Article Text
Abstract
Objective To determine whether occupational exposure to silica dust causes an increased risk of developing Crohn’s disease (CD) and ulcerative colitis (UC).
Design Case–control study of CD (K50) and UC (K51) from 2007 through 2016. Controls were matched to cases (2:1) based on age, sex and county at the time of diagnosis. A job exposure matrix was used to estimate the occupational silica exposure of all cases and controls.
Setting Medical and occupational data from the National Outpatient Register were used to implement a case–control analysis, while the two controls used for each case were selected from the National Register of the Total Population.
Participants All men and women aged 20–65 years old who were diagnosed with CD (K50) and UC (K51) during the years of study were included and assigned two controls, resulting in 58 136 cases and 116 272 controls.
Main outcomes Silica dust exposure correlates with an increased risk of developing UC in men and CD in women.
Results The prevalence of UC was significantly higher in the group exposed to silica dust (OR 1.13, 95% CI 1.06 to 1.21) than in controls, particularly in individuals with over 5 years exposure. When stratified by sex, a significantly increased OR was detected for men (OR 1.33, 95% CI 1.05 to 1.22). This trend was also consistent with longer exposure times. The prevalence of UC was not increased in exposed women. The prevalence of CD was significantly increased among exposed women (OR 1.29, 95% CI 1.01 to 1.65), but not for exposed men.
Conclusions Silica dust exposure correlates with an increased risk of developing UC, especially in men, and the risk seems to increase with the duration and degree of exposure. Conversely, silica dust exposure correlates positively with the risk of developing CD in women.
- inflammatory bowel disease
- occupational & industrial medicine
- public health
- epidemiology
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Strengths and limitations of this study
This case–control study includes anyone who was diagnosed with Crohn’s disease and ulcerative colitis in Sweden in the years 2007–2016, and not just a selection.
Sweden maintains high-quality registers that cover the entire population, together with unique personal identification numbers that can link patient data across different nationwide registers.
The diagnoses were based on the data recorded in the national non-primary outpatient visits register, which is significantly more accurate than diagnoses based on questionnaires.
This study lacks information on potential confounders such as smoking habits; however, cases and controls are matched based on age, sex and geographical area and therefore one could assume the distribution of these confounders among the cases and controls.
Introduction
Silica is a mineral composed of oxygen and silicon that is highly abundant in the earth’s crust and is a component of most types of rocks. Exposure to silica dust is therefore common during, for example, mining operations. Silica’s high melting point, hardness and chemical inertness make it ideal for several industrial applications. During the latter half of the 20th century, extensive efforts were made to reduce the amount of silica dust in Swedish workplaces.1 Nevertheless, there is still significant occupational silica exposure in Sweden and abroad.2 3
Exposure to silica dust is known to cause silicosis which is predominantly a pulmonary disease.4 However, recent studies suggest a correlation between silica and several inflammatory diseases.5 6 The pathological mechanisms underpinning these relationships are poorly understood.7 8 No relationship between silicosis and the inflammatory bowel diseases (IBDs), Crohn’s disease (CD) or ulcerative colitis (UC), has previously been reported.
The term IBD encompasses many diseases of the gastrointestinal tract. However, in this paper IBD refers only to CD or UC. CD and UC share several pathological features: they are both chronic and have symptoms including diarrhoea, hematochezia, abdominal pain, fever, weight loss and lesion formation. However, the distribution of lesions differs between the two diseases. CD may affect the entire gastrointestinal tract but the distribution of lesions is often discontinuous. Conversely, the UC lesions start in the rectum and are continuous. While these two IBDs share several risk factors, smoking has opposing effects on the risks of their development: it increases the risk of developing CD but reduces that of UC.9 10
The aetiology of IBD has been studied extensively but remains somewhat unclear. There is strong evidence of a familial component to its development,9–12 and a significant effect of environmental factors.12 This is further supported by the observation that the global prevalence of IBD has risen as increasing numbers of countries outside of Europe and North America have adopted western lifestyles.13
The identity of the environmental factors that may influence IBD pathogenesis is a subject of ongoing debate.11 14 The intake of various airborne particles has been posited to contribute to IBD development despite a lack of direct evidence.15 16 Two early studies suggested silica dust particles may contribute to the aetiology of IBD because a fraction of the inhaled dust is swallowed with the mucus.17 One of these studies have reported that silica dust was detected in the intestinal epithelium and distributed systemically after mice were fed a diet enriched with silica dust.18 The other study showed that injecting silica dust directly into the intestinal lymphatic system of dogs causes the formation of enteric lesions similar to those associated with IBD.19
IBD has a significant impact on patients’ quality of life and an economic impact on society because it increases hospitalisation and mortality rates.20 Furthermore, the prevalence of IBD in Sweden is increasing more rapidly than previously expected,21 making it important to identify agents that cause or exacerbate this disease.
Aim
To determine whether silica dust exposure increases the risk of developing CD or UC.
Material and methods
Registers, administrative authorities and job exposure matrix (JEM)
The Swedish administrative authority known as the National Board of Health and Welfare (NBHW) maintains and validates the ‘National Non-Primary Outpatient Care Register’ (OPR). This register is part of the National Patient Register and has been maintained since 2001. It contains medical data from registered outpatients of Swedish healthcare facilities. The NBHW also maintains the Swedish Cause of Death Register.
Another administrative authority, the Swedish Central Bureau of Statistics (SCB), maintains the Longitudinal Integration Database for Health Insurance and Labor Market Studies (LISA) as well as the National Register of the Total Population (RTB) and the Multi-Generation Register (MGR). LISA holds records going back to 1990 including occupational and domicile information on all individuals resident in Sweden aged 16 years or above, registered as of 31 December in the relevant reference year. The RTB is an extract from the civil register of the Swedish Tax Agency and contains residential data. MGR contains information on individuals’ biological relatives.
All data preparation and register matching for this paper was done by the SCB and NBHW with guidance from a statistician working for the Department of Occupational and Environmental Medicine. The processed data were de-identified such that the only personally identifying information remaining in the processed dataset were dates of birth and years of employment. Other information relating to the identities of individuals represented in the data was replaced with a serial number. All statistical analyses were performed using these de-identified data. A JEM was used to estimate the rates of occupational exposure to silica based on individuals’ job assignments.22 23
Study design
Data from the registers were used to conduct a case–control analysis. The cases were selected from the OPR by the NBHW. The inclusion criteria for the cases were a single diagnosis of one of the following conditions (ICD-10-CM codes are given in parentheses): CD (K50) and UC (K51). The study participants were required to be between 20 and 65 years old at the time of diagnosis and must have been diagnosed between 2005 and 2016.
Matching controls for the case–control study were randomly selected by the SCB using data from RTB and MGR. The inclusion criteria for the controls were having no previous diagnosis of any condition used to select the positive cases as well as not having sarcoidosis (D86), ankylosing spondylitis (M45), seropositive rheumatoid arthritis (M05) or other rheumatoid arthritis (M06) and not being a first-degree relative of the corresponding case. Controls were also required to share the gender, age and county of residence at the time of diagnosis of the matched case (figure 1).
Occupational data for all case and control individuals were obtained from LISA. The JEM was then used to determine whether the occupational history of each case and control would have resulted in silica dust exposure as well as to assess the cumulative exposure. Cases not exposed to silica any time the last 5 years before diagnosis were excluded (figure 1). Jobs among the cases and controls that according to the JEM were classified as containing exposure to silica included concrete workers, casters, masons, ceramic and glass manufacturers, miners and so on.
The cases were selected from the years 2005 to 2016, with a washout period from 2005 to 2006 in order to include recent diagnoses and avoid including data from follow-up medical examinations.
Statistical analysis
The odds of being exposed to silica dust before the time of diagnosis were calculated for cases and controls and expressed as ORs. Being exposed to silica dust was defined based on JEM data as being employed in an environment where silica dust was present. A conditional logistic regression was used to find the 95% CI for the OR. The OR was considered to be significantly greater than 1.00 if the lower limit of the 95% CI was above 1.00.
The study population was stratified according to the duration of exposure and sex. The stratification for duration of exposure was divided into the following time frames: 1–5 years, 5–10 years and more than 10 years. Based on the JEM, a cumulative exposure to silica was also calculated with regards to both prevalence and level of exposure for the different jobs. The cumulative exposure was stratified into 0.01–0.99 mg/m3 and more than 1 mg/m3.
Standardised mortality ratios (SMRs) were used to compare UC and CD cases and controls to the general Swedish population. The data were stratified according to gender.
Results
Study population
A total of 58 136 cases and 116 272 controls where included in this study. Figure 1 shows a flowchart of the study population. For UC, 19 830 cases were included in this study with a mean age of 42.9±12.9 years at the time of diagnosis. Of these 48.9% were men and 51.13% were women. The mean ages of the men and women were 41.4±13.6 years and 42.0±13.1 years, respectively. Of these 40.0% of the men and 36.3% of the women were 35 years or younger at the time of diagnosis. A total of 39 660 matched controls were randomly selected for these cases.
For CD, 10 261 cases were included with a mean age of 41.1±13.7 at the time of diagnosis. Of these 45.3% were men and 54.7% were women. The mean ages of the men and women were 40.9±13.8 years and 41.3±13.6 years, respectively. Of these 40.9% of the men and 39.7% of the women were 35 years or younger at the time of diagnosis. A total of 20 522 matched controls were randomly selected for these cases.
Exposure to silica dust, UC and CD
Table 1 shows the number of cases and controls for UC and CD along with the OR of exposure and the corresponding 95% CI. For men, regardless of age and duration of occupation, the odds of developing UC while exposed to silica dust were 33% (OR 1.33, 95% CI 1.05 to 1.22). The risk of developing UC increases as a function of the duration of occupational exposure across the entire study population as well as in men (table 1).
When data on exposure levels from the JEM was added, a cumulative exposure level could be calculated. The cumulative exposure showed a dose-dependent increase in risk for developing UC for both the total study population (among the highest exposed OR was 1.4, 95% CI 1.12 to 1.74), as well as for both men and women (OR of respectively 1.34, 95% CI 1.05 to 1.71 and 1.71, 95% CI 1.04 to 2.08 among the highest exposed men and women; table 1)
The risk of developing CD was significantly increased among women (OR 1.29, 95% CI 1.01 to 1.65) and particularly in women exposed to silica for >5–10 years (OR 1.62, 95% CI 1.03 to 2.54). However, this trend was not observed across the patient population as a whole, or in men.
Mortality rates for UC and CD
The SMRs for men and women with UC or CD are shown in table 2. Men diagnosed with either UC or CD exhibit an elevated mortality rate. The mortality rate of UC and CD was increased in both exposed and non-exposed individuals, although there seems to be a higher SMR for those who were identified by the JEM as having been exposed to silica dust (SMR for UC=1.64, 95% CI 1.16 to 2.24; SMR for CD=1.88, 95% CI 1.08 to 3.06; table 2). For women, an increased SMR was found for both UC and CD. However, the SMR for women exposed to silica dust could not be calculated due to a lack of fatalities in the extracted medical data. For both UC and CD neoplasms (ICD 10 C00–D48) was the main cause of death, with the highest SMR for cases exposed to silica (SMR for UC=2.21, 95% CI 1.35 to 3.41; SMR for CD=2.95, 95% CI 1.41 to 5.42). For cases not exposed to silica the SMR was lower, but still statistically significant (SMR for UC=1.41, 95% CI 1.20 to 1.63; SMR for CD=1.45, 95% CI 1.17 to 1.79).
Discussion and conclusions
The OR for developing UC was significantly increased among men with an exposure duration of more than 5 years. No significant difference was observed between individuals exposed for >5–10 years and those exposed for over 10 years. The OR also increased with an increased cumulative exposure (table 1). This suggests that there is a threshold beyond which further exposure no longer affects the risk of developing UC.
The results obtained indicate that silica dust exposure correlates with an increased risk of developing UC among men. Previous reports suggested that silica exposure among men correlates with an increased incidence of another inflammatory disease, sarcoidosis.6 24 Among women, increased risk for developing UC was seen when data were analysed according to cumulative exposure.
Both sarcoidosis and UC are less prevalent in smokers.10 25 It is possible that smoking offers a protective action that interacts with the pathological action of silica dust. The protective effects of smoking are not currently understood but appear to impact men and women differently.26 27 Helper T cells are thought to play a critical role in the pathology of UC10 as well as other inflammatory diseases including sarcoidosis, rheumatoid arthritis and systemic lupus erythematosus.28–31 Silica dust exposure appears to increase the incidence of all these conditions. The immune responses to UC and CD reportedly involve different T cell populations,10 32 which may respond in different ways to silica dust exposure. This may explain the different OR values observed for men and women in this work because there appear to be gender-based differences in immune responses.33–35 These differences are not well understood but similar effects have been seen for asthma.36 37
The results from this study show that SMR for UC and CD was increased in both men and women, which is consistent with previous reports.38 Furthermore, for men, exposure to silica dust seems to increase the SMR for CD and UC compared with unexposed, however not statistically significant.
The large patient population and the validity of the medical data extracted from the registers represent major strengths of this study. The OPR maintains a thorough nationwide database of outpatients, so all individuals who reside in Sweden and are diagnosed with UC or CD were potentially eligible as cases for this study. Swedish law requires all publicly and privately funded physicians to report data to the OPR; the medical data contained in the OPR has been affirmed by practicing physicians and is therefore considered highly valid.
The RTB database from which the controls were selected is also nationwide, making the entire population of Sweden available as potential controls. Consequently, the results presented here are representative of the Swedish population. No medical data originally gathered before 2005 were used in this work. In this study, the time of diagnosis was operationally defined as the first date of entry in the National Register from which medical data was extracted for each case. It is possible that some individuals were actually diagnosed prior to the date recorded in the national registry, so we did not consider diagnoses made between 2005 and 2006 to reduce the risk of including previously diagnosed cases. Another limitation of the study was that the individuals excluded from the dataset were not further evaluated.
A limitation of this study is that silica dust exposure was defined on the basis of JEM data as being employed in an environment, where silica dust is present when duration of exposure was calculated. However, the presence of silica dust at a job site does not by itself mean that all employees would have been exposed to the particles, so this definition may have exaggerated the number of exposed individuals. Any such inflation of exposure rates was likely mitigated by the large sample size. However, if the JEM overestimated the detrimental level of silica exposure in the non-exposed cases, the results obtained would be skewed towards a null hypothesis, reducing their significance. Another limitation of this study is that as this is a register study there is a lack in information on potential confounders. Among these possible confounders are smoking habits and socioeconomic factors, both of which correlate with the incidence of IBD.9 10 However, since the sample was large and the cases and controls were matched, it is reasonable to assume that the incidence of cofounding factors in the two groups is similar. Nevertheless, it is possible that the results presented here do not reflect an intrinsic capability of silica dust to induce inflammation but a masked confounding factor associated with silica dust exposure.
Habitual smokers have a reduced risk of developing UC.10 While this is interesting, it is unlikely that including data on patients who smoke increased the possibility of detecting an increased risk of OR. However, a history of smoking is a known strong risk factor for developing CD, so not controlling for smoking cessation may have introduced a bias towards an alpha error.
In conclusion, this study suggests a positive correlation between silica dust exposure and the risk of developing UC.The risk increases with the duration of exposure as well as with increased cumulative exposure, especially for men. Conversely, silica dust exposure in women increases the risk of developing CD. No comparable findings have been reported previously to the author’s knowledge. Both UC and CD also appear to increase mortality, and exposure to silica dust seems to be an aggravating factor.
References
Footnotes
Contributors PV, ILB and PG conceived and designed the study. PW constructed the adopted JEM. ILB did the main data analysis and AW, PV, ILB and PG interpreted the results. AW, LF, PV, ILB and PG participated in the writing of the manuscript. All authors approved the final version.
Funding This study was done with support from Region Örebro County.
Competing interests None declared.
Patient consent for publication Not required.
Ethics approval The study was approved by the Regional Ethical Review Board in Uppsala, Sweden (Swedish Ethical Review Authority); DNR 2017/252.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Data are available upon reasonable request.