Predictors of household food insecurity in the United States during the COVID-19 pandemic

Objective: To examine associations between sociodemographic and mental health characteristics with household food insecurity as a result of the COVID-19 outbreak. Design: Cross-sectional online survey analyzed using univariable tests and a multivariable logistic regression model. Setting: The United States during the week of March 30, 2020. Participants: Convenience sample of 1,965 American adults using Amazon's Mechanical Turk (MTurk) platform. Participants reporting household food insecurity prior to the pandemic were excluded from analyses. Results: 1,517 participants reported household food security before the COVID-19 outbreak. Among this subset, 30% reported food insecurity after the COVID-19 outbreak, 53% were women and 72% were white. On multivariable analysis, race, income, relationship status, anxiety, and depression were significantly associated with incident household food insecurity. Black respondents, Hispanic/Latino respondents, and respondents with annual income less than $100,000 were significantly more likely to experience incident household food insecurity. Individuals experiencing incident household food insecurity were 2.09 (95% CI 1.58-2.83) times more likely to screen positively for anxiety and 1.88 (95% CI 1.37-2.52) times more likely to screen positively for depression. Conclusions: Food insecurity due to the COVID-19 pandemic is common, and certain populations are particularly vulnerable. There are strong associations between food insecurity and anxiety/depression. Public health interventions to increase the accessibility of healthful foods, especially for Black and Hispanic/Latino communities, are crucial to relieving the economic stress of this pandemic.


Introduction
Methods 73 We conducted a cross-sectional survey of American adults through Amazon's 74 Mechanical Turk (MTurk), an online labor market of over 225,000 U.S. workers who complete 75 online tasks and surveys.(13) Participants were invited to take part in an online survey, 76 administered using Qualtrics, that included questions regarding demographics, social distancing, 77 food security before and after the COVID-19 outbreak, and anxiety and depression (described in 78 detail below in Measures and in the Supplementary Material). Surveys were completed between 79 March 30, 2020 -April 2, 2020, after many states had imposed stay-at-home orders.(14) 80 Participants were compensated with $0.50 for completing the survey, which took 10-15 minutes 81 to complete.  Household Food Insecurity. We assessed household food insecurity using a validated 87 two-item screen. (15) Household food insecurity was defined as responses of "Sometimes true" or 88 "Often true" for both items. Participants reported answers to each of these questions for the 89 periods before and after the COVID-19 outbreak. analyses as they were unrelated to our primary aims.

99
Analysis 100 Univariable Analysis. The study population was restricted to respondents with household 101 food security before the COVID-19 outbreak. We compared baseline demographic 102 characteristics and anxiety and depression across respondents with and without incidence 103 household food insecurity. We used Chi-square tests of independence for categorical variables, 104 and Fisher's exact test when expected values were less than or equal to five. Respondents with 105 missing data for a particular analysis were excluded. Statistical significance was defined as p < 106 0.05.

107
Multivariable Analysis. We constructed a multivariable logistic regression model 108 assessing incident household food insecurity as our outcome variable. Gender and age were 109 included as covariables a priori. We included additional demographic characteristics if p < 0.20 110 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 12, 2020. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 12, 2020. . https://doi.org/10.1101/2020.06.10.20122275 doi: medRxiv preprint

114
A total of 1,965 participants across the United States completed our survey between 115 March 30th and April 2nd, 2020. Among the total survey respondents, 1,527 individuals (78%) 116 reported household food security before the outbreak. Subsequent analyses were based on this 117 subset of the population.

118
Within the population of food-secure participants pre-COVID-19, 44% identified as 119 male, 72% were between the ages of 25 and 55, 72% were white, 10% were Asian, 7% were 120 Black, and 7% were Hispanic. In addition, 48% reported an annual household income of over 121 $50,000 and 64% reported full-time employment at the time of the survey. Roughly 30% of 122 participants reported incident household food insecurity after the outbreak (Table 1). Figure 1  In this cross-sectional survey of American adults, we found that 30% of respondents 154 became food-insecure following the COVID-19 outbreak. By contrast, the USDA reported that 155 only 11% of total U.S. households were food-insecure in 2018. (18) Risk for incident household 156 food insecurity was significantly higher among Black and Hispanic participants, as well as 157 participants who were married or cohabitating, or reported household incomes below $100,000 at  Our study has limitations. Firstly, our cross-sectional approach precludes causality 190 inferences, and relies on retrospective reports of food insecurity prior to the COVID-19 outbreak.

191
Secondly, we collected data through Amazon's Mechanical Turk, potentially creating a response 192 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted June 12, 2020. . https://doi.org/10.1101/2020.06.10.20122275 doi: medRxiv preprint bias and limiting the generalizability of our findings to the larger US population. Finally, our 193 data were collected in late March, when the economic impacts of the outbreak were just 194 beginning to affect many nationwide. We hypothesize that this may partially explain the 195 unexpected result that unemployment was not a predictor of incident food insecurity. Although 196 the validated food insecurity screening questions that we used in our study specifically 197 referenced an ability to afford food (rather than an inability to buy food due to shortages), our 198 finding of increased risk of food insecurity among those with annual income between $50,000-199 $100,000 is also surprising. Because we did not specify a time frame for these earnings,  CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted June 12, 2020. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted June 12, 2020. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted June 12, 2020. . https://doi.org/10.1101/2020.06.10.20122275 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted June 12, 2020. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted June 12, 2020. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 12, 2020. . https://doi.org/10.1101/2020.06.10.20122275 doi: medRxiv preprint CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 12, 2020. . https://doi.org/10.1101/2020.06.10.20122275 doi: medRxiv preprint CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 12, 2020. . https://doi.org/10.1101/2020.06.10.20122275 doi: medRxiv preprint