sensitivity of 96.3% and specificity of 87.5%9, 14–16). In the present study, the Cronbach's alpha was 0.70. Therefore, the presence of depressive symptoms was defined as a score of 5 or higher in line with findings from validation studies9, 16).
Anxiety disorder was identified with a question about excessive worries or tense or anxious feelings lasting for at least one month during the previous 12 months answered with a yes or no. The question asked was "In the previous 12 months, did you have any of the following: feeling worried, tense or anxious most of the time over a period of at least one month".
Other information and measurementsData for sociodemographic characteristics and psychosocial variables included age, gender, education, marital status, years after retirement and personal annual income in the last year. Smoking history was classified into former smokers, current smokers and never smokers. Physical activity was assessed using the International Physical Activity Questionnaire (IPAQ) short form, which was validated in the GBCS17). Self-rated health status was assessed by a subjective rating scale (very good, good, poor or very poor), which was dichotomized into good or poor. Chronic cough or phlegm was defined as usually having cough or phlegm first thing in the morning or either during the day or at night. Dyspnea was defined as being troubled by shortness of breath when hurrying on level ground or walking up a slight slope.
Participants were asked to report past medical illnesses diagnosed and treated by a doctor including psychiatric illness, pulmonary disease, coronary artery disease, hypertension, dyslipidemia, diabetes mellitus, stroke, hypothyroidism, hip fracture, arthritis, cataract and any other conditions they had.
Potential confounders were age, years after retirement, sex, educational level, marital status, smoking status, physical activity, current annual personal income, chronic cough or phlegm, dyspnea, chemical exposure, physical exposure and health status because of their known associations with mental health status9, 18–21).
Statistical analysisWe used chi-square tests to compare categorical variables between the exposed and unexposed workers and t-tests and one-way analysis of variance for continuous variables. Multivariable linear regression was used to assess the association of occupational exposure with GDS score, and adjusted means and 95% confidence intervals (CIs) were reported. Logistic regression models were used to examine the association of occupational exposure with depressive symptoms and anxiety disorder. Sensitivity analysis was performed by excluding former and current smokers. Statistical analyses were performed using SPSS version 20.0. All tests were two-sided, and a p value of <0.05 was considered as statistically significant.
The mean age of the 359 participants exposed to at least one of the 4 dusts was 60.9 (8.3) (mean (SD)) years, and the mean duration of employment was 23.2 (9.1) years. Table 1 shows that dust exposure was associated with sex (men>women), lower educational level, smoking, physical inactivity, chronic cough or phlegm and dyspnea. The prevalences of depressive symptoms and anxiety disorder were 20.6% and 14.5% among the exposed group and 14.8% and 7.9% among the unexposed group, respectively (p<0.01).
| Characteristics | Unexposed workers | Workers exposed to dust | p |
|---|---|---|---|
| N=1,253 | N=359 | ||
| Sex; n (% ) | <0.001 | ||
| Women | 1,035 (82.6) | 239 (66.6) | |
| Men | 218 (17.4) | 120 (33.4) | |
| Age (years); mean (SD) | 60.3 (8.2) | 60.9 (8.3) | 0.07 |
| 50− | 705 (56.3) | 188 (52.4) | 0.34 |
| 60− | 342 (27.3) | 102 (28.4) | |
| 70− | 206 (16.4) | 69 (19.2) | |
| Years after retirement (years); mean (SD) | 9.8 (7.9) | 10.0 (8.1) | 0.61 |
| Educational level; n (%) | 0.004 | ||
| <Primary | 132 (10.5) | 58 (16.2) | |
| Primary | 400 (31.9) | 128 (35.7) | |
| Junior middle | 426 (34.0) | 106 (29.5) | |
| ≥Senior middle | 295 (23.5) | 67 (18.7) | |
| Marital status; n (%) | 0.72 | ||
| Married | 1,005 (80.2) | 291 (81.1) | |
| Separated, divorced, widow, never | 248 (19.8) | 68 (18.9) | |
| Smoking; n (%) | <0.001 | ||
| Never | 1,081 (86.3) | 263 (73.3) | |
| Ex | 80 (6.4) | 43 (12.0) | |
| Current | 92 (7.3) | 53 (14.8) | |
| Physical activity (IPAQ)a; n(%) | 0.03 | ||
| Inactive | 104 (8.3) | 19 (5.3) | |
| Minimally active | 344 (27.5) | 119 (33.1) | |
| Active | 805 (64.2) | 221 (61.6) | |
| Current annual personal income (Yuan); (%) | 0.85 | ||
| ≤10,000 | 312 (24.9) | 95 (26.5) | |
| 10.15,000 | 712 (56.8) | 200 (55.7) | |
| >15,000 | 177 (14.1) | 52 (14.5) | |
| Don't know | 52 (4.2) | 12 (3.3) | |
| Hypertension; n (% yes) | 480 (38.3) | 157 (43.7) | 0.06 |
| Diabetes; n (% yes) | 129 (10.3) | 27 (7.5) | 0.12 |
| Hyperlipidemia; n (% yes) | 617 (49.2) | 182 (50.7) | 0.63 |
| Self-rated health; n (% poor) | 272 (21.7) | 72 (20.1) | 0.50 |
| Chronic cough/phlegm; n (% yes) | 130 (10.4) | 62 (17.3) | <0.001 |
| Dyspnea; n (% yes) | 433 (34.6) | 152 (42.3) | 0.01 |
| GDS score*; mean(SD) | 2.4 (2.3) | 2.9 (2.6) | 0.001 |
| Depressive symptoms (GDS.5); n (% yes) | 185 (14.8) | 74 (20.6) | 0.01 |
| Self-rated anxietyb; n (% yes) | 99 (7.9) | 52 (14.5) | <0.001 |
Table 2 shows that larger percentages of men were exposed for ≥26 years (45.8% vs. 25.1%) and had greater cumulative dust exposures compared with women (76.7% vs. 66.1%).
| Total | Women | Men | p | |
|---|---|---|---|---|
| Any dust exposure; n | 359 | 239 | 120 | |
| Other exposure | ||||
| Chemical exposure; n | 113 | 61 | 52 | |
| Physical exposure; n | 170 | 106 | 64 | |
| Dust exposure intensity; n (%) | 0.68 | |||
| Mild | 150 | 98 (41.0) | 52 (43.3) | |
| Moderate | 115 | 75 (31.4) | 40 (33.3) | |
| Heavy | 94 | 66 (27.6) | 28 (23.3) | |
| Dust exposure duration; n (%) | 0.01 | |||
| <26 years | 252 | 179 (74.9) | 73 (60.8) | |
| ≥26 years | 107 | 60 (25.1) | 47 (45.8) | |
| Dust cumulative exposure*; n (%) | 0.04 | |||
| Low | 109 | 81 (33.9) | 28 (23.3) | |
| High | 250 | 158 (66.1) | 92 (76.7) | |
| Dust category | ||||
| Silica; n | 89 | 57 | 32 | |
| Cement; n | 155 | 112 | 43 | |
| Coal; n | 83 | 45 | 38 | |
| Asbestos; n | 32 | 25 | 7 |
Despite the sex difference in exposure, there was no evidence that the relationship between any dusts exposure and mental health outcomes varied with sex (p for interaction for depressive symptoms and anxiety disorder=0.80 and 0.54, respectively). The results are therefore presented for both sexes together.
Table 3 shows a dose-response relationship between GDS score and silica exposure after adjusting for age, years of retirement, sex, educational level, marital status, smoking, physical activity, personal income, chronic cough or phlegm, dyspnea, chemical exposure, physical exposure and self-rated health (p for trend: 0.002). Sensitivity analysis on never smokers only showed significant trends for silica after adjusting for multiple potential confounders.
| No exposure | Low exposure | High exposure | p for trend | |
|---|---|---|---|---|
| Number of subjects, any dust* | 1,253 | 109 | 250 | |
| Crude model | 2.44 (2.32–2.57) | 2.72 (2.26–3.19) | 2.97 (2.63–3.30) | 0.004 |
| Adjusted model† | 2.47 (2.33–2.60) | 2.62 (2.18–3.06) | 2.88 (2.52–3.24) | 0.16 |
| Dust category | ||||
| Number of subjects: silica | 1,253 | 41 | 48 | |
| Crude model | 2.44 (2.32–2.57) | 2.44 (1.76–3.12) | 3.94 (3.04–4.84) | <0.001 |
| Adjusted model† | 2.43 (2.31–2.56) | 2.50 (1.77–3.23) | 4.08 (3.16–5.00) | 0.002 |
| Number of subjects: cement | 1,253 | 45 | 110 | |
| Crude model | 2.44 (2.32–2.57) | 3.20 (2.34–4.06) | 2.67 (2.25–3.09) | 0.06 |
| Adjusted model† | 2.47 (2.35–2.60) | 3.01 (2.31–3.71) | 2.40 (1.88–2.93) | 0.26 |
| Number of subjects: coal | 1,253 | 14 | 69 | |
| Crude model | 2.44 (2.32–2.57) | 2.64 (1.27–4.01) | 2.64 (1.89–3.38) | 0.75 |
| Adjusted model† | 2.44 (2.32–2.57) | 2.31 (1.09–3.53) | 2.68 (1.89–3.48) | 0.79 |
| Number of subjects: asbestos | 1,253 | 9 | 23 | |
| Crude model | 2.44 (2.32–2.57) | 1.78 (0.71–2.85) | 3.35 (2.41–4.28) | 0.11 |
| Adjusted model† | 2.43 (2.31–2.55) | 2.06 (0.53–3.58) | 3.89 (2.54–5.25) | 0.06 |
| Number of never smokers, any dust | 1,081 | 86 | 177 | |
| Crude model | 2.41 (2.27–2.54) | 2.83 (2.28–3.38) | 3.14 (2.72–3.56) | <0.001 |
| Adjusted model# | 2.44 (2.30–2.59) | 2.72 (2.21–3.22) | 2.96 (2.53–3.39) | 0.10 |
| Number of never smokers, silica | 1,081 | 31 | 34 | |
| Crude model | 2.41 (2.27–2.54) | 2.39 (1.57–3.21) | 4.44 (3.27–5.61) | <0.001 |
| Adjusted model# | 2.42 (2.28–2.56) | 2.32 (1.49–3.16) | 4.03 (2.94–5.13) | 0.01 |
a, b, c, in one column indicate a significant difference (P < 0.05).
Table 4 shows that the adjusted ORs for depressive symptoms in the highest exposure category were 3.12 (95% CI 1.17–8.31) for silica and 6.90 (1.29–36.75) for asbestos.
| Exposure index | n | Depressive disorder | Anxiety disorder | ||||
|---|---|---|---|---|---|---|---|
| N (%) | Crude OR (95% CI) | Adjusted OR (95% CI) | N (%) | Crude OR (95% CI) | Adjusted OR(95% CI) | ||
| No exposure | 1,253 | 185 (14.8) | 1.00 (reference) | 1.00 (reference) | 99 (7.9) | 1.00 (reference) | 1.00 (reference) |
| Any dust† | |||||||
| Low | 109 | 22 (20.2) | 1.46 (0.89–2.39) | 1.38 (0.79–2.43) | 15 (13.8) | 1.86 (1.04–3.33)* | 2.01 (1.04–3.87)* |
| High | 250 | 52 (20.8) | 1.52 (1.08–2.14)* | 1.47 (0.89–2.42) | 37 (14.8) | 2.03 (1.35–3.04)** | 2.29 (1.30–4.03)** |
| p for trend | 0.03 | 0.27 | 0.001 | 0.01 | |||
| Dust category | |||||||
| Silica† | |||||||
| Low | 41 | 9 (22.0) | 1.62 (0.76–3.46) | 1.54 (0.64–3.69) | 4 (9.8) | 1.26 (0.44–3.61) | 1.42 (0.44–4.57) |
| High | 48 | 17 (35.4) | 3.17 (1.72–5.84)** | 3.12 (1.17–8.31)* | 10 (20.8) | 3.07 (1.48–6.34)** | 5.29 (1.76–15.92)** |
| p for trend | 0.001 | 0.07 | 0.01 | 0.01 | |||
| Cement† | |||||||
| Low | 45 | 11 (24.4) | 1.87 (0.93–3.75) | 1.41 (0.59–3.36) | 8 (17.8) | 2.52 (1.14–5.56)* | 3.20 (1.27–8.07)* |
| High | 110 | 19 (17.3) | 1.21 (0.72–2.02) | 0.79 (0.36–1.72) | 16 (14.5) | 1.98 (1.12–3.50)* | 2.30 (1.09–4.87)* |
| p for trend | 0.18 | 0.46 | 0.01 | 0.02 | |||
| Coal† | |||||||
| Low | 14 | 2 (14.3) | 0.96 (0.21–4.33) | 0.96 (0.19–4.93) | 3 (21.4) | 3.18 (0.87–11.58) | 1.90 (0.38–9.58) |
| High | 69 | 11 (15.9) | 1.09 (0.56–2.13) | 1.37 (0.49–3.79) | 6 (8.7) | 1.11 (0.47–2.63) | 0.48 (0.08–2.89) |
| p for trend | 0.96 | 0.82 | 0.21 | 0.33 | |||
| Asbestos† | |||||||
| Low | 9 | 0 (0.0) | 0.00 (N/A) | 0.00 (N/A) | 0 (0.0) | 0.00 (N/A) | 0.00 (N/A) |
| High | 23 | 5 (21.7) | 1.60 (0.59–4.37) | 6.90 (1.29–36.75)* | 5 (21.7) | 3.24 (1.18–8.91)* | 4.52 (0.77–26.55) |
| p for trend | 0.65 | 0.08 | 0.08 | 0.25 | |||
| Never smokers only#, any dust | |||||||
| No exposure | 1,081 | 155 (14.3) | 1.00 (reference) | 1.00 (reference) | 88 (8.1) | 1.00 (reference) | 1.00 (reference) |
| Low | 86 | 19 (22.1) | 1.69 (0.99–2.90) | 1.49 (0.80–2.77) | 12 (14.0) | 1.83 (0.96–3.50) | 2.03 (0.98–4.22) |
| High | 177 | 44 (24.9) | 1.98 (1.35–2.89)** | 1.62 (0.94–2.82) | 28 (15.8) | 2.12 (1.34–3.36)** | 2.33 (1.23–4.40)** |
| p for trend | 0.001 | 0.16 | 0.002 | 0.02 | |||
| Never smokers silica# | |||||||
| Low | 31 | 7 (22.6) | 1.74 (0.74–4.11) | 1.50 (0.56–4.01) | 2 (6.5) | 0.78 (0.18–3.32) | 0.95 (0.20–4.45) |
| High | 34 | 15 (44.1) | 4.72 (2.35–9.48)** | 2.69 (0.85–8.51) | 8 (23.5) | 3.47 (1.53–7.89)** | 5.61 (1.49–21.07)* |
| p for trend | <0.001 | 0.23 | 0.01 | 0.03 | |||
†Adjusted for age, years after retirement, sex, education level, marital status, smoking, physical activity, personal income, chronic cough or phlegm, dyspnea, chemical exposure, physical exposure and self-rated health. #Adjusted for age, years after retirement, sex, education level, marital status, physical activity, personal income, chronic cough or phlegm, dyspnea, chemical exposure, physical exposure and self-rated health. Logistic regression (enter) was used to assess the association of the exposure to dust with depressive symptoms and anxiety disorder. *p<0.05; **p<0.01.
The risks of anxiety were higher in those with low or high exposures to any dust (OR of 2.01 and CI of 1.04–3.87 and OR of 2.29 and CI of 1.30–4.03, respectively) and cement (OR of 3.20 and CI of 1.27–8.07 and OR of 2.30 and CI of 1.09–4.87, respectively) and with high exposure to silica (OR of 5.29, CI of 1.76–15.92). Sensitivity analysis showed an adjusted OR of 2.33 (95% CI 1.23–4.40) for any dust and 5.61 (1.49–21.07) for silica in never smokers.
To our knowledge, this is the first study to show that depressive symptoms and anxiety disorder were associated with past occupational dust exposure in retired Chinese factory workers without pneumoconiosis after adjusting for multiple confounders.
For depressive symptoms, the risk was the greatest with the highest estimated cumulative exposure to asbestos dust, followed by silica dust, which suggested that asbestos or silica dusts were more likely to lead to, or be followed by, depression compared with other kinds of dusts.
For anxiety disorders, the adjusted odds ratios were the greatest for silica, followed by any of the 4 dusts, and cement, suggesting that silica or cement dusts were more likely to lead to, or be followed by, anxiety disorders.
Smoking is recognized widely as a major risk factor for respiratory impairment in dust-exposed workers and is strongly associated with depression and anxiety disorders20). Our sensitivity analysis, by excluding current and ex-smokers, also showed in significantly increased odds for anxiety disorder with high exposure to silica. Due to the small number of cases, we could not examine reliably other specific dust exposures in never smokers only.
The mechanism of the effect of dusts on the central nervous system and psychiatric symptoms remains unclear. There are some possible explanations for more depressive symptoms and anxiety among our subjects with greater cumulative exposure. The etiological factors may be attributed to the complex work settings. Simple consideration of single dust hazards is inadequate. Many workers are simultaneously exposed to multiple hazards, such as lead, arsenic, radon, cadmium, aluminum, polycyclic aromatic hydrocarbons and other impurities, resulting in synergistic effects that may have greater influences on health than if confronted with single hazards2, 22).
Another explanation could be the result of social and psychological factors. Our subjects were worked under very hazardous conditions, with exposure to silica, cement, coal or asbestos, before retirement. According to Chinese occupational health regulations, workers are required or entitled to participate in periodic medical surveillance, early preclinical detection of adverse health effects and services by proper medical facilities for disease diagnosis, treatment and rehabilitation23). National regulations also specify the process for compensation and pensions for workers with occupational diseases. Workers who are diagnosed with an occupational disease should be fully covered by medical insurance and should be compensated by their employers according to national law23). Hence, workers who are exposed to hazardous dusts have periodic medical examinations, including chest radiography for early detection of pneumoconiosis, and prompt treatment or removal from further exposure as a secondary preventive measure2). Susceptible workers may change jobs or leave the workforce permanently after heavy brief exposure, and they may develop chronic respiratory disease after leaving hazardous jobs or may die sooner than workers in other occupations13). Of note, our subjects were found to be free of abnormal signs on chest X-ray examination when they left state-owned or municipal collective enterprises. Generally, chest X-ray was done about every 2 or 3 years for workers exposed to dust, which was usually funded by the enterprises. Nevertheless, with the recent nationwide economic reforms, to save costs or because of financial hardships, many enterprises have discontinued periodic medical examinations, and early detection or early diagnosis of occupational disease has become impracticable2); also, follow-up medical examinations for retired workers are not done. Thus, it is possible that some exposed workers might feel abandoned, show anxiety, with fear of occupational lung diseases and worries about welfare benefits for their illnesses, and become depressed even after exposure ended. However, whether the participants had knowledge of the adverse health effects of dust exposure, such as occupational lung cancer, was not assessed during the interview. Of the 359 participants with past occupational exposure to dusts, most were exposed to silica (25%), cement (43%) or coal (23%) dusts. Because lung cancer caused by these dusts has not been recognized as a statutory occupational lung disease in China23), our participants' awareness or knowledge of the potential carcinogenic effects would be low and could not be the reason for their depressive and anxiety symptoms. Further studies are warranted to clarify the exact mechanism.
Depression and anxiety are associated with impaired health-related quality of life and social functioning24), as well as excess disability8, 25). Given these associations, using an integrated approach to assess mental health and to provide health education and mental health services, including counseling and other therapies, dust-exposed workers is needed.
The present study had several limitations. First, the cross-sectional design of the study could not establish the direction of causality between dust and depressive and anxiety disorders and the effect of behavioral mediators. Thus, we cannot rule out the possibility that exposure to dusts led to behavioral changes that increased depressive or anxiety disorders. Our sensitivity analysis on never smokers could exclude the confounding effect of smoking on the association. Second, the classification of occupational exposure was crude. There could have been random misclassification of exposure arising from self reporting, and this could have attenuated the strength of the association. Third, bias could have arisen from differential reporting of occupational exposure due to presence of depression or anxiety. However, it was unlikely that our participants with mood disorders exaggerated the intensity and duration of dust exposure specifically, as we did not observe such an association for other chemical and physical exposures (data not shown). Fourth, we used simple measures for the outcomes, with no psychiatric confirmation. Finally, misclassification would be more likely to have come from underestimation of mood disorders, which would lead to an underestimation of the strength of the association.
The relationships between occupational dust exposure and ill mental health are complex. To understand the complex pattern of interactions leading to adverse mental health outcomes among silica-, cement-, coal- and asbestos-exposed workers, further research with more detailed and accurate information about occupational environmental dust levels, individual exposures and psychiatric diagnoses is warranted.
Our results showed a significant association between past dust exposure in the workplace and depressive symptoms and anxiety disorder with dose-response relationships in older retired Chinese factory workers. Occupational exposures to these dusts may result later in life in adverse mental health outcomes. However, the mechanism of the relationship between occupational exposures and psychological symptoms in later life is unclear, and further studies are needed to clarify and confirm our results. Nevertheless, occupational health physicians should enquire about such symptoms during periodic examination.
Acknowledgments: The Guangzhou Biobank Cohort Study was funded by the University of Hong Kong Foundation for Educational Development and Research, Hong Kong, the Guangzhou Public Health Bureau and the Guangzhou Science and Technology Bureau, Guangzhou, P. R. China, and the University of Birmingham, UK. We thank the Guangzhou Health and Happiness Association for the Respectable Elders for recruiting the participants.
