Understanding the malaria-poverty vicious circle 6. Project Summary/Abstract The long-term goals of this research are (a) to develop a conceptual framework for understanding the potential bi-directional malaria-poverty causality;(b) to empirically test the proposed conceptual framework;(c) to provide evidence regarding the magnitude of the dual relationship between malaria and socioeconomic status (SES);(d) to design new survey instruments that facilitate the measurement of mechanisms included in the proposed conceptual framework but not yet available in standard surveys;and (e) to evaluate the economic impacts of successful malaria control interventions. As a first step toward these long-term goals, the objectives of this R21 application are to develop analytical and empirical assessments of the bi-directional malaria- poverty causality. We will propose a conceptual framework based on literature review of factors that impact the links between malaria and poverty, and we will add a layer of complexity by proposing possible variations of these links at multiple spatial scales. The framework builds upon a systemic view of malaria;it will allow for study replication and facilitate comparison of findings. The empirical model will be based on a system of simultaneous equations and will utilize Demographic and Health Surveys (DHS) data for Tanzania, Rwanda, and Angola. We will estimate the model using two-stage least squares with instrumental variables. Malaria will be quantified by parasitaemia (as measured by microscopy), reducing the potential bias observed in studies that considered fever as a proxy for malaria. SES will be measured by a wealth index based on household ownership of durable assets and on housing conditions. Instruments for estimating the effect of malaria on SES include pregnancy status, elevation, and precipitation;instruments for estimating the effect of SES on malaria include gender and ethnicity;distance to major towns;and total rainfall during the agricultural season(s). We will conduct statistical tests of instrument validity using a Sargan's test for over-identifying restrictions. Depending on the suitability of the instruments, we may need to use other methods for causal analysis (e.g. propensity score matching). We choose Tanzania, Rwanda, and Angola because recent DHS surveys included malaria parasite testing by microscopy, reducing the potential bias observed in studies that consider fever as a proxy for malaria. The DHS are nationally representative surveys and therefore comprise diverse settings: urban and rural, low and high malaria risk, poor and relatively wealthy;they also contain geographical information that allows for the assessment of spatial variations, facilitating the targeting of malaria control interventions. Our analytical and empirical models will allow for the evaluation of economic impacts of on- going control interventions that significantly reduced malaria incidence, which is a research priority for the current goals of malaria elimination.
proposed study will provide a framework for increased scientific investigation of the bi-directional malaria-poverty causality, facilitating cross-country comparisons. It has the potential to allow malaria-endemic countries to better select and target interventions that both reduce the malaria burden and promote poverty alleviation. The research will also prepare the way to assess the economic impact of malaria control efforts in different countries, a research priority toward current malaria elimination goals.
| de Castro, Marcia Caldas; Fisher, Monica G (2012) Is malaria illness among young children a cause or a consequence of low socioeconomic status? evidence from the united Republic of Tanzania. Malar J 11:161 |
| Duncan, Dustin T; Castro, Marcia C; Gortmaker, Steven L et al. (2012) Racial differences in the built environment--body mass index relationship? A geospatial analysis of adolescents in urban neighborhoods. Int J Health Geogr 11:11 |
| Duncan, Dustin T; Castro, Marcia C; Blossom, Jeffrey C et al. (2011) Evaluation of the positional difference between two common geocoding methods. Geospat Health 5:265-73 |
