Call for emergency action to restore dietary diversity and protect global food systems in times of COVID-19 and beyond: Results from a cross-sectional study in 38 countries

Background: The COVID-19 pandemic has revealed the fragility of the global food system, sending shockwaves across countries' societies and economy. This has presented formidable challenges to sustaining a healthy and resilient lifestyle. The objective of this study is to examine the food consumption patterns and assess diet diversity indicators, primarily focusing on the food consumption score (FCS), among households in 38 countries both before and during the first wave of the COVID-19 pandemic. Methods: A cross-sectional study with 37 207 participants (mean age: 36.70 ± 14.79, with 77 % women) was conducted in 38 countries through an online survey administered between April and June 2020. The study utilized a pre-tested food frequency questionnaire to explore food consumption patterns both before and during the COVID-19 periods. Additionally, the study computed Food Consumption Score (FCS) as a proxy indicator for assessing the dietary diversity of households. Findings: This quantification of global, regional and national dietary diversity across 38 countries showed an increment in the consumption of all food groups but a drop in the intake of vegetables and in the dietary diversity. The household's food consumption scores indicating dietary diversity varied across regions. It decreased in the Middle East and North Africa (MENA) countries, including Lebanon (p < 0.001) and increased in the Gulf Cooperation Council countries including Bahrain (p = 0.003), Egypt (p < 0.001) and United Arab Emirates (p = 0.013). A decline in the household's dietary diversity was observed in Australia (p < 0.001), in South Africa including Uganda (p < 0.001), in Europe including Belgium (p < 0.001), Denmark (p = 0.002), Finland (p < 0.001) and Netherland (p = 0.027) and in South America including Ecuador (p < 0.001), Brazil (p < 0.001), Mexico (p < 0.0001) and Peru (p < 0.001). Middle and older ages [OR = 1.2; 95 % CI = [1.125–1.426] [OR = 2.5; 95 % CI = [1.951–3.064], being a woman [OR = 1.2; 95 % CI = [1.117–1.367], having a high education (p < 0.001), and showing amelioration in food-related behaviors [OR = 1.4; 95 % CI = [1.292–1.709] were all linked to having a higher dietary diversity. Conclusion: The minor to moderate changes in food consumption patterns observed across the 38 countries within relatively short time frames could become lasting, leading to a significant and prolonged reduction in dietary diversity, as demonstrated by our findings.