Inter-rater reliability of occupational exposure assessment in a case-control study of female breast cancer.

The objective of this paper was to estimate the inter-rater reliability of expert assessments of occupational exposures. An inter-rater reliability sub-study was conducted within a population-based case-control study of postmenopausal breast cancer. Detailed information on lifetime occupational histories was obtained from participants and two industrial hygienists assigned exposures to 185 jobs using a checklist of 293 agents. Experts rated exposure for each job-agent combination according to exposure status (unexposed/exposed), confidence that the exposure occurred (possible/probable/definite), intensity (low/medium/high), and frequency (% time per week). The statistical unit of observation was each job-agent assessment (185 jobs × 293 agents = 54,205 assessments per expert). Crude agreement, Gwet AC1/2 statistics, and Cohen's Kappa were used to estimate inter-rater agreement for confidence and intensity; for frequency, the intra-class correlation coefficient (ICC) was used. The majority of job-agent combinations were evaluated by the two experts to be not exposed (crude agreement >98% of decisions). The degree of agreement between the experts for the confidence of exposure status was Gwet AC1/2 = 0.99 (95% CI: 0.99-0.99), and for intensity, a Gwet AC2 = 0.99 (95% CI: 0.99-0.99). For frequency, an ICC of 0.31 (95% CI: 0.26-0.35) was found. A sub-analysis restricted to job-agent combinations for which the two experts agreed on exposure status revealed a moderate agreement for confidence of exposure (Gwet AC2 = 0.66) and high agreement for intensity (Gwet AC2 = 0.96). For frequency, the ICC was 0.52 (95% CI: 0.47-0.57). A high level of inter-rater agreement was found for identifying exposures and for coding intensity, but agreement was lower for the coding of frequency of exposure.

Risk of lung cancer in relation to various metrics of smoking history: a case-control study in Montreal.

BACKGROUND: Few epidemiologic findings are as well established as the association between smoking and lung cancer. It is therefore somewhat surprising that there is not yet a clear consensus about the exposure-response relationships between various metrics of smoking and lung cancer risk. In part this is due to heterogeneity of how exposure-response results have been presented and the relative paucity of published results using any particular metric of exposure. The purposes of this study are: to provide new data on smoking-lung cancer associations and to explore the relative impact of different dimensions of smoking history on lung cancer risk. METHODS: Based on a large lung cancer case-control study (1203 cases and 1513 controls) conducted in Montreal in 1996-2000, we estimated the lifetime prevalence of smoking and odds ratios in relation to several smoking metrics, both categorical and continuous based on multivariable unconditional logistic regression. RESULTS: Odds ratios (ORs) for ever vs never smoking were 7.82 among males and 11.76 among females. ORs increased sharply with every metric of smoking examined, more so for duration than for daily intensity. In models using continuous smoking variables, all metrics had strong effects on OR and mutual adjustment among smoking metrics did not noticeably attenuate the OR estimates, indicating that each metric carries some independent risk-related information. Among all the models tested, the one based on a smoking index that integrates several smoking dimensions, provided the best fitting model. Similar patterns were observed for the different histologic types of lung cancer. CONCLUSIONS: This study provides many estimates of exposure-response relationships between smoking and lung cancer; these can be used in future meta-analyses. Irrespective of the histologic type of lung cancer and the smoking metric examined, high levels of smoking led to high levels of risk, for both men and women.

Probabilistic Multiple-Bias Modeling Applied to the Canadian Data From the Interphone Study of Mobile Phone Use and Risk of Glioma, Meningioma, Acoustic Neuroma, and Parotid Gland Tumors.

We undertook a re-analysis of the Canadian data from the 13-country case-control Interphone Study (2001-2004), in which researchers evaluated the associations of mobile phone use with the risks of brain, acoustic neuroma, and parotid gland tumors. In the main publication of the multinational Interphone Study, investigators concluded that biases and errors prevented a causal interpretation. We applied a probabilistic multiple-bias model to address possible biases simultaneously, using validation data from billing records and nonparticipant questionnaires as information on recall error and selective participation. In our modeling, we sought to adjust for these sources of uncertainty and to facilitate interpretation. For glioma, when comparing those in the highest quartile of use (>558 lifetime hours) to those who were not regular users, the odds ratio was 2.0 (95% confidence interval: 1.2, 3.4). After adjustment for selection and recall biases, the odds ratio was 2.2 (95% limits: 1.3, 4.1). There was little evidence of an increase in the risk of meningioma, acoustic neuroma, or parotid gland tumors in relation to mobile phone use. Adjustments for selection and recall biases did not materially affect interpretation in our results from Canadian data.

Perineal use of talc and risk of ovarian cancer.

Ovarian cancer is one of the most common gynaecological neoplasms, especially in industrialised countries. The aetiology of the disease is not well understood, except that inherited mutations in the breast cancer genes BRCA-1 and BRCA-2 account for up to 10% of all cases, and child-bearing, oral contraceptive use and breast-feeding reduce the risk. Some environmental exposures, notably talc and asbestos, have been suspected as ovarian carcinogens.

Occupational exposure to lead compounds and risk of cancer among men: a population-based case-control study.

The International Agency for Research on Cancer recently classified inorganic lead as a probable carcinogen, while organic lead remained unclassifiable. Uncertainty persists because of limited epidemiologic evidence. The authors addressed the relation between occupational exposure to lead and the risk of 11 types of cancer among men in a case-control study conducted in Montreal, Quebec, Canada, in the 1980s. Incident cases (n = 3,730) and general population controls (n = 533) were interviewed to elicit information on job history and potential confounders. Expert chemists translated each job into a list of substances to which the subject had potentially been exposed. Exposure to lead was classified into three categories: organic lead (3% of subjects ever exposed), inorganic lead (17%), and lead in gasoline emissions (39%). Odds ratios and 95% confidence intervals were estimated by logistic regression using two control groups: general population controls and cancer controls. Stomach cancer was associated with organic lead when the authors used population controls (odds ratio (OR) = 3.0, 95% confidence interval (CI): 1.2, 7.3) and cancer controls (OR = 2.0, 95% CI: 1.1, 3.8) and with substantial exposure to lead in gasoline emissions when they used cancer controls (OR = 2.9, 95% CI: 1.4, 5.9). There was no association with inorganic lead and little evidence for associations with other cancer types.

Comparison of self-reported height and weight by cancer type among men from Montreal, Canada.

Among men there is epidemiological evidence for an association between obesity and increased risk of renal cell carcinoma, colon cancer and adenocarcinoma of the oesophagus. The evidence for other cancer sites remains inconsistent. We conducted a large population-based, multi-site, case-control study of environmental causes of cancer among males in Montreal, Canada. Among the many questionnaire items collected by interview were height and usual weight. We compared height, weight and body mass index (BMI) among individuals with 11 different cancer types (combined N=3016) and population-based controls (N=509). Linear regression was used to model the relationship of the disease status with each of three dependent continuous variables (height, weight and BMI), while adjusting for covariates. For most cancer groups, weight and BMI were lower than among population controls. Because of potential information bias and reverse causality bias, we focused on the comparisons among cancer types. The lowest BMI values were observed among men with squamous cell carcinoma of the oesophagus, lung and stomach cancers. The highest BMIs were reported by men with prostate and kidney cancers, and oesophageal adenocarcinoma. Inconsistencies in the epidemiological literature on obesity and cancer risk could be related to the difficulties in obtaining unbiased reports of pre-disease weight and to publication bias.

Reanalysis of the Harvard Six Cities Study, part I: validation and replication.

Because the results of the Harvard Six Cities Study played a critical role in the establishment of the current U.S. ambient air quality objective for fine particles (PM(2.5)), the U.S. Environmental Protection Agency, industry, and nongovernmental organizations called for an independent reanalysis of this study to validate the original findings reported by Dockery and colleagues in the New England Journal of Medicine (vol. 329, pp. 1753-1759) in 1993. Validation of the original findings was accomplished by a detailed statistical audit and replication of original results. With the exception of occupational exposure to dust (14 discrepancies of 249 questionnaires located for evaluation) and fumes (15/249), date of death (2/250), and cause of death (2/250), the audit identified no discrepancies between the original questionnaires and death certificates in the audit sample and the analytic file used by the original investigators. The data quality audit identified a computer programming problem that had resulted in early censorship in 5 of the 6 cities, which resulted in the loss of approximately 1% of the reported person-years of follow-up; the reanalysis team updated the Six Cities cohort to include the missing person-years of observation, resulting in the addition of 928 person-years of observation and 14 deaths. The reanalysis team was able to reproduce virtually all of the original numerical results, including the 26% increase in all-cause mortality in the most polluted city (Stubenville, OH) as compared to the least polluted city (Portage, WI). The audit and validation of the Harvard Six Cities Study conducted by the reanalysis team generally confirmed the quality of the data and the numerical results reported by the original investigators. The discrepancies noted during the audit were not of epidemiologic importance, and did not substantively alter the original risk estimates associated with particulate air pollution, nor the main conclusions reached by the original investigators.

Reanalysis of the Harvard Six Cities Study, part II: sensitivity analysis.

Following the validation and replication of the Harvard Six Cities Study (Krewski et al., this issue), we conducted a wide range of sensitivity analyses to explore the observed associations between long-term exposure to fine particle or sulfate air pollution and mortality. We examined the impact of alternative risk models on estimates of risk, taking into account covariates not included in the original analyses. These risk models provided a basis for identifying covariates that may confound or modify the association between fine particle or sulfate air pollution and mortality, and for identifying sensitive population subgroups. The possibility of confounding due to occupational exposures was also investigated. Residence histories were coded for the study subjects and were used to examine temporal patterns of exposure and risk. Our sensitivity analyses showed the mortality risk estimates for fine particle and sulfate air pollution to be highly robust against alternative risk models of the Cox proportional hazards family, including models with additional covariates from the original questionnaires not included in the original published analyses. There was limited evidence of departures from the proportional hazards assumption. Flexible exposure-response models provided some evidence of departures from linearity at both low and high sulfate concentrations. Incorporating information on changes over time in cigarette smoking and body mass index had little effect on the association between fine particles and mortality. There was limited evidence of variation in risk with attained age, gender, smoking status, occupational exposure to dust and fumes, marital status, heart or lung diseases, or lung function. However, air pollution risk did appear to decreasing with increasing educational attainment. Extensive adjustment for occupation using aggregate indices of occupational "dirtiness" and occupational exposure to known lung carcinogens had little impact on the mortality risks associated with particulate air pollution. Our evaluation of population mobility indicated that relatively few subjects moved from their original city of residence. Attempts to identify critical exposure time windows were limited by the lack of marked interindividual variation in temporal exposure patterns throughout the study period. Overall, this extensive sensitivity analysis both supported the conclusions reached by the original investigators and demonstrated the robustness of these conclusions to alternative analytic approaches.

A flexible modeling approach to estimating the component effects of smoking behavior on lung cancer.

OBJECTIVE: Despite the established causal association between cigarette smoking and lung cancer, the relative contributions of age started, duration, years since quitting, and daily amount smoked have not been well characterized. We estimated the contribution of each of these aspects of smoking behavior. STUDY DESIGN AND SETTING: A case-control study was conducted in Montreal on the etiology of lung cancer. There were 640 cases and 938 control subjects for whom lifetime smoking histories were collected. We used generalized additive models, incorporating cubic smoothing splines to model nonlinear effects of various smoking variables. We adopted a multistep approach to deal with the multicollinearity among time-related variables. RESULTS: The main findings are that (1) risk increases independently by daily amount and by duration; (2) among current smokers, lung cancer risk doubles for every 10 cigarettes per day up to 30 to 40 cigarettes per day and tails off thereafter; (3) among ex-smokers, the odds ratio decreases with increasing time since quitting, the rate of decrease being sharper among heavy smokers than among light smokers; and (4) absolute risks demonstrate the dramatic public health benefits of long-term smoking cessation. CONCLUSION: Our results reinforce some previous findings on this issue.

Estimating the distribution of lag in the effect of short-term exposures and interventions: adaptation of a non-parametric regression spline model.

Information on the distribution of lag duration between exposure or intervention and the subsequent changes in risk can help in assessing the impact of exposure, predicting cost-effectiveness of intervention, and understanding the underlying biological mechanisms. Previous approaches focused more on optimizing the strength of the exposure-disease association than on directly estimating lag duration. We propose an alternative approach applicable to the analysis of the lagged effects of binary exposure variables. The density function of the distribution of lags is estimated based on flexible modelling of changes in hazard ratio of exposed versus unexposed subjects. The methodology is evaluated in a simulation study and is applied to the Framingham data to investigate the lagged effect of smoking cessation on coronary heart disease risk.

Consumption of non-alcoholic beverages and prostate cancer risk.

Our objective was to investigate the relations between the consumption of coffee, tea and carbonated beverages and the development of prostate cancer. The design was a population-based case-control study set in Montreal. The analysis was restricted to the subset of men, aged 45-70 years, who underwent interviews in which aspects of lifelong consumption of non-alcoholic beverages were ascertained. There were 399 incident cases of prostate cancer, 476 population controls and 621 cancer controls. There was no association between the consumption of either coffee or carbonated beverages and the development of prostate cancer. Among daily tea drinkers, the odds ratio associated with the highest tertile of cumulative consumption was 2.0 (95% confidence interval (CI) 1.3-3.0) when using population controls and 1.6 (95% CI 1.0-2.4) when using cancer controls. In conclusion, the consumption of coffee or carbonated beverages does not influence the risk of prostate cancer. Our findings provide no support to the hypothesis that tea consumption may be protective. While tea consumption may increase prostate cancer risk, we were unable to rule out alternative explanations for the positive association that we observed.

Case-control study of alcohol consumption and prostate cancer risk in Montréal, Canada.

OBJECTIVES: to estimate the risk of prostate cancer associated with alcohol consumption. METHODS: Between 1979 and 1985 a population-based case-control study was carried out in Montréal, which accrued over 4000 men in total, including cases of prostate cancer, other cancers, and population controls. The present analysis was restricted to the subset, aged 45-70 years. who underwent face-to-face interviews, in which aspects of lifelong alcohol consumption were ascertained. The cancer control series was further restricted to men whose tumor types were considered unrelated to alcohol consumption. There were 399 incident cases of prostate cancer, 476 population controls, and 674 cancer controls. RESULTS: When using the population controls, risk increased with increasing cumulative consumption of alcohol. There was no decrease in risk after quitting. Risk was particularly high among those who reported having started before age 15 years (odds ratio = 3.8; 95% confidence interval: 1.6-9.3). The results obtained using the cancer controls were less pronounced, but still indicated an excess risk associated with alcohol consumption. Beer was the most prevalent type of alcohol consumed in this population and showed the strongest association with prostate cancer. CONCLUSIONS: The results are consistent with an increase in the risk of prostate cancer due to alcohol consumption.

Exposure to titanium dioxide and risk of lung cancer in a population-based study from Montreal.

OBJECTIVES: This study assessed the lung cancer risk from exposure to titanium dioxide, an important pigment with limited evidence of carcinogenicity in experimental animals but sparse data for humans. METHODS: The risk of lung cancer among residents in Montreal, Canada, was analyzed, including 857 histologically confirmed cases of lung cancer diagnosed during 1979-1985 among men aged 35-70 years and a group of referents comprising 533 randomly selected, healthy residents and 533 persons with cancer in organs other than the lung. Exposure to titanium dioxide and other titanium compounds was assessed by a team of industrial hygienists on the basis of a detailed occupational questionnaire. RESULTS: Thirty-three cases and 43 referents were classified as exposed to titanium dioxide. The odds ratio was 0.9 [95% confidence interval (95% CI) 0.5-1.5]. No trend was apparent according to the estimated frequency, level, or duration of exposure. The odds ratio was 1.0 (95% CI 0.3-2.7) for medium or high exposure for at least 5 years. Few subjects were classified as exposed to titanium dioxide fumes or to other titanium compounds, but the risk of lung cancer was nonsignificantly increased for exposure to these agents. CONCLUSIONS: Although misclassification of exposure and low exposure prevalence might have resulted in false negative results, this study does not suggest that occupational exposure to titanium dioxide increases the risk of lung cancer.

Joint effects of smoking and body mass index on prostate cancer risk.

Between 1979 and 1985, a population-based case-control study of cancer at multiple sites was carried out in Montréal. A total of 399 cases with histologically confirmed prostate cancer and 476 population controls, 45-70 years of age, gave face-to-face interviews and provided adequate smoking histories. We analyzed the effects of smoking cigarettes only and of smoking cigars, or pipes, or both, with or without cigarettes, on the risk of prostate cancer. Overall, the associations between smoking cigarettes and prostate cancer were weak and compatible with no effect; the associations with cigar and pipe smoking were stronger. Among men with high body mass index, however, we found appreciable associations between cigarette smoking and prostate cancer risk. A history of ever smoking daily was associated with an odds ratio of 2.31 (95% confidence interval = 1.09-4.89). Risk increased with the amount smoked per day and with the duration of smoking. Taken together, the findings of increased risk associated with cigar and pipe smoking and the findings of increased risk associated with cigarette smoking among obese men suggest that tobacco smoking may be a risk factor for prostate cancer.

Activities and exposures during leisure and prostate cancer risk.

The objective of this study was to identify activities and exposures during leisure that might be associated with the development of prostate cancer. We analyzed data derived from a population-based case-control study that was carried out in Montreal between 1979 and 1985. Men (>4000) were interviewed, including cases of prostate cancer, other cancers, and population controls. The present analysis was restricted to the subset, aged 45-70 years, who underwent face-to-face interviews in which aspects of activities and exposures during leisure were ascertained. There were 400 incident cases of prostate cancer and 476 population controls. We calculated odds ratios (OR) for prostate cancer, adjusted for age, ethnic origin, respondent status, family income, body mass index, cigarette smoking, and alcohol consumption. Home or furniture maintenance was associated with an increased risk [OR, 1.4; 95% confidence interval (CI), 1.0-1.9], as was painting, stripping, or varnishing furniture (OR, 2.1; 95% CI, 0.7-6.7). Exposure during leisure to metal dust was associated with prostate cancer (OR, 3.2; 95% CI, 1.0-9.9), as was exposure to lubricating oils or greases (OR, 2.2; 95% CI, 1.2-3.7) and exposure to pesticides or garden sprays (OR, 2.3; 95% CI, 1.3-4.2). These findings are consistent with results derived from studies of occupational exposures.

A case-control study of the relationship between the risk of colon cancer in men and exposures to occupational agents.

BACKGROUND: We conducted a population-based case-control study in Montreal, Canada, to explore associations between hundreds of occupational circumstances and several cancer sites, including colon. METHODS: We interviewed 497 male patients with a pathologically confirmed diagnosis of colon cancer, 1514 controls with cancers at other sites, and 533 population-based controls. Detailed job histories and relevant potential confounding variables were obtained, and the job histories were translated by a team of chemists and industrial hygienists into a history of occupational exposures. RESULTS: We found that there was reasonable evidence of associations for men employed in nine industry groups (adjusted odds ranging from 1.1 to 1.6 per a 10-year increase in duration of employment), and in 12 job groups (OR varying from 1.1 to 1.7). In addition, we found evidence of increased risks by increasing level of exposures to 21 occupational agents, including polystyrene (OR for "substantial" exposure (OR(subst)) = 10.7), polyurethanes (OR(subst) = 8.4), coke dust (OR(subst) = 5.6), mineral oils (OR(subst) = 3.3), polyacrylates (OR(subst) = 2.8), cellulose nitrate (OR(subst) = 2.6), alkyds (OR(subst) = 2.5), inorganic insulation dust (OR(subst) = 2.3), plastic dusts (OR(subst) = 2.3), asbestos (OR(subst) = 2.1), mineral wool fibers (OR(subst) = 2.1), glass fibers (OR(subst) = 2.0), iron oxides (OR(subst) = 1.9), aliphatic ketones (OR(subst) = 1.9), benzene (OR(subst) = 1.9), xylene (OR(subst) = 1.9), inorganic acid solutions (OR(subst) = 1.8), waxes, polishes (OR(subst) = 1.8), mononuclear aromatic hydrocarbons (OR(subst) = 1.6), toluene (OR(subst) = 1.6), and diesel engine emissions (OR(subst) = 1.5). Not all of these effects are independent because some exposures occurred contemporaneously with others or because they referred to a group of substances. CONCLUSIONS: We have uncovered a number of occupational associations with colon cancer. For most of these agents, there are no published data to support or refute our observations. As there are few accepted risk factors for colon cancer, we suggest that new occupational and toxicologic studies be undertaken focusing on the more prevalent substances reported herein.

Occupational risk factors for renal cell carcinoma in Montreal.

BACKGROUND: Little is known about the role of workplace exposures on the risk of renal cell cancer. METHODS: A population-based case-control study was undertaken in Montreal to assess the association between hundreds of occupational circumstances and several cancer sites, including the kidney. A total of 142 male patients with pathologically confirmed renal cell carcinoma, 1900 controls with cancer at other sites and 533 population-based controls were interviewed. Detailed job histories and relevant data on potential confounders were obtained. A group of chemists-hygienists evaluated each job reported and translated them into a history of occupational exposures using a checklist of 294 substances. Multivariate logistic regression models using either population, cancer controls, or a pool of both groups were used to estimate odds ratios. RESULTS: There were some indications of excess risks among printers, nursery workers (gardening), aircraft mechanics, farmers, and horticulturists, as well as in the following industries: printing-related services, defense services, wholesale trade, and retail trade. Notwithstanding the low precision of many of the odds ratio estimates, the following workplace exposures showed some evidence of excess risk: chromium compounds, chromium (VI) compounds, inorganic acid solutions, styrene-butadiene rubber, ozone, hydrogen sulphide, ultraviolet radiation, hair dust, felt dust, jet fuel engine emissions, jet fuel, aviation gasoline, phosphoric acid and inks. CONCLUSIONS: For most of these associations there exist no, or very little, previous data. Some associations provide suggestive evidence for further studies.

Occupational risks for lung cancer among nonsmokers.

We conducted a case-control study in 12 European study centers to evaluate the role of occupational risk factors among nonsmokers. We obtained detailed occupational histories from 650 nonsmoking cases (509 females/141 males) and 1,542 nonsmoking controls (1,011 females/531 males). On the basis of an a priori definition of occupations and industries that are known (list A) or suspected (list B) to be associated with lung carcinogenesis, we calculated odds ratios (ORs) for these occupations, using unconditional logistic regression models and adjusting for sex, age, and center effects. Among nonsmoking men, an excess relative risk was observed among those who had worked in list-A occupations [OR = 1.52; 95% confidence interval (C) = 0.78-2.97] but not in list-B occupations (OR = 1.05; 95%), CI = 0.60-1.83). Among nonsmoking women, there was an elevation of risk for list-A occupations (OR = 1.50; 95% CI = 0.49-4.53), although this estimate was imprecise, given that less than 1% of cases and controls were exposed. Exposure to list-B occupations was associated with an increase in relative risk (OR = 1.69; 95% CI = 1.09-2.63) in females, but not in males. Women who had been laundry workers or dry cleaners had an OR of 1.83 (95% CI = 0.98-3.40). Our findings confirm that certain occupational exposures are associated with an increased risk for lung cancer among both female and male nonsmokers; however, knowledge on occupational lung carcinogens is biased toward agents to which mainly men are exposed.