Evolution of household availability of regional foods in Amazonas

OBJECTIVE: To evaluate the evolution of household availability of regional foods in the state of Amazonas, their distribution according to sociodemographic characteristics, and potential differences when compared to the remaining areas of Brazil. METHODS: Data on food acquisition for home consumption from the 2002-2003, 2008-2009, and 2017-2018 Pesquisa de Orçamentos Familiares (POFs – Consumer Expenditure Surveys) were analyzed, covering, respectively, 48,470, 55,970, and 57,920 households in Brazil, of which 1,075, 1,344, and 1,833 are in Amazonas. Foods were categorized into three groups: cassava and its derivatives, freshwater fish, and regional fruits. The study analyzed the amount of regional food purchased, expressed in relative household caloric share, for the entire area of Amazonas. Additionally, the data was stratified and analyzed according to sociodemographic variables, with differences assessed through the overlapping of 95% confidence intervals. FINDINGS: The household caloric share of the total regional foods in Amazonas was 22.54% in 2002-2003, 18.18% in 2008-2009, and 6.49% in 2017-2018. Across Brazil, those percentages were much lower in the same period: 3.67%, 3.34%, and 1.82%, respectively. Changes in Amazonas were primarily attributed to the steep drop in the cassava and derivatives group, which decreased from 14.30% in 2002-2003 to 12.74% in 2008-2009 and further declined to 3.09% in 2017-2018. Additionally, there was a gradual decline in household availability of freshwater fish, decreasing from 7.30% in 2002-2003 to 4.85% in 2008-2009 and reaching 2.90% in 2017-2018. Households in rural areas and with lower per capita income presented a higher proportion of calories from total regional foods; this particular stratum also experienced the most significant reductions in their consumption. CONCLUSION: During the study period, there was a significant decrease in the consumption of regional foods in Amazonas, particularly in lower income households in rural areas. Among them, the family reference person was typically a younger male with a lower educational background.