The issue of confounding in epidemiological studies of ambient air pollution and pregnancy outcomes

BACKGROUND

Relationships between ambient air pollution levels during pregnancy and adverse pregnancy outcomes have been investigated using one of three analytic approaches: ambient pollution levels have been contrasted over space, time or both space and time. Although the three approaches share a common goal, to estimate the causal effects of pollution on pregnancy outcomes, they face different challenges with respect to confounding.

METHODS

A framework based on counterfactual effect definitions to examine issues related to confounding in spatial, temporal, and spatial-temporal analyses of air pollution and pregnancy outcomes is presented, and their implications for inference are discussed.

RESULTS

In spatial analyses, risk factors that are spatially correlated with pollution levels are confounders; the primary challenges relate to the availability and validity of risk factor measurements. In temporal analyses, where smooth functions of time are commonly used to control for confounding, concerns relate to the adequacy of control and the possibility that abrupt changes in risk might be systematically related to pollution levels. Spatial-temporal approaches are subject to challenges faced in both spatial and temporal analyses.

CONCLUSION

Each approach faces different challenges with respect to the likely sources of confounding and the ability to control for that confounding because of differences in the type, availability, and quality of information required. Thoughtful consideration of these differences should help investigators select the analytic approach that best promotes the validity of their research.

A case-control study of tobacco use and other non-occupational risk factors for t(14;18) subtypes of non-Hodgkin's lymphoma (United States)

OBJECTIVE

Non-Hodgkin's lymphoma (NHL) encompasses diverse subtypes, and analyzing NHL as a single outcome may mask associations. In a new approach we evaluated associations with subtypes defined by the t(14;18) translocation, reasoning that cases within these subtypes would have more common risk factors than all NHL combined.

METHODS

Archival biopsies from cases in a population-based NHL study were assayed for t(14;18) using polymerase chain reaction amplification. Exposures in 68 t(14;18)-positive and 114-negative cases were compared with 1245 controls. The expectation-maximization algorithm was used to fit polytomous regression models based on all available information, including data from 440 unclassified cases.

RESULTS

Family history of hemolymphatic cancer was associated with t(14;18)-negative NHL (odds ratio (OR) 2.4, 95% confidence interval (CI) 1.4 3.9). but not t(14;18)-positive NHL. Cigarette smoking was weakly associated with t(14;18)-positive NHL (OR 1.7, CI 0.9-3.3), but ORs decreased as smoking increased. Chewing tobacco was associated with t(14;18)-positive NHL, particularly when used before age 18 (OR 2.5. CI 1.0-6.0, 13 exposed cases). Odds ratios for both case-subtypes were doubled among hair-dye users.

CONCLUSIONS

Cigarette smoking was not clearly associated with t(14;18)-positive NHL. Family history may be a marker for factors that act specifically through t(14;18)-negative pathogenic mechanisms.

Breast-feeding among women exposed to polybrominated biphenyls in Michigan

In the early 1970s, the largest industrial accident in the United States resulted in widespread contamination of the food supply in Michigan with polybrominated biphenyls (PBBs). The chemical similarity of PBBs to compounds implicated as endocrine disruptors has raised the question of whether PBBs could affect the reproductive system. In the present analysis we examine the relation between serum measurements of PBBs and the frequency and duration of lactation. Persons who lived on or received food from farms exposed to PBBs were enrolled in a registry by the Michigan Department of Public Health. Female members of the cohort were invited to participate in a telephone survey of reproductive outcomes. The three outcomes of interest in the present analysis were a) the decision to breast-feed (yes/no); b) the duration, in months, of breast-feeding as the main source of nutrition; and c) the total duration, in months, of breast-feeding. None of the three outcomes was significantly associated with serum PBB levels, even after controlling for maternal age, previous history of breast-feeding, body mass index, maternal education, household income, history of smoking in the year before pregnancy, consumption of alcohol during the first trimester of pregnancy, history of thyroid disorder, gestational age of the infant in weeks, time to pregnancy, and year of birth.

Agricultural risk factors for t(14;18) subtypes of non-Hodgkin's lymphoma

The t(14;18) translocation is a common somatic mutation in non-Hodgkin's lymphoma (NHL) that is associated with bcl-2 activation and inhibition of apoptosis. We hypothesized that some risk factors might act specifically along t(14;18)-dependent pathways, leading to stronger associations with t(14;18)-positive than t(14;18)-negative non-Hodgkin's lymphoma. Archival biopsies from 182 non-Hodgkin's lymphoma cases included in a case-control study of men in Iowa and Minnesota (the Factors Affecting Rural Men, or FARM study) were assayed for t(14;18) using polymerase chain reaction amplification; 68 (37%) were t(14;18)-positive. We estimated adjusted odds ratios (OR) and 95% confidence intervals (CI) for various agricultural risk factors and t(14;18)-positive and -negative cases of non-Hodgkin's lymphoma, based on polytomous logistic regression models fit using the expectation-maximization (EM) algorithm. T(14;18)-positive non-Hodgkin's lymphoma was associated with farming (OR 1.4, 95% CI = 0.9-2.3), dieldrin (OR 3.7, 95% CI = 1.9-7.0), toxaphene (OR 3.0, 95% CI = 1.5-6.1), lindane (OR 2.3, 95% CI = 1.3-3.9), atrazine (OR 1.7, 95% CI = 1.0-2.8), and fungicides (OR 1.8, 95% CI = 0.9-3.6), in marked contrast to null or negative associations for the same self-reported exposures and t(14;18)-negative non-Hodgkin's lymphoma. Causal relations between agricultural exposures and t(14;18)-positive non-Hodgkin's lymphoma are plausible, but associations should be confirmed in a larger study. Results suggest that non-Hodgkin's lymphoma classification based on the t(14;18) translocation is of value in etiologic research.

Agricultural exposure history among African-American farmers in Georgia

Agricultural exposures differ across the United States by region, calendar time period, and agricultural practice, but most of the published literature focuses on white men in the Midwest. A pilot study was conducted to explore the breadth and diversity of farming practices over time among African-American farmers in Georgia whose exposures may differ in important ways. Using a comprehensive life events calendar questionnaire, 17 male African-American farmers aged 36 to 86 yr residing in southeastern Georgia were interviewed regarding their agricultural history in July 1997. Most men (15/17) reported working on multiple farms in their lifetime; 3 men worked on 5 different farms during their lifetime. These farmers reported using more chemicals during their lifetime than farmers in the Midwest. Used motor oil was the most frequently reported insecticide applied to animals; this apparently common practice has not been described in the literature and should be better understood since its use may result in dermal exposure to polyaromatic hydrocarbons. Better characterization of regionally specific farming history and individual farming practices will facilitate studies of the health effects of farming.

Age at menarche and tanner stage in girls exposed in utero and postnatally to polybrominated biphenyl

Accidental contamination of the Michigan food chain with polybrominated biphenyls (PBBs) led to the exposure of more than 4,000 individuals in 1973. Because PBB exposure is suspected to disrupt endocrine function, we assessed pubertal development in females 5-24 years of age (N = 327) who were exposed to PBB in utero and, in many cases, through breastfeeding. We estimated in utero PBB exposure using maternal serum PBB measurements taken after exposure (1976-1979) and extrapolated to time of pregnancy using a model of PBB decay. We found that breastfed girls exposed to high levels of PBB in utero (> or =7 parts per billion) had an earlier age at menarche (mean age = 11.6 years) than breastfed girls exposed to lower levels of PBB in utero (mean age = 12.2-12.6 years) or girls who were not breastfed (mean age = 12.7 years). This association persisted after adjustment for potential confounders (menarche ratio = 3.4, 95% confidence interval = 1.3-9.0). Perinatal PBB exposure was associated with earlier pubic hair stage in breastfed girls, but little association was found with breast development. The associations observed here lend support to the hypothesis that pubertal events may be affected by pre- and postnatal exposure to organohalogens.

K-ras and p53 in pancreatic cancer: association with medical history, histopathology, and environmental exposures in a population-based study

Pancreatic cancer is a highly fatal cancer with few identified risk factors. Increased risk of pancreatic cancer in tobacco smokers and among diabetic patients is well established, and some reports have suggested associations with coffee consumption and occupational exposure to organochlorines. At present, there is little information regarding the possible association of these risk factors with the known genetic alterations found in pancreatic cancers, such as activation of the K-ras oncogene and inactivation of the p53 tumor suppressor gene. Knowledge of such relationships may help to understand the molecular pathways of pancreatic tumorigenesis. We investigated the association between these molecular defects and risk factors for pancreatic cancer in 61 newly diagnosed patients identified through an ongoing study of pancreatic cancer in the San Francisco Bay Area. Interview information was obtained regarding environmental exposures, medical history, and demographic factors. Serum levels of dichlorodiphenyltrichloroethylene (DDE) and polychlorinated biphenyls were available on a subset of 24 patients. Tumor blocks were located from local hospitals and used for K-ras mutational analysis at codon 12 and for p53 protein immunohistochemistry. The molecular analyses were facilitated through the use of laser capture microdissection, which provides a reliable method to obtain almost pure populations of tumor cells. Mutations in K-ras codon 12 were found in 46 (75%) of 61 pancreatic cancers. A prior diagnosis of diabetes was significantly associated with K-ras negative tumors (P = 0.002, Fisher's exact test). The absence of this mutation was also associated with increased serum levels of DDE, although this association was not statistically significant (P = 0.16, Wilcoxon's test). There was no difference in polychlorinated biphenyl levels between the K-ras wild-type and mutant groups. Immunohistochemical staining for p53 protein did not differ by patient characteristics or clinical history, but significant associations were found with poor glandular differentiation (P = 0.002, chi2 trend test), severe nuclear atypia (P = 0.0007, chi2 trend test), and high tumor grade (P = 0.004, chi2 trend test). Our results are suggestive of the presence of K-ras codon 12 mutation-independent tumorigenesis pathways in patients with prior diabetes and possibly in patients with higher serum levels of DDE. Our results also support a role for the p53 tumor suppressor protein in the maintenance of genomic integrity.

Measurement of PCBs, DDE, and hexachlorobenzene in cord blood from infants born in towns adjacent to a PCB-contaminated waste site

There are limited data on the concentrations of common contaminants--polychlorinated biphenyls (PCBs), dichlorodiphenyl dichloroethene (pp'-DDE) and hexachlorobenzene (HCB)--in umbilical cord blood. Cord blood provides the primary direct measure of prenatal exposure to these contaminants, the key determinant of PCBs' neurodevelopmental toxicities. The objective of this study was to characterize cord blood levels of PCBs, pp'-DDE, and HCB among 751 infants who were born between 1993 and 1998 to mothers residing adjacent to a PCB-contaminated harbor in New Bedford, Massachusetts, and for whom the neurodevelopmental toxicities of these compounds are being studied. We refined standard analytic methods to optimize the sensitivity and precision of trace-level PCB, p,p'-DDE, and HCB measurements in blood. Using these methods, we measured the concentrations of 51 individual PCBs, their sum (sum(PCB)), p,p'-DDE, and HCB in cord serum. With correction for background contamination, the respective mean+/-SD cord serum concentrations of sum(PCB), p,p'-DDE, and HCB were 0.54+/-0.83, 0.48+/-0.94, and 0.03+/-0.04 ng/g serum. These concentrations were generally lower than those in most of the few published studies with congener-specific measures of PCBs in cord blood. However, for less-chlorinated PCB congeners (e.g., congeners 99 and 118), study samples had concentrations comparable to those in other populations, including groups at risk for high dietary PCB exposure. Of note, the contaminated harbor sediment has a relatively high proportion of less-chlorinated PCB congeners. Thus, although the sum(PCB) in study infants was not higher than concentrations in infants studied elsewhere, the relative predominance of less-chlorinated congeners was generally consistent with the characteristics of the contaminated site.

Interim results of the study of particulates and health in Atlanta (SOPHIA)

Substantial evidence supports an association of particulate matter (PM) with cardiorespiratory illnesses, but little is known regarding characteristics of PM that might contribute to this association and the mechanisms of action. The Atlanta superstation sponsored by the Electric Power Research Institute as part of the Aerosol Research and Inhalation Epidemiology Study (ARIES) study is monitoring chemical composition of ambient particles by size fraction, as well as a comprehensive suite of other pollutants, at a site in downtown Atlanta during the 25-month period, August 1, 1998-August 31, 2000. Our investigative team is making use of this unique resource in several morbidity studies, called the "Study of Particulates and Health in Atlanta (SOPHIA)". The study includes the following components: (1) a time series investigation of emergency department (ED) visits for the period during which the superstation is operating; (2) a time series investigation of ED visits during the 5 years prior to implementation of the superstation; and (3) a study of arrhythmic events in patients equipped with automatic implantable cardioverter defibrillators (AICDs) for the period January 1, 1993-August 31, 2000. Thirty-three of 39 Atlanta area EDs are participating in the ED studies, comprising over a million annual ED visits. In this paper, we present initial analyses of data from 18 of the 33 participating EDs. The preliminary data set includes 1,662,713 ED visits during the pre-superstation time period and 559,480 visits during the superstation time period. Visits for four case groupings--asthma, chronic obstructive pulmonary disease (COPD), dysrhythmia, and all cardiovascular diseases (CVDs) combined--have been assessed relative to daily air quality indices, controlling for long-term temporal trends and meteorologic variables, using general linear models, generalized estimating equations and generalized additive models. Single-pollutant models predicting case visitation rates using moving averages of 0-, 1-, and 2-day lagged air quality variables were run. For the pre-superstation period, PM10 (24-h), ozone (8-h), SO2 (1-h), NO2 (1-h) and CO (1-h) were studied. For the first 12 months of superstation operation, the following air quality variables of a priori interest were available: ozone (8-h), NO2 (1-h), SO2 (1-h), CO (1-h), and 24-h measurements of PM10, coarse PM (PM 2.5-10 microm), PM2.5, polar VOCs, 10-100 nm particulate count and surface area, and in the PM2.5 fraction: sulfates, acidity, water-soluble metals, organic matter (OM), and elemental carbon (EC). During the pre-superstation time period, statistically significant, positive associations were observed for adult asthma with ozone, and for COPD with ozone, NO2 and PM10. During the superstation time period, the following statistically significant, positive associations were observed: dysrhythmia with CO, coarse PM, and PM2.5 EC; and all CVDs with CO, PM2.5 EC and PM2.5 OM. While covariation of many of the air quality indices limits the informativeness of this analysis, the study provides one of the first assessments of PM components in relation to ED visits.

Electric blanket use and breast cancer in the Nurses' Health Study

Electric and magnetic fields (EMFs) have been hypothesized to increase the risk of breast cancer, and electric blankets represent an important source of exposure to EMFs. The authors examined the relation between electric blanket use and invasive breast cancer in the Nurses' Health Study. On the biennial questionnaire in 1992, 87,497 women provided information on this exposure during three consecutive time periods. In a prospective analysis with 301,775 person-years of follow-up through 1996 (954 cases), the relative risk for any electric blanket use was not elevated (relative risk (RR) = 1.08, 95% confidence interval (CI): 0.95, 1.24) after controlling for breast cancer risk factors. There was a weak association between breast cancer and electric blanket use at least 16 years before diagnosis and long-term use in age-adjusted analyses but not in multivariate models. In a retrospective analysis of 1,318,683 person-years of follow-up (2,426 cases), the multivariate relative risk associated with use before disease follow-up began was null (RR = 1.05, 95% CI: 0.95, 1.16). Similar results were obtained in analyses stratified by menopause and restricted to estrogen receptor-positive breast cancers. While 95% confidence intervals for these estimates did not exclude small risks, overall, results did not support an association between breast cancer risk and exposure to EMFs from electric blankets.

Case-control assessment of the association between non-Hodgkin's lymphoma and occupational radiation with doses assessed using a job exposure matrix

BACKGROUND

Epidemiologic data for an association between radiation exposure and non-Hodgkin's lymphoma (NHL) have been inconclusive though the strongest evidence has been provided by studies of patients treated with radiotherapy.

METHODS

We evaluated the association between occupational radiation exposure and non-Hodgkin's lymphoma in men using a population-based case-control study with 1,056 case and 1,860 control subjects sampled from eight geographic areas in the United States. Because dosimetry data were not available, doses were estimated for individuals who reported occupational radiation exposure using a radiation job exposure matrix developed for this purpose. Conditional logistic regression was used to model the association between reported occupational radiation exposure and NHL incidence.

RESULTS

We found that most men (> 90%) did not report exposure to occupational sources of radiation. Among those who reported exposure, estimated cumulative doses were low, with an estimated mean of less than 0.02 Gray and a maximum of 0.12 Gray. The risk for NHL was not associated with ever having reported an occupational radiation exposure (OR = 0.90, 95% CI = 0.74-1.10) nor was there evidence of a dose-response relationship between risk and either the estimated cumulative doses or duration of exposure.

CONCLUSIONS

The findings in this study are consistent with results from most current research on occupational radiation and NHL risk that have found no increased risk of NHL at low levels of occupational radiation exposure. While it should be noted that exposure misclassification likely biased our results toward the null, this large population-based case-control study adds to existing evidence which suggests that there is little to no increased risk for NHL associated with exposure to low levels of radiation such as that commonly found in many occupational settings.

Occupational exposure to chlorophenol and the risk of nasal and nasopharyngeal cancers among U.S. men aged 30 to 60

BACKGROUND

Elevated rates of nasal and nasopharyngeal cancers have been associated with wood-related occupational exposures, including chlorophenols, formaldehyde, and wood dust.

METHODS

Occupational information was obtained from 43 nasal carcinoma cases, 92 nasopharyngeal carcinoma cases, and 1909 controls, by interview. Exact conditional logistic regression was used to evaluate the association of these cancers with chlorophenol exposure, estimated from a review of verbatim responses.

RESULTS

Both nasal and nasopharyngeal cancers were significantly associated with estimated duration of chlorophenol exposure. For nasopharyngeal cancer, elevated risk was observed among those who held jobs assigned medium or high intensity chlorophenol exposure (n(exposed)=18, OR=1.94, 95% CI=1.03-3.50) and among those with 10+ years in jobs assigned high intensity with high certainty (n(exposed)=3, OR=9.07, 95% CI=1.41-42. 9). Controlling for estimated formaldehyde and wood dust exposure did not alter these findings, as much of the estimated chlorophenol exposure was among machinists.

CONCLUSIONS

These findings support the hypothesis that occupational exposure to chlorophenol is a risk factor for nasal and nasopharyngeal cancer, although the role of machining-related exposures warrants further assessment.

Air quality and pediatric emergency room visits for asthma in Atlanta, Georgia, USA

Pediatric emergency room visits for asthma were studied in relation to air quality indices in a spatio-temporal investigation of approximately 130,000 visits (approximately 6,000 for asthma) to the major emergency care centers in Atlanta, Georgia, during the summers of 1993-1995. Generalized estimating equations, logistic regression, and Bayesian models were fitted to the data. In logistic regression models comparing estimated exposures of asthma cases with those of the nonasthma patients, controlling for temporal and demographic covariates and using residential zip code to link patients to spatially resolved ozone levels, the estimated relative risk per 20 parts per billion (ppb) increase in the maximum 8-hour ozone level was 1.04 (p < 0.05). The estimated relative risk for particulate matter less than or equal to 10 microm in aerodynamic diameter (PM10) was 1.04 per 15 microg/m3 (p < 0.05). Exposure-response trends (p < 0.01) were observed for ozone (>100 ppb vs. <50 ppb: odds ratio = 1.23, p = 0.003) and PM10 (>60 microg/m3 vs. <20 microg/m3: odds ratio = 1.26, p = 0.004). In models with ozone and PM10, both terms became nonsignificant because of collinearity of the variables (r= 0.75). The other analytical approaches yielded consistent findings. This study supports accumulating evidence regarding the relation of air pollution to childhood asthma exacerbation.

Determinants of polybrominated biphenyl serum decay among women in the Michigan PBB cohort

Accidental contamination of the food chain in Michigan in 1973 with polybrominated biphenyls (PBBs) led to the establishment of a registry of exposed individuals in 1976. Serum was collected and analyzed for PBB at the time of enrollment and for targeted studies in the following years. We used the archived PBB data to study the elimination of PBB and to identify factors associated with elimination. A total of 380 women >= 16 years of age who had an initial PBB level of 2 ppb and at least two serum samples drawn when they were not pregnant were included in the analysis. The mean initial PBB level was 20.9 ppb (median 4) and mean time between the first and last measurement was 4.2 years (range 0.5-11.1). PBB was assumed to reach equilibrium in the body before substantial amounts were eliminated and before the first serum measurements were taken; therefore, the entire body was modeled as a single compartment for PBB with exponential decay. Subject-specific decay rate estimates were regressed on predictor variables including initial age, body mass index (BMI), smoking history, breast-feeding duration, and parity. In women with an initial PBB level < 10 ppb, the median half-life was 12.9 years; in those with > 10 ppb, the median half-life was 28.7 years. Decay was significantly slower among women with an initial BMI at or above the median (BMI >= 23). The calculated half-life values are estimates of decay and can be used to estimate body burden of PBB at various points in time other than at the time of serum collection.

Pancreatic cancer and serum organochlorine levels

Occupational exposure to p,p'-dichlorodiphenyltrichloroethane (DDT) has been associated with increased pancreatic cancer risk. We measured organochlorine levels in serum obtained at the study enrollment from 108 pancreatic cancer cases and 82 control subjects aged 32-85 years in the San Francisco Bay Area between 1996 and 1998. Cases were identified using rapid case-ascertainment methods; controls were frequency-matched to cases on age and sex via random digit dial and random sampling of Health Care Financing Administration lists. Serum organochlorine levels were adjusted for lipid content to account for variation in the lipid concentration in serum between subjects. Median concentrations of p,p'-dichlorodiphenyldichloroethylene (DDE, 1290 versus 1030 ng/g lipid; P = 0.05), polychlorinated biphenyls (PCBs; 330 versus 220 ng/g lipid; P<0.001), and transnonachlor (54 versus 28 ng/g lipid; P = 0.03) were significantly greater among cases than controls. A significant dose-response relationship was observed for total PCBs (P for trend <0.001). Subjects in the highest tertile of PCBs (> or =360 ng/g lipid) had an odds ratio (OR) of 4.2 [95% confidence interval (CI) = 1.8-9.4] compared to the lowest tertile. The OR of 2.1 for the highest level of p,p'-DDE (95% CI = 0.9-4.7) diminished (OR = 1.1; 95% CI = 0.4-2.8) when PCBs were included in the model. Because pancreatic cancer is characterized by cachexia, the impact of this on the serum organochlorine levels in cases is difficult to predict. One plausible effect of cachexia is bioconcentration of organochlorines in the diminished lipid pool, which would lead to a bias away from the null. To explore this, a sensitivity analysis was performed assuming a 10-40% bioconcentration of organochlorines in case samples. The OR associated with PCBs remained elevated under conditions of up to 25% bioconcentration.

Estimating occupational radiation doses when individual dosimetry information is not available: a job exposure matrix

BACKGROUND

A job exposure matrix (JEM) was developed for a population based case control study to assess the possible association between occupational radiation and non-Hodgkin's lymphoma.

METHODS

Using published radiation monitoring data, we developed a radiation JEM composed of estimated annual dose distributions, categorized by time period, for a broad range of occupational and industrial groups.

RESULTS

When information is available to correctly assess an individual's exposure status, the annual dose distributions in the JEM can be used in conjunction with job histories to estimate the distribution of possible cumulative doses for individuals. The median of the cumulative dose distribution can then be used in standard epidemiologic analysis. In addition, methods can be applied that incorporate the uncertainty about each individual's true dose into risk estimates and associated confidence intervals.

CONCLUSIONS

The JEM can be useful in estimating occupational radiation exposures in other studies, particularly population based case control studies which include detailed occupational histories.

Occupational risk factors for sarcoma subtypes

Herbicides, chlorophenols, and other occupational exposures are suspected risk factors for soft-tissue sarcoma, but the epidemiologic evidence is inconsistent. Given that soft-tissue sarcomas represent a heterogeneous mix of cancer subtypes and that these subtypes have different disease patterns by race, sex, and age at diagnosis, studying all soft-tissue sarcomas combined may mask subtype-specific associations. Using the Selected Cancers Study, a large population-based case-control study of sarcoma conducted among U.S. men aged 30 to 60 in 1984 to 1988, we explored the occupational risk factors for soft-tissue sarcoma subtypes and skeletal sarcoma. The analysis included 251 living sarcoma cases (48 dermatofibrosarcoma protuberans, 32 malignant fibrohistiocytic sarcoma, 67 leiomyosarcoma, 53 liposarcoma, and 51 skeletal sarcoma) and 1908 living controls. Exact conditional logistic regression models suggested patterns of subtype specificity for occupational exposures. Self-reported herbicide use was associated with malignant fibrohistiocytic sarcoma (OR = 2.9, 95% CI = 1.1-7.3). We found elevated risks for chlorophenol exposure and cutting oil exposure and malignant fibrohistiocytic sarcoma and leiomyosarcoma. We found no occupational risk factor for liposarcoma. Polytomous regression models identified different odds ratios across subtypes for plywood exposure and exposure to wood and saw dust. Although exploratory, this analysis suggests that occupational risk factors for sarcoma are not uniform across subtypes.

Occupational chlorophenol exposure and non-Hodgkin's lymphoma

Occupational exposure to chlorophenols is suspected to increase non-Hodgkin's lymphoma (NHL) risk. This association was examined using data on 995 NHL cases and 1783 controls from the Selected Cancers Study, a population-based case-control study of men aged 32 to 60 years from eight population-based cancer registries conducted from 1984 to 1988. Potential chlorophenol exposure was characterized by an industrial hygienist using intensity estimates and confidence ratings, based upon review of verbatim job histories. Cases with substantial chlorophenol exposure had a significantly greater number of years of chlorophenol exposure (median years: cases, 4.0; controls, 2.0; P = 0.046); however, in conditional logistic regression models, the odds ratio for more than 8 years of substantial exposure was 1.51 (95% CI, 0.88 to 2.59). Overall, the findings do not provide strong support for an association with NHL risk. Chlorophenol exposure in this study is not based upon measured values and, therefore, may fail to characterize actual chlorophenol exposures accurately. Because of the large presence of machinists in the potentially chlorophenol-exposed group, these results may be underestimated by exposure misclassification if these subjects were not exposed to chlorophenolic biocides. However, these results are consistent with other findings, which suggest that chlorophenol exposure is not likely to be a strong risk factor for NHL.

Patterns of genetic alterations in pancreatic cancer: a pooled analysis

Both K-ras and p53 gene mutations are found commonly in pancreatic tumors. Analysis of the mutational patterns may provide insight into disease etiology. To further describe the mutational patterns of pancreatic cancer and to assess the evidence to date, we performed a pooled analysis of the published data on genetic mutations associated with pancreatic ductal adenocarcinoma. We included data from studies that evaluated point mutations in the two genes most studied in pancreatic cancer, K-ras and p53. A majority of the 204 tumors had mutations in at least one gene, with 29% having both K-ras and p53 mutations, 39% with K-ras mutation alone, and 16% having p53 mutation alone. Sixteen percent of tumors lacked mutation in either gene. K-ras mutations were present in high frequencies in all tumor grades (>69%). A statistically significant trend was observed for p53 mutation with higher tumor grade (P = 0.04). For K-ras, G2 and G3 grades, combined, had notably higher prevalences of mutation than G1 (P = 0.004). CGT mutations in K-ras codon 12 were marginally associated with lower tumor grade (P for trend = 0.09), and these tumors were somewhat less likely to have a p53 mutation than tumors with other K-ras mutations (P = 0.06). In the 59 K-ras+/p53+ tumors, 64% had the same type of mutation (transition or transversion) in both genes, suggesting a common mechanism. The mutational pattern of p53 in pancreatic cancer is similar to bladder cancer, another smoking-related cancer, but not to lung cancer. Analyses of molecular data, such as that performed here, present new avenues for epidemiologists in the study of the etiology of specific cancers.

Use of a life events calendar approach to elicit occupational history from farmers

BACKGROUND

Precise and valid exposure assessment is generally the primary challenge in retrospective occupational epidemiology studies, particularly when the only available method for exposure characterization is a personal interview. Agricultural workers may represent a particular challenge; for example, whereas many farmers have worked from childhood at the same location, raising the same crops and animals, they may have used different equipment, chemicals, and protective gear over time. One method to assist in recall is the "life events calendar," a cognitive tool based on the subject's own life history to help anchor occupational activities in time.

METHODS

Unstructured interviews of farmers, focus groups, and pilot interviews among rural men, primarily African-Americans, were conducted to create a questionnaire for obtaining farm history information within the context of personal life events.

RESULTS

Farmers used both personal events and national events (as well as events relating directly to farming) to recall their activities. These subjects had extensive history of farming (10-75 years) and chemical use (median lifetime chemicals = 13).

CONCLUSION

The life events calendar provided a useful tool to facilitate the recall of a lifetime of agricultural activity. Life events calendars are useful additions to the tools available for retrospective occupational exposure assessment.

Relation of cigarette smoking to non-Hodgkin's lymphoma among middle-aged men

Because of previous inconsistencies in the observed relation of cigarette smoking to non-Hodgkin's lymphoma, this association was investigated in the Selected Cancers Study, a population-based case-control study of 1,193 non-Hodgkin's lymphoma cases and 1,903 controls, conducted between 1984 and 1988. Study subjects were men, and the median age of non-Hodgkin's lymphoma cases was 50 years (range, 32-60 years). As compared with the risk among men who had never smoked cigarettes, the risk among ever smokers was not increased (odds ratio (OR) = 1.05, p approximately 0.50), but the risk was significantly elevated among men who reported smoking > or = 2 1/2 packs per day and among men who had smoked for 30-39 years (OR = 1.45 in each group, p < 0.05). The estimated odds ratio among the 350 heavy smokers (> or = 50 pack-years) was 1.41 (95% confidence interval 1.08-1.85) after controlling for educational achievement, various occupational and medical exposures, and other potential confounders. The observed associations, however, tended to vary by age, with the odds ratio among heavy smokers decreasing from 2.8 among 32- to 44-year-olds to 1.1 among men over 55 years of age. These age-related differences, which may account for some of the inconsistencies seen in previous studies of cigarette smoking and non-Hodgkin's lymphoma, should be considered in future investigations.

Occupational chlorophenol exposure and soft tissue sarcoma risk among men aged 30-60 years

To evaluate the association of chlorophenol exposure with soft tissue sarcoma risk independent of phenoxyherbicide exposure, the authors analyzed data from the Selected Cancers Study, a population-based case-control study that included 295 male soft tissue sarcoma cases, aged 32-60 years, from eight population-based cancer registries and 1,908 male controls. Chlorophenol exposure was assigned using both an intensity and a confidence estimate by an industrial hygienist based on verbatim job descriptions. Seventeen percent of the jobs rated as high intensity involved wood preservation, while 82% involved cutting oils. Soft tissue sarcoma risk, modeled using conditional logistic regression, was significantly associated with ever having high-intensity chlorophenol exposure (odds ratio = 1.79, 95% confidence interval 1.10-2.88). A duration-response trend was evident among more highly exposed subjects (p for trend < 0.0001). For subjects with 10 or more years of substantial exposure, the odds ratio was 7.78 (95% confidence interval 2.46-24.65). These results suggest that chlorophenol exposure independent of phenoxyherbicides may increase the risk of soft tissue sarcoma. Because of the large number of machinists in the exposed group and the complex composition of cutting fluids, it is possible that another exposure involved in machining is responsible for the observed excess risk.

Mortality studies of metalworking fluid exposure in the automobile industry: VI. A case-control study of esophageal cancer

BACKGROUND

Results are reported from a nested case-control study of 60 esophageal cancer deaths among 46,384 automobile manufacturing workers potentially exposed to metalworking fluids (MWF) in machining and grinding operations.

METHODS

By using incidence-density sampling, controls were selected with a sampling ratio of 20:1 from among co-workers who remained at risk by the age of death of the case, matched on race, gender, plant, and year of birth. Conditional logistic regression was used to evaluate the risk associated with cumulative exposure (mg/m3-years) to each of three types of metalworking fluid (straight, soluble, and synthetic MWF), as well as with years of exposure to selected components of MWF, including nitrosamines, sulfur, biocides, and several metals.

RESULTS

Esophageal cancer was found to be significantly associated with exposure to both soluble and synthetic MWF in grinding operations. The odds ratios (ORs) for grinding with soluble MWF were elevated at 2.5 or greater in all categories of cumulative exposure, although the exposure-response trend was statistically significant only when exposure was measured as duration. Those with 12 or more years exposure to soluble MWF in grinding operations experienced a 9.3-fold relative risk of esophageal cancer mortality (95% CI = 2.1-42.1). The OR for ever grinding with synthetic MWF was 4.1 (95% CI = 1.1-15.0). Elevated risk was also associated with two agents found in both synthetic and soluble fluids, nitrosamines, and biocides. For exposure to nitrosamines, the OR was 5.4 (95% CI = 1.5-19.9); for biocides the OR was 3.8 (95% CI = 0.8-18.9). However, because the same workers were exposed to grinding with synthetics, nitrosamines and biocides, it was not possible to separate the specific risks associated with these components.

Temporal and spatial distributions of ozone in Atlanta: regulatory and epidemiologic implications

Relationships between ambient levels of selected air pollutants and pediatric asthma exacerbation in Atlanta were studied retrospectively. As a part of this study, temporal and spatial distributions of ambient ozone concentrations in the 20-county. Atlanta metropolitan area during the summers of 1993, 1994, and 1995 were assessed. A universal kriging procedure was used for spatial interpolation of aerometric monitoring station data. In this paper, the temporal and spatial distributions of ozone are described, and regulatory and epidemiologic implications are discussed. For the study period, the Atlanta ozone nonattainment area based on the 1-h, exceedance-based standard of 0.12 ppm is estimated to expand--from 56% of the Atlanta MSA by area and 71% by population to 88% by area and 96% by population--under the new 8-h, concentration-based standard of 0.08 ppm. Regarding asthma exacerbation, a 4% increase in pediatric asthma rate per 20-ppb increase in ambient ozone concentration was observed (p-value = 0.001), with ambient ozone level representing a general indicator of air quality due to its correlations with other pollutants. The use of spatial ozone estimates in the epidemiologic analysis demonstrates the need for control of demographic covariates in spatiotemporal assessments of associations of ambient air pollutant concentrations with health outcome.

Reproductive toxins and alligator abnormalities at Lake Apopka, Florida

The alligator population at Lake Apopka in central Florida declined dramatically between 1980 and 1987. Endocrine-disrupting chemicals and specifically DDT metabolites have been implicated in the alligators' reproductive failure. The DDT metabolite hypothesis is based largely on the observation of elevated concentrations of p,p-DDE and p,p-DDD in alligator eggs obtained from Lake Apopka in 1984 and 1985. In the following commentary, we draw attention to two nematocides that are established reproductive toxins in humans, dibromochloropropane (DBCP) and ethylene dibromide (EDB), which could also have played a role in the reproductive failure observed in alligators from Lake Apopka in the early 1980s.

Mortality studies of machining fluid exposure in the automobile industry. V: A case-control study of pancreatic cancer

Results are presented from a case-control study of 97 cases of pancreatic cancer nested in a cohort of workers from three automobile manufacturing plants. Risk was examined for lifetime exposure to straight, soluble, and synthetic metalworking fluids, as used in specific machining or grinding operations, as well as for constituents of the fluids. Pancreatic cancer was associated with exposure to synthetic fluids in grinding operations, with an odds ratio of 3.0 (95% CI: 1.2-7.5) among those with more than 1.4 mg/m3-years of exposure. We were unable to examine synthetic exposure in the absence of grinding because there was virtually no exposure to synthetics in machining operations in this study population. Although a disproportionately high percent of the cases were black, no black workers had any exposure to synthetic fluids, and no other measured exposure was found to be related to risk. Thus, the previously documented excess risk of pancreatic cancer among blacks in this cohort remains unexplained.

Mortality studies of machining fluid exposure in the automobile industry. V: A case-control study of pancreatic cancer

Results are presented from a case-control study of 97 cases of pancreatic cancer nested in a cohort of workers from three automobile manufacturing plants. Risk was examined for lifetime exposure to straight, soluble, and synthetic metalworking fluids, as used in specific machining or grinding operations, as well as for constituents of the fluids. Pancreatic cancer was associated with exposure to synthetic fluids in grinding operations, with an odds ratio of 3.0 (95% CI: 1.2-7.5) among those with more than 1.4 mg/m3-years of exposure. We were unable to examine synthetic exposure in the absence of grinding because there was virtually no exposure to synthetics in machining operations in this study population. Although a disproportionately high percent of the cases were black, no black workers had any exposure to synthetic fluids, and no other measured exposure was found to be related to risk. Thus, the previously documented excess risk of pancreatic cancer among blacks in this cohort remains unexplained.

Mortality studies of machining fluid exposure in the automobile industry. IV: A case-control study of lung cancer

Machining fluids are diverse products that contain numerous additives and contaminants, including polycyclic aromatic hydrocarbons. Studies treating machining fluids as an aggregate exposure have found both positive and negative associations with lung cancer. In this nested case-control study of automotive workers (667 cases and 3,041 matched controls), individual estimates of exposure quantity and duration for specific classes of machining fluids were derived. An inverse dose-response relationship was found between synthetic machining fluids and lung cancer mortality, with an odds ratio of 0.6 (95% CI = 0.4, 0.8) for the highest level of lifetime exposure. The relationship was strongest for recent exposures. There was little evidence of an association with soluble or straight oil machining fluids. Risks were inconsistently elevated in workers exposed to aluminum. Results from this study provide strong evidence that exposure to machining fluids is not associated with an increased risk of lung cancer mortality in automotive workers.

Oils and cancer

Epidemiologic evidence on the relationship between mineral oil exposure and cancer is reviewed. The review is restricted to occupations involving substantial dermal and inhalational exposure and for which an epidemiologic literature exists: metal machining, print press operating, and cotton and jute spinning. Mineral oils are complex mixtures of aliphatic hydrocarbons, naphthenics, and aromatics, the relative distribution of which depends on the source of the oil and the method of refinement. End-use products contain a variety of additives, and contamination by other agents generally occurs during use. Suspect agents include polycyclic aromatic hydrocarbons (PAH) (particularly benz[a]pyrene), nitrosamines, chlorinated paraffins, long-chain aliphatics, sulfur, N-phenyl-2-naphthylamine, and formaldehyde. The heterogeneity of this exposure makes epidemiologic study difficult and meta-analysis inappropriate. Nonetheless, several associations emerge from the literature with varying degrees of support. There is clear evidence that early formulations of mineral oils used in cotton and jute spinning and in metal machining were carcinogenic to the skin. Associations of mineral oil exposure with laryngeal and rectal cancer have received some support in the literature, particularly with respect to straight oils. Evidence is suggestive that grinding operations (which can entail either mineral oil-based or ethanolamine-based fluids) are associated with excess risk of cancer of the esophagus, stomach, and pancreas. A number of bladder cancer case-control studies have noted an association with work as a machinist. There is limited evidence of an association with cancer of the colon, prostate, and sinonasal region. Several studies of printers have yielded positive findings for lung cancer, whereas studies in metal machinists have been generally negative. The PAH and nitrosamine content of current formulations is lower than in the past and the implications of these changes in composition to the carcinogenicity of the formulations are not yet known.

Mortality studies of machining fluid exposure in the automobile industry. III: A case-control study of larynx cancer

A case-control study of larynx cancer was conducted within a cohort of automobile workers exposed to metal working fluids, commonly referred to as machining fluids (MF). Results are based on 108 cases of larynx cancer and 5:1 matched controls. Risks associated with specific types of MF, as well as specific components of the fluids were evaluated. Based on a retrospective exposure assessment, lifetime exposures to straight and soluble fluids, grinding particulate, biocides, selected metals, sulfur, and chlorine were examined. Exposure to asbestos and acid mists at two of the three study sites was also characterized. Results suggest that straight mineral oils are associated with almost a two-fold excess in larynx cancer risk. There was also evidence of an association with elemental sulfur, commonly added to straight MF to improve the integrity of the materials under extreme pressure and heat. It is not clear whether sulfur is causally related to an excess relative risk of larynx cancer or whether the observed association is the result of unmeasured confounding by another contaminant or process feature. For example, the high stress operations that require MF enriched with sulfur are also more likely to produce polycyclic aromatic hydrocarbons (PAHs) during the process. Thus, the observed association with sulfur may be due to an association with PAH. The finding of excess risk of laryngeal cancer associated with MF is consistent with several previous reports in the literature. This is the first study, however, to distinguish straight mineral oils from other types of MF. Based on these findings, a general reduction in concentrations of straight mineral oil particulate in occupational environments would be prudent.

Mortality studies of machining-fluid exposure in the automobile industry. II. Risks associated with specific fluid types

Excesses of digestive and respiratory cancers have been reported previously in association with exposure to machining fluids, agents in widespread use as coolants and lubricants in machining operations. Previous studies have had limited power to distinguish the effects of the different types of machining fluids in use. In a cohort of over 30,000 workers employed at two automotive plants in Michigan, mortality patterns were studied in relation to exposure to each of the three major fluid types--straight oils, soluble oils, and synthetic fluids. Standardized mortality ratios were estimated for subgroups of the cohort ever exposed to each of the three fluid types, and Poisson regression analyses were used to assess trends in risk with duration of exposure. The data suggest modest positive associations between exposure to straight oils and rectal, laryngeal, and prostatic cancer and a negative association between soluble and synthetic fluid exposure and lung cancer.

Micronuclei and other nuclear anomalies in buccal smears: methods development

Laboratory work aimed at improving the epidemiologic utility of an innovative genotoxicity assay is described. The exfoliated cell micronucleus assay involves microscopic analysis of epithelial smears to determine the prevalence of micronucleation, an indicator of structural or numerical chromosome aberrations. While the assay holds promise for the study of epithelial carcinogens, it is hampered by the fact that exfoliated cells are moribund and undergo degenerative phenomena that can produce extranuclear objects difficult to distinguish from classical micronuclei. Modifications in the protocol were assessed in sample buccal smears from several study populations: radiotherapy patients, nonusers of tobacco, and snuff users. Refinements in micronucleus scoring criteria and the inclusion of other nuclear anomalies in the scoring system are proposed. We demonstrate that our criteria are successful in detecting excess micronucleation in positive controls. We also provide evidence that other nuclear anomalies are at least as common as micronucleation and that therefore there is the potential for extensive misclassification. Reliability was assessed in duplicate readings.

Mortality studies of machining fluid exposure in the automobile industry I: A standardized mortality ratio analysis

Machining fluids are widely used in a variety of common industrial metalworking operations to lubricate and cool both the tool and the working surfaces. Previous studies have suggested elevated respiratory, digestive, and skin cancers in exposed populations. This cohort study was initiated to assess whether long-term exposure to machining fluids in the course of machining, grinding, and other cutting operations is associated with excess cancer mortality. The cohort includes more than 45,000 automobile production workers from 3 plants, almost 1 million years of follow-up, over 10,000 deaths, and an extensive exposure assessment component. Standardized mortality ratios (SMRs) have been estimated for each of the 3 plants, using both U.S. as well as local populations as reference. Relative risks of 1.2-3.1 have been observed for several specific respiratory and digestive cancers of a priori interest, including cancer of the stomach, large intestine, pancreas, lung, and larynx. In addition, elevated risks for leukemia and asthma were noted. Future exposure-response analyses will provide the opportunity to identify relatively modest excesses in cause-specific mortality risk associated with exposure to specific types (straight, soluble, or synthetic), additives, or components of machining fluids.

Micronuclei and other nuclear anomalies in buccal smears: a field test in snuff users

A revised protocol for the exfoliated cell micronucleus assay was field-tested in a population exposed to a genotoxic agent, snuff, at levels associated with a significant increase in cancer risk. The standard assay involves examination of epithelial smears to determine the prevalence of micronucleated cells, an indication of chromosome breakage or mitotic interference. The assay was revised to increase specificity and to include separate scoring of other nuclear anomalies associated with cytotoxicity and genotoxicity. The modified assay was applied to buccal smears of 38 female snuff users and 15 female nonusers recruited from a North Carolina clinic in 1987. The prevalence of micronucleation was elevated in the snuff users as compared with the nonusers (prevalence ratio = 2.4, 95% confidence interval 1.1-5.2) and, to a lesser extent, at the usual contact site as compared with a distal buccal site in the snuff users (prevalence ratio = 1.5, 95% confidence interval 0.9-2.5). The pattern of relative frequencies of several nuclear anomalies provided strong evidence of a cytotoxic effect, the prevalence ratios ranging from 2 to 13. Nuclear degenerative phenomena can be difficult to distinguish from classical micronuclei; thus, the observed association of indicators of cytotoxicity with exposure introduces the possibility of bias away from the null in micronucleus findings due to differential misclassification. Until methods to better distinguish extranuclear bodies of different origins become available, investigators should use the revised protocol and should focus on agents not thought to be cytotoxic.

Cancer and other causes of death among a cohort of dry cleaners

Mortality among 5365 members of a dry cleaning union in St. Louis, Missouri, was less than expected for all causes combined (SMR = 0.9) but slightly raised for cancer (SMR = 1.2). Among the cancers, statistically significant excesses occurred for oesophagus (SMR = 2.1) and cervix (SMR = 1.7) and non-significant excesses for larynx (SMR = 1.6), lung (SMR = 1.3), bladder (SMR = 1.7), thyroid (SMR = 3.3), lymphosarcoma and reticulosarcoma (SMR = 1.7), and Hodgkin's disease (SMR = 2.1). Mortality from emphysema was also significantly raised (SMR = 2.0). Eleven of the 13 deaths from oesophageal cancer occurred among black men. The risk of this cancer showed a significant association with estimated cumulative exposure to dry cleaning solvents (rising to 2.8-fold in the highest category) but not with level or duration of exposure. Mortality from kidney cancer was not excessive as reported in other studies. Excesses for emphysema and cancers of the larynx, lung, oesophagus, bladder, and cervix may be related to socioeconomic status, tobacco, or alcohol use. Although the number of deaths was small, the greatest risk for cancers of the lymphatic and haematopoietic system (fourfold) occurred among workers likely to have held jobs where exposures were the heaviest. Small numbers and limited information on exposure to specific substances complicates interpretation of this association but is unlikely to be due to confounding by tobacco use. It was not possible to identify workers exposed to specific dry cleaning solvents but mortality among those entering the union after 1960, when use of perchloroethylene was predominant, was similar to those entering before 1960.