Menu
Menu
Menu
Menu
Background: Food insecurity (FI) is the lack of physical, social, and economic access to sufficient food for dietary needs and food preferences. We examined the association between FI and women’s uptake of services to prevent mother-to-child HIV transmission (MTCT) in Zimbabwe. Methods: We analyzed cross-sectional data collected in 2012 from women living in five of ten provinces. Eligible women were ≥16 years old, biological mothers of infants born 9-18 months before the interview, and were randomly selected using multi-stage cluster sampling. Women and infants were tested for HIV and interviewed about health service utilization during pregnancy, delivery, and post-partum. We assessed FI in the past four weeks using a subset of questions from the Household Food Insecurity Access Scale and classified women as living in food secure, moderately food insecure, or severely food insecure households. Results: The weighted population included 8,790 women. Completion of all key steps in the PMTCT cascade was reported by 49%, 45%, and 38% of women in food secure, moderately food insecure, and severely food insecure households, respectively (adjusted prevalence ratio (PRa)=0.95, 95% confidence interval (CI): 0.90, 1.00 (moderate FI vs. food secure), PRa=0.86, 95% CI: 0.79, 0.94 (severe FI vs. food secure)). Food insecurity was not associated with maternal or infant receipt of ART/ARV prophylaxis. However, in the unadjusted analysis, among HIV-exposed infants, 13.3% of those born to women who reported severe household food insecurity were HIV-infected compared to 8.2% of infants whose mothers reported food secure households (PR=1.62, 95% CI: 1.04, 2.52). After adjustment for covariates, this association was attenuated (PRa=1.42, 95% CI: 0.89, 2.26). There was no association between moderate food insecurity and MTCT in unadjusted or adjusted analyses (PRa=0.68, 95% CI: 0.43, 1.08). Conclusions: Among women with a recent birth, food insecurity is inversely associated with service utilization in the PMTCT cascade and severe household food insecurity may be positively associated with MTCT. These preliminary findings support the assessment of FI in antenatal care and integrated food and nutrition programs for pregnant women to improve maternal and child health.
We analyzed data from a 2012 cross-sectional survey of mother/caregiver-infant pairs conducted as part of the impact evaluation of Zimbabwe’s Accelerated National PMTCT Program [5,37]. The survey targeted women who were ≥16 years old and biological mothers or caregivers of infants (alive or deceased) born 9–18 months earlier in order to capture MTCT during pregnancy, delivery and breastfeeding. The primary outcomes of the impact evaluation were MTCT and HIV-free infant survival. Because the analyses presented in this paper use data from the 2012 survey, which were the baseline data for the parent impact evaluation, we describe the association between food security and engagement in PMTCT services before the Ministry of Health and Child Care’s (MoHCC) implementation of PMTCT strategy ‘Option A’. We restricted the sample of 9,018 mothers and caregivers to 8,662 biological mothers and their eligible infants by excluding 356 (3.9%) caregiver/infant pairs. The two-stage sampling strategy has been previously described [37,38]. Five provinces (Harare, Mashonaland West, Mashonaland Central, Manicaland, and Matabeleland South) were purposefully selected to include Zimbabwe’s capital, rural communities with higher and lower HIV prevalence, and both Shona and Ndebele ethnic groups. In the first stage, we randomly selected 157 catchment areas from 699 health facilities offering PMTCT services, proportionate to the number of facilities per district. In the second stage, in each catchment area, a pre-determined proportion of eligible infants was randomly sampled, depending on the size of the catchment area. Potentially eligible infants and their mothers/caregivers were identified by pooling information from: 1) community health workers, 2) immunization registers from both sampled and nearby health facilities, and 3) peer referral. Together, this approach efficiently identified eligible participants without screening all households and captured mother-infant pairs who did not utilize any health services and those who accessed care outside of their area of residence. Mothers providing written informed consent completed an anonymous interviewer-administered survey about maternal and household demographics, health services accessed during pregnancy and after delivery, and behaviors germane to MTCT (e.g., breastfeeding). Household food security was determined with a subset of questions from the Household Food Insecurity Access Scale (HFIAS) [30]. Due to interview time constraints, we selected three questions for inclusion, one from each domain of food access of the HFIAS: 1) anxiety and uncertainty about household food supply; 2) insufficient quality, including food variety and preferences; and 3) insufficient food intake and its physical consequences [30,39]. Women were asked how often, in the last 4 weeks, they worried that their household would not have enough food (anxiety/uncertainty), how often they were not able to eat preferred foods because of lack of resources (insufficient quality), and whether anyone in the household went to bed hungry (insufficient intake). Based on the distribution of these responses, consideration of the recommendations for categorizing responses to the full HFIAS, and examination of other food security scales [40], we determined an algorithm to classify households into three mutually exclusive groups: food secure, moderately food insecure, and severely food insecure. Severe food insecurity was defined as ≥1 household member going to bed hungry (even if infrequently or rarely) or “often” worrying (more than 10 times in the last month) about food access or food quality. Households were classified as having moderate food insecurity if they “sometimes” (3–10 times in the last month) worried about food access or quality. Food secure households experienced either none of the food insecurity conditions or they only rarely worried about food access or quality. We assumed that household food security status in the previous 4 weeks was strongly correlated with what food security status would have been during pregnancy, 9–18 months prior. We excluded seven women without food security information from the analysis. Living mothers and infants provided blood spot samples for HIV testing, which were air-dried onto filter papers and stored at room temperature until biweekly transport to the laboratory. Maternal samples were tested for HIV-1 antibody using AniLabsytems EIA kit (AniLabsystems Ltd, OyToilette 3, FIN-01720, Vantaa, Finland) with positive specimens confirmed using Enzygnost Anti-HIV 1/2 Plus ELISA (Dade Behring, Marburg, Germany) and discrepant results resolved by Western Blot. Samples from HIV-exposed infants and infants of mothers with unavailable samples were tested for HIV with DNA polymerase chain reaction (Roche Amplicor HIV-1 DNA Test, version 1.5). Results were available for 97.8% and 97.2% of women and HIV-exposed infants, respectively. Women were able to receive their HIV test results at the local health facility up to 3 months after the survey using a card with a barcode of their unique identification number. We first compared socio-demographic characteristics and service utilization stratified by food security status. We examined the following maternal health services: ANC (any and the WHO-recommended ≥4 visits [41]), gestational age in weeks at ANC registration (WHO recommends the first visit should occur in first trimester [41]), HIV testing during ANC or labor and delivery (or prior knowledge of HIV-positive serostatus), facility-based delivery, and postnatal visit attendance (6–8 weeks postpartum). Among HIV-infected women, we examined reported use of maternal and infant ART/ARV prophylaxis, infant co-trimoxazole prophylaxis, exclusive breastfeeding (≥1 month), and MTCT, stratified by food security status. We also examined a combined category indicating “completion” of the cascade including the following key services: ≥4 ANC visits, HIV testing, facility-based delivery, postnatal visit attendance, and among HIV-infected women, report of maternal and infant ART or ARV prophylaxis and co-trimoxazole prophylaxis. Missing values of PMTCT services were <1%; in those few cases, women were classified as not having received the service. We conducted an exploratory analysis to describe the association between food security and completion of the PMTCT cascade and MTCT using Poisson regression models. With cross-sectional data, the exponentiated parameter estimates represent prevalence ratios (PR) [42-44]. The fully adjusted models contain all covariates specified a priori for inclusion (see below) and key services or behaviors not hypothesized to lie on the causal pathway between food insecurity and the outcome. Covariates with variance inflation factors >10 (indicating multicollinearity) were examined for correlation with food security status and if necessary, excluded [45]. We present PRs and 95% confidence intervals (CI) computed with linearized standard errors to account for the sample design. Several covariates, which likely preceded pregnancy, were considered for inclusion in models as potential confounders: province, mother’s age, religion, tribe, being married or having a regular sexual partner, mother’s highest educational level, household size, lifetime births, and the building materials of the best building on the homestead. Additionally, we created a household asset index, divided into quartiles, using principal component analysis with a polychoric correlation matrix [46-48]. We also included a variable to indicate the infant’s age in months at the time of the survey (or age the infant would have been, if deceased), indicative of the time elapsed between the pregnancy and the interview to account for recall bias. No more than 1% of any covariate was missing. All analyses were conducted with STATA 12 (College Station, Texas) and were weighted to account for the varying sampling fraction by catchment area and 1.1% survey non-response. The Medical Research Council of Zimbabwe and the ethical review boards at the University of California, Berkeley and University College London approved this study.
N/A
