Professor Tufts University Boston, Massachusetts, United States
Disclosure(s):
Rachel D. Gilbert, MSc: No relevant financial relationship(s) with ineligible companies to disclose.
Objectives: We test the hypothesis that the relationship between diet quality and income is non-linear and non-monotonic. We hypothesize that diet quality improves as low-income households get wealthier, while the relationship levels off and then becomes negative at sufficiently high incomes when food choice is driven by other factors such as taste, convenience, and aspirations.
Methods: This analysis employs four rounds of the Malawi Integrated Household Panel Survey, implemented by the National Statistical Office and the World Bank in 2010, 2013, 2016, and 2019. We examine the quality of household diets across three dimensions: diet-related risk factors (Global Diet Quality Score), food groups (mean food group adequacy), and nutrients (mean adequacy ratio). We measure mean food group adequacy based on the proportionality of calories across food groups as recommended in food-based dietary guidelines. We combine nonparametric, semiparametric, and parametric regression methods to test the nature and form of the relationship between income and diet quality, as well as the consumption of recommended and unhealthy food groups.
Results: We find evidence that higher incomes lead to improved diet quality amongst some of the world's lowest-income households in Malawi. We find suggestive evidence of a nonmonotonic relationship between diet quality and income, whereby higher incomes are associated with lower diet quality in terms of risk factors for disease and food group adequacy. This threshold occurs at levels well below the international poverty line. The displacement of healthy food groups with sugar and sweets drives declining diet quality. We also see declines in calorie intake from vegetables and legumes, nuts, and seeds among wealthier households. Income gains in households with high diet quality are associated with larger diet quality increases.
Conclusions: Our results are consistent with the finding that diets improve as incomes increase, but large increases in household incomes would be needed to achieve marginal improvements in diet quality for most households in Malawi. Approaches that target low-income households or specific food groups may be needed for significant diet quality improvements.
Funding Sources: This study is part of the first author’s dissertation, conducted with funding from the Bill & Melinda Gates Foundation jointly with the UK government.