ABSTRACT
OBJECTIVE: To determine the reliability, validity and correlates of measures of food insecurity (FI) obtained using an individually focused food insecurity access scale (IFIAS) among pregnant women of mixed HIV status in northern Uganda. DESIGN: A mixed-methods study involving cognitive interviews nested within a cross-sectional survey. SETTING: The antenatal care clinic of Gulu Regional Referral Hospital. SUBJECTS: Survey respondents included 403 pregnant women, recruited in a ratio of one HIV-infected to two HIV-uninfected respondents, twenty-six (nine of them HIV-infected) of whom were asked to participate in the cognitive interviews. RESULTS: Over 80% of cognitive interview participants reported understanding the respective meanings of six of the nine items (i.e. items 4 to 9) on the IFIAS. Two main factors emerged from rotated exploratory factor analysis of the IFIAS: mild to moderate FI (IFIAS items 1-6) and severe FI (items 7-9). Together, they explained 90·4% of the FI measure's variance. The full IFIAS and the two subscales had moderate to high internal consistency (Cronbach's α ranged from 0.75 to 0.87). Dose-response associations between IFIAS scores, and measures of socio-economic status and women's diet quality, were observed. Multivariate linear regression revealed significant positive associations between IFIAS scores and HIV infection, maternal age, number of children and a history of internal displacement. IFIAS scores were negatively associated with women's diet diversity score, asset index and being employed. CONCLUSIONS: The IFIAS showed strong reliability, validity and contextual relevance among women attending antenatal care in northern Uganda.
