O etanol de cana-de-açúcar se consolida como combustível renovável, o que promove nova expansão da cultura da cana-de-açúcar no Brasil e, principalmente, no Estado de São Paulo. Como a queima prévia controlada da palha de cana-de-açúcar ainda é considerada uma prática agrícola necessária para a viabilização econômica da colheita, em mais de 70% dos municípios do Estado de São Paulo a população é obrigada a conviver com essa poluição. Para estudar a distribuição desse fator de risco e sua relação com a saúde, realizei um estudo epidemiológico ecológico nos 645 municípios de São Paulo. Usei um modelo Bayesiano de regressão multivariada relacionando os efeitos na saúde com a exposição à queima prévia da palha de cana-de-açúcar, sendo controlados os efeitos das variações socioeconômicas (saneamento, educação e renda) e climáticas (temperatura máxima, umidade mínima e precipitação), através da inserção das mesmas no modelo. O efeito sobre a saúde foi medido por meio da Razão de Mortalidade e Morbidade Padronizada (RMP) dos desfechos: óbitos por doenças respiratórias nas faixas etárias acima de 65 anos e internações por doença respiratória, nas faixas etárias menores de 5 anos e acima de 65 anos de cada um dos municípios. Usei como medida de exposição à queima prévia dados obtidos no INPE: percentual da área de cana colhida com queima (PMQ), níveis médios de Aerossol e Focos de queima, testadas separadamente. Para resolver a autocorrelação entre os dados, estes foram considerados conforme sua disposição espacial, através da construção de uma matriz de vizinhança dos 645 municípios do Estado. Utilizei o método de simulação de Monte Carlo via Cadeias de Markov (MCMC) para \'suavizar\' as estimativas da RMP. A análise demonstrou que existe associação entre a queima prévia da palha de cana-de-açúcar e a ocorrência de doenças respiratórias, porque o aumento nos focos de queima (Focos) esteve associado significativamente com o aumento das internações por doenças respiratórias, na faixa etária de menores de cinco anos. Os resultados mostraram que a queima prévia da palha da cana-de-açúcar oferece efetivamente risco à saúde da população e, adicionados aos mapas coropléticos gerados, oferecem subsídios para a vigilância epidemiológica e contribuem para o estabelecimento de políticas públicas para controle da poluição do ar, que contemplem além dos grandes centros urbanos, os pequenos municípios. A eliminação desse fator de risco deve fazer parte das medidas primordiais de prevenção à saúde a serem adotadas no Estado / Ethanol from sugarcane is consolidated as a renewable fuel which promotes further expansion of the culture of sugarcane in Brazil and especially in the State of São Paulo. As the controlled pre-harvest burning of sugarcane straw is still considered an agricultural practice necessary for the economic viability of this crop in more than 70% of municipalities in the State of São Paulo the population is forced to live with this pollution. To study the distribution of this risk factor and its relationship with the health of the population, I conducted an ecological study in the 645 municipalities of São Paulo. I used a Bayesian multivariate regression model relating the health effects and the exposure to previous straw burning of sugarcane, controlling the effects of socioeconomic factors (sanitation, education and income) and climate (maximum temperature, minimum humidity and precipitation) by the insertion of these variables in the model. The effect on health was measured by Standardized Mortality and Morbidity Ratio (SMR) of the outcomes: deaths from respiratory diseases in the age group above 65 years old and admissions for respiratory disease in children less than 5 years old and above 65 years old of each of the municipalities. I used as a measure of exposure to the pre-harvest burning data obtained at INPE: percent of sugarcane area harvest with burning (PMQ), levels of Aerosol and Spotlights of burning, tested separately. To solve the autocorrelation in the data these were considered as their spatial arrangement, by building a neighborhood matrix of the 645 municipalities in the state. I used the Markov Chain-Monte Carlo simulation method (MCMC) to \'soften\' the estimates of the SMR. The analysis showed that there is an association between previous straw burning of sugarcane and respiratory diseases, because the increase in outbreaks of burning (Spotlights) was significantly associated with increased hospital admissions for respiratory diseases in children aged under five years old. The results show that the previous straw burning of sugarcane effectively offers health risk to the population and added to the choropleth maps generated provide valuable information for epidemiological surveillance and contribute to the establishment of public policies for the control of air pollution, which should contemplate beyond the major urban centers, the small towns. The elimination of this risk factor should be part of a primordial prevention measure to be taken in the state

Text in English with abstracts and keywords in English and Setswana / Aim: The purpose of the research was to determine the magnitude and factors associated with gestational diabetes mellitus (GDM), and to explore the experiences of gynaecologists/obstetricians and midwives in the monitoring and prevention of GDM- related adverse maternal outcomes in order to propose best practice guidelines which may be implemented to overcome the problem. Methods: A concurrent mixed methods design was used. Participants for the quantitative study were selected using systematic random sampling, with purposive sampling being used for the qualitative part of the study. A total of 2000 medical records were reviewed using a checklist, in addition to which 7 gynaecologists/obstetricians and 12 midwives were interviewed using an in-depth interview guide. Descriptive and inferential statistics were used for the quantitative part, while Colaizzi’s manual qualitative data analysis method wasused for the qualitative part of the study. Findings: The magnitude of GDM was found to be 2.2%. Age and family history of diabetes mellitus were found to be factors associated with GDM (at p < 0.001). Other factors such as obesity, previous GDM, previous history of fetal macrosomia and multiple gestations were identified by respondents as factors related with GDM. In addition, the study explored the experiences of health professionals (HPs) in the monitoring and prevention of adverse maternal outcomes related to GDM, with the results showing some differences in screening and diagnostic techniques. It was also shown that lifestyle modification (physical exercise, diet management) and medication were utilised for managing women with GDM. In this regard, all the HPs agreed that creating awareness is the best intervention for preventing GDM as well as its adverse maternal outcomes. Conclusions: The magnitude of GDM is increasing, and much needs to be done to draw attention to the burden that GDM places on the health of pregnant women and the public. Since GDM is not considered a public health problem, little is being done to monitor the condition and its adverse maternal outcomes. It is hoped that the best practice guidelines developed from this research study may assist in reducing the adverse maternal outcomes of GDM in Ethiopia / Maikaelelo: Lebaka la patlisiso e ne e le go tlhotlhomisa go nna teng le mabaka a a golaganeng le bolwetsi jwa sukiri jwa baimana (GDM), le go tlhotlhomisa maitemogelo a dingaka tsa malwetsi a basadi (gynaeologists/ obstetricians) le babelegisi mo go tlhokomeleng le go thibeleng ditlamorago tse di maswe mo baimaneng tse di amanang le GDM gore go tshitshinngwe dintlhakaelo tse di gaisang tse di ka diragadiwang go fenya bothata. Mekgwa: Go dirisitswe thadiso ya mekgwa e e tlhakantsweng. Banni-le-seabe ba thutopatlisiso e e lebelelang dipalopalo ba ne ba tlhophiwa go diriswa go tlhopha sampole ka go se latele thulaganyo, mme go tlhopha sampole ka maikaelelo go ne ga diriswa mo karolong ya thutopatlisiso e e lebelelang mabaka. Go sekasekilwe palogotlhe ya direkoto tsa kalafi tse 2 000 go diriswa lenanetshekatsheko, mme mo godimo ga moo, go ne ga nna le dipotsolotso le dingaka tsa malwetsi a basadi di le supa le babelegisi ba le 12 go diriswa kaedi ya dipotsolotso tse di tseneletseng. Dipalopalo tse di tlhalosang le tse go sweditsweng ka tsona di ne tsa diriswa mo karolong ya dipalopalo ya thutopatlisiso, fa go dirisitswe mokgwa wa ga Colaizi wa tokololo ya data ya mabaka mo karolong e e lebelelang mabaka. Diphitlhelelo: Go nna teng ga GDM go ne ga fitlhelwa e le 2.2%. Dingwaga le hisetori ya bolwetsi jwa sukiri ya baimana mo lelapeng di fitlhetswe e le dintlha tse di golaganeng le GDM (ka p < 0.001). Dintlha dingwe, jaaka go nona phetelela, GDM mo nakong e e fetileng, go nna teng ga macrosomia ya masea mo nakong e e fetileng le boimana jwa masea a feta bongwe di ne tsa supiwa ke batsibogi jaaka dintlha tse di golaganeng le GDM. Go tlaleletsa foo, thuto e ne ya sekaseka maitemogelo a baporofešenale ba boitekanelo (HPs) mo tlhokomelong le thibelo ya ditlamorago tse di sa siamang mo baimaneng tse di golaganeng le GDM, mme dipholo di bontshitse dipharologano dingwe mo dithekeniking tsa go sekirina le go phekola. Go bonagetse gape gore phetolo ya mokgwa wa botshelo (katiso ya mmele, tsamaiso ya mokgwa wa go ja) le kalafi di ne tsa diriswa go laola bolwetsi jwa basadi ba ba nang le GDM. Mo lebakeng le, baporofešenale botlhe ba boitekanelo ba ne ba dumelana gore go dira temoso ke tsereganyo e e gaisang ya go thibela GDM ga mmogo le ditlamorago tsa yona tse di sa siamang mo baimaneng. Ditshwetso: Go nna teng ga GDM go a oketsega, mme go tshwanetse go dirwa go le gontsi go lemosa ka mokgweleo o bolwetse jono bo o bayang mo boitekanelong jwa baimana le setšhaba. Ka ntlha ya gore GDM ga e kaiwe jaaka bothata jwa boitekanelo jwa setšhaba, ga go dirwe go le kalo go tlhokomela bolwetsi le ditlamorago tsa jona tse di sa siamang mo baimaneng. Go solofelwa gore dintlhakaelo tsa tiriso e e gaisang tse di dirilweng mo thutopatlisisong eno di ka thusa go fokotsa ditlamorago tse di sa siamang tsa GDM mo baimaneng kwa Ethiopia. / Health Studies / D. Litt. et Phil. (Public Health)

Although the female menopause has been extensively characterized as a well-defined symptomatic state of oestrogen deficiency, which responds relatively well to oestrogen replacement therapy, the symptomatic state of androgen deficiency in men is poorly defined and uncertainty exists whether it responds to testosterone replacement. It has been proposed that hypothalamic-pituitary-testicular (HPT)-axis function (responsible for the production of androgens) and regulation could be viewed as a ‘barometer’ of health status in older men and that potential alterations in HPT-axis function and regulation reflect subclinical and clinical deficits in function and health, which may result in an aged phenotype of human health and disease in older men. The HPT-axis constitutes a well-defined, tractable, clinically-relevant, biological system, which may permit insight into the mechanisms underlying the expression of ageing-related phenotypes of human health and disease. By using a different lens – such as the genetic background; the compensatory responses within the HPT-axis; the syndromes of androgen deficiency; the ethnic background of an individual or the life course trajectory of function and health from conception into older age – to magnify potential dysregulation in the HPT-axis will it be possible to visualize and understand the phenotypic expression of human male ageing as a gradient of functional and health outcomes. This will allow for a better understanding of the physiological mechanics underlying symptomatic expression of dysregulation in the HPT-axis.

Problématique : L’amélioration de l’accessibilité financière aux soins de santé est essentielle pour réduire la morbidité et de la mortalité infantile dans les pays à ressources limitées. Cependant, les preuves disponibles sur la relation entre un accès accru aux soins et l’amélioration la santé infantile, dans le long terme, demeurent insuffisantes et parfois inconnues. Dans le contexte spécifique de la région du Sahel au Burkina Faso où les niveaux élevés de morbidité et de malnutrition coïncident avec un faible recours aux soins, une intervention de santé publique associant la gratuité des soins à l’amélioration de la qualité des soins et à la prise en charge de la malnutrition dans la communauté a été mise en œuvre en septembre 2008. Objectifs : En utilisant des approches statistiques et épidémiologiques appliquées aux données transversales et de séries chronologiques, cette thèse vise à apporter une meilleure compréhension de la façon dont la présence de l’intervention dans les communautés peut augmenter et maintenir l’utilisation des services de santé à long terme et améliorer la santé des enfants de moins de cinq ans. Les objectifs spécifiques sont : 1) évaluer le maintien à long terme des effets de l’intervention sur l’utilisation des services de santé chez les enfants de moins de cinq ans, 2) évaluer l’effet contextuel de l’intervention, quatre ans après le début de sa mise en œuvre, sur la probabilité de survenue d’une maladie et sur la probabilité d’utilisation des services de santé chez les enfants de moins de cinq ans, et 3) évaluer l’effet contextuel de l’intervention, quatre ans après le début de sa mise en œuvre, sur le retard de croissance chez les enfants de moins de cinq ans. Méthodes : Les données proviennent du système national d’information sanitaire, d’une enquête rétrospective sur les services de santé ainsi que d’une enquête de ménages réalisée quatre ans après le début de l’intervention dans 41 villages du district d’intervention et 51 villages du district de comparaison. Nous avons utilisé un plan quasi expérimental à séries temporelles interrompues avec groupe de comparaison pour évaluer les effets immédiats et à long terme de l’intervention sur les taux d’utilisation des services de santé. Ensuite, un plan d’étude transversale post-intervention avec un groupe de comparaison nous a permis d’évaluer l’effet contextuel de l’intervention sur la probabilité de survenue d’une maladie, sur la probabilité d’utilisation des services de santé et sur le retard de croissance chez les enfants de moins de cinq ans. La stratégie analytique a combiné la méthode de pondération par les scores de propension pour équilibrer les covariables entre les deux groupes, la modélisation binomiale négative à effets mixtes, les régressions linéaire et logistique multiniveaux. Résultats : L’intervention de gratuité des soins associée à l’amélioration de la qualité des soins et à la prise en charge de la malnutrition dans la communauté était associée à l’augmentation et au maintien de l’utilisation des services de santé au-delà de quatre ans (ratio des taux d’incidence = 2,33 ; IC 95 % = 1,98 – 2,67). En outre, comparativement aux enfants vivant dans le district de contrôle, la probabilité d’utiliser les services de santé était de 17,2 % plus élevée chez les enfants vivant dans le district d’intervention (IC 95 % = 15,01–26,6) ; et de 20,7 % plus élevée lorsque l’épisode de maladie était sévère (IC 95 % = 9,9–31,5). Ces associations étaient significatives, quels que soient la distance par rapport aux centres de santé et le statut socio-économique du ménage. Par ailleurs, alors que le contexte de résidence expliquait 9,36 % de la variance du retard de croissance (corrélation intraclasse = 9,36 % ; IC 95 % = 6,45–13,38), la présence de l’intervention dans les villages n’explique que 2 % de la variance du retard de croissance. Cependant, nous n’avons pas pu démontrer que la présence de l’intervention dans les communautés était associée à une réduction de la probabilité de survenue d’un épisode de maladie (Différentiel des probabilités = 4.4 ; IC 95% = -1.0 – 9.8), ni à une amélioration significative de l’état nutritionnel des enfants de moins de cinq ans (RC = 1,13 ; IC 95 % = 0,83–1,54). Conclusion : Cette thèse souligne que la gratuité des soins associée à l’amélioration de la qualité des soins et à la prise en charge de la malnutrition dans la communauté est efficace pour augmenter et maintenir l’utilisation des services de santé et réduire les inégalités géographiques de recours aux soins. Cependant, cette intervention n’était pas associée à une amélioration des résultats de santé infantile. Bien que des études longitudinales rigoureuses soient nécessaires pour comprendre pleinement l’influence potentielle de cette intervention sur la morbidité, cette thèse plaide pour la nécessité d’agir simultanément sur les autres déterminants sociaux de la santé et d’intégrer, de manière synergique, des interventions spécifiques à la nutrition pour plus d’impact sur la santé infantile. / Introduction: Improving financial access to health care is believed to be essential for reducing the burden of child morbidity and mortality in resource-limited settings, but the available evidence on the relationship between increased access and health remains scarce and the long-term issues are still unknown. In the specific context of the Sahel region in Burkina Faso where high levels of morbidity and malnutrition coincide with low health care use, a pilot intervention for free health care including quality of care improvement and management of malnutrition at the community level was implemented in September 2008. Objectives: Using statistical and epidemiological approaches applied to cross-sectional and time series data, this thesis aims to provide a better understanding of how the presence of intervention in communities can increase and maintain long-term use of health services and improve the health of children under five years. The specific objectives are: 1) to evaluate the long-term effects of the intervention on the use of health services in children under the age of five, 2) to estimate the contextual effect of intervention on the probability of occurrence of and the likelihood of health services being used by children under five, four years after the start of its implementation, and 3) to evaluate the contextual effect of the intervention on stunting in children under five, four years after the start of its implementation. Methods: The data for the analyses were provided from a variety of sources including the national health information system, a retrospective health services survey, and a household survey conducted four years after the intervention onset in 41 villages in the intervention district and 51 villages in the comparison district. We used a quasi-experimental controlled interrupted time-series design group to analyze the immediate and long-term effects of the intervention on the rate of health services utilization in children under five. Then, a quasi-experimental post-test-only design that included a control group allowed us to evaluate the contextual effect of the intervention on the probability of occurrence of a disease, on the probability of use of health services, and stunting in children under five. The analytic strategy combined the propensity score weighting method to balance the covariates between the two groups, two-level mixed-effects negative binomial, and linear and logistic regression models to account for the hierarchical structure of data. Results: The intervention for free health care including quality of care improvement and management of malnutrition at the community level was associated with an increased and maintained use of health services beyond four years after the onset of intervention (incidence rate ratio = 2.33; 95% CI = 1.98–2.67). In addition, compared to children living in the comparison district, the probability of using health services was 17.2% higher among those living in the intervention district (95% CI = 15.0–26.6); and 20.7% higher when the illness episode was severe (95% CI = 9.9–31.5). These associations were significant regardless of the distance to health centers and the socio-economic status of households. In addition, inequalities in the use of care were less pronounced in the intervention villages compared to those in the control village. Finally, the results also showed that the residence context accounted for 9.36% of the variance in stunting (intra-class correlation = 9.36% ; 95% CI = 6.45–13.38), and only 2% of the variance in stunting was explained by the intervention. However, we could not demonstrate that the intervention in these communities was associated with a reduced probability of an illness occurring (AME=4.4 (95% CI: -1.0 – 9.8), nor with a significant improvement in the nutritional status among children under five (OR = 1.13; 95% CI = 0.83–1.54). Conclusion: This thesis underlines the importance that affordable health care, including quality of care, as well as improving the management of malnutrition at the community level, are effective in increasing and maintaining the use of health services and reduce geographical inequalities in the use of care. However, this intervention was not associated with improved child health outcomes. Although rigorous longitudinal studies are necessary to fully understand the potential influence of this intervention on morbidity, this thesis highlights the need to simultaneously act on other social determinants of health and to synergistically integrate nutrition-specific interventions for greater impact on child health.