Please note the following RSS meeting featuring talks on spatial/environmental statistical methods as applied to Environmental Epidemiology ---------------------- Environmental Epidemiology 29 October 2010 12:00 - 16:00 Mid afternoon tea/coffee Errol Street, London EC1Y 8LX This meeting considers a number of current developments in the investigation of topics concerning environmental epidemiology. Please click here to register for the event There is no charge for this event. A Bayesian analysis of the impact of air pollution episodes on cardio-respiratory hospital admissions in the Greater London area Marta Blangiardo, Sylvia Richardson, John Gulliver and Anna Hansell (Department of Epidemiology and Biostatistics, School of Public Health, Imperial College and MRC-HPA Centre for Environment and Health) In this talk the speaker will present a Bayesian hierarchical model to evaluate the effect of long-range and local range PM10 during air pollution episodes on hospital admissions for cardio-respiratory diseases in Greater London. These episodes in 2003 are matched with the same periods during the previous year, used as a control. A baseline dose-response function is estimated for the controls and carried forward in the episodes, which are characterised by an additional component that estimates their specific effect on the health outcome. Dirichlet process mixture models used to measure the association between deprivation indicators and air pollution John Molitor (Imperial College) Standard regression analyses are often plagued with problems encountered when one tries to make inference going beyond main effects, using datasets that contain dozens of variables that are potentially correlated. This situation arises, for example, in environmental deprivation studies where a large number of deprivation scores are used as covariates yielding a potentially unwieldy set of inter-related data from which teasing out the joint effect of multiple deprivation indices is difficult. We propose a method, based on Dirichlet-process mixture models, that addresses these problems by using, as its basic unit of inference, a profile formed from a sequence of continuous deprivation measures. These deprivation profiles are clustered into groups and associated via a regression model to an air pollution outcome. The Bayesian clustering aspect of the proposed modeling framework has a number of advantages over traditional clustering approaches in that it allows the number of groups to vary, uncovers clusters and examines their association with an outcome of interest and fits the model as a unit, allowing a region's outcome potentially to influence cluster membership. The method is demonstrated with an analysis indicators of deprivation and measures of multiple air pollution exposures in Los Angeles County. Bayesian profile regression with an application to the study of lung cancer in a large cohort study and in a GWA study Michail Papathomas, John Molitor, Sylvia Richardson, Clive Hoggart, Elio Riboli and Paolo Vineis (Imperial College) Standard regression analyses are often plagued with problems encountered when one tries to make meaningful inference going beyond main effects, using datasets that contain dozens of potentially correlated variables . We propose a method that addresses these problems by using, as its basic unit of inference, a profile, formed from a sequence of covariate values. These covariate profiles are clustered into groups using the Dirichlet process, and are associated via a regression model to a relevant outcome. The Bayesian clustering aspect of the proposed modelling framework has a number of advantages over traditional clustering approaches in that it allows the number of groups to vary, allows comparison of arbitrary subgroups of the data, can incorporate a priori known structures, uncovers subgroups based on their association with an outcome of interest and fits the model as a unit, allowing an individual's outcome to influence cluster membership. Different variable selection approaches are introduced and compared. Profile regression has been applied to a large cohort study in order to examine the effect of environmental carcinogens and explore possible gene-environment interactions. It has also been applied to a GWA study on lung cancer, in order to explore gene-gene and gene-environment interactions. Zero-inflated Binomial models for the analysis of amphybian decline Susan F Walker, Jaime Bosch, Virgilio Gomez (University of Barcelona), Trenton W J Garner, Andrew A Cunningham, Dirk S Schmeller, Miguel Ninyerola, Daniel A Henk, Cedric Ginestet, Christian-Philippe Arthur and Matthew C Fisher In this talk the speakers will show how zero-inflated Binomial models can be used to model amphybian epidemics. In particular, they will focus on the analydes of survey data on mortality by Amphibian chytridiomycosis, which is a disease caused by the fungus Batrachochytrium dendrobatidis. These data have been collected at different points of the Iberian Peninsula over a period of several years. The aim of this work is to dissentangle the effects of different environmental factors that may drive the spread of the fungus and the disease. Contributors will formulate their models under the Bayesian framework. They aim at modelling not only the spread of the disease but also, given that the disease is present at a site, what are the main environmental factors driving it. Meeting Contact: Paul Hewson ([log in to unmask]) or John Molitor ([log in to unmask]) Organising Group(s): General Applications Section You may leave the list at any time by sending the command SIGNOFF allstat to [log in to unmask], leaving the subject line blank.