Search

Geospatial methods are increasingly used to evaluate environmental exposures in epidemiologic studies. Geospatial data are often acquired from different sources and time periods and at different spatial resolutions. This dissertation presents methods for combining datasets that do not overlap in space and, or time and are essential for the accurate quantification of metrics used in Environmental Health Sciences. In three chapters, we demonstrated and compared advances in small-area geospatial methods used to reduce misalignment and misclassification of key variables (e.g., exposure).First, we showed how small-area analyses of COVID-19 can benefit from spatial aggregation to account for areal misalignment. In an analysis of the association between air pollution levels and COVID-19 incidence and mortality, we used residential building footprints to combine misaligned COVID-19 outcomes recorded at the neighborhood level, population demographics recorded at the census-tract level, and NO2 interpolated surfaces recorded at 30-meter grids. We found NO2 to be positively associated with COVID-19 incidence and mortality for neighborhoods in Los Angeles.Second, we used data acquired through the PASTA-LA study to attribute daily green space exposure for physically-active spaces (PASs) where activity, location, and green space data were misaligned. We used tracking data from accelerometers and smartphones to attribute green-space exposure using 21 geospatial methods. We found that exposures attributed to home-address buffers, a commonly used method, can result in exposure misclassification. We also found a large range in correlation depending on tracking device used and the way physical activity was categorized across the 21 methods, suggesting that the method selected is key to the findings.Finally, third, also using PASTA-LA data, we studied the association between heat exposure and physical activity in which temperature, green space, and participant-level covariates were misaligned. We showed that heat exposure was associated with increased physical activity and that exposure to green space modified this association in only some models. In this chapter, and in all chapters of this dissertation, we demonstrated how choice in geospatial approach to spatiotemporal misalignment can yield different study results.

This scoping review aims to synthesise the evidence on the associations between place-based risk factors and multimorbidity.

This is the preregistration of the protocol for a scoping review, which aims to synthesise the evidence on the associations between place-based risk factors and multimorbidity. This scoping review examines peer-reviewed observational studies published between January 2010 to March 2023.

Subnational mortality estimation allows measuring spatial disparities in mortality and their evolution over time. It thus reflects a fundamental aspect of health inequalities. This however means estimating mortality indicators for small population sizes where the stochasticity in death counts is high. This, in turn, can lead to unclear underlying mortality levels. In these contexts, Bayesian Hierarchical Models (BHM) offer good performances, finding an appropriate balance between robustness and sensitivity. This dissertation is articulated around three applications where the research questions related to subnational mortality estimation are addressed thanks to statistical opportunities offered by BHM. First, I assess the possibility to estimate the probability of dying for children aged 5 to 14 years old at a subnational level in a sample of Sub-Saharan countries using survey data. Second, I measure the heterogeneity of the mortality shock in the context of the COVID-19 pandemic at the district level in Belgium. Third, I compare the performances of models allowing to estimate mortality age schedules at a subnational level. These three applications allow defining general guidelines for subnational mortality estimation. (POLS - Sciences politiques et sociales) -- UCL, 2023

We consider disease mapping of early- and late-stage cancer, in order to identify and monitor inequalities in early detection. Our method is demonstrated by mapping cancer incidence at high geographical resolution using data on 10,302 cutaneous malignant melanoma (CMM) cases within the 3.7 million population of South-West Sweden. The cases were geocoded into small-areas, each with a population size between 600 and 2600 and accessible socio-demographic data. Using the disease mapping application Rapid Inquiry Facility (RIF) 4.0, we produced regional maps to visualise spatial variations in stage I, II and III–IV CMM incidences, complemented by local maps to explore the variations within two urban areas. Pronounced spatial disparities in stage I CMM incidence were revealed by the regional and local maps. Stage I CMM incidence was markedly higher in wealthier small-areas, in particular within each urban area. A twofold higher stage I incidence was observed, on average, in the wealthiest small-areas (upper quintile) than in the poorest small-areas (lower quintile). We identified in the regional map of stage III–IV CMM two clusters of higher or lower than expected late-stage incidences which were quite distinct from those identified for stage I. In conclusion, our analysis of CMM incidences supported the use of this method of cancer stage incidence mapping for revealing geographical and socio-demographic disparities in cancer detection. Electronic supplementary material The online version of this article (10.1007/s10654-020-00637-0) contains supplementary material, which is available to authorized users.

The turn of the millennium has been accompanied by a rapid growth in data collection along with an increasing ability to store, manipulate and analyze it. In tandem with this development in technology and surveillance, there has been a growing understanding of the importance of the socio-physical environment on population behavior and human health. We capitalize on this progress to develop a methodology by which we develop two macro scale datasets, one national and one state-wide, to support the efforts of the Centers for Disease Control and Prevention's Environmental Public Health Tracking Network (EPHTN) in understanding two important health risks: smoking and obesity. Moreover, we use new geographic information science techniques, spatial statistics methodologies and machine learning algorithms to gain a better understanding of the relationship between spatial patterns in physical and socio-demographic characteristics and health risks. Specifically, we address three specific aims: (1) To use current data systems to develop national small-area predictions of adult smoking and obesity for the EPHTN and research; (2) To evaluate current data systems and spatial analysis tools available to describe the within-school environment and the school-neighborhood socio-physical characteristics of California's public schools thought to influence childhood obesity, and use these to develop a comprehensive multilevel dataset for the EPHTN and research; (3) To test the utility of the dataset developed in aim 2 by applying it to an analysis used to increase the understanding of the relationship between the school environment and childhood obesity by examining the relative importance of school attributes. This analysis determines that current data collection systems provide a valuable resource which we combine for the ease of future research use. We demonstrate that using the spatial structure and socio-demographic patterns of health risks, we are able to downscale adult smoking and obesity prevalence to the ZIP code and census tract levels for the conterminous United States, and develop five-year predictions for these for the four quinquennia in the 1991-2010 time period. Lastly, we confirm the utility of the school dataset and determine through the third aim that individual demographics and the social environment seem to be the predominant determinants of childhood obesity.

Long-term exposure to primary traffic pollutants may be harmful for health but few studies have investigated effects on mortality. We examined associations for six primary traffic pollutants with allcause and cause-specific mortality in 2003e2010 at small-area level using linear and piecewise linear Poisson regression models. In linear models most pollutants showed negative or null association with allcause, cardiovascular or respiratory mortality. In the piecewise models we observed positive associations in the lowest exposure range (e.g. relative risk (RR) for all-cause mortality 1.07 (95% credible interval (CI) ¼ 1.00e1.15) per 0.15 mg/m 3 increase in exhaust related primary particulate matter � 2.5 mm (PM2.5)) whereas associations in the highest exposure range were negative (corresponding RR 0.93, 95% CI: 0.91 e0.96). Overall, there was only weak evidence of positive associations with mortality. That we found the strongest positive associations in the lowest exposure group may reflect residual confounding by unmeasured confounders that varies by exposure group. © 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license

Background: The emergence and evolution of socioeconomic inequalities in health involves multiple factors interacting with each other at different levels. Simulation models are suitable for studying such complex and dynamic systems and have the ability to test the impact of policy interventions in silico. Objective: To explore how simulation models were used in the field of socioeconomic inequalities in health. Methods: An electronic search of studies assessing socioeconomic inequalities in health using a simulation model was conducted. Characteristics of the simulation models were extracted and distinct simulation approaches were identified. As an illustration, a simple agent-based model of the emergence of socioeconomic differences in alcohol abuse was developed. Results: We found 61 studies published between 1989 and 2013. Ten different simulation approaches were identified. The agent-based model illustration showed that multilevel, reciprocal and indirect effects of social determinants on health can be modeled flexibly. Discussion and Conclusions: Based on the review, we discuss the utility of using simulation models for studying health inequalities, and refer to good modeling practices for developing such models. The review and the simulation model example suggest that the use of simulation models may enhance the understanding and debate about existing and new socioeconomic inequalities of health frameworks.

There has long been a demand for cancer incidence data at a fine geographic resolution for use in etiologic hypothesis generation and testing, methodological evaluation and teaching. In this paper we describe a public domain dataset containing data for 23 anatomic sites of cancer diagnosed in New York State, USA between 2005 and 2009 at the census block group level. The dataset includes 524,503 tumours distributed across 13,823 block groups with an average population of about 1400. In addition, the data have been linked with race/ethnicity and with socioeconomic indicators such as income, educational attainment and language proficiency. We demonstrate the application of the dataset by confirming two well-established relationships: that between breast cancer and median household income and that between stomach cancer and Asian race. We foresee that this dataset will serve as the basis for a wide range of spatial analyses and as a benchmark for evaluating spatial methods in the future.

Cette thèse comporte trois articles dont un est publié et deux en préparation. Le sujet central de la thèse porte sur le traitement des valeurs aberrantes représentatives dans deux aspects importants des enquêtes que sont : l’estimation des petits domaines et l’imputation en présence de non-réponse partielle. En ce qui concerne les petits domaines, les estimateurs robustes dans le cadre des modèles au niveau des unités ont été étudiés. Sinha & Rao (2009) proposent une version robuste du meilleur prédicteur linéaire sans biais empirique pour la moyenne des petits domaines. Leur estimateur robuste est de type «plugin», et à la lumière des travaux de Chambers (1986), cet estimateur peut être biaisé dans certaines situations. Chambers et al. (2014) proposent un estimateur corrigé du biais. En outre, un estimateur de l’erreur quadratique moyenne a été associé à ces estimateurs ponctuels. Sinha & Rao (2009) proposent une procédure bootstrap paramétrique pour estimer l’erreur quadratique moyenne. Des méthodes analytiques sont proposées dans Chambers et al. (2014). Cependant, leur validité théorique n’a pas été établie et leurs performances empiriques ne sont pas pleinement satisfaisantes. Ici, nous examinons deux nouvelles approches pour obtenir une version robuste du meilleur prédicteur linéaire sans biais empirique : la première est fondée sur les travaux de Chambers (1986), et la deuxième est basée sur le concept de biais conditionnel comme mesure de l’influence d’une unité de la population. Ces deux classes d’estimateurs robustes des petits domaines incluent également un terme de correction pour le biais. Cependant, ils utilisent tous les deux l’information disponible dans tous les domaines contrairement à celui de Chambers et al. (2014) qui utilise uniquement l’information disponible dans le domaine d’intérêt. Dans certaines situations, un biais non négligeable est possible pour l’estimateur de Sinha & Rao (2009), alors que les estimateurs proposés exhibent un faible biais pour un choix approprié de la fonction d’influence et de la constante de robustesse. Les simulations Monte Carlo sont effectuées, et les comparaisons sont faites entre les estimateurs proposés et ceux de Sinha & Rao (2009) et de Chambers et al. (2014). Les résultats montrent que les estimateurs de Sinha & Rao (2009) et de Chambers et al. (2014) peuvent avoir un biais important, alors que les estimateurs proposés ont une meilleure performance en termes de biais et d’erreur quadratique moyenne. En outre, nous proposons une nouvelle procédure bootstrap pour l’estimation de l’erreur quadratique moyenne des estimateurs robustes des petits domaines. Contrairement aux procédures existantes, nous montrons formellement la validité asymptotique de la méthode bootstrap proposée. Par ailleurs, la méthode proposée est semi-paramétrique, c’est-à-dire, elle n’est pas assujettie à une hypothèse sur les distributions des erreurs ou des effets aléatoires. Ainsi, elle est particulièrement attrayante et plus largement applicable. Nous examinons les performances de notre procédure bootstrap avec les simulations Monte Carlo. Les résultats montrent que notre procédure performe bien et surtout performe mieux que tous les compétiteurs étudiés. Une application de la méthode proposée est illustrée en analysant les données réelles contenant des valeurs aberrantes de Battese, Harter & Fuller (1988). S’agissant de l’imputation en présence de non-réponse partielle, certaines formes d’imputation simple ont été étudiées. L’imputation par la régression déterministe entre les classes, qui inclut l’imputation par le ratio et l’imputation par la moyenne sont souvent utilisées dans les enquêtes. Ces méthodes d’imputation peuvent conduire à des estimateurs imputés biaisés si le modèle d’imputation ou le modèle de non-réponse n’est pas correctement spécifié. Des estimateurs doublement robustes ont été développés dans les années récentes. Ces estimateurs sont sans biais si l’un au moins des modèles d’imputation ou de non-réponse est bien spécifié. Cependant, en présence des valeurs aberrantes, les estimateurs imputés doublement robustes peuvent être très instables. En utilisant le concept de biais conditionnel, nous proposons une version robuste aux valeurs aberrantes de l’estimateur doublement robuste. Les résultats des études par simulations montrent que l’estimateur proposé performe bien pour un choix approprié de la constante de robustesse., This thesis focuses on the treatment of representative outliers in two important aspects of surveys: small area estimation and imputation for item non-response. Concerning small area estimation, robust estimators in unit-level models have been studied. Sinha & Rao (2009) proposed estimation procedures designed for small area means, based on robustified maximum likelihood parameters estimates of linear mixed model and robust empirical best linear unbiased predictors of the random effect of the underlying model. Their robust methods for estimating area means are of the plug-in type, and in view of the results of Chambers (1986), the resulting robust estimators may be biased in some situations. Biascorrected estimators have been proposed by Chambers et al. (2014). In addition, these robust small area estimators were associated with the estimation of the Mean Square Error (MSE). Sinha & Rao (2009) proposed a parametric bootstrap procedure based on the robust estimates of the parameters of the underlying linear mixed model to estimate the MSE. Analytical procedures for the estimation of the MSE have been proposed in Chambers et al. (2014). However, their theoretical validity has not been formally established and their empirical performances are not fully satisfactorily. Here, we investigate two new approaches for the robust version the best empirical unbiased estimator: the first one relies on the work of Chambers (1986), while the second proposal uses the concept of conditional bias as an influence measure to assess the impact of units in the population. These two classes of robust small area estimators also include a correction term for the bias. However, they are both fully bias-corrected, in the sense that the correction term takes into account the potential impact of the other domains on the small area of interest unlike the one of Chambers et al. (2014) which focuses only on the domain of interest. Under certain conditions, non-negligible bias is expected for the Sinha-Rao method, while the proposed methods exhibit significant bias reduction, controlled by appropriate choices of the influence function and tuning constants. Monte Carlo simulations are conducted, and comparisons are made between: the new robust estimators, the Sinha-Rao estimator, and the bias-corrected estimator. Empirical results suggest that the Sinha-Rao method and the bias-adjusted estimator of Chambers et al (2014) may exhibit a large bias, while the new procedures offer often better performances in terms of bias and mean squared error. In addition, we propose a new bootstrap procedure for MSE estimation of robust small area predictors. Unlike existing approaches, we formally prove the asymptotic validity of the proposed bootstrap method. Moreover, the proposed method is semi-parametric, i.e., it does not rely on specific distributional assumptions about the errors and random effects of the unit-level model underlying the small-area estimation, thus it is particularly attractive and more widely applicable. We assess the finite sample performance of our bootstrap estimator through Monte Carlo simulations. The results show that our procedure performs satisfactorily well and outperforms existing ones. Application of the proposed method is illustrated by analyzing a well-known outlier-contaminated small county crops area data from North-Central Iowa farms and Landsat satellite images. Concerning imputation in the presence of item non-response, some single imputation methods have been studied. The deterministic regression imputation, which includes the ratio imputation and mean imputation are often used in surveys. These imputation methods may lead to biased imputed estimators if the imputation model or the non-response model is not properly specified. Recently, doubly robust imputed estimators have been developed. However, in the presence of outliers, the doubly robust imputed estimators can be very unstable. Using the concept of conditional bias as a measure of influence (Beaumont, Haziza and Ruiz-Gazen, 2013), we propose an outlier robust version of the doubly robust imputed estimator. Thus this estimator is denoted as a triple robust imputed estimator. The results of simulation studies show that the proposed estimator performs satisfactorily well for an appropriate choice of the tuning constant.

Evidence on the effects of long-term exposure to traffic pollution on health is inconsistent. In Greater London we examined associations between traffic pollution and emergency hospital admissions for cardio-respiratory diseases by applying linear and piecewise linear Poisson regression models in a small-area analysis. For both models the results for children and adults were close to unity. In the elderly, linear models found negative associations whereas piecewise models found non-linear associations characterized by positive risks in the lowest and negative risks in the highest exposure category. An increased risk was observed among those living in areas with the highest socioeconomic deprivation. Estimates were not affected by adjustment for traffic noise. The lack of convincing positive linear associations between primary traffic pollution and hospital admissions agrees with a number of other reports, but may reflect residual confounding. The relatively greater vulnerability of the most deprived populations has important implications for public health.

Background There is growing evidence that areas characterised by high levels of social fragmentation have higher suicide rates. Previous ecological studies have focused on relatively large geographic areas and/or examined associations in all age groups combined. Methods Negative binominal regression was used to assess age- and sex-specific suicide rate ratios for a range of census-derived indicators of the social, health and economic characteristics of small areas (mean population aged ≥ 15: 4500) in England and Wales. Results Indicators of social fragmentation (e. g. proportion of people living alone or population mobility) were most consistently associated with suicide risk. For example, across quartiles of wards ranked according to increasing proportions of single-person households, age- and sexadjusted suicide rate ratios were: 1.00, 1.05 (1.00, 1.11), 1.14 (1.08, 1.19) and 1.42 (1.36, 1.49). Associations were strongest in 15 to 44 and 45 to 64 year-olds.Associations with social fragmentation persisted after controlling for the effect of other area characteristics. Conclusions Targeted mental health promotion and social policy initiatives to reduce area-health inequalities in suicide might usefully focus on areas with high levels of social fragmentation.

BACKGROUND: Transport-related air and noise pollution, exposures linked to adverse health outcomes, varies within cities potentially resulting in exposure inequalities. Relatively little is known regarding inequalities in personal exposure to air pollution or transport-related noise. OBJECTIVES: Our objectives were to quantify socioeconomic and ethnic inequalities in London in 1) air pollution exposure at residence compared to personal exposure; and 2) transport-related noise at residence from different sources. METHODS: We used individual-level data from the London Travel Demand Survey (n = 45,079) between 2006 and 2010. We modeled residential (CMAQ-urban) and personal (London Hybrid Exposure Model) particulate matter £75,000) had lower residential NO2 (-1.3 (95% CI -2.1, -0.6) μg/m3 in the 95th exposure quantile) but higher personal NO2 exposure (1.9 (95% CI 1.6, 2.3) μg/m3 in the 95th quantile), which was driven largely by transport mode and duration. Inequalities in residential exposure to NO2 with respect to area-level deprivation were larger at lower exposure quantiles (e.g. estimate for NO2 5.1 (95% CI 4.6, 5.5) at quantile 0.15 versus 1.9 (95% CI 1.1, 2.6) at quantile 0.95), reflecting low-deprivation, high residential NO2 areas in the city centre. Air pollution exposure at residence consistently overestimated personal exposure; this overestimation varied with age, household income, and area-level income deprivation. Inequalities in road traffic noise were generally small. In logistic regression models, the odds of living within a 50 dB contour of aircraft noise were highest in individuals with the highest household income, white ethnicity, and with the lowest area-level income deprivation. Odds of living within a 50 dB contour of rail noise were 19% (95% CI 3, 37) higher for black compared to white individuals. CONCLUSIONS: Socioeconomic inequalities in air pollution exposure were different for modeled residential versus personal exposure, which has important implications for environmental justice and confounding in epidemiology studies. Exposure misclassification was dependent on several factors related to health, a potential source of bias in epidemiological studies. Quantile regression revealed that socioeconomic and ethnic inequalities in air pollution are often not uniform across the exposure distribution. This work was supported by the UK Natural Environment Research Council, Medical Research Council, Economic and Social Research Council, Department of Environment, Food and Rural Affairs, and Department of Health (NE/I007806/1, NE/I00789X/1, NE/I008039/1) through the cross-research council Environmental Exposures & Health Initiative. The work of the UK Small Area Health Statistics Unit is funded by Public Health England as part of the MRC-PHE Centre for Environment and Health, funded also by the UK Medical Research Council. CT was funded through a Ramón y Cajal fellowship (RYC-2015-17402) awarded by the Spanish Ministry of Economy and Competitiveness