Article Text
PDF
Abstract
Background Wegener's granulomatosis is a systemic vasculitis of unknown aetiology. Previous studies have presented environmental exposures such as silica and farming as potential risk factors.
Objective To investigate the potential risk for Wegener's granulomatosis associated with occupations involving contact with animals and various airway exposures, using a population-based approach.
Methods In the Swedish Register of inpatient care 2288 cases with Wegener's granulomatosis were identified. Ten matched controls for every case were selected from the Swedish Population Register. By linking the cases and controls to the Swedish population censuses, information on employments before the diagnosis of Wegener's granulomatosis was collected. Relative risks were assessed as odds ratios using conditional logistic regression.
Results Odds ratios for specific occupations ranged from 0.6 to 1.9, and centred symmetrically around 1. No statistically significant increased risk was noted for the investigated occupations.
Statistics from Altmetric.com
Introduction
Wegener's granulomatosis is a systemic vasculitis predominantly affecting the upper and lower airways, and the kidneys. The aetiology of Wegener's granulomatosis is unknown, but is supposed to result from interplay between genetic susceptibility and environmental triggers. A genetic predisposition to Wegener's granulomatosis is supported by some findings, such as α1-antitrypsin deficiency1 and functional polymorphism in the gene PTPN22.2
There are a few case reports on familiar clustering of Wegener's granulomatosis, 3 4 but in a recent large population-based study of first-degree relatives of patients with Wegener's granulomatosis from our group, no strong familial association was found.5
As possible environmental triggers, infectious agents such as Staphylococcus aureus,6 and silica dust, pesticides and solvents have been investigated.7,–,10 In a previous study of 75 cases of primary systemic vasculitis, including 47 cases with Wegener's granulomatosis from Great Britain,9 farming was identified as a possible risk factor for primary systemic vasculitis (relative risk 2.7, 95% CI 1.2 to 5.8). In the same study, occupational silica exposure was non-statistically significantly associated with risk for Wegener's granulomatosis (relative risk 2.5, 95% CI 0.8 to 8.5).9 Based on previous observations and on our own clinical experience we therefore chose to investigate further possible occupational risk factors, focusing on two occupational groups: those involving exposure to various forms of inhaled particles, and those including exposure to livestock/animals.
Subjects and methods
Cases and controls
In Sweden (nine million inhabitants) inpatient care is public- and population-based. Referral to hospital is based on geography rather than financial capacity or insurance. The Swedish nationwide and population-based inpatient register contains individual-based information on all inpatient care since 1964, with nationwide coverage since 1987.11 For each person the discharging doctor's medical diagnoses are recorded according to the International Classification of Disease (ICD), versions 7, 8, 9 and 10, together with information on date of admission and discharge. In this register, we identified as cases 2288 patients discharged with a registered diagnosis of Wegener's granulomatosis between 1970 and 2003 (ICD 8 446.2, ICD 9 446.4 (446E in Sweden) and ICD 10 M31.3). Mean age at entry was 61 years, 53% were male and the median year of first hospitalisation was 1995 (10th and 90th centiles: 1980–2003). Previous assessments of the register-based diagnosis of Wegener's granulomatosis in the inpatient register has indicated that close to 90% of all discharges listing Wegener's granulomatosis were correct, that an equally high proportion of patients received their diagnosis during their first hospitalisation and suggested a high coverage of the register for all true Wegener's granulomatosis during our study period.12 13
For every case with Wegener's granulomatosis, we randomly selected from the Registry of Total Population 10 general population controls (total N=22 883), matched for sex, age, county of residence and marital status (fig 1). Controls had to be alive, and with no Wegener's granulomatosis, at the time of entry of their corresponding case.
Register linkage procedure. (1) Identification of patients with Wegener's granulomatosis. (2) Identification of 10 matched population controls for each subject. (3) Identification of the occupations of both subjects and controls as registered in population census register of 1960, 1970, 1980 and 1990.
Occupational exposure
In the Swedish national and virtually complete population censuses of 1960, 1970, 1980 and 199014 information on gainful employment is registered according to a predefined comprehensive list of (>1000) occupations for Swedish citizens above 16 years of age. By linking cases and their controls to this register, the occupation registered for each person at each census, was obtained. For this study, we identified 32 occupations involving substantial exposure to inhaled particles or animal contact (fig 2). Exposure was defined as stating one of the above occupations in at least one census before the date of entry with Wegener's granulomatosis. Separate exposure assessments were also performed according to whether the subject was registered with the same occupation in one, two, or more censuses before the date of entry with Wegener's granulomatosis. The median time between stated occupations and entry with Wegener's granulomatosis was 24 years (interquartile range 16–32).
The 32 studied occupations and their relative risk for Wegener's granulomatosis presented with block bars in the graph and in the right column (95% CI). Marked in black bars and circles are the adjusted relative risks after validation of the Wegener's granulomatosis diagnosis (95% CI).
Validation of Wegener's granulomatosis
For all occupations in our predefined list for which we noted increased relative risks, and for a selected set of other occupations, whether statistically significant or not but based on >2 exposed cases, we validated each case's diagnosis of Wegener's granulomatosis by retrieving and scrutinising the medical files against the American College of Rheumatology (ACR) criteria.15 Of a total of 61 cases with Wegener's granulomatosis selected for validation, the complete medical files of 59 cases were found. The diagnosis Wegener's granulomatosis met the ACR criteria in 49 (83%) of these patients, 32 (65%) of these had their diagnosis confirmed by histology. (The majority of the incorrectly registered cases had been investigated for suspected Wegener's granulomatosis, but turned out to have various other diseases including infections, pulmonary malignancies, alveolitis, nephritis.) After exclusion of cases not meeting the ACR criteria, the relative risks were recalculated (table 1).
Relative risks for Wegener's granulomatosis in selected occupations before and after validation of the registered diagnosis of (and subsequent elimination of cases not fulfilling the American College of Rheumatology criteria for) Wegener's granulomatosis
Statistical methods
Odds ratios (ORs) and two-sided 95% CIs were calculated to estimate the relative risk for Wegener's granulomatosis associated with the occupations of interest. Conditional logistic regression models for each of the occupations, taking the matched design into account, were used. A two-sided 95% CI excluding 1.0 was used to define statistical significance. Models were also fitted including information on years between first recorded occupation and first discharge listing Wegener's granulomatosis. The statistical model was calculated using the SAS 9.1/Linux statistical software.
Results
When ranked according to the magnitude of the relative risk, the 32 occupations centred around 1.0, with individual occupational ORs ranging from 0.6 to 1.9, and an OR of 1.1 (95% CI 1.0 to 1.3) for the combined group of the 32 occupations (fig 2). Although no statistically significant associations were observed, borderline (here defined as lower CIs between 0.9 and 1) increased risks were noted for bakers (OR=1.6, 95% CI 1.0 to 2.6), paper workers (OR=1.8, 95% CI 0.9 to 3.5), miners (OR=1.9, 95% CI 1.0 to 3.5) and for animal keepers (OR=1.8, 95% CI 0.9 to 3.5), the magnitude of which persisted following validation (fig 2). Farmers or farm workers were not at increased risk of Wegener's granulomatosis. After exclusion of exposed cases not fulfilling the ACR criteria for Wegener's granulomatosis, all ORs were, as expected, attenuated such that little indication of any increased risk remained.
Discussion
In this large population-based nationwide study of 32 occupations as putative risk factors for Wegener's granulomatosis, modest risks were observed for a series of occupations, none of which attained formal statistical significance. Farming was not associated with an increased risk of Wegener's granulomatosis.
Considering the frequent and initial involvement of the upper airways, exposures to inhaled substances have been previously studied, and suggested high rates of self-reported exposure to inhaled substances but failed to identify any particular inhalant,9 increased risks with silica exposure (OR=1.6, 95% CI 0.9 to 1.3),8 with occupational solvent exposure (OR=3.4, 95% CI 1.3 to 8.9) and with farming (OR=2.7, 95% CI 1.3 to 5.7).7 However, these and other previous studies on environmental hazardous exposures in Wegener's granulomatosis7,–,10 have been based on questionnaires or interviews of subjects with already confirmed vasculitic disease, the samples have been small (7–129 cases with WG) and exposures have all been self-reported. Against this background, we chose to perform a large, national, population-based case–control study involving more than 2000 patients registered with Wegener's granulomatosis and their matched controls, where the information on occupational exposure was collected before the disease, irrespective of disease status, thereby avoiding recall bias. However, occupational exposure is not necessarily synonymous with actual exposure, as specific exposures may vary between people registered under the same occupation, and does not take into account exposures outside work. Although this ‘insensitivity’ was similar for cases and controls, and thus unlikely to bias the relative risk, we cannot exclude the possibility that the general lack of association between, for example, farming and risk of Wegener's granulomatosis would overlook an association with a particular exposure (yet to be defined) relevant to only a subset of all people with a particular occupation.
To avoid including as ‘cases’ people who (on the basis of specific occupational exposures) had developed conditions that might have been mistaken for Wegener's granulomatosis, we validated all Wegener cases with occupations for which we noted increased risks. This validation suggested a somewhat lower level of correct diagnosis (81%) than previously (close to 90% and based on a larger set of validated cases) seen.12 Selective exclusion of exposed cases who did not meet the Wegener's granulomatosis definition (but not of unexposed cases) will inevitably lead to an underestimation of the OR, so it is reasonable to assume that the ‘true’ OR for these occupations would fall somewhere between the non-validated and the validated OR.
We assessed a total of 32 occupations, and would therefore expect to find one or two statistically significant associations by chance alone (we found one association with a lower 95% CI bond=1.0). Further, given the symmetrical distributions of OR around unity, we cannot exclude the possibility that the signals detected for mining and baking steel, and animal-specific occupations were due to chance. With all ORs <2 the risk conferred by each such occupation would be lower than the ‘background’ risk in the general population.
In conclusion, our results suggest that there is no general association between these occupationally inhaled exposures, or between livestock farming and Wegener's granulomatosis.
References
Footnotes
-
Ethics approval This study was conducted with the approval of the regional ethics committee in Stockholm, Sweden.
-
Provenance and peer review Not commissioned; externally peer reviewed.