Deletion diagnostics for the generalised linear mixed model with independent random effects.

The Generalised linear mixed model (GLMM) is widely used for modelling environmental data. However, such data are prone to influential observations, which can distort the estimated exposure-response curve particularly in regions of high exposure. Deletion diagnostics for iterative estimation schemes commonly derive the deleted estimates based on a single iteration of the full system holding certain pivotal quantities such as the information matrix to be constant. In this paper, we present an approximate formula for the deleted estimates and Cook's distance for the GLMM, which does not assume that the estimates of variance parameters are unaffected by deletion. The procedure allows the user to calculate standardised DFBETAs for mean as well as variance parameters. In certain cases such as when using the GLMM as a device for smoothing, such residuals for the variance parameters are interesting in their own right. In general, the procedure leads to deleted estimates of mean parameters, which are corrected for the effect of deletion on variance components as estimation of the two sets of parameters is interdependent. The probabilistic behaviour of these residuals is investigated and a simulation based procedure suggested for their standardisation. The method is used to identify influential individuals in an occupational cohort exposed to silica. The results show that failure to conduct post model fitting diagnostics for variance components can lead to erroneous conclusions about the fitted curve and unstable confidence intervals.

Childhood exposure to ambient polycyclic aromatic hydrocarbons is linked to epigenetic modifications and impaired systemic immunity in T cells.

BACKGROUND: Evidence suggests that exposure to polycyclic aromatic hydrocarbons (PAHs) increases atopy; it is unclear how PAH exposure is linked to increased severity of atopic diseases. OBJECTIVE: We hypothesized that ambient PAH exposure is linked to impairment of immunity in atopic children (defined as children with asthma and/or allergic rhinitis) from Fresno, California, an area with elevated ambient PAHs. METHODS: We recruited 256 subjects from Fresno, CA. Ambient PAH concentrations (ng/m(3) ) were measured using a spatial-temporal regression model over multiple time periods. Asthma diagnosis was determined by current NHLBI criteria. Phenotyping and functional immune measurements were performed from isolated cells. For epigenetic measurements, DNA was isolated and pyrosequenced. RESULTS: We show that higher average PAH exposure was significantly associated with impaired Treg function and increased methylation in the forkhead box protein 3 (FOXP3) locus (P < 0.05), conditional on atopic status. These epigenetic modifications were significantly linked to differential protein expression of FOXP3 (P < 0.001). Methylation was associated with cellular functional changes, specifically Treg dysfunction, and an increase in total plasma IgE levels. Protein expression of IL-10 decreased and IFN-γ increased as the extent of PAH exposure increased. The strength of the associations generally increased as the time window for average PAH exposure increased from 24 hr to 1 year, suggesting more of a chronic response. Significant associations with chronic PAH exposure and immune outcomes were also observed in subjects with allergic rhinitis. CONCLUSIONS AND CLINICAL RELEVANCE: Collectively, these results demonstrate that increased ambient PAH exposure is associated with impaired systemic immunity and epigenetic modifications in a key locus involved in atopy: FOXP3, with a higher impact on atopic children. The results suggest that increased atopic clinical symptoms in children could be linked to increased PAH exposure in air pollution.

Possible pro-carcinogenic association of endotoxin on lung cancer among Shanghai women textile workers.

BACKGROUND: Endotoxin (lipopolysaccharide) is a widespread contaminant in many environmental settings. Since the 1970s, there has been generally consistent evidence indicating reduced risks for lung cancer associated with occupational endotoxin exposure. METHODS: We updated a case-cohort study nested within a cohort of 267,400 female textile workers in Shanghai, China. We compared exposure histories of 1456 incident lung cancers cases diagnosed during 1989-2006 with those of a reference subcohort of 3022 workers who were free of lung cancer at the end of follow-up. We applied Cox proportional hazards modelling to estimate exposure-response trends, adjusted for age and smoking, for cumulative exposures lagged by 0, 10, and 20 years, and separately for time windows of ⩽15 and >15 years since first exposure. RESULTS: We observed no associations between cumulative exposure and lung cancer, irrespective of lag interval. In contrast, analyses by exposure time windows revealed modestly elevated, but not statistically significant relative risks (∼1.27) at the highest three exposure quintiles for exposures that occurred >15 years since first exposure. CONCLUSIONS: The findings do not support a protective effect of endotoxin, but are suggestive of possible lung cancer promotion with increasing time since first exposure.

Cotton dust, endotoxin and cancer mortality among the Shanghai textile workers cohort: a 30-year analysis.

BACKGROUND: Although occupational exposure to cotton dust and endotoxin is associated with adverse respiratory health, associations with cancer are unclear. We investigated cancer mortality in relation to cotton dust and endotoxin exposure in the Shanghai textile workers cohort. METHODS: We followed 444 cotton textile and a reference group of 467 unexposed silk workers for 30 years (26 777 person-years). HRs for all cancers combined (with and without lung cancer) and gastrointestinal cancer were estimated in Cox regression models as functions of cotton textile work and categories of cumulative exposure (low, medium, high), after adjustment for covariates including pack-years smoked. Different lag years accounted for disease latency. RESULTS: Risks of mortality from gastrointestinal cancers and all cancers combined, with the exclusion of lung cancer, were increased in cotton workers relative to silk workers. When stratified by category of cumulative cotton exposure, in general, risks were greatest for 20-year lagged medium exposure (all cancers HR=2.7 (95% CI 1.4 to 5.2); cancer excluding lung cancer HR=3.4 (1.7-7.0); gastrointestinal cancer HR=4.1 (1.8-9.7)). With the exclusion of lung cancer, risks of cancer were more pronounced. When stratified by category of cumulative endotoxin exposure, consistent associations were not observed for all cancers combined. However, excluding lung cancer, medium endotoxin exposure was associated with all cancers and gastrointestinal cancer in almost all lag models. CONCLUSIONS: Cotton dust may be associated with cancer mortality, especially gastrointestinal cancer, and endotoxin may play a causative role. Findings also indirectly support a protective effect of endotoxin on lung cancer.

Circulating adhesion molecules after short-term exposure to particulate matter among welders.

BACKGROUND: Studies from several countries indicate that welders experience increased risk of mortality and morbidity from ischaemic heart disease. Although the underlying mechanisms are unclear, vascular responses to particulate matter contained in welding fumes may play a role. To investigate this, we studied the acute effects of welding fume exposure on the endothelial component of vascular function, as measured by circulating adhesion molecules involved in leukocyte adhesion (sICAM-1 and sVCAM-1) and coagulation (vWF). METHODS: A panel of 26 male welders was studied repeatedly across a 6 h work-shift on a high exposure welding day and/or a low exposure non-welding day. Personal PM(2.5) exposure was measured throughout the work-shift. Blood samples were collected in the morning (baseline) prior to the exposure period, immediately after the exposure period, and the following morning. To account for the repeated measurements, we used linear mixed models to evaluate the effects of welding (binary) and PM(2.5) (continuous) exposure on each blood marker, adjusting for baseline blood marker concentration, smoking, age and time of day. RESULTS: Welding and PM(2.5) exposure were significantly associated with a decrease in sVCAM-1 in the afternoon and the following morning and an increase in vWF in the afternoon. CONCLUSIONS: The data suggest that welding and short-term occupational exposure to PM(2.5) may acutely affect the endothelial component of vascular function.

Chronic obstructive pulmonary disease mortality in railroad workers.

BACKGROUND: There is little information describing the risk of non-malignant respiratory disease and occupational exposure to diesel exhaust. METHODS: US railroad workers have been exposed to diesel exhaust since diesel locomotives were introduced after World War II. In a retrospective cohort study we examined the association of chronic obstructive pulmonary disease (COPD) mortality with years of work in diesel-exposed jobs. To examine the possible confounding effects of smoking, multiple imputation was used to model smoking history. A Cox proportional hazards model was used to estimate an incidence rate ratio, adjusted for age, calendar year, and length of follow-up after leaving work (to reduce bias due to a healthy worker survivor effect). RESULTS: Workers in jobs with diesel exhaust exposure had an increased risk of COPD mortality relative to those in unexposed jobs. Workers hired after the introduction of diesel locomotives had a 2.5% increase in COPD mortality risk for each additional year of work in a diesel-exposed job. This risk was only slightly attenuated after adjustment for imputed smoking history. CONCLUSIONS: These results support an association between occupational exposure to diesel exhaust and COPD mortality.

Sister chromatid exchange in chronic ethylene oxide-exposed primates: unexpected effects of in vitro culture duration, incubation temperature, and serum supplementation.

Ethylene oxide (EtO), a potent monofunctional DNA alkylating agent, has been shown to induce sister chromatid exchanges (SCE) in the peripheral blood lymphocytes of animals and workers exposed to it in vivo. We have previously reported that elevations of SCE persist for 6 years after cessation of EtO exposure in cynomolgus monkeys chronically exposed to EtO; the elevation in mean SCE was entirely attributable to a subpopulation of high SCE frequency cells (HFCs). We now report that the detection of persistent HFCs is dependent on the conditions of cell growth, and that EtO exposure increases the replication indices of lymphocytes from the exposed animals when these cells are examined at early cytogenetic harvest times. Culture of lymphocytes in differing serum supplements, changes in cytogenetic harvest times, and alterations in in vitro incubation temperature all markedly affected mean SCE frequency by influencing the detection of HFCs. The frequency of EtO-induced HFCs was independent of 5-bromodeoxyuridine concentration, used for differential staining of sister chromatids. These observations indicate that the detection of persistent alkylation-induced chromosomal changes, observed long after cessation of in vivo chronic exposure of these animals, is highly dependent upon factors affecting cell growth.

Persistently elevated sister chromatid exchanges in ethylene oxide-exposed primates: the role of a subpopulation of high frequency cells.

Ethylene oxide (EtO) is a potent DNA-alkylating agent which has been shown to induce sister chromatid exchanges (SCE) in the peripheral blood lymphocytes of exposed workers. To study further the persistence of EtO-induced SCE, we have examined lymphocytes from a group of cynomolgus monkeys exposed to EtO in control, 50-ppm, and 100-ppm concentrations for 7 h/day, 5 days/week over the years 1979-1981. The data collected in 1987 were compared with those generated immediately prior to the cessation of exposure in 1981. EtO-induced SCE persisted at levels significantly above those of the nonexposed controls. Comparison of the distributions of SCE between 1979 and 1987 shows that, although mean SCE decreased from 1981 to 1987, the mean SCE in the top 10% of the distribution has not diminished over time. Consequently, the increased level of SCE is entirely attributable to a subpopulation of cells with high frequencies of SCE. These findings suggest that long-lived lymphocytes may inefficiently repair EtO-induced lesions which produce SCE. The results also have important implications for the proper use of SCE analytical techniques in the epidemiological study of cytogenetic damage after chronic exposure to DNA-alkylating agents.

Acute and chronic respiratory effects of sodium borate particulate exposures.

This study examined work-related chronic abnormality in pulmonary function and work-related acute irritant symptoms associated with exposure to borate dust in mining and processing operations. Chronic effects were examined by pulmonary function at the beginning and end of a 7-year interval. Time-specific estimates of sodium borate particulate exposures were used to estimate cumulative exposure during the study interval. Change in pulmonary function over the 7 years was found unrelated to the estimate of cumulative exposure during that interval. Exposure-response associations also were examined with respect to short-term peak exposures and incidence of five symptoms of acute respiratory irritation. Hourly measures of health outcome and continuous measures of particulate exposure were made on each subject throughout the day. Whenever a subject reported one of the irritant symptoms, a symptom intensity score was also recorded along with the approximate time of onset. The findings indicated that exposure-response relationships were present for each of the specific symptoms at several symptom intensity levels. The associations were present when exposure was estimated by both day-long and short-term (15-min) time-weighted average exposures. Associations persisted after taking account of smoking, age, and the presence of a common cold. No significant difference in response rate was found between workers exposed to different types of sodium borate dusts.

Pulmonary function among cotton textile workers. A study of variability in symptom reporting, across-shift drop in FEV1, and longitudinal change.

Longitudinal variability in respiratory responses, including symptom reporting and across-shift change in ventilatory function, were examined in relation to long-term loss of ventilatory function in a group of 447 cotton textile workers in Shanghai, China. The study used a standardized respiratory questionnaire and standardized spirometric testing before and after a work shift on the first day of the workweek. Prediction equations for FEV1 were generated from a group of silk textile workers from the same city. Environmental samples included both vertical elutriated cotton dust and endotoxin levels. There was considerable variability in symptom reporting between the baseline and 5-year follow-up survey for all symptoms. However, subjects who consistently reported symptoms had a significantly accelerated 5-year loss in FEV1 compared with those who never reported symptoms. Subjects with symptoms of chest tightness or dyspnea at one survey lost FEV1 at a rate intermediate between the never or both groups. Moreover, subjects with an across-shift change in FEV1 of more than 5 percent at both surveys had the greatest loss in FEV1 over 5 years (-267 ml) when compared with one-time responders (-224 ml), and nonresponders (-180 ml), though the differences were not significant. Workers with chest tightness and chronic bronchitis in both surveys were overrepresented in the high dust and endotoxin areas. Our results indicate that even with substantial survey-to-survey variability in responses, there is important information contained in both questionnaires and across-shift spirometry. Among cotton workers, consistent responders to either symptom questionnaire or across-shift FEV1 decrements of > or = 5 percent appear to be at increased risk for lung function impairment.

Estimation of ventilatory capacity in subjects with unacceptable lung function tests.

Based on pulmonary function data collected annually for six years on 540 Vermont granite workers, FEV1 in survey 1 was estimated by extrapolating back from subsequent measurements. The extrapolation method was found to fit the observed data of subjects with reproducible initial values very well (R2 = 0.87). Extrapolated FEV1s for workers unable to perform an adequate pulmonary function test according to the standards of the American Thoracic Society were compared to extrapolated values in the rest of the cohort. After adjusting for confounding, subjects with test failure in survey 1 had a lower extrapolated FEV1 than the rest of the cohort (p = 0.07). The mean extrapolated FEV1 of the 71 workers with an initial test failure was only 95% of a predicted value derived from the group with reproducible data, and the per cent predicted decreased from 98% to 71% as the number of test failures in the follow-up surveys increased (p = 0.0004). The American Thoracic Society and the Epidemiology Standardization Project currently recommend that test failures be excluded from the analysis of epidemiological data. Our findings suggest that alternative strategies for handling non-reproducible lung function may need to be explored in order to avoid selection bias.

Healthy worker effect in a longitudinal study of one-second forced expiratory volume (FEV1) and chronic exposure to granite dust.

BACKGROUND: Low level effects of granite dust on one-second forced expiratory volume (FEV1) are estimated in 618 Vermont granite workers followed for 5 years with annual pulmonary function tests. Reduced pulmonary function has already been reported for the subset of subjects lost to follow-up (dropouts) suggesting possible bias in analyses based only on survivors. METHOD: Healthy worker selection bias is directly assessed by comparing the dose-response associations between survivors who remained in the study for the full 5-year observation period and the dropouts. RESULTS: The 353 survivors had an FEV1 of 96% of predicted at baseline and were losing FEV1 at an average rate of 44 ml/yr. No association was found in this group between the rate of FEV1 decline and lifetime dust exposure. However, the 265 workers with incomplete follow-up, 'dropouts', had a lower FEV1 at baseline (94%) and were losing FEV1 at an average rate of 69 ml/yr. The dose-response parameter in this group was estimated to be 4 ml/yr loss per mg/m3-year and was statistically significant. CONCLUSIONS: These results provide an illustration of bias due to the healthy worker effect and an example of the failure to detect a true work-related health effect in a study based only on a 'survivor' population.

Defining measurement precision for effort dependent tests: the case of three neurobehavioural tests.

For effort dependent tests, the estimation of measurement precision (reproducibility) is complicated by learning effects and submaximal efforts which inflate the variance of repeated trials. To illustrate an approach to the estimation problem, precision was evaluated for three neurobehavioural tests based on the responses of 76 boatbuilders tested on four separate test occasions within a one-week period. The average Coefficient of Variation (CV) for repeated trials within a test session was 6%, 16% and 13% for the Continuous Performance Test (CPT), Symbol-Digit substitution test (SDS) and the Hand-Eye motor-coordination test (HEM), respectively. In order to adjust for the effects of learning, the first trial(s) of a session were excluded from the calculation of performance level and its precision. This adjustment for learning significantly improved the precision for SDS and HEM to a CV of 8%. Inspection of the distributions of best efforts by trial number indicated that dropping the early trial(s) eliminated the best efforts of 34%, 22% and 7% of the subjects on the three tests respectively. When the worst two trials were excluded regardless of order, precision improved significantly to less than 5% for all three tests. On the basis of these results, a 5% precision rule for CPT and a 10% precision rule for SDS and HEM are provisionally recommended. The test results of subjects unable to meet this criterion should be identified, but in order to avoid selection bias, they should be analysed separately rather than excluded.

Risk of upper aerodigestive tract cancers in a case-cohort study of autoworkers exposed to metalworking fluids.

AIMS: To re-examine aerodigestive cancer risk in a cohort of autoworkers exposed to metal working fluids (MWF), using improved case definition and more recently diagnosed cases. METHODS: The autoworker cohort included 31 100 hourly workers alive on 1 January 1985 who worked at three automobile plants in Michigan. A case-cohort design was carried out that included incident cases of cancers of the larynx, oesophagus, and stomach, and a 10% sample of the cohort. A Cox proportional hazards model was used to estimate MWF exposure effects. The smoothing method of penalised splines was used to explore the shape of the underlying exposure-response curves. RESULTS: The most important finding was the association between larynx cancer incidence and cumulative straight MWF exposure. The results for oesophageal cancer were less consistent. For stomach cancer there was no evidence of excess risk. CONCLUSION: This association between larynx cancer and straight MWF exposures was consistent with a previous finding in this cohort, providing further support for a causal relation.

Smoothing in occupational cohort studies: an illustration based on penalised splines.

AIMS: To illustrate the contribution of smoothing methods to modelling exposure-response data, Cox models with penalised splines were used to reanalyse lung cancer risk in a cohort of workers exposed to silica in California's diatomaceous earth industry. To encourage application of this approach, computer code is provided. METHODS: Relying on graphic plots of hazard ratios as smooth functions of exposure, the sensitivity of the curve to amount of smoothing, length of the exposure lag, and the influence of the highest exposures was evaluated. Trimming and data transformations were used to down-weight influential observations. RESULTS: The estimated hazard ratio increased steeply with cumulative silica exposure before flattening and then declining over the sparser regions of exposure. The curve was sensitive to changes in degrees of freedom, but insensitive to the number or location of knots. As the length of lag increased, so did the maximum hazard ratio, but the shape was similar. Deleting the two highest exposed subjects eliminated the top half of the range and allowed the hazard ratio to continue to rise. The shape of the splines suggested a parametric model with log hazard as a linear function of log transformed exposure would fit well. CONCLUSIONS: This flexible statistical approach reduces the dependence on a priori assumptions, while pointing to a suitable parametric model if one exists. In the absence of an appropriate parametric form, however, splines can provide exposure-response information useful for aetiological research and public health intervention.

A case-crossover study of transient risk factors for occupational acute hand injury.

BACKGROUND: Workers with acute hand injuries account for over 1 000 000 emergency department visits annually in the United States. AIMS: To determine potential transient risk factors for occupational acute hand injury. METHODS: Subjects were recruited from 23 occupational health clinics in five northeastern states in the USA. In a telephone interview, subjects were asked to report the occurrence of seven potential risk factors within a 90-minute time period before an acute hand injury. Each case also provided control information on exposures during the month before the injury. The self-matched feature of the study design controlled for stable between-person confounders. RESULTS: A total of 1166 subjects were interviewed (891 men, 275 women), with a mean age (SD) of 37.2 years (11.4). The median time interval between injury and interview was 1.3 days. Sixty three per cent of subjects had a laceration. The relative risk of a hand injury was increased when working with equipment, tools, or work pieces not performing as expected (11.0, 95% CI 9.4 to 12.8), or when using a different work method to do a task (10.5, 95% CI 8.7 to 12.7). Other transient factors in decreasing order of relative risk were doing an unusual task, being distracted, and being rushed. Wearing gloves reduced the relative risk by 60% (0.4, 95% CI 0.3 to 0.5). Occupational category, job experience, and safety training were found to alter several of these effects. CONCLUSION: The results suggest the importance of these transient, potentially modifiable factors in the aetiology of acute hand injury at work. Attempts to modify these exposures by various strategies may reduce the incidence of acute hand injury at work.

Respiratory symptoms and cotton dust exposure; results of a 15 year follow up observation.

AIMS: To determine chronic effects of long term exposure to cotton dust and endotoxin on incidence of respiratory symptoms and the effect of cessation of exposure. METHODS: Respiratory health in 429 Chinese cotton textile workers (study group) and 449 silk textile workers (control group) was followed prospectively from 1981 to 1996. Byssinosis, chest tightness, and non-specific respiratory symptoms were assessed by means of identical standardised questionnaires at four time points. Exposures to cotton dust and endotoxin were estimated using area samples collected at each survey. Incidence and persistence of symptoms were examined in relation to cumulative exposure and exposure cessation using generalised estimating equations (GEE). RESULTS: Among cotton workers, the cumulative incidence of byssinosis and chest tightness was 24% and 23%, respectively, and was significantly more common in smokers than in non-smokers. A high proportion of symptoms was found to be intermittent, rather than persistent. Among silk workers, no typical byssinosis was identified; the incidence of chest tightness was 10%. Chronic bronchitis, cough, and dyspnoea were more common and persistent in the cotton group than in the silk group. Significantly lower odds ratios for symptoms were observed in cotton workers who left the cotton mills; risk was also related to years since last worked. Multivariate analysis indicated a trend for higher cumulative exposure to endotoxin in relation to a higher risk for byssinosis. CONCLUSION: Chronic exposure to cotton dust is related to both work specific and non-specific respiratory symptoms. Byssinosis is more strongly associated with exposure to endotoxin than to dust. Cessation of exposure may improve the respiratory health of cotton textile workers; the improvement appears to increase with time since last exposure.

Methodology for analyzing episodic events.

Many health outcomes of interest to occupational epidemiologists are common recurrent health events. Epidemiologic approaches to the study of such health events are reviewed. Episodic events are considered to be events that occur at a distinct point in time - either with sudden onset or the sudden crossing of a threshold of detection - and they must be reversible events that can recur in the same person in response to a proximate trigger. Studying such health events poses 4 challenges to existing methods: (i) key epidemiologic concepts, such as incidence, do not naturally accommodate recurrent events, (ii) study designs must capture time-varying exposures, (iii) statistical models must be able to handle correlated outcomes, and (iv) feedback bias must be addressed. In response, methods such as longitudinal studies, case-crossover designs and generalized estimating equations are identified as appropriate tools.

Across-shift changes in the pulmonary function of meat-wrappers and other workers in the retail food industry.

Pulmonary function was measured before, during, and after the end of the workshift in 83 workers in the retail food industry. All acute changes in forced expiratory volume in 1 s were standardized for lung size before the magnitude of the changes were compared between the workers exposed and unexposed to the use of hot wires for cutting plastic film. No association was found between acute drop in pulmonary function and either direct or indirect exposure in the absence of a history of asthma or allergy to inhaled materials. The borderline significance of an interaction term between exposure and asthma/allergy in a regression analysis suggests that workers with a history of asthma or atopy may have an acute response to hot-wire wrapping emissions.