Aplicação de um sistema de indicadores intraurbano à cidade de São Luís para mensurar a sua desigualdade socioambiental. Como instrumentos para medir e informar sobre as cidades, os indicadores urbanos podem ser classificados em: indicadores intermunicipais e intramunicipais. Com base nesta divisão, utiliza-se a escala intraurbana para propor um sistema de indicadores aplicado à capital maranhense, subdividido em quatro dimensões: habitabilidade, saneamento, educação e renda. Para compor os indicadores das respectivas dimensões, foram utilizados os dados secundários do censo IBGE/2000, desagregados ao nível dos 780 setores censitários de São Luís (MA). A análise estatística dos dados é fundamentada em técnicas multivariadas, a saber: Análise Fatorial (AF) e Análise de Clusters (AC). A primeira técnica permitiu resumir e reduzir os dados censitários, sem perda significativa das informações originais. A segunda técnica proporcionou a definição e a agregação dos setores censitários em quatro clusters distintos. Com os resultados estatísticos dos indicadores que componham cada uma das quatro dimensões, os clusters identificados foram mapeados e comparados entre si para dimensionar a desigualdade intraurbana. Na dimensão habitabilidade, avaliou-se dois aspectos: padrão de moradia e o acesso à propriedade. Observou-se que não há grandes desigualdades intraurbanas na dimensão habitabilidade. A segunda dimensão foi o saneamento, com ênfase em três subdimensões: abastecimento de água, rede coletora de esgotos e de lixo. Nesse tópico a maior desigualdade intraurbana está relacionada ao serviço de coleta de esgotos, seguido do serviço de coleta de lixo e abastecimento de água. Na dimensão educação, abordaram-se duas subdimensões: alfabetização da população por faixa etária e escolaridade do responsável pelas residências. As desigualdades intraurbanas não revelaram grandes discrepâncias quanto aos indicadores educacionais entre os quatro clusters na cidade, exceto para a subdimensão, escolaridade do responsável com 15 anos de estudo, sobretudo, quando comparado às áreas norte e sul do espaço urbano. A dimensão renda do responsável foi o tópico que revelou a maior discrepância intraurbana em relação às demais dimensões avaliadas e entre os clusters. Através do sistema de indicadores intraurbanos, proposto para São Luís, concluiu-se que espaço urbano é desigual social e ambientalmente. As diferenças internas da cidade foram avaliadas e cartograficamente representadas, utilizou-se para elaboração dos mapas, o programa ArcGIS versão 9.3. / The aim of this work is to establish and apply an interurban system of indices to the city of São Luís in order to measure its social and environmental inequality. As instruments to measure and inform about cities, urban indices may be classified as: intermunicipal and intramunicipal indices. Based on this division, the interurban scale is used to propose a system of indices applied to the capital of the State of Maranhão, subdivided into four dimensions: habitability, sanitation, education and income. In order to compose the indices for the respective dimensions, the secondary data of the Demographic Census 2000 by IBGE (Brazilian Institute of Geography and Statistics) were utilized, disaggregated to the level of the 780 sectors surveyed in São Luís (MA). The statistical data analysis is based on multivaried techniques, viz: factorial analysis (FA) and clusters Analysis (CA). The first technique allowed to abridge and to reduce the data of the census statistics, without significant loss of the original pieces of information. The second technique made possible the definition and the aggregation of the surveys sectors in four distinct clusters. With the statistical results of the indices which composed each one of the four dimensions, the clusters identified were mapped out and compared among each of them in order to measure interurban inequality. In the habitability dimension, two aspects were evaluated: housing standard and access to property. It was observed that there are not significant intraurban inequalities in the dimension habitability. The second dimension was sanitation, with emphasis on three subdimensions: water supply, sewage system and garbage. In this topic the greatest intraurban inequality is related to the sewage system, along with the garbage collection and water supply. In the dimension of education, two subdimensions were approached: literacy of population by age group and schooling of the responsible for the houses. Intraurban inequalities did not reveal great discrepancies in relation to the educational indices among the four clusters in the city, except for the subdimension schooling of the responsible for the house with 15 years of study, mostly, when compared to the northern and southern areas of the urban space. The dimension guarantors income was the topic which revealed the greatest intraurban discrepancy in relation to the other dimensions evaluated and among clusters. Through intraurban system of indices proposed to São Luís, it was concluded that the urban space is not equal either in social affairs or in environmental affairs. The inner differences in the city were revealed and represented in chart, was used for mapping, the program ArcGIS version 9.3.

abstract: Cities are hubs for economic and social development, but they are increasingly becoming hotspots of environmental problems and socio-economic inequalities. Because cities result from complex interactions among ecological, social and economic factors, environmental problems and socio-economic inequalities are often spatially interconnected, generating emergent environmental inequity issues due to the unfair distribution of environmental quality among socioeconomic groups. Since urban environmental quality is tightly related to the capacity of urban landscapes to provide ecosystem services, optimizing the allocation of ecosystem services within cities is a main goal for moving towards more equitable and sustainable cities. Nevertheless, we often lack the empirical data and specific methods for planning urban landscapes to optimize the provision of ecosystem services. Therefore, the development of knowledge and methods to optimize the provision of ecosystem services is essential for tackling urban environmental problems, reducing environmental inequities, and promoting sustainable cities. The main goal of this dissertation is to generate actionable knowledge for helping decision-makers to optimize the allocation of urban vegetation for reducing environmental inequities through the provision of ecosystem services. The research uses the city of Santiago de Chile as a case study from a Latin-American city. To achieve this goal, I framed my dissertation in four linked research chapters, each of them providing methodological approaches to help link environmental inequity problems with the development of urban planning interventions promoting an equitable provision of urban ecosystem services. These chapters are specifically aimed at providing actionable knowledge for: (1) Identifying the level, distribution, and spatial scales at which environmental inequities are more relevant; (2) Identifying the areas and administrative units where environmental inequities interventions should be prioritized; (3) Identifying optimal areas to allocate vegetation for increasing the provision of urban ecosystem services; (4) Evaluating the role that planned urban vegetation may have in the long-term provision of ecosystem services by natural remnants within the urban landscape. Thus, this dissertation contributes to urban sustainability science by proposing methods and frameworks to address urban environmental inequities through the provision of ecosystem services, but it also provides place-based information that can be readily used for planning urban vegetation in Santiago. / Dissertation/Thesis / Doctoral Dissertation Sustainability 2017

Cette thèse étudie les inégalités environnementales dans un objectif de justice sociale. La première partie propose une définition interdisciplinaire de ces inégalités afin de capter son caractère multidimensionnel et de montrer son utilité en économie de l’environnement et en économie écologique. À partir de la littérature en philosophie économique, le caractère juste ou injuste des inégalités environnementales est ensuite analysé en mobilisant différentes conceptions égalitaristes de la justice et en considérant l’existence d’autres objectifs sociétaux. À la suite de ce travail théorique, la seconde partie s’articule autour d’une application aux enjeux de l’accès à l’eau et à l’assainissement au Brésil. Après avoir présenté la situation relative à ces services dans le contexte brésilien, les déterminants socioéconomiques et institutionnels de cet accès sont identifiés par des analyses économétriques sur des données à l’échelle des ménages. Ces analyses permettent de mettre en lumière des déterminants au cœur des problématiques de justice environnementale comme le revenu, l’ethnicité, mais aussi le rôle des États en ayant recours à des modèles multi-niveaux. En utilisant un indicateur d’accès à l’eau et à l’assainissement construit à partir d’une analyse factorielle, les inégalités associées à ces services sont ensuite mesurées par différentes méthodes correspondant à chacune des conceptions de la justice développées dans la première partie / This thesis studies the environmental inequalities in an objective of social justice. The first part offers an interdisciplinary definition of these inequalities in order to catch their multidimensional nature and to show their usefulness in environmental economics and ecological economics. Based on a philosophy of economics perspective, the fairness of environmental inequalities is analysed by using different egalitarian conceptions of justice with a consideration of the existence of other societal objectives. Following this theoretical development, the second part is devoted to an analysis of the issues of access to water and sanitation services in the context of Brazil. After a discussion over the situation related to the water and sanitation sector in Brazil, the socioeconomic and institutional determinants of the access to these services are identified using econometric analyses at the household level. These analyses enable to expose key determinants in the perspective of environmental justice including income level and ethnicity. In addition, they show the role of the state level thanks to the use of multilevel models. By using an indicator of the access to water and sanitation services elaborated through a factor analysis, the inequalities associated to these services are then measured by different methods which correspond to each conception of justice developed in the first part.

Aplicação de um sistema de indicadores intraurbano à cidade de São Luís para mensurar a sua desigualdade socioambiental. Como instrumentos para medir e informar sobre as cidades, os indicadores urbanos podem ser classificados em: indicadores intermunicipais e intramunicipais. Com base nesta divisão, utiliza-se a escala intraurbana para propor um sistema de indicadores aplicado à capital maranhense, subdividido em quatro dimensões: habitabilidade, saneamento, educação e renda. Para compor os indicadores das respectivas dimensões, foram utilizados os dados secundários do censo IBGE/2000, desagregados ao nível dos 780 setores censitários de São Luís (MA). A análise estatística dos dados é fundamentada em técnicas multivariadas, a saber: Análise Fatorial (AF) e Análise de Clusters (AC). A primeira técnica permitiu resumir e reduzir os dados censitários, sem perda significativa das informações originais. A segunda técnica proporcionou a definição e a agregação dos setores censitários em quatro clusters distintos. Com os resultados estatísticos dos indicadores que componham cada uma das quatro dimensões, os clusters identificados foram mapeados e comparados entre si para dimensionar a desigualdade intraurbana. Na dimensão habitabilidade, avaliou-se dois aspectos: padrão de moradia e o acesso à propriedade. Observou-se que não há grandes desigualdades intraurbanas na dimensão habitabilidade. A segunda dimensão foi o saneamento, com ênfase em três subdimensões: abastecimento de água, rede coletora de esgotos e de lixo. Nesse tópico a maior desigualdade intraurbana está relacionada ao serviço de coleta de esgotos, seguido do serviço de coleta de lixo e abastecimento de água. Na dimensão educação, abordaram-se duas subdimensões: alfabetização da população por faixa etária e escolaridade do responsável pelas residências. As desigualdades intraurbanas não revelaram grandes discrepâncias quanto aos indicadores educacionais entre os quatro clusters na cidade, exceto para a subdimensão, escolaridade do responsável com 15 anos de estudo, sobretudo, quando comparado às áreas norte e sul do espaço urbano. A dimensão renda do responsável foi o tópico que revelou a maior discrepância intraurbana em relação às demais dimensões avaliadas e entre os clusters. Através do sistema de indicadores intraurbanos, proposto para São Luís, concluiu-se que espaço urbano é desigual social e ambientalmente. As diferenças internas da cidade foram avaliadas e cartograficamente representadas, utilizou-se para elaboração dos mapas, o programa ArcGIS versão 9.3. / The aim of this work is to establish and apply an interurban system of indices to the city of São Luís in order to measure its social and environmental inequality. As instruments to measure and inform about cities, urban indices may be classified as: intermunicipal and intramunicipal indices. Based on this division, the interurban scale is used to propose a system of indices applied to the capital of the State of Maranhão, subdivided into four dimensions: habitability, sanitation, education and income. In order to compose the indices for the respective dimensions, the secondary data of the Demographic Census 2000 by IBGE (Brazilian Institute of Geography and Statistics) were utilized, disaggregated to the level of the 780 sectors surveyed in São Luís (MA). The statistical data analysis is based on multivaried techniques, viz: factorial analysis (FA) and clusters Analysis (CA). The first technique allowed to abridge and to reduce the data of the census statistics, without significant loss of the original pieces of information. The second technique made possible the definition and the aggregation of the surveys sectors in four distinct clusters. With the statistical results of the indices which composed each one of the four dimensions, the clusters identified were mapped out and compared among each of them in order to measure interurban inequality. In the habitability dimension, two aspects were evaluated: housing standard and access to property. It was observed that there are not significant intraurban inequalities in the dimension habitability. The second dimension was sanitation, with emphasis on three subdimensions: water supply, sewage system and garbage. In this topic the greatest intraurban inequality is related to the sewage system, along with the garbage collection and water supply. In the dimension of education, two subdimensions were approached: literacy of population by age group and schooling of the responsible for the houses. Intraurban inequalities did not reveal great discrepancies in relation to the educational indices among the four clusters in the city, except for the subdimension schooling of the responsible for the house with 15 years of study, mostly, when compared to the northern and southern areas of the urban space. The dimension guarantors income was the topic which revealed the greatest intraurban discrepancy in relation to the other dimensions evaluated and among clusters. Through intraurban system of indices proposed to São Luís, it was concluded that the urban space is not equal either in social affairs or in environmental affairs. The inner differences in the city were revealed and represented in chart, was used for mapping, the program ArcGIS version 9.3.

Depuis la fin du 19ème siècle, l’espace minier de Gafsa (Sud-Ouest de Tunisie) s’est spécialisé dans la mono-activité de l’extraction et l’enrichissement du phosphate pour des raisons de profits économiques certains. Avec ces exportations, l’espace a subi une dégradation intense des ressources du sol, de la couverture végétale et une détérioration avancée des zones agricoles. La gestion des infrastructures minières n’a pas su répondre aux normes environnementales et a provoqué ainsi des inégalités sociales et spatiales et a accentué le niveau des risques environnementaux.L’activité minière a été à l’origine de la naissance et de la genèse des villes minières dans le Bassin Minier de Gafsa (BMG). La formation de cités spontanées a provoqué la désarticulation du tissu urbain. Les extensions urbaines notamment l’expansion des quartiers anarchiques, s’étendant vers les laveries et vers les mines, ont créé ainsi des conflits avec le domaine phosphatier. Tenant compte de la complexité des multiples aspects sociaux, spatiaux et environnementaux du BMG, nous avons utilisé à la fois l’approche géographique classique et l’approche géomatique qui nous a semblé appropriée pour aider à la prise de décision. L’intérêt de cette dernière approche permet d’identifier, d’analyser, et de modéliser l’état des lieux, les mutations urbaines et environnementales en relation avec la croissance spatiale-minière. Les principaux résultats permettent d’avoir une banque de données statistiques et cartographiques, à différentes échelles, pour une vision synthétique et une meilleure compréhension du BMG et de réaliser un modèle d’aide à la décision. / Since the end of the 19th century, the mining area of Gafsa (south-west Tunisia) has specialized in the single-activity extraction and enrichment of phosphate for reasons of economic profits. With these exports, this area has suffered an intense degradation of soil resources, vegetation cover and an advanced deterioration of agricultural areas. The management of the mining infrastructure has failed to meet environmental standards, causing social and spatial inequalities and increasing the level of environmental risks.The mining activity was at the origin of the birth and genesis of the mining towns in the Basin of Gafsa (BMG). The formation of spontaneous cities has caused the disarticulation of the urbanization. Urban extensions, including the expansion of anarchic cities, extending to laundries and to mines, thus created conflicts with the phosphatier domain. All these factors make costly and complicated attempts at development and rehabilitation that mining companies do not dare to commit.Taking into account the complexity of the many social, spatial and environmental aspects of the BMG, we used both the classical geographical approach and the geomatics approach that seemed appropriate for better decision-making. The interest of this latter approach makes it possible to identify, analyze, and model the state of the sites, the urban and environmental mutations in relation to the spatial-mining growth.

Encore peu explorée à ce jour, la thématique des inégalités environnementales fait l’objet d’un intérêt grandissant en France. Parmi les multiples définitions proposées et les quelques tentatives d’évaluation réalisées, une acception semble toutefois plus fréquemment rencontrée. Cette dernière renvoie à l’existence d’inégalités sociales dans l’accès aux ressources environnementales et l’exposition aux nuisances, risques et pollutions. Cette thèse s’est donc proposée, à l’échelle de la région Ile-de-France, d’explorer les relations qu’entretiennent les caractéristiques physiques de l’environnement et les caractéristiques socio-économiques des individus et des territoires. Partant d’indicateurs statistiques usuels, sociaux et environnementaux, deux typologies multicritères ont été réalisées à l’échelle des 1300 communes franciliennes. Leur croisement permet de confirmer l’existence d’une correspondance globale entre les caractéristiques environnementales et sociales des communes et ce, à la fois à l’échelle régionale ainsi qu’à l’échelle plus réduite des départements de première couronne francilienne. Nos résultats révèlent toutefois des liaisons plus ou moins prononcées selon les facteurs environnementaux renseignés et plus particulièrement selon la nature même de ces facteurs, porteurs d’aménités ou de désaménités environnementales. Cette première lecture est alors complétée par une enquête par questionnaires auprès des habitants de 6 communes franciliennes (600 personnes). L’enquête, centrée sur le vécu environnemental des habitants, réaffirme en premier lieu l’importance de la qualité de l’environnement en tant que facteur de choix et d’arbitrage résidentiel des ménages. Parmi les variables renseignées, deux registres explicatifs apparaissent plus particulièrement structurants des différences de satisfaction constatées: (i) l’ancrage résidentiel et la relation affective qui s’instaure au lieu de vie ; (ii) la confiance que les individus accordent aux moyens d’action collectifs pour prendre en charge les attentes qu’ils expriment en matière d’environnement. Aussi, l’analyse des déterminants de la satisfaction environnementale révèle que la satisfaction serait peut être moins directement dépendante des variables socio-économiques stricto sensu, des caractéristiques « objectives » de l’environnement, que des capacités et aptitudes différenciées des individus de contrôler leur environnement et d’agir sur celui-ci. Ce faisant, les informations livrées par le ressenti de la population in situ adressent des questions à la mesure empirique des inégalités environnementales et au cadre d’analyse, encore aujourd’hui en débat, de telles situations inégalitaires. / Still little investigated until now, the issue of environmental inequalities is a subject of growing interest in France. Among numerous definitions and a few evaluation attempts, one understanding seems to be more common. It refers to social inequalities in regard of both the access to environmental resources and the exposure to nuisances, risks and pollution. This thesis’ purpose was to explore the relations between the physical characteristics of the environment and the socio-economic characteristics of the individuals and the territories on the scale of the Paris metropolitan area. Two multi-criteria typologies were carried out on the scale of the 1300 cities belonging to the Paris metropolitan area, based on usual statistical indicators – both social and environmental. The crossing of these typologies confirmed the existence of an overall match between the environmental and social characteristics of the cities. This appeared to be true both on the regional scale and on a smaller scale, the departments of Paris’ inner suburbs. Our results revealed connections that are more or less distinct depending on the examined environmental factors and particularly on the very nature of these factors, i.e. whether they encompass environmental amenities or desamenities. The empirical research in form of questionnaires and distributed among inhabitants of 6 cities of the Paris metropolitan area (600 people) completed this first approach. The survey, focusing on the personal environmental experience of the inhabitants, initially reaffirmed the importance of environmental quality as selection criteria and residential arbitration for the households. Among the examined variables, two explanatory registers appeared especially structurant of the recorded differences in satisfaction: (i) the residential anchorage and the emotional relation taking place in the living space ; (ii) the trust individuals grant to means of collective action to convey their environmental expectations. Therefore, the analysis of the determinants of environmental satisfaction unveiled that satisfaction would be be less directly related to neither sole socio-economic variables or “objective” environmental data than to the individuals’ differentiated capacities and aptitudes to control their environment and act upon it. Hence, the information collected on site driven by the population’s feeling questions the empirical measurement of environmental inequalities as well as the framework of analysis for such unequal situations– at this stage still an ongoing discussion.

La réduction des inégalités d’exposition environnementale constitue un axe majeur en santé publique en France comme en témoignent les priorités des différents Plan Nationaux Santé Environnement (PNSE). L’objectif de cette thèse est de développer une approche intégrée pour la caractérisation des inégalités environnementales et l’évaluation de l’exposition spatialisée de la population aux HAP en France.Les données produites dans le cadre des réseaux de surveillance de la qualités des milieux environnementaux sont le reflet de la contamination réelle des milieux et de l’exposition globale des populations. Toutefois, elles ne présentent généralement pas une représentativité spatiale suffisante pour caractériser finement les expositions environnementales, ces réseaux n’ayant pas été initialement conçus dans cet objectif. Des méthodes statistiques sont développées pour traiter les bases de données d’entrée (concentrations environnementales dans l’eau, l’air et le sol) et les rendre pertinentes vis à vis des objectifs définis de caractérisation de l’exposition. Un modèle multimédia d’exposition, interfacé avec un Système d’Information Géographique pour intégrer les variables environnementales, est développé pour estimer les doses d’exposition liées à l’ingestion d’aliments, d’eau de consommation, de sol et à l’inhalation de contaminants atmosphériques. La méthodologie a été appliquée pour trois Hydrocarbures Aromatiques Polycycliques (benzo[a]pyrène, benzo[ghi]pérylène et indéno[1,2,3-cd]pyrène) sur l’ensemble du territoire français. Les résultats permettent de cartographier des indicateurs d’exposition, d’identifier les zones de surexposition et de caractériser les déterminants environnementaux. Dans une logique de caractérisation de l’exposition, la spatialisation des données issues des mesures environnementales pose un certain nombre de questions méthodologiques qui confèrent aux cartes réalisées de nombreuses incertitudes et limites relatives à l’échantillonnage et aux représentativités spatiales et temporelles des données. Celles-ci peuvent être réduites par l’acquisition de données supplémentaires et par la construction de variables prédictives des phénomènes spatiaux et temporels considérés.Les outils de traitement statistique de données développés dans le cadre de ces travaux seront intégrés dans la plateforme PLAINE pour être déclinés sur d’autres polluants en vue de prioriser les mesures de gestion à mettre en œuvre. / Reducing environmental exposure inequalities has become a major focus of public health efforts in France, as evidenced by the French action plans for health and the environment. The aim of this thesis is to develop an integrated approach to characterize environmental inequalities and evaluate the spatialized exposure to PAH in France.The data produced as part of the monitoring quality networks of environmental media reflect the actual contamination of the environment and the overall exposure of the populations. However they do not always provide an adequate spatial resolution to characterize environmental exposures as they are usually not assembled for this specific purpose. Statistical methods are employed to process input databases (environmental concentrations in water, air and soil) in the objective of characterizing the exposure. A multimedia model interfaced with a GIS, allows the integration of environmental variables in order to yield exposure doses related to ingestion of food, water and soil as well as atmospheric contaminants' inhalation.The methodology was applied to three Polycyclic Aromatic Hydrocarbon substances, (benzo[a]pyrene, benzo[ghi]perylene and indeno[1,2,3-cd]pyrene), in France. The results obtained, allowed to map exposure indicators and to identify areas of overexposure and characterize environmental determinants. In the context of exposure characterization, the direct spatialization of available data from environmental measurement datasets poses a certain number of methodological questions which lead to uncertainties related to the sampling and the spatial and temporal representativeness of data. These could be reduced by acquiring additional data or by constructing predictive variables for the spatial and temporal phenomena considered.Data processing algorithms and calculation of exposure carried out in this work, will be integrated in the French coordinated integrated environment and health platform-PLAINE in order to be applied on other pollutants and prioritize preventative actions.

La grossesse est une période sensible et déterminante pour le développement de l’enfant et l’état de santé à l’âge adulte. L'adaptation biologique et physiologique de l'organisme face à des « stresseurs » physiques et psychosociaux au cours de cette période peut ainsi exercer ses effets à l'âge adulte (et possiblement sur plusieurs générations). Cette toxicité différée suppose intrinsèquement la nécessité d’étudier les conséquences des expositions environnementales au cours de la vie foetale et ce suivant une approche holistique intégrant autant les facteurs de risque à des niveaux micro (caractéristiques individuelles) et macro (expositions physiques externes et contexte psycho-social). Etant donné l’hétérogénéité du territoire breton dans lequel s’inscrit ce travail de thèse, nous avons cherché à explorer l’impact des inégalités sociales et de l’environnement physique (pollution atmosphérique) sur l’issue de la grossesse (croissance foetale et risque de prématurité) indépendamment chez des femmes enceintes résidant dans des zones urbaines et rurales. Ce travail s’est appuyé sur les données issues de la cohorte bretonne mère-enfant PELAGIE, qui a inclus 3421 femmes enceintes entre 2002- 2006. Les paramètres anthropométriques et l’âge gestationnel à la naissance ont été renseignés à l’accouchement par le personnel médical. Le contexte urbain et rural breton a été caractérisé à partir de la définition des unités urbaines de l’INSEE. Les concentrations annuelles de pollution atmosphérique (dioxyde d’azote (NO2)) ont été modélisées à une échelle de 100 m à partir d’un modèle de "land-use regression" développé à l’échelle européenne. Enfin, le niveau socio-économique des IRIS a été estimé à l’aide d’un indice de désavantage social - construit à partir des données du recensement de l’INSEE et dont la validité et l’adaptabilité à des territoires à la fois urbains et ruraux a été examinée au préalable. Nos résultats suggèrent une influence délétère d’un contexte de vie socioéconomique défavorable sur la croissance intra-utérine, spécifiquement chez les femmes résidant en milieu rural. Nous avons également observé une augmentation du risque de prématurité associée à des niveaux > 16.4 μg.m-3 de NO2 dans l’air, à l’inverse, uniquement chez les femmes résidant dans des zones urbaines. Les associations entre l’exposition à la pollution atmosphérique et les marqueurs de la croissance intra-utérine, bien que sexe-spécifiques, ne semblent en revanche pas varier sensiblement suivant le gradient urbain-rural. Ce travail confirme la nécessité d’évaluer l’influence des inégalités sociales et environnementales sur le développement intra-utérin et de considérer l’importance et le rôle du contexte de vie, notamment urbain-rural, dans la formation de ces inégalités. / Pregnancy is a sensitive and critical period for the development of the child and the health of adults-to-be. The biological and physiological adaptation of the body dealing physical and psychosocial stressors during this period may exert its effects in adulthood (and possibly over several generations). This delayed toxicity presupposes intrinsically the need to study the effects of exposure to environmental risk factors during fetal life using a holistic approach involving risk factors at both the micro (individual characteristics) and the macro level (physical and psycho-social context). Given the heterogeneity of the Breton territory in which this work was conducted, we explored whether the impact of social inequalities and the physical environment (air pollution) on birth outcomes (fetal growth and the risk of prematurity) could be modified according to an urban or rural place of residence. This work was based on data collected as part of the Breton mother-child cohort PELAGIE, which had included 3421 pregnant women between 2002- 2006. The anthropometric parameters and gestational age at birth were measured by medical personnel at delivery. We defined urban and rural areas according to the definition of “urban units” from the National Census Bureau (INSEE). The annual concentrations of air pollution (nitrogen dioxide (NO2)) were estimated using a land-use regression modeled at a 100 m scale and developed as part of an European project. Finally, neighbourhood deprivation was estimated using a composite index developed at census blocks level and whose use was legitimated over both urban and rural areas. Neighbourhood deprivation was associated with an increased risk of infants with fetal growth restriction, only for women living in rural areas. We also observed an increased risk of preterm birth associated with NO2 concentrations > 16.4 μg.m-3, only among women residing in urban areas. The associations between air pollution and fetal growth, although sex-specific, did not seem on the other hand, to vary significantly according to the urban-rural spectrum. This work confirms the need to explore the influence of both social and environmental inequalities on intrauterine development, and to assess the role of place-based factors, such as the urban-rural context, in shaping these inequalities.