PhD-Thesis
Although the prevalence of HIV/AIDS and food insecurity were relatively high in Rufiji
District in the mid-2000s, the extent to which they were linked was empirically unknown.
Therefore, a research for this thesis was done with the ultimate objective to determine the
linkage between HIV/AIDS and food security at the household level. The main indicator
of HIV/AIDS was a household having lost an adult member due to AIDS from January
2003 to December 2005; the main indicator of food security was dietary energy
consumed (DEC) per adult equivalent per day. Data were collected among 225
households between November 2005 and October 2006 through Participatory Rural
Appraisal (PRA), Household Income and Expenditure Survey (HIES) and structured
interviews. Binary logistic regression was used for analysis and the dependent variable
was food security in terms of food insecure (0) and food secure (1) based on kilocalories
consumed per adult equivalent per day. The independent variables included having been
affected by HIV/AIDS in terms of not affected (0) and affected (1). The results reveal
that the odds for households affected by HIV/AIDS to be food secure were 0.705 times as
high as the odds for households not affected by HIV/AIDS to be food secure. This means
that households affected by HIV/AIDS were less likely to be food secure in comparison
with those not affected by HIV/AIDS. The B statistic for having been affected by
HIV/AIDS was negative (B —0.350) meaning that being affected by HIV/AIDS had
negative impact on food security. However, the Wald statistic that shows the magnitude
of impact was small (0.251) and not significant (p = 0.617) implying little impact of
HIV/AIDS on food security. Based on these findings, it is concluded that although being
affected by HIV/AIDS has negative impact on food security, it does not automatically
make households food insecure, especially in a short run, and that some non-HIV/AIDS
factors have bigger impact than that of HIV/AIDS on food security. On the basis of theiii
conclusion, it is recommended that efforts to improve food security among households
affected by HIV/AIDS should consider both HIV/AIDS and non-HIV/AIDS factors.