|Home | About | Journals | Submit | Contact Us | Français|
C. J. Turner, S. M. Kennedy. School of Environment and Health, University of British Columbia
ObjectivesOur objective is to examine the natural history, risk factors and early indicators for the development of chronic respiratory morbidity among young adults in “blue collar” trades. This analysis focuses on the hypothesis that changes in lung function recorded in the first 2 years of work may be useful markers of subsequent, more chronic measures of impaired lung functioning.
MethodsIn 1988 we enrolled a cohort of 356 first year apprentices in four industrial trades and examined respiratory health at baseline (v1: 1988–90), after 2 years (v2: 1990–2) and after 16 years (v3: 2004–7), using standardised questionnaires, tests of airway calibre (forced expired volume in 1 s, FEV1) and airway responsiveness (a marker of asthma). Individuals were retained in the cohort whether or not they remained in their original trade.
ResultsA total of 209 workers were examined at visit 3. There were no significant differences in baseline lung health measures among participants and those lost to follow‐up. Participant who had changed jobs between v1 and v3 for health reasons were significantly more likely to have asthma at follow‐up (38% vs 12%, p<0.01) and have increased bronchial responsiveness between v1 and v2 (p<0.01). FEV1 at v3 was significantly associated with both FEV1 and bronchial responsiveness level at v1, and with the new development of clinically relevant bronchial hyper‐responsiveness between v1 and v2, but not with a rapid decline in FEV1 between v1 and v2. Similarly, the severity of bronchial hyper‐responsiveness at v3 was significantly associated with both baseline responsiveness and change in responsiveness between v1 and v2. These effects were present even after controlling for the diagnosis of asthma at baseline and at follow‐up.
ConclusionWe conclude that lung function level at entry into the workforce is an important determinant of subsequent lung function as much as 16 years later, and that a rapid change in bronchial responsiveness early in the working career (but not a rapid change in simple measures of lung function) are also predictive of subsequent airflow obstruction and bronchial hyper‐responsiveness.
Key wordsoccupational lung disease; young workers; asthma
S. F. Buechte1, P. Morfeld1, H. J. Bicker2, H. Lenaerts2, B. Kalkowsky2, J. Pohlplatz1, A. Koesters2, C. Piekarski1. 1Institute for Occupational Medicine, Cologne University, Germany; 2Occupational Medical Service, Deutsche Steinkohle AG, Herne, Germany
ObjectivesMultiple exposures receive an increasing amount of attention. Recent publications (MacLehose et al. Epidemiology 2007;18(2):199–207) have demonstrated the pitfalls epidemiologists should be aware off when dealing with several highly correlated exposures. In our longitudinal study we find two correlated exposures: respirable quartz and non‐quartz coal mine dust. First analyses exhibited instabilities in the observed associations depending on variable transformation. By following a Bayesian approach (data augmentation) we were able to apply ridge regression in conjunction with GEE regression models to our data.
MethodsFor 1369 subjects who started working underground between 1974 and 1979 at two coal mines (A, B) response data (FVC), data on covariates (simple model: smoking, height, weight; final model: simple model extended with year of examination, spirometric device and device operator, nationality, mine (MI)) and cumulative exposure data (respirable quartz (cRQD), non‐quartz dust (cRNQD)) were repeatedly measured and collected. The follow‐up period ended on 31 December 1998. Following a linear regression analysis, variance inflation factors are calculated. After data augmentation for exposure and MI, we performed GEE and ridge regression analyses applying a biasing constant c ranging from 0.001 to 100. All analyses were performed for the complete study group and restricted to each mine.
ResultsLinear regression and subsequent calculation of VIF yield factors close to 10 for YOE, cRQD and cRNQD and even greater factors when restricting analysis to each mine. The GEE regression analysis with the simple model for FVC returned a positive association for cRQD (p<0.0005) and a negative association for cRNQD (p<0.0005) while in the final model these associations were no longer observed GEE ridge regression analyses with the final model for the complete study group yeild no hint for correlation‐induced bias in the final model results.
ConclusionThe counter‐intuitive associations observed in this study in connection with ridge regression results demonstrate that a naive approach is not advisable when aiming at the effect dissection of complex mixtures. Further, even a potent dataset does not guarantee a simple interpretation of results.
Key wordscorrelated exposures; GEE ridge regression analysis; quartz dust, coal mine dust
K. Radon1, A. Schulze1, V. Ehrenstein1, R. T. van Strien2, G. Praml1, D. Nowak1. 1Institute for Occupational and Environmental Medicine, LMU Munich; 2Municipal Health Service Amsterdam, Amsterdam, The Netherlands
ObjectivesPotential adverse health effects of concentrated animal feeding operations (CAFOs) are of public concern. However, there is a lack of objectively assessed data on environmental exposure to CAFOs and objective data on respiratory health. We aimed to assess respiratory health in neighbours of CAFOs using objective measurements of exposure and outcome.
MethodsA cross‐sectional survey was done in 2002–2004 among all adults (18–44 years old) living in four rural German towns with a high density of CAFOs. Questionnaire data were available for 6937 (68%) eligible subjects. In a random sample we measured specific IgE to common and farm‐specific allergens, lung function and bronchial hyper‐responsiveness to methacholine (BHR) (participation 60%). Exposure was assessed by (1) geo‐coded data on the number of animal houses within 500 m of the home and (2) self‐reported odour annoyance. LOESS plots were used to model the association between exposure and outcome. Data were stratified by private or occupational contact with farm environments (farm subjects).
ResultsThe prevalence of atopic diseases was low in the population under study (eg, prevalence of nasal allergies 14%). In farm and non‐farm subjects, the prevalence of self‐reported asthma symptoms and nasal allergies increased with self‐reported odour annoyance. In contrast, self‐reported odour annoyance and results of the clinical measurements correlated poorly with one another. In non‐farm subjects, the number of animal houses within 500 m was a predictor of self‐reported wheeze (odds ratio for >12 animal houses within 500 m of the home: 2.7; 95% CI 1.4 to 5.4) and decreased forced expiratory volume in 1 s (mean decrease 0.26 litres, 95% CI 0.04 to 0.48). Neither for farm or non‐farm subjects was the number of animal houses associated with allergic rhinitis or with specific sensitisation.
ConclusionOur study indicates that confined animal feeding operations may contribute to the burden of respiratory disease among their neighbours. In addition, our findings underline the importance of objective assessment of exposure and outcome in environmental epidemiology.
Key wordsasthma; confined animal feeding operations; exposure assessment
J. P. Bonde1, S. W. Christensen2, Ø. Omland1. 1Department of Occupational Medicine, Aarhus University Hospital; 2Department of Occupational Medicine, Aalborg University Hospital
ObjectivesNumerous cross‐sectional studies have reported reduced lung function among welders but limitations of exposure assessment and design preclude causal inference. This study takes advantage of a 17‐year follow‐up of lung function measurements and a comprehensive assessment of individual exposure to welding fume particulates. The aim of this study was to investigate if long‐term exposure to welding fume particulates impairs lung function.
MethodsLung function was measured by spirometry in 1987 and again in 2004 among 68 mild steel and stainless welders and 32 non‐welding production workers recruited from six workplaces. The decline in forced expiratory volume (FEV1 and FVC) was analysed in relation to the accumulated exposure to fume particulates among welders during the follow‐up period with adjustment for lung function at baseline, age, height, body weight and tobacco smoking. The exposure assessment combined a Danish welding exposure matrix based upon more than 1000 workplace measurements and individual interview data on welding work obtained in 2004. The individual exposure estimate accounted for different levels of particulate exposure according to calendar year, type of steel, welding method and intermittence, use of exhaust ventilation and work in confined spaces.
ResultsWelders were not at increased risk of accelerated decline of lung function during follow‐up in comparison with non‐welding metal workers. Within welders the decline in lung function was not related to the cumulated welding particulate exposure during the follow‐up period in either smokers or non‐smokers. On the contrary, the decline of lung function during follow‐up was highly significantly related to the total amount of tobacco pack‐years during follow‐up.
ConclusionOur data do not support the hypothesis that the age‐related decline in lung function is substantially accelerated by long‐term exposure to welding fume particulates at geometric mean ambient air levels of 1.5–6.5 mg/m3. However, our findings are compatible with minor effects on lung function and do not rule out that higher exposure levels may impair lung function.
Key wordsobstructive lung disorder; welding; lung function
A. Pronk1, L. Preller2, G. Doekes3, I. Wouters3, J. Rooijackers4, J‐W. Lammers5, D. Heederik3. 1IRAS‐TNO; 2TNO Quality of Life; 3Institute for Risk Assessment Sciences (IRAS); 4Netherlands Expertise Centre for Occupational Respiratory Disorders; 5University Medical Centre Utrecht
ObjectivesMost exposure response information for diisocyanates involves monomer exposure. Associations between oligomeric isocyanate exposure and respiratory disease have received little attention, despite the extensive use of diisocyanate oligomers. In a cross‐sectional study of spray painters exposed to mainly hexamethylene diisocyanate (HDI) oligomers, an association between exposure, respiratory symptoms and isocyanate specific sensitisation was demonstrated. In a subset of that population we assessed bronchial hyper‐responsiveness (BHR), lung function and exhaled NO (eNO) and investigated associations with exposure and specific antibody response.
MethodsBaseline spirometry, metacholine challenge and eNO measurements were performed in 229 workers. Questionnaires and blood samples were obtained. Personal exposure was estimated by combining over 500 personal task‐based inhalatory exposure measurements (LC‐MS) and time activity information. Specific serum IgE and IgG to hexamethylene diisocyanate (HDI) were assessed by ImmunoCAP assay and enzyme immunoassay using vapour and liquid phase HDI‐human serum albumin (HSA) and HSA‐conjugates prepared with oligomeric HDI.
ResultsA strong positive association was found between exposure and BHR (prevalence ratio (PR) (95% CI interquartile range): 1.8 (1.1 to 3.0)). Smoothed spline plots corroborated a log linear association without any clear indication for an exposure threshold. In addition, indications of exposure related obstructive lung function changes independent of BHR were found (FEV1, FEV1/FVC, PEF and MMEF associated with exposure, p<0.05). Overall sensitised (specific IgE or IgG) workers were more often hyper‐responsive. This was statistically significant for IgG positives measured by CAP assay only (PR (95% CI): 3.1 (1.1 to 8.2)). Although eNO was not associated with exposure, workers with IgG to the oligomeric HDI conjugates had higher levels of eNO.
ConclusionExposure to isocyanate oligomers showed strong associations with BHR and, independently, lung function parameters. This indicates that both asthma and chronic respiratory effects may be involved. The mechanisms behind the development of respiratory effects remain unclear. Different phenotypical subgroups with respect to BHR, FEV1/FVC<70%, antibodies and increased eNO were observed. These may reflect different health end‐points, underlying mechanisms or disease stages.
Key wordsisocyanate; asthma; BHR