Adult mortality and its impact on children in two informal settlements in Nairobi, Kenya

This thesis examines the impact of adult deaths on children in two slums in Nairobi city. Over the last two decades, there has been a marked increase in adult mortality in Sub-Saharan Africa. Data on adult mortality in the region are scanty and this makes assessment of its impact on child well-being hard. The thesis analyses data from a longitudinal demographic surveillance system that monitors births, migration, and deaths and identifies causes of death using verbal autopsy. Other data collected include: household characteristics, schooling and health care utilisation. It investigates: i) levels, trends and causes of adult deaths; ii) the impact of adult deaths on children’s household circumstances, and iii) the impact of adult deaths on children’s health and social outcomes. Measures of adult mortality were estimated using life-table and survival analysis techniques. Regression techniques were used to assess impact of adult death on children’s migration, living arrangements, survival, immunisation and schooling. Life expectancy in the two slum populations was low. Adult mortality was higher in women than men. Ethnicity, gender, wealth status were associated with the risk of adult death. Overall, HIV/AIDS was the leading cause of adult death, followed by injuries and tuberculosis. The risk of death from HIV/AIDS was highest in Korogocho slum and the Luo ethnic group. Child mobility in the slums was high. After death of a mother, the risk of child out-migration increased. Death of a father increased average household size while death of a mother resulted in a reduction in household size. Households that experienced adult deaths were more likely to be headed by an older person. Death of a mother, especially from HIV/AIDS, but not that of a father, increased the risk of child death. The risk was highest in the 6 months before and after maternal death. The effect of adult deaths on children’s education depended on slum of residence. While Viwandani children had better educational outcomes overall, death of a mother in Viwandani resulted into poorer schooling outcomes. Interventions aimed at the leading causes of adult deaths need to be scaled up. The results here confirm that adult deaths negatively impact child well-being in this urban setting. Child survival can benefit from scaling up existing interventions, while mitigation of social impacts may require a mix of family and institution-based support for orphaned and vulnerable children.