Investing in prospective cohorts for etiologic study of occupational exposures.

Prospective cohorts have played a major role in understanding the contribution of diet, physical activity, medical conditions, and genes to the development of many diseases, but have not been widely used for occupational exposures. Studies in agriculture are an exception. We draw upon our experience using this design to study agricultural workers to identify conditions that might foster use of prospective cohorts to study other occupational settings. Prospective cohort studies are perceived by many as the strongest epidemiologic design. It allows updating of information on exposure and other factors, collection of biologic samples before disease diagnosis for biomarker studies, assessment of effect modification by genes, lifestyle, and other occupational exposures, and evaluation of a wide range of health outcomes. Increased use of prospective cohorts would be beneficial in identifying hazardous exposures in the workplace. Occupational epidemiologists should seek opportunities to initiate prospective cohorts to investigate high priority, occupational exposures.

Fungicide application practices and personal protective equipment use among orchard farmers in the agricultural health study.

Fungicides are routinely applied to deciduous tree fruits for disease management. Seventy-four private orchard applicators enrolled in the Agricultural Health Study participated in the Orchard Fungicide Exposure Study in 2002-2003. During 144 days of observation, information was obtained on chemicals applied and applicator mixing, application, personal protective, and hygiene practices. At least half of the applicators had orchards with <100 trees. Air blast was the most frequent application method used (55%), followed by hand spray (44%). Rubber gloves were the most frequently worn protective equipment (68% mix; 59% apply), followed by respirators (45% mix; 49% apply), protective outerwear (36% mix; 37% apply), and rubber boots (35% mix; 36% apply). Eye protection was worn while mixing and applying on only 35% and 41% of the days, respectively. Bivariate analyses were performed using repeated logistic or repeated linear regression. Mean duration of mixing, pounds of captan applied, total acres sprayed, and number of tank mixes sprayed were greater for air blast than for hand spray (p < 0.05). Spraying from a tractor/vehicle without an enclosed cab was associated with wearing some type of coverall (p < 0.05). Applicators often did not wash their hands after mixing (77%), a finding not explained by glove use. Glove use during mixing was associated with younger age, while wearing long-sleeve shirts was associated with older age (p < 0.05 each). Self-reported unusually high fungicide exposures were more likely on days applicators performed repairs (p < 0.05). These data will be useful for evaluating fungicide exposure determinants among orchard applicators.

Characterization of endotoxin and 3-hydroxy fatty acid levels in air and settled dust from commercial aircraft cabins.

Endotoxin was measured in air and dust samples collected during four commercial aircraft flights. Samples were analyzed for endotoxin biological activity using the Limulus assay. 3-hydroxy fatty acids (3-OH FA) of carbon chain lengths C10:0-C18:0 were determined in dust by gas chromatography-ion trap tandem mass spectrometry. The geometric mean (geometric standard deviation) endotoxin air level was 1.5 EU/m3 (1.9, n = 28); however, significant differences were found by flight within aircraft type. Mean endotoxin levels were significantly higher in carpet dust than in seat dust (140 +/- 81 vs. 51 +/- 25 EU/mg dust, n = 32 each, P < 0.001). Airborne endotoxin levels were not significantly related to either carpet or seat dust endotoxin levels. Mean 3-OH FA levels were significantly higher in carpet dust than in seat dust for C10:2, C12:0, and C14:0 (P < 0.001 for each), while the mean level of C16:0 was significantly higher in seat dust than in carpet dust (P < 0.01). Carpet dust endotoxin was significantly, but moderately, correlated with 3-OH-C12:0 and 3-OH-C14:0 (Pearson r = 0.52 and 0.48, respectively), while correlation of seat dust endotoxin with individual 3-OH FAs depended on the test statistic used. Mean endotoxin potency was significantly higher for carpet dust than for seat dust (6.3 +/- 3.0 vs. 3.0 +/- 1.4 EU/pmol LPS, P < 0.0001). Mean endotoxin levels in the air and dust of commercial aircraft cabins were generally higher than mean levels reported in homes and office buildings. These results suggest that exposure route and dust source are important considerations when relating endotoxin exposure to specific health outcomes.

Evaluation of acute sensory--motor effects and test sensitivity using termiticide workers exposed to chlorpyrifos.

Sensory and motor testing was performed on a group of termiticide workers primarily using chlorpyrifos-containing products to evaluate both the acute effects from current exposure and sensitivity of the measures to detect effects. The study group comprised 106 applicators and 52 nonexposed participants. Current exposure was measured by urinary concentrations of 3,5,6-trichloro-2-pyridinol (TCP) collected the morning of testing. The mean TCP value for the 106 applicators was 200 microg/g creatinine. Participants received 4--5 h of testing and were evaluated using a sensory--motor test battery recommended by a National Institute for Occupational Safety and Health (NIOSH)-sponsored advisory panel to be appropriate for testing effects from pesticide exposures. Measurements testing olfactory dysfunction, visual acuity, contrast sensitivity, color vision, vibrotactile sensitivity, tremor, manual dexterity, eye--hand coordination, and postural stability were analyzed. Study results indicated limited acute effects from exposure to chlorpyrifos using urinary TCP as a measure of current exposure. The effects occurred primarily on measures of postural sway in the eyes closed and soft-surface conditions, which suggests a possible subclinical effect involving the proprioceptive and vestibular systems. Several other tests of motor and sensory functions did not show any evidence of acute exposure effects, although statistically significant effects of urinary TCP on the Lanthony color vision test scores and one contrast sensitivity test score were found. The visual measures, however, were not significant when a step-down Bonferroni correction was applied. Information also is presented on the sensitivity of the measures to detect effects in an occupationally exposed population using standard error of the parameter estimates.

Development of sampling and analytical methods for concerted determination of commonly used chloroacetanilide, chlorotriazine, and 2,4-D herbicides in hand-wash, dermal-patch, and air samples.

Sampling and analytical methods were developed for commonly used chloroacetanilide, chlorotriazine, and 2,4-D herbicides in hand washes, on dermal patches, and in air. Eight herbicides selected for study were alachlor, atrazine, cyanazine, 2,4-dichlorophenoxyacetic acid (2,4-D), metolachlor, simazine, and two esters of 2,4-D, the 2-butoxyethyl ester (2,4-D, BE) and the 2-ethylhexyl ester (2,4-D, EH). The hand-wash method consisted of shaking the worker's hand in 150 mL of isopropanol in a polyethylene bag for 30 seconds. The dermal-patch method entailed attaching a 10-cm x 10-cm x 0.6-cm polyurethane foam (PUF) patch to the worker for exposure; recovery of the herbicides was achieved by extraction with 40 mL of isopropanol. The air method involved sampling with an OVS-2 tube (which contained an 11-mm quartz fiber filter and two beds of XAD-2 resin) and recovery with 2 mL of 10:90 methanol:methyl t-butyl ether. Analysis of each of the three sample types was performed by gas chromatography with an electron-capture detector. Diazomethane in solution was employed to convert 2,4-D as the free acid to the methyl ester in each of the three methods for ease of gas chromatography. Silicic acid was added to sample solutions to quench excess diazomethane. Limits of detection for all eight herbicides were matrix-dependent and, generally, less than 1 microgram per sample for each matrix. Sampling and analytical methods met NIOSH evaluation criteria for all herbicides in hand-wash samples, for seven herbicides in air samples (all herbicides except cyanazine), and for six herbicides in dermal-patch samples (all herbicides except cyanazine and 2,4-D). Speciation of 2,4-D esters and simultaneous determination of 2,4-D acid were possible without losses of the esters or of other herbicides (acetanilides and triazines) being determined.

Determinants of chlorpyrifos exposures and urinary 3,5,6-trichloro-2-pyridinol levels among termiticide applicators.

The exposures and work activities of 41 applicators in North Carolina using chlorpyrifos-containing termiticides were characterized. Personal air and urine samples were collected on multiple days within one week. Detailed information about chemical use, tasks, personal protective equipment and hygiene was recorded. During the 202 applicator-days monitored, 415 treatment jobs were performed. Full-shift chlorpyrifos exposures ranged from <0.048 to 110 microg/m(3) (N=184), with a geometric mean (GM) of 10 microg/m(3). Urinary 3,5,6-trichloro-2-pyridinol (TCP) levels ranged from 9.42 to 1960 microg/g creatinine (N=271) and varied significantly by day of the week (GM range: 169-262 microg/g creatinine). Predictive models for chlorpyrifos air exposure and urinary TCP levels were developed using mixed-effects stepwise linear regression. Determinants of airborne chlorpyrifos exposure included minutes chlorpyrifos applied and enclosed crawl space treated (yes/no). Determinants of TCP levels (depending on the model) included day-of-the-week, the chlorpyrifos air concentration one and two days before urine collection, minutes of chlorpyrifos applied one and two days before urine collection, enclosed crawl space treated (yes/no), and commercial structure treated (time-weighted). Within- and between-worker variability was similar for airborne chlorpyrifos; however, for TCP, between-worker variability exceeded within-worker variability by six times. The elimination half-life of TCP (26.9 h) and possibly the short sampling interval (one week) may explain the low TCP within-worker variability. Applicators' weekly mean ln(TCP levels) and weekly mean ln(chlorpyrifos air concentrations) were highly and positively linearly correlated (r(2)=0.73, P<0.0001). In summary, mixed-effects models were successfully constructed to predict airborne chlorpyrifos exposure and urinary TCP levels.

Distributions and determinants of pre-emergent herbicide exposures among custom applicators.

Custom applicators intensively apply herbicides to corn and soybean fields each spring. The primary objective of this study was to characterize the exposure distributions of the herbicides alachlor, atrazine, 2,4-D 2-ethylhexyl ester (2,4-D EH), and metolachlor among a group of applicators during the spring pre-emergent spray season. A secondary objective was to evaluate determinants of exposure and to estimate within- and between-worker variance components. Fifteen applicators were sampled using a systematic design that included spray and non-spray days and multiple measurements (five to seven) on each applicator. Air, patch, and handwash samples were collected on 89 applicator-days. Applicator-days were classified into three categories: target herbicide sprayed, non-target herbicide sprayed, and no herbicide sprayed. Mixed-model regression analysis was used. For all exposure metrics, adjusted mean herbicide exposures were significantly higher on days when target herbicides were sprayed as compared to non-spray days. For 2,4-D EH only, adjusted mean exposures on non-target herbicide spray days were significantly higher than on non-spray days. Wearing gloves significantly reduced adjusted mean hand exposure for all herbicides (4-20 fold) and adjusted mean thigh exposure for three herbicides (8-53 fold) on days the herbicides were sprayed; however, wearing gloves significantly increased adjusted mean atrazine hand and thigh exposures (9 and 7 fold, respectively) on days that non-atrazine herbicides were sprayed. Few of the other covariates were consistent determinants of exposure. For all exposure metrics, the within-worker variability (GSD(W) 2.1-5.6) was greater than the between-worker variability (GSD(B) 1.2-2.7).

Saliva biomonitoring of atrazine exposure among herbicide applicators.

A field study was conducted in which saliva samples were collected from a cohort of herbicide applicators during the pre-emergent spray season in Ohio in 1996. Atrazine concentrations were detected in human saliva samples using an enzyme-linked immunosorbent assay (ELISA) method. Trend due to atrazine exposure and subsequent elimination in the body were evidenced by the temporal pattern of decreasing atrazine concentrations in saliva over time. Median salivary concentrations of atrazine on non-spray days were significantly lower than on spray days for each sampling time (Mann-Whitney U-Wilcoxon rank sum test, P < 0.01). Within spray days, median salivary atrazine concentrations were significantly higher on days atrazine was sprayed than on days herbicides other than atrazine were sprayed for each sampling time (Mann-Whitney U-Wilcoxon rank sum test, P = 0.02 for 4 6 p.m. samples, P = 0.04 for bedtime samples, P = 0.03 for next-morning samples). Median salivary atrazine concentrations on days atrazine was sprayed were higher than the median concentration for the corresponding sampling time on non-spray days and on days when other herbicides were sprayed. Salivary concentration of atrazine is a plausible indicator of those days in which atrazine spraying was likely to have occurred. Salivary concentrations of atrazine not only reflect exposures resulting from spraying atrazine, but also exposures from other field activities where applicators may come in contact with atrazine. The results of this study confirmed data from animal experiments that atrazine is able to cross the cell membranes of salivary glands, and can be measured in human saliva with high sensitivity. The sampling method itself is convenient and easy to use in the field, with a high compliance rate, and analytical procedures are rapid and inexpensive. It is, therefore, concluded that saliva sampling of atrazine exposure among herbicide applicators is a feasible biomonitoring method.

Characterization and variability of endotoxin and 3-hydroxy fatty acids in an office building during a particle intervention study.

Air and dust samples were collected on two floors of an office building during a double-blind particle intervention study to examine spatial and temporal variability of airborne endotoxin over a period of weeks, and to characterize endotoxin activity and lipopolysaccharide (LPS) content in carpet and chair dust. Air samples were collected on multiple days within and across weeks. Dust samples were collected from carpets and chairs one day per week for three weeks. Endotoxin was measured using a Limulus assay. Dust samples were analyzed for LPS by determination of 3-hydroxy fatty acids (3-OHFAs) using gas chromatography-mass spectrometry. The geometric mean (geometric standard deviation) for 96 indoor air samples was 0.24 (1.6) EU/m3. Significant within-floor spatial variation of airborne endotoxin was found (P < 0.0001, n = 80). Temporal variability of airborne endotoxin was not significant across weeks. Mean (+/- SD) endotoxin levels in carpet dust (59 +/- 9.3 EU/mg dust, n = 12) and in chair dust (38 +/- 7.7 EU/mg dust, n = 10) were significantly different (P < 0.001). Carbon chain length-dependent differences in 3-OHFA levels by dust source and floor were found. Enhanced air filtration did not significantly affect airborne endotoxin (P = 0.62); however, total dust mass and total endotoxin in carpet dust samples increased significantly after enhanced surface cleaning (P < 0.01). These findings suggest that spatial variability, dust source, and surface cleaning may influence building occupant exposures to endotoxin.

Neurologic function among termiticide applicators exposed to chlorpyrifos.

Chlorpyrifos is a moderately toxic organophosphate pesticide. Houses and lawns in the United States receive a total of approximately 20 million annual chlorpyrifos treatments, and 82% of U.S. adults have detectable levels of a chlorpyrifos metabolite (3,5, 6-trichloro-2-pyridinol; TCP) in the urine. The U.S. Environmental Protection Agency has estimated that there are 5,000 yearly reported cases of accidental chlorpyrifos poisoning, and approximately one-fourth of these cases exhibit symptoms. Organophosphates affect the nervous system, but there are few epidemiologic data on chlorpyrifos neurotoxicity. We studied neurologic function in 191 current and former termiticide applicators who had an average of 2.4 years applying chlorpyrifos and 2.5 years applying other pesticides, and we compared them to 189 nonexposed controls. The average urinary TCP level for 65 recently exposed applicators was 629.5 microg/L, as compared to 4.5 microg/L for the general U.S. population. The exposed group did not differ significantly from the nonexposed group for any test in the clinical examination. Few significant differences were found in nerve conduction velocity, arm/hand tremor, vibrotactile sensitivity, vision, smell, visual/motor skills, or neurobehavioral skills. The exposed group did not perform as well as the nonexposed group in pegboard turning tests and some postural sway tests. The exposed subjects also reported significantly more symptoms, including memory problems, emotional states, fatigue, and loss of muscle strength; our more quantitative tests may not have been adequate to detect these symptoms. Eight men who reported past chlorpyrifos poisoning had a pattern of low performance on a number of tests, which is consistent with prior reports of chronic effects of organophosphate poisoning. Overall, the lack of exposure effects on the clinical examination was reassuring. The findings for self-reported symptoms raise some concern, as does the finding of low performance for those reporting prior poisoning. Although this was a relatively large study based on a well-defined target population, the workers we studied may not be representative of all exposed workers, and caution should be exercised in generalizing our results.

Personal fluoride and solvent exposures, and their determinants, in semiconductor manufacturing.

Personal air sampling for fluorides and solvents was done at 35 semiconductor fabrication facilities in the United States. Fluoride compounds were used in etching and cleaning operations, and solvents were used in photoresist and developing operations. All personal solvent and fluoride levels were less than 2 percent of current Occupational Safety and Health Administration (OSHA) standards. Statistical models of the exposure determinants for the target agents found production level, as indicated by number of semiconductor wafer cassettes loaded/unloaded from the target machines or baths, was predictive of fluoride, xylene and 1-methoxy-2-propyl acetate exposures. The percent of fresh air ventilation and the percent of xylene in the photoresist were also significant determinants in the statistical model predicting personal xylene exposure levels.

Comparison of three methods for assessment of hand exposure to azinphos-methyl (Guthion) during apple thinning.

Hand exposures of apple thinners to the pesticide azinphos-methyl (Guthion) were measured using three methods (glove, handwash, and wipe). Hand exposure sampling for each method was conducted following apple thinning work for a period of two hours for six to eight workers. Foliar residue samples were collected on each day of hand exposure sampling in the same orchard blocks that were thinned; foliar residues are considered to have been constant during the four-day study, which took place, on average, six days after pesticide application. Hand exposure estimates derived from each of the three methods differed significantly (ANOVA: p < 0.001). Mean measured exposure rates for the glove, handwash, and wipe methods were 6.48, 1.83, and 0.28 mg/hr, respectively. A corrected estimate of hand exposure, 2.7 mg/hr, was calculated from the handwash measurements and the handwash removal efficiency factor from a laboratory study. Comparison with this hand exposure estimate suggests that the glove method produced a 2.4-fold overestimate of exposure, whereas the wipe method produced a 10-fold underestimate. Studies that measure hand exposure to pesticides should include a careful description of sampling methods and should recognize the potential for measurement bias. Furthermore, the standardization and validation of dermal exposure assessment methods are critical to developing more comparable and more accurate pesticide exposure estimates.

Tiered exposure-assessment strategy in the Semiconductor Health Study.

The Semiconductor Health Study included 14 U.S. companies with 82 silicon-based wafer-fabrication rooms and a complex array of processes and chemicals. Its epidemiologic components were a historical examination of spontaneous abortion rates among 902 women, a prospective evaluation of reproductive outcomes in 483 women, and a cross-sectional review of male fertility and of respiratory, ergonomic, neurologic, or gastrointestinal problems among 3,175 men and women. Designing an exposure assessment strategy presented unique problems, and multiple agents had to be evaluated. A three-tiered approach to exposure assessment was developed to reflect increasing specificity of exposures. At the first tier, employees were divided into fabrication and nonfabrication groups. At the second tier, work groups with qualitatively different exposures were determined. At the third tier, intensity of exposures to specific chemical, physical, and ergonomic agents was evaluated. Evaluations were based on worker reports of tasks performed, moderated by fabrication-specific factors observed by study industrial hygienists during site visits.

Patterns of chemical use and exposure control in the Semiconductor Health Study.

Information on chemical use and exposure control between 1986 and 1990 was collected from 14 companies participating in the Semiconductor Health Study. Questionnaires and site visits provided data used to develop exposure categories for three epidemiological studies: prospective, historical, and cross-sectional. Patterns of use of target chemicals were compiled for 82 silicon-wafer fabrication rooms (fabs), including 47 from which subjects were selected for study. Chemical use was examined by operation, year, and epidemiological component. Target agents for epidemiological analyses were present in more than 50% of fabs. Use of these agents was fairly constant from 1986 to 1990, except for a moderate increase in use of propylene glycol monomethyl ether acetate, a solvent being substituted for ethylene-based glycol ethers (EGE) in photoresists. The distribution of personal protective equipment, engineering controls, and other factors potentially affecting employee exposure was also examined. Controls designed to manage processes or high acute toxicity were present in most fabs; their prevalence remained unchanged from 1986 through 1990. Controls designed to reduce exposures to chemicals with low acute toxicity were less widely distributed; their prevalence increased moderately from 1986 to 1990.

Algorithms for estimating personal exposures to chemical agents in the Semiconductor Health Study.

For exposure assessment in a series of epidemiologic studies of semiconductor industry workers, estimates of potential personal exposures to selected chemical agents were developed using a set of algorithms. A worker's total exposure to each agent was defined as the sum of task scores for each task using that agent. Task scores were calculated as the intensity and frequency score for each task, modified by a factor indicating degree of presence of the target agent and a process-specific emission factor representing degree of emission control present for that process. The algorithms used multiple sources of information to generate an exposure score for each worker-process-agent combination. Exposure scores were then placed into ordinal categories of exposure (0-3) for use in the epidemiologic analysis.

Hierarchical cluster analysis for exposure assessment of workers in the Semiconductor Health Study.

The fabrication of integrated circuits in the semiconductor industry involves worker exposures to multiple chemical and physical agents. The potential for a high degree of correlation among exposure variables was of concern in the Semiconductor Health Study. Hierarchical cluster analysis was used to identify groups or "clusters" of correlated variables. Several variations of hierarchical cluster analysis were performed on 14 chemical and physical agents, using exposure data on 882 subjects from the historical cohort of the epidemiological studies. Similarity between agent pairs was determined by calculating two metrics of dissimilarity, and hierarchical trees were constructed using three clustering methods. Among subjects exposed to ethylene-based glycol ethers (EGE), xylene, or n-butyl acetate (nBA), 83% were exposed to EGE and xylene, 86% to EGE and nBA, and 94% to xylene and nBA, suggesting that exposures to EGE, xylene, and nBA were highly correlated. A high correlation was also found for subjects exposed to boron and phosphorus (80%). The trees also revealed cluster groups containing agents associated with work-group exposure categories developed for the epidemiologic analyses.

Historical cohort study of spontaneous abortion among fabrication workers in the Semiconductor Health Study: agent-level analysis.

Risk of spontaneous abortion (SAB) was examined in relation to chemical and physical agents in a retrospective study of employees of 14 seminconductor manufacturers: After screening over 6,000 employees, 506 current and 385 former workers were eligible. If a woman had multiple eligible pregnancies, one was selected at random. Telephone interviews provided data on demographics and occupational and other exposures during the first trimester. Two groups of chemicals accounted for the 45% excess risk of SAB among fabrication-room (fab) workers: photoresist and developed solvents (PDS), including glycol ethers, and fluoride compounds used in etching. Women exposed to high levels of both these agents were at greater risk (RR = 3.21, 95% confidence interval [CI] = 1.29-5.96). In fab workers without these exposures, SAB rates were not elevated (adjusted relative risk [RR] = 0.98, 95% CI = 0.55-1.69). An association was seen with workplace stress, which was not limited to women exposed to PDS or fluoride, nor did stress explain the associations between these chemicals and SAB.