Economic and Psychosocial Impact of COVID-19 in the Hispanic Community Health Study/Study of Latinos

Objectives: To examine the prevalence and correlates of economic hardship and psychosocial distress experienced during the initial phase of the coronavirus disease 2019 (COVID-19) pandemic in a large cohort of Hispanic/Latino adults. Methods: The Hispanic Community Health Study/Study of Latinos (HCHS/SOL), an ongoing multicenter study of Hispanic/Latino adults, collected information about COVID-19 illness and psychosocial and economic distress that occurred during the pandemic (N=11,283). We estimated the prevalence of these experiences during the initial phase of the pandemic (May 2020 to May 2021) and examined the prepandemic factors associated with pandemic-related economic hardship and emotional distress using multivariable log linear models with binomial distributions to estimate prevalence ratios. Results: Almost half of the households reported job losses and a third reported economic hardship during the first year of the pandemic. Pandemic-related household job losses and economic hardship were more pronounced among noncitizens who are likely to be undocumented. Pandemic-related economic hardship and psychosocial distress varied by age group and sex. Contrary to the economic hardship findings, noncitizens were less likely to report pandemic-related psychosocial distress. Prepandemic social resources were inversely related to psychosocial distress. Conclusions: The study findings underscore the economic vulnerability that the pandemic has brought to ethnic minoritized and immigrant populations in the United States, in particular noncitizens. The study also highlights the need to incorporate documentation status as a social determinant of health. Characterizing the initial economic and mental health impact of the pandemic is important for understanding the pandemic consequences on future health. Clinical Trial Registration Number: NCT02060344


Introduction
In the United States, the pandemic of coronavirus disease 2019 , caused by severe acute respiratory syndrome coronavirus-2 (SARS-COV-2) infection, has been characterized by striking racial and ethnic disparities in overall prevalence, disease severity, and mortality. 1,2 The impact of COVID-19 has been especially striking in Black/African American, Hispanic/Latino, and American Indian/Alaskan Native populations. 1,2 Adverse social determinants of health have driven inequities in the burden and outcomes of COVID-19 among these populations, 3 by shaping the ability to practice physical distancing and to obtain access to timely testing, vaccination, and high-quality health care. The economic consequences of the pandemic also appear to be driven by these structural inequities. 1,4 Hispanic/Latino families have faced disproportionate job losses, as the service and retail industries were decimated, and/or were more likely to work in essential jobs characterized by heightened exposure risks and no guaranteed sick time or time off for testing and vaccination. [5][6][7] Although economic improvements have been noted in the United States, Hispanic/Latino individuals and other underserved groups continue to experience disproportionate food insufficiency, an inability to cover rent, and trouble meeting household expenses. 8 In addition, the pandemic and efforts to contain it appear to have fostered a new mental health crisis. 9,10 A Centers for Disease Control and Prevention (CDC) survey in June 2020 found that more than 4 in 10 adults had at least 1 mental or behavioral health concern, with 31% reporting symptoms of anxiety or depression. 11 These concerns occurred more often among young adults, racial/ethnic minorities, and those with preexisting mental health concerns. 12,13 Additional research is needed to understand the consequences of the pandemic on Hispanic/Latino families to guide targeted prevention and intervention programs and predict future physical and mental health consequences of COVID-19 and future pandemics. The current study attempted to address this gap in knowledge using data from the Hispanic Community Health Study/Study of Latinos (HCHS/SOL). HCHS/SOL represents a unique opportunity to study the initial impact of the COVID-19 pandemic on the Hispanic/Latino population given the well-characterized cohort from different areas of the U.S. and representing multiple heritage groups. Characterizing the initial experience is important, as the first year of the pandemic was the most disruptive socially and economically. Furthermore, the cohort has been followed since 2008-2011, allowing examination of prepandemic influences on economic and psychosocial distress.
The goals of the study were to examine economic hardship, emotional distress (depression and anxiety symptoms), and social isolation during the COVID-19 pandemic, and to explore prepandemic social determinants of health, including U.S. citizenship status as key social determinant of health in this large and predominantly immigrant cohort.

Methods
HCHS/SOL is an ongoing population-based cohort study of 16,415 Hispanic/Latino adults who were selected using a multistage probability sampling design from four U.S. metropolitan areas (Chicago, IL; Miami, FL; Bronx, NY; San Diego, CA) with representation of the major Hispanic/Latino subpopulations (i.e., Mexican, Puerto Rican, Cuban, Dominican, Central or South American). Details about the aims and methodology of HCHS/SOL are published elsewhere. 14,15 In brief, at the baseline in-person examination (Visit 1; 2008-2011), participants completed a comprehensive assessment of cardiometabolic risk or protective factors, questionnaires in their language of preference, and a blood draw for DNA extraction and other biomarkers. A second in-person examination occurred *6 years later (Visit 2; 2014-2017). A third examination is ongoing.
The HCHS/SOL team conducted a 15-min phone survey to capture COVID-19 economic and psychosocial consequences. The survey was completed by 11,283 individuals (72.5% response rate) between May 2020 and May 2021; 59% of surveys were completed between May and August 2020, 30% between September and December of 2020, 9% between January and March 2021, and 2% during April and May 2021. The study was conducted with Institutional Review Board approval from each of the institutions involved in the study.
Measures COVID-19 experiences and consequences COVID-19 pandemic-related job loss and economic hardship. Individuals were asked to report whether they or someone in the household lost their job at any point since the start of the COVID-19 pandemic in the United States (March 2020). If so, they were categorized as experiencing a household job loss. Specific scores for financial (difficulty paying for utilities, phone, or internet, range = 0-3), housing (range = 0-2), and food insecurity (range = 0-1) were also calculated. The primary economic hardship variable was a composite score of report of difficulties paying for basic needs (housing, food, phone or internet, utilities). A numeric score ranging from 0 to 6 was assigned based on the number of hardships reported. For analysis, hardship scores were dichotomized as 0 versus ‡ 1.

COVID-19 pandemic-related psychosocial distress.
Anxiety and depression symptoms were assessed using the validated 2-item Patient Health Questionnaire (PHQ-2) screeners for depression 16,17 and Generalized Anxiety Disorders 2-item (GAD-2). 18 Respondents indicate how often they experienced cardinal depression and anxiety symptoms on a scale of 0 (not at all) to 3 (nearly every day). PHQ-2 and GAD-2 scale scores of ‡ 3 (range = 0-6) were used to indicate elevated depression or anxiety symptoms. 17,18 Individuals with scores ‡ 3 on either or both scales were categorized as having elevated psychological distress. Social isolation was evaluated using a single item assessing how often the respondent felt isolated from others in the past 2 weeks, using the same response format as the PHQ-2 and GAD-2 (range = 0-3). A scale score of ‡ 2 (felt isolated most or all days) was used to indicate elevated social isolation.

Prepandemic factors
Prepandemic psychological distress. At HCHS/SOL Visit 2, depression symptoms were assessed with the Centers for Epidemiological Studies in Depression-10 item scale, 19 which has been shown to be valid and reliable in the HCHS/SOL. 20 Anxiety symptoms were assessed with the GAD-7, a reliable and valid screener for probable clinically significant anxiety in the general population. 21,22 Recommended cut scores of ‡ 10 were used to indicate prior moderate/severe symptoms of depression or anxiety. 19,22 Individuals were classified as having prepandemic elevated psychological distress if their Visit 2 CESD-10 and/or GAD-7 scores were ‡ 10.
Prepandemic social resources. At HCHS/SOL Visit 2, participants completed the Interpersonal Support Evaluation List-12 item (ISEL-12), 23 a measure of available functional (i.e., perceived) social support, which has been shown to be reliable and valid in the HCHS/SOL cohort. The upper quartile was used to define individuals with ''high'' social resources before the start of the COVID-19 pandemic.
Citizenship. During the HCHS/SOL Visit 2, participants reporting that they were not U.S. citizens were asked whether they were legal permanent residents, had applied for legal permanent residency, held another type of visa, or if none of these situations applied. Consistent with prior methods applied in HCHS/SOL, 24,25 citizen status was classified as: U.S.born citizen (i.e., all individuals born in the 50 U.S. states/DC or U.S. territories); naturalized citizen; documented noncitizen (i.e., holding or applying for legal permanent residency or another visa); or likely undocumented noncitizen (not holding or applying for a visa).
Sociodemographic factors, including sex, Hispanic/ Latino background, and place of birth were collected at baseline. Duration of U.S. residence and age were calculated at the time of COVID-19 interview. Annual household income and educational attainment data were obtained from HCHS/SOL Visit 2.

Data analyses
Unadjusted prevalence of self-reported economic hardship, and psychosocial distress were summarized using complex survey procedures to account for the HCHS/ SOL design. Multivariable log linear models with binomial distributions were used to estimate the prevalence ratio (PR) to examine the prepandemic factors associated with each of the outcomes of interest. These models included the following covariates: age groups at the time of COVID-19 interview, sex, field center, Hispanic/ Latino background, income, education attainment, citizenship, preexisting psychosocial distress, and preexisting social resources. The entire HCHS/SOL cohort at baseline, excluding those who died during the follow-up period (N = 851), was included. Missing covariates and outcomes were handled using multiple imputation. 26 The imputation model included all covariates in the log linear model and variables potentially related to missingness.
These auxiliary variables included: language preference, marital status, health insurance status, cardiovascular disease, hypertension, diabetes, physical activity, smoking, alcohol intake, mental health status, physical health status, social acculturation, anxiety and depression symptoms, time between baseline visit and COVID-19 interview, and baseline exam sampling weight. The fully conditional specification method was used for imputation. 27 Logistic or linear regression methods were used depending on whether variables were categorical or continuous. Each variable with missingness was imputed sequentially until all covariates were complete.
This imputation process was repeated 10 times resulting in 10 full complete datasets. 26 Ninety-five percent confidence intervals (95% CIs) for the regression coefficients were constructed based on the combined estimates and the associated standard errors. PRs and the associated 95% CIs were then calculated by exponentiating the combined estimates of the regression coefficients and their associated 95% CIs. All the analyses were performed using SAS 9.4 software (SAS Institute, Cary, NC) and SUDAAN software Release 11 (RTI International, Research Triangle Park, NC).

Study population
The characteristics of the COVID-19 interview sample and the entire HCHS/SOL cohort are presented in Table 1. In brief, the COVID-19 interview sample included 7143 females and 4140 males between the ages of 27 and 88. The majority of the sample was born outside of the United States (states/DC/U.S. territories); about 30% were naturalized citizens, 29% were documented noncitizens, and 17% were likely undocumented noncitizens. Participants were predominantly of lower socioeconomic status; 39% reported an annual household income < $20,000 before the pandemic and about 36% did not graduate from high school. The distributions of characteristics in the COVID-19 interview sample was similar to those for the entire HCHS/SOL cohort at baseline.

COVID-19 pandemic-related economic hardship
The prevalence of household job loss and economic hardship during the COVID-19 pandemic overall and by sociodemographic factors are depicted in Table 2.
As predicted, individuals in HCHS/SOL were deeply impacted by the economic consequences of the pandemic. A high percentage (46%) reported household job loss and elevated economic hardship (35%). Among the types of economic hardship, 19% reported financial insecurity, 19% reported housing insecurity, and 23% reported food insecurity. The oldest age group ( ‡ 65 years) reported less household job loss and economic hardship than younger groups. The highest economic hardship was reported among noncitizens who are likely undocumented (54%). This group also reported the highest percentage of household job losses, financial, housing, and food insecurity.
As displayed in  COVID-19 pandemic-related psychosocial distress As shown in Table 4, elevated depression and anxiety symptoms varied by sex and age group, with women and older groups most affected. About 14% of the total population reported feelings of social isolation, with a higher prevalence of elevated social isolation among women and lower prevalence among those < 35 years old. There were also differences across field centers, with those in the Bronx and Miami appearing to experience more psychosocial distress relative to those in Chicago and San Diego. Higher prevalence of psychosocial distress and social isolation were observed among the lowest household income group. While U.S.-born citizens showed the least economic hardship; they showed the highest prevalence of elevated anxiety and depression symptoms, and social isolation, relative to noncitizens. In multivariable analyses (

Discussion
This study demonstrated high levels of COVID19related economic hardship in the longest ongoing cohort of Hispanic/Latino adults, with almost half of the households reporting job losses and a third rep-orting economic hardship in the first year of the pandemic. Important variations were noted by citizenship, which highlights documentation status as a key social determinant of health. [28][29][30] Specifically, pandemicrelated household job losses and economic hardship were more pronounced among noncitizens who are likely to be undocumented. These findings underscore the economic vulnerability that the pandemic has brought to immigrant populations in the United States, and in particular noncitizens. These findings are not entirely surprising given that many federal and state economic relief measures excluded undocumented individuals. 5,31 Even among documented immigrants, the fear of losing documented status or a path to citizenship may prevent Hispanic/Latino adults from accessing the financial aid for which they are eligible. Furthermore, the study revealed that individuals with lower household incomes, who are likely to have been financially insecure before the pandemic, were the most likely to experience economic hardship during the pandemic. Of all the factors that define economic hardship, food insecurity was the most prevalent, affecting noncitizens the most. Our results are in line with previous reports. The Urban Institute reported that greater economic losses and material hardship following the COVID-19 pandemic occurred among households with a noncitizen in the family. 32 Another national survey also found that those with the lowest prepandemic incomes were more likely to experience economic hardship during the initial pandemic period. 33 Taken together, these findings are worrisome indicators of how the U.S. is failing to protect the most vulnerable in our society.
Younger adults were more likely than adults ‡ 65 years of age to report household job losses and economic hardship. This finding is not surprising since most individuals over 65 are retired, but it may have important implications for the future. In addition, nonskilled and service work are common occupations among HCHS/SOL individuals, occupations that are less conducive for remote work and may have offered less job security during the initial phases of the pandemic. On the other hand, young adulthood represents a period when individuals start families and have their most productive years. Facing economic adversity as a young adult may have important long-term socioeconomic and health consequences later in life. Ultimately, the long-term health and socioeconomic consequences may represent another means through which the COVID-19 pandemic has magnified health disparities experienced by Additional support was provided by the Collaborative Cohort of Cohorts for COVID-19 Research (C4R)(1OT2HL156812) and New York Regional Center for Diabetes Translation Research (P30 DK111022) through funds from the National Institute of Diabetes and Digestive and Kidney Diseases. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NHLBI or the National Institutes of Health.