Food insecurity and health outcomes in Mozambique: in what way does the socioeconomic position of the household head matter

Abstract

Background and objectives: Food insecurity (FI) is a global concern and is one of the major causes of malnutrition among vulnerable populations in low- and middle-income countries (LMICs). Food insecurity has been linked to a range of negative health outcomes, of which non-communicable diseases (NDCs) have become the world’s leading causes of mortality and morbidity and are significant contributors to the global burden of disease. In addition, there is a growing recognition of mental health as an intrinsic component of general health and well-being. For a variety of reasons, however, many people in LMICs still suffer from mental health conditions in silence. Food security (FS) is paramount to maintaining not only physical but also mental health. In the context of multiple risk factors, there is an increasing need to understand the interplay between FI and NCDs, and mental health outcomes, especially among sensitive populations. The burden of FI in southern Africa including Mozambique and how several factors impact FI, is not well known, although FI remains an important public health concern in the region. There is shortage of scientific data on the relationship between socioeconomic position (SEP), FI and health outcomes in southern Africa and specifically Mozambique. Furthermore, especially when formulating and implementing policies and health programmes aimed to alleviate FI and promote better health outcomes, it is crucial to understand the specific circumstances that force food-insecure households to resort to different coping strategies. The overall objective of this thesis was to assess the impact of SEP on FI and physical and psychological health outcomes among adults in Maputo City, southern Mozambique. Specifically, the thesis objectives were to systematically review empirical evidence on the relationship between FI and health outcomes among adults in southern Africa (including Mozambique) (Study I); to estimate the prevalence of FI and assess its associated factors among households in Maputo City (Study II); to examine the association between SEP, FI and hypertension and type 2 diabetes (Study III); to examine the association between SEP, FI and anxiety and depression (Study IV), and to understand the perceptions and coping strategies used by household heads in situations of FI (Study V). Methods: Study I was based on 14 peer-reviewed journal articles that met the inclusion criteria. The literature search was conducted and reported using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Most studies assessed FI using either contextually adapted versions of the US Department of Agriculture (USDA) Household Food Security Survey Module (HFSSM) or the Household Food Insecurity Access Scale (HFIAS). Physical health outcomes (e.g. hypertension, diabetes and HIV) were assessed based on self-reports of actual diagnoses performed at hospitals, health centres or medical clinics. Mental health outcomes (e.g. anxiety and depression) were measured using various self- reporting scales with some focus on the Self-Reporting Questionnaire (SRQ). By contrast, Study II, III and IV used cross-sectional data from a sample of 1,842 household heads in Maputo City. In Study II, FI was assessed using a contextually adapted version of the USDA HFSSM, and the relationship between FI and socioeconomic and demographic factors was explored through multiple regressions. In Study III, the assessment of hypertension and diabetes relied on self-reports, by heads of households, of the actual diagnoses performed at hospitals, health centres or medical clinics. For study purposes, 1,820 self-reports were included in the data analysis. Multinomial logistic regression was used to analyse the association between FI, SEP, hypertension, and diabetes, and interaction terms were used to assess the effects of SEP on this association. In Study IV, the Hospital Anxiety and Depression Scale (HADS) was used to measure anxiety and depression. A composite variable for psychological health was created. Propensity score matching and interaction effect analyses were employed to examine the effects of FI on psychological health, as well as the moderating role of SEP. In Study V, a qualitative descriptive design was employed, and based on data saturation criteria, a total of 16 in- depth interviews with heads of households experiencing FI were conducted, audio-recorded, and transcribed verbatim. Accordingly, a qualitative content analysis was performed using an inductive approach. Results: In Study I, a broad range of prevalence and severity of FI was registered (18–91%), depending on the sociodemographic characteristics of the studied population and the measurement instruments. Food insecurity was frequently associated with hypertension, diabetes, increased risk of HIV acquisition, anxiety and depression. In Study II, 79% of the households were in a situation of FI; of these, about 21% had mild FI, 35.5% moderate and 43.5% severe FI. The study showed that low income, low education, low food diversity, and reduced number of meals per day were consistently and significantly associated with increased odds of FI. In Study III, the findings revealed a significant association between FI, SEP (especially education and income), hypertension, and type 2 diabetes. Furthermore, the interaction analyses highlighted the influence of SEP on the relationship between FI and hypertension, and consistently showed a nuanced influence on type 2 diabetes. Specifically, food- insecure individuals with a higher SEP were more likely to develop diabetes than their counterparts with a lower SEP. In Study IV, of the 1,174 participants randomly assigned for propensity score matching, 787 were exposed to FI while 387 were unexposed. The analysis revealed stark disparities in psychological health outcomes associated with FI. The risk of poor psychological health among those exposed to FI was 25.79%, which was significantly higher than the 0.26% in the unexposed group, with a risk ratio of 99.82. The attributable fractions revealed that nearly all the risk for poor psychological health in the exposed group could be ascribed to FI, particularly moderate and severe FI. The interaction effects analysis revealed that SEP greatly modifies this relationship. Specifically, the heads of food-insecure households with a lower SEP tended to report less favourable mental health compared with their food-secure counterparts with a higher SEP. Finally, in Study V, the results were summarized into five themes: experiences and perceptions of FI; coping strategies applied in situations of FI; food choices; effects of climate change on FS; and effects of FI on perceived health. A broad spectrum of experiences and coping strategies were described, starting from cooking any food available, skipping meals, receiving remittance from relatives and friends, consuming unsafe foods, and cooking least favourite foods, to having a repetitive and less-nutritious diet. The heads of households also reported emotional distress, anxiety and depression, substance use, and other adverse health outcomes as consequences of FI. Some had been diagnosed with hypertension, diabetes or HIV/AIDS. Conclusions: Food insecurity is a great concern in southern Africa and is associated with various negative health outcomes. The studies point to the need for future research on the relationship between FI and health outcomes, to help standardize measures of FI and psychological health, and to inform government policies and interventions aimed to alleviate FI and promote better health outcomes in the region. More than three-quarters of households in Maputo City suffer from FI, and several factors (e.g. SEP, household size and structure) appear to play a significant role, emphasizing the need for decent work and employment creation, as well as the need for women’s empowerment in the country. Furthermore, the heads of households applied various coping strategies to acquire and manage food, some of which are considered risky to health. This highlights their extreme hardships and vulnerability. Lastly, FI was found to have a positive and significant association with anxiety and depression, and hypertension and type 2 diabetes. In addition, the SEP of household heads appears to modify these associations. Therefore, addressing household FI and improving the SEP of the most sensitive groups may be crucial measures in reducing the risk factors associated with NCDs and poor mental health in the country. These findings highlight the significance of an all- inclusive approach to health promotion and disease prevention. Future longitudinal studies are needed to gain deeper insight into the pathways linking socioeconomic and demographic factors to household FI and negative health outcomes, and to establish causal inferences.