Investigation of excess environmental risk around putative sources: Stone's test with covariate adjustment

Association between proximity to industrial chemical installations and cancer mortality in Spain

It is likely that pollution from chemical facilities will affect the health of any exposed population; however, the majority of scientific evidence available has focused on occupational exposure rather than environmental. Consequently, this study assessed whether there could have been an excess of cancer-related mortality associated with environmental exposure to pollution from chemical installations - for populations residing in municipalities in the vicinity of chemical industries. To this end, we designed an ecological study which assessed municipal mortality due to 32 types of cancer in the period from 1999 to 2008. The exposure to pollution was estimated using distance from the facilities to the centroid of the municipality as a proxy for exposure. In order to assess any increased cancer mortality risk in municipalities potentially exposed to chemical facilities pollution (situated at a distance of ≤5 km from a chemical installation), we employed Bayesian Hierarchical Poisson Regression Models. This included two Bayesian inference methods: Integrated Nested Laplace Approximations (INLA) and Markov Chain Monte Carlo (MCMC, for validation). The reference category consisted of municipalities beyond the 5 km limit. We found higher mortality risk (relative risk, RR; estimated by INLA, 95% credible interval, 95%CrI) for both sexes for colorectal (RR, 1.09; 95%CrI, 1.05-1.15), gallbladder (1.14; 1.03-1.27), and ovarian cancers (1.10; 1.02-1.20) associated with organic chemical installations. Notably, pleural cancer (2.27; 1.49-3.41) in both sexes was related to fertilizer facilities. Associations were found for women, specifically for ovarian (1.11; 1.01-1.22) and breast cancers (1.06; 1.00-1.13) in the proximity of explosives/pyrotechnics installations; increased breast cancer mortality risk (1.10; 1.03-1.18) was associated with proximity to inorganic chemical installations. The results suggest that environmental exposure to pollutants from some types of chemical facilities may be associated with increased mortality from several different types of cancer.

Industrial pollution and cancer in Spain: An important public health issue

Cancer can be caused by exposure to air pollution released by industrial facilities. The European Pollutant Release and Transfer Register (E-PRTR) has made it possible to study exposure to industrial pollution. This study seeks to describe the industrial emissions in the vicinity of Spanish towns and their temporal changes, and review our experience studying industrial pollution and cancer. Data on industrial pollutant sources (2007-2010) were obtained from the E-PRTR registries. Population exposure was estimated by the distance from towns to industrial facilities. We calculated the amount of carcinogens emitted into the air in the proximity (<5km) of towns and show them in municipal maps. We summarized the most relevant results and conclusions reported by ecological E-PRTR-based on studies of cancer mortality and industrial pollution in Spain and the limitations and result interpretations of these types of studies. There are high amounts of carcinogen emissions in the proximity of towns in the southwest, east and north of the country and the total amount of emitted carcinogens is considerable (e.g. 20Mt of arsenic, 63Mt of chromium and 9Mt of cadmium). Although the emissions of some carcinogens in the proximity of certain towns were reduced during the study period, emissions of benzene, dioxins+furans and polychlorinated biphenyls rose. Moreover, the average population of towns lying within a 5km radius from emission sources of carcinogens included in the International Agency for Research on Cancer list of carcinogens was 9 million persons. On the other hand, the results of the reviewed studies suggest that those Spanish regions exposed to the pollution released by certain types of industrial facilities have around 17% cancer excess mortality when compared with those unexposed. Moreover, excess mortality is focused on digestive and respiratory tract cancers, leukemias, prostate, breast and ovarian cancers. Despite their limitations, ecological studies are a useful tool in environmental epidemiology, not only for proposing etiological hypotheses about the risk of living close to industrial pollutant sources, but also for providing data to account for situations of higher mortality in specific areas. Nevertheless, the reduction of emissions should be a goal, with special relevance given to establishing limits for known carcinogens and other toxic substances in the environs of population centers, as well as industry-specific emission limits.

Childhood cancer in small geographical areas and proximity to air-polluting industries

AIM

Pediatric cancer has been associated with exposure to certain environmental carcinogens. The purpose of this work is to analyse the relationship between environmental pollution and pediatric cancer risk.

METHOD

We analysed all incidences of pediatric cancer (<15) diagnosed in a Spanish region during the period 1998-2015. The place of residence of each patient and the exact geographical coordinates of main industrial facilities was codified in order to analyse the spatial distribution of cases of cancer in relation to industrial areas. Focal tests and focused Scan methodology were used for the identification of high-incidence-rate spatial clusters around the main industrial pollution foci.

RESULTS

The crude rate for the period was 148.0 cases per 1,000,0000 children. The incidence of pediatric cancer increased significantly along the period of study. With respect to spatial distribution, results showed significant high incidence around some industrial pollution foci group and the Scan methodology identify spatial clustering. We observe a global major incidence of non Hodgkin lymphomas (NHL) considering all foci, and high incidence of Sympathetic Nervous System Tumour (SNST) around Energy and Electric and organic and inorganic chemical industries foci group. In the analysis foci to foci, the focused Scan test identifies several significant spatial clusters. Particularly, three significant clusters were identified: the first of SNST was around energy-generating chemical industries (2 cases versus the expected 0.26), another of NHL was around residue-valorisation plants (5 cases versus the expected 0.91) and finally one cluster of Hodgkin lymphoma around building materials (3 cases versus the expected 2.2) CONCLUSION: Results suggest a possible association between proximity to certain industries and pediatric cancer risk. More evidences are necessary before establishing the relationship between industrial pollution and pediatric cancer incidence.

Identifying pediatric cancer clusters in Florida using loglinear models and generalized lasso penalties

We discuss the identification of pediatric cancer clusters in Florida between 2000 and 2010 using a penalized generalized linear model. More specifically, we introduce a Poisson model for the observed number of cases on each of Florida's ZIP Code Tabulation Areas (ZCTA) and regularize the associated disease rate estimates using a generalized Lasso penalty. Our analysis suggests the presence of a number of pediatric cancer clusters during the period over study, with the largest ones being located around the cities of Jacksonville, Miami, Cape Coral/Fort Meyers and Palm Beach.

L’analyse de cluster en épidémiologie géographique : utilisation de plusieurs méthodes statistiques et comparaison de leurs résultats

BACKGROUND

The increasing interest in environmental epidemiology has been followed by the development of many statistical tests for detecting disease clustering near a point source. The objectives of this study were to compare several tests to detect disease clustering, among which modelisation using Markov Chain Monte Carlo methods.

METHODS

We compared six statistical methods for detecting disease clustering of bladder cancer around an industrial centre of Isère (France) for the period 1983-1997: Stone's test, score test, and two log-linear modelisations (with and without corrections for extra-Poisson variations) using two ways of parameters estimation (maximum likelihood and Markov Chain Monte Carlo methods).

RESULTS

The results of the Stone test and the score test are not in favour of a higher risk of bladder cancer around the considered point source. The conclusions brought by the log linear modelisations are the same, but the results obtained using the Markov Chain Monte Carlo Method are very dependant of prior distributions determined for the different parameters.

CONCLUSION

Markov Chain Monte Carlo methods, which allow taking into account complex geographical effects, seem well adapted to cluster analysis in geographical epidemiology. However, they remain difficult to implement.

Risk of respiratory cancer around a sewage plant in Prato, Italy

In this study, the authors evaluated the risk of respiratory cancer related to environmental pollutants among a population that resided near a sewage plant in Prato, Italy. Subjects included lung cancer deaths (1987-1996) and incident cases of lung and laryngeal cancers (1987-1994) among residents of Prato. The authors used the mortality or incidence rates for the entire population of Prato (by gender and by 5-yr age group) to calculate the expected cases in each census unit. Data were analyzed and adjusted for an index of social deprivation (Stone test). Among males, the excess risk of lung cancer mortality decreased as distance from the plant increased for 2 time periods (1987-1996 [p = .008] and 1990-1996 [p = .030]) and for lung cancer incidence during 1987-1994 (p = .011). Similar results were obtained when sewage plant workers were excluded from the analysis. A similar, but not statistically significant, trend was observed among female incident lung cancer cases, as well as among male incident laryngeal cancer cases. Despite methodological limitations common to geographic studies, the results were consistent with those previously published on mortality excesses for lung cancer among plant workers under study. The role of environmental pollutants as a risk for respiratory cancer must be further clarified with additional epidemiological studies and an environmental monitoring program.

Additive isotonic regression models in epidemiology

Stone's method for assessing disease risk around a point source through isotonic regression is routinely used in spatial epidemiology. It is useful in situations where the relationship of risk with exposure (distance being commonly used as a surrogate variable) is assumed monotonic but otherwise of unknown form. This paper extends this method to non-spatial epidemiology, where typically two or more risk factors are present. The methodology described is based on the additive isotonic model approach of Bacchetti; versions appropriate to count (Poisson) data and case-control (binomial) data are described. In both cases, adjustment for covariates is incorporated, and a Monte Carlo method of hypothesis testing and interval estimation is presented. The methodology is illustrated through a case-control example concerning the analysis of the possible effect of preconceptional external ionizing radiation doses on the sex ratio at birth among children of fathers working at the Sellafield nuclear installation, Cumbria, U.K.

Issues in the statistical analysis of small area health data

The availability of geographically indexed health and population data, with advances in computing, geographical information systems and statistical methodology, have opened the way for serious exploration of small area health statistics based on routine data. Such analyses may be used to address specific questions concerning health in relation to sources of pollution, to investigate clustering of disease or for hypothesis generation. We distinguish four types of analysis: disease mapping; geographic correlation studies; the assessment of risk in relation to a prespecified point or line source, and cluster detection and disease clustering. A general framework for the statistical analysis of small area studies will be considered. This framework assumes that populations at risk arise from inhomogeneous Poisson processes. Disease cases are then realizations of a thinned Poisson process where the risk of disease depends on the characteristics of the person, time and spatial location. Difficulties of analysis and interpretation due to data inaccuracies and aggregation will be addressed with particular reference to ecological bias and confounding. The use of errors-in-variables modelling in small area analyses will be discussed.

Case-control isotonic regression for investigation of elevation in risk around a point source

Stone's isotonic regression method for analysing count data to estimate disease risk in relation to a point source of environmental pollution is now routinely used. This paper develops the corresponding procedure for case-control data consisting of the locations of individual cases with controls with associated covariate information. In this setting, the generalized likelihood ratio statistic to test the null hypothesis of constant risk against the alternative that risk is a monotone non-increasing function of distance from the point source is intractable. An approximate Monte Carlo test is described, extending an exact test proposed by Bithell for the situation in which there are no covariates. Interval estimates of risk as a function of distance from the point source are constructed by simulation of the sampling distribution of the isotonic regression estimator. The methodology is illustrated by two applications: one to the relative risk of larynx cancers and lung cancers near a now-disused industrial incinerator; the other to the risk of asthma in children in relation to distance of residence from the nearest main road.