Spectrum of Disease Burden in Urban Informal Settlements of Brazil
Robert Eugene Snyder
Doctor of Philosophy in Epidemiology
University of California, Berkeley
Professor Lee W. Riley, Chair
There are more than one billion urban informal slum residents worldwide, comprising almost one-third of the global urban population. Slums are characterized by their abundance of risk factors for communicable and non-communicable diseases. Some of these include poor water and sanitation infrastructure, in addition to poor structural housing quality, overcrowding and insecure residential status (i.e. lack of land tenure). Residents of these communities shoulder a greater burden of biologic, spatial and social health determinants of disease and their corresponding disease outcomes than residents of the same city who do not live in slums. Due to systematic disenfranchisement and the complicated logistics involved in studying these populations, our knowledge about their health conditions and determinants of disease occurrence is sparse. This dissertation describes three approaches to begin to address this gap in knowledge.
Chapter 1 sets the tone for this dissertation by describing the concept of slums and lays out the evidence for the disproportionate burden of disease borne by slum residents. While not being representative of all slums worldwide - Brazil is now considered an upper-middle-income country by the World Bank– the country’s slums serve as an important case study. Brazil has a relatively robust healthcare system, a functional democracy, and an active research sector, all of which facilitate the collection and comparison of data inside and outside of the country’s slums. Estimates of Brazil’s slum population range from 11,425,644 (6% of the total population) in the 2010 Brazilian Census (the official government estimate) to 45.7 million (28%) estimated by the United Nations Human Settlements Program. In 2010, the Brazilian government carried out a census with the goal of systematically enumerating and describing the physical and demographic characteristics of the country’s slums, coining the technical term aglomerados subnormais (AGSN) to describe slums.
Perhaps the most recognizable of all slums in Brazil are the favelas of Rio de Janeiro; these architecturally colorful and culturally vibrant communities accentuate the city’s magnificent physical geography and beauty. Unfortunately, due to their precarious physical location, the quasi-legal land tenure of slum residents, the blatant political corruption, and the persistently poor implementation of the national healthcare system (in 2010 approximately 50% of Rio de Janeiro’s population had access to the Sistema Único de Saúde – SUS), these communities’ residents suffer from stark health inequalities. Chapter 2 uses the 2010 Census to describe the spatial distribution of social determinants of health in Rio de Janeiro. The findings from this analysis highlight differences in age, income, and access to electricity, sanitation, water, and solid waste disposal throughout the city, and point to the possibility that differences in these characteristics contribute to an inequitable distribution of disease between the city’s slum and non-slum areas.
Chapter 3 delves more deeply into these inequalities. Given our group’s previously published evidence suggesting that the burden of tuberculosis is greater among residents of Rio de Janeiro’s slums than among residents of non-slum areas, we seek to fulfill a principle public health axiom: that we must seek and apply solutions to improve the health of populations. Chapter 3 evaluates the effectiveness of the directly observed treatment (DOT) program for tuberculosis treatment outcomes among TB cases inside and outside of Rio de Janeiro’s slums. Using the Brazilian Notifiable Disease System, Sistema de Informação de Agravos de Notificação (SINAN), we compare treatment outcomes between TB cases on DOT in AGSN and non-AGSN census tracts. While we found that DOT coverage was low inside and outside of AGSNs, we report that DOT had a greater impact on the cure rate for TB cases in AGSN areas compared to TB cases in non-AGSN areas.
Chapter 4 shifts the lens to non-communicable diseases (NCD) and to a different city, Salvador, where thirty-three percent of the population lives in slums. We compare the burden of several NCD (i.e. diabetes mellitus, hypertension, dyslipidemia) and the prevalence of risk factors (i.e. overweight, obesity and smoking) in a convenience sample of residents of Pau da Lima, an urban slum, with residents of the entire city as captured in a telephone survey. This telephone survey is an annual landline-based survey undertaken by the Brazilian Ministry of Health with the express purpose of monitoring the prevalences of NCD and NCD risk factors in Brazil’s capital cities. We age- and sex- standardized the prevalences of these conditions and risk factors and compared them between these two populations. We found that the age- and sex-adjusted prevalences of diabetes mellitus, smoking, being overweight/obese, and being obese, were higher among residents of Pau da Lima than in residents of the entire city of Salvador. A striking finding was that women living in slums suffered a disproportionate burden of being overweight or obese, and men in slums smoked at almost twice the frequency as men in the city as a whole.
While these observational studies and analyses do not provide causal evidence for a link between slum residence and adverse health outcomes, they provide preliminary data in support of the hypothesis that residents of slums in several major Brazilian cities have unique social and spatial determinants of health, and subsequently distinct disease profiles, compared to residents of non-slum neighborhoods in the same city.
Because of poor access to healthcare and a host of other slum-specific obstacles, slum residents rarely appear in official disease or mortality estimates. Consequently, slum-specific analyses such as those presented in this dissertation that describe the burden of disease among slum residents can be used to design policies aimed at mitigating the inequitable distribution of disease in these communities, also serving as a baseline to evaluate the impact of these policies on the health of this population. The findings from this dissertation should be used to encourage further epidemiologic analyses of slum populations, not only in Brazil, but also among slum populations of megacities elsewhere in the world.