HIV/AIDS Mortality and the Role of Woodland Resources in the Maintenance of Household Fod Security in a Rural District of South Africa

SUMMARY This study examined food security among HIV/AIDS-impacted households (compared to non-HIV/AIDS-impacted households) in rural South Africa, with a particular focus on the role of savanna woodland resources (e.g. wild foods) in shaping household resilience following the death of a prime-age adult. The study was conducted in the Agincourt health and demographic surveillance site in the rural northeast of South Africa. A cross-sectional survey was conducted in 290 rural households in May and June 2006. Households were stratified by their experience of an HIV-related death of a prime-age adult in the previous two years as follows: HIV death (n=109), quick non-HIV death (n=71) and no adult death (control) (n=110). Experience of a mortality, as well as household socio-demographic data, were provided by the Agincourt Health and Demographic Surveillance System which is run by the University of the Witwatersrand/Medical Research Council's Rural Public Health and Health Transitions Research Unit (Agincourt Unit). A survey questionnaire was used to quantify household food security, livelihoods, use of woodland resources, and impacts of the experience of an adult death on the household. Food security was assessed in terms dietary diversity, experience of hunger, short-term coping strategies, and longer term adaptive strategies. Survey interviews were conducted in the local language by experienced local fieldworkers from the Agincourt Unit. Detailed qualitative interviews were also conducted by the researchers, with assistance from local interpreters, in 17 mortality-impacted households. Satellite imagery was used to quantify woodland cover around each of the study villages. HIV-impacted households tended to be poorer than those impacted by a non-HIV death, although their socioeconomic status, indexed by wealth ranking, did not decline during the two years subsequent to the adult HIV death. Dietary diversity was significantly lower in HIV-impacted households than in control households with no death. However, this pattern largely disappeared when socioeconomic status was factored into the analysis. Mortality-impacted households were more likely than households in the control group to have experienced hunger in the last 30 days. However, this was more pronounced in households which had experienced a quick non-HIV adult death. Similarly, although both categories of mortality-impacted households were more likely to have engaged in coping strategies in response to food shortages in the last seven days, this was more prevalent in 'non-HIV death' households. Both experience of hunger and engaging in coping strategies were also positively related to poverty. The deceased household member among mortality-impacted households had more often …