POPULATION HEALTH MEETS EVOLUTIONARY ANTHROPOLOGY: FAMILY STRUCTURE, RURAL LIVELIHOODS AND CHILD HEALTH IN TANZANIA

Over recent decades, many countries have witnessed substantial improvements in child wellbeing, with for example global child mortality reducing by nearly 50% in the last 20 years. Nevertheless, persistent and dramatic international inequalities remain, with children living in rural settings in Sub-Saharan Africa at particular risk of both ill health and low educational attainment. The proposed research seeks to improve our current understanding of the role of family structure as a determinant of child wellbeing in this context. Population health scientists are in wide agreement that 'traditional' African family structures, (specifically high fertility, early age at reproduction and high levels of polygynous marriage), contribute to poor child wellbeing, because they lead to apparent lower parental investments in human capital. However, frameworks for understanding the role of family structure as a determinant of child wellbeing, particularly in the context of the developing world, have also been criticised for their relative theoretical and methodological simplicity. The proposed fellowship will address these shortcomings. This will be achieved by grounding research in the fresh theoretical perspective of Evolutionary Anthropology; a framework notable for emphasizing: i) the role of contextual factors in determining relationships between family structure and child wellbeing; and ii) the potential for contextual variation in such relationships to feedback into cultural norms determining family formation. Furthermore, the project will utilise innovative 'multilevel' statistical methods to consider the role of family structure amid the broader individual, household and community-level determinants of wellbeing.

The proposed research will be use data from a new research collaboration initiated during the fellowship with the 'Whole Village Project', a monitoring and evaluation development project run by the NGO 'Savannas Forever Tanzania', which collects longitudinal demographic, economic and development indicator data from 56 villages in the north of the country. This remarkable source of data provides detailed information on household and family structure, socioeconomic profile, livelihood productivity and losses (through agriculture, pastoralism, or wage-labour), along with levels of food insecurity and nutrition. Child wellbeing is assessed by: i) child health, as measured by physical assessments of heights and weights of children under 5yrs, and ii) child education, as measured by attendance and school level-for-age for all children under 15yrs. Furthermore, the nature of data collection enables villages to be contrasted on the basis of community service provision, such as proximity to health clinics, schools and safe sources of water. To consider the generalisability of the study population, the Tanzanian 2010 Demographic and Health Survey will also be used as a source of comparative data.

To meet the fellowship's aims, planned analyses will consider how variation in (i) family size, (ii) age at reproduction and (iii) polygynous vs. monogamous marriage, predict observed inequalities in child health and education, and how these relationships interact with other identified household and community-level determinants of wellbeing. As such the project will generate new policy relevant knowledge about specifically when and where particular patterns of family structure place children at risk. It will also provide new data to consider how variation in rural livelihoods and community-level service provision may create shifting incentives for particular family formation behaviours, by changing their impact on child wellbeing. Research findings will be discussed in articles published in both population health and anthropology journals. They will also be summarized in open-access 'Summary Reports', with explicit consideration of their relevance to the design of public policy and development initiatives in the developing world.

OBJECTIVES: The central objective of the research is to improve current understandings of the impact of family structure on child wellbeing in the developing world. The project will be grounded in the theoretical perspective of Evolutionary Anthropology, emphasizing i) the role of contextual factors in determining relationships between family structure and child wellbeing; and ii) the potential for contextual variation in such relationships to feedback into cultural norms determining family formation.
DATA: Data will come from the NGO-run Whole Village Project, which collects demographic, economic and development data from 56 villages in northern Tanzania (3500+ households). Comparative analyses will also be carried out using the Tanzanian 2010 DHS. Child wellbeing will be assessed by health (heights & weights of children under 5yrs) and educational attainment (attendance & school level-for-age for children under 15yrs). Family structure is recorded in terms of i) family size; ii) age at reproduction; and iii) polygynous vs. monogamous marriage.
METHODS: Multilevel regression analyses will be used to determine the individual, household and community-level determinants of child wellbeing. Within this framework, particular focus will be given to quantifying interactions between family structure and i) household-level factors, such as wealth, livelihood type and food insecurity; and ii) community-level factors, focusing on service provision (proximity to health clinics, water source, schools).
SCIENTIFIC & MEDICAL OPPORTUNITIES: The project will generate policy relevant knowledge about specifically when and where patterns of family structure place children at risk. It will also consider how variation in rural livelihoods and service provision shifts incentives for family formation. By considering crosscutting issues at the intersection of population health and evolutionary anthropology, the project will forge links between these divergent fields of research.

INDIRECT IMPACT
The proposed research addresses key themes in population health science regarding child wellbeing with strong relevance to the design of public policy and development projects in the developing world. In this vein, it is projected that the research will make indirect impacts by advancing our current knowledge regarding which children, households and villages are most at risk of poor child wellbeing. This will be achieved, for example, by using rigorous statistical methods to compare levels of child wellbeing in households with different family structures, those practicing pastoralism vs. agriculture, and contrasting villages with and without key services such as health clinics or safe water sources. Such findings are of high relevance to any policy or development initiatives that seek to deliver resources with priority to those children most likely to suffer poor health or poor educational outcomes.

Planned open-access 'Summary Reports' to be disseminated by my NGO collaborators, will aim to make clear the implications of the research in this regard, identifying relevant results that apply specifically to the study population, but also with discussion of how they might apply more generally across Tanzania and similar populations in East Africa. The project is well situated to disseminate research findings to relevant policy-makers and public health professionals in Tanzania and beyond, as Savannas Forever Tanzania (SFTZ) prepare and disseminate District and Sector Reports on key issues based on their data (e.g. Food Insecurity, Maternal Health, Malaria and HIV/AIDS, Natural Resources and Climate Change) and also report the results of their surveys directly back to the communities included in data collection. Thus they have demonstrated a strong commitment to disseminating research findings in a way most likely to deliver real impact.

DIRECT IMPACT
In addition, by collaborating closely with the NGO SFTZ, carrying out analysis in-house in Tanzania and collaborating with local researchers, the project will directly support local capacity building for research at SFTZ. Ultimately, it is hoped that my research will lead to future refinements in the way that SFTZ collect data and boost their ability with regard to data analysis and research publication.