Increasing Trends in Physical and Mental Health Problems among US Workers in Different Job Sectors During the COVID-19 Pandemic

Data

We used data from the April, August, and December 2020 rounds of the US Census Bureau's Household Pulse Survey (HPS).²¹ The HPS is a national sample household survey in which data on socioeconomic, demographic, self-assessed health, mental health, and healthcare characteristics during the COVID-19 pandemic are collected in near real-time via email and internet. All information collected in the survey are based on self-reports. The HPS uses a systematic sample design and is representative of the civilian non-institutionalized population of the United States.

The HPS was developed as a rapid response survey in order to track the social and economic impacts of the COVID-19 pandemic on American households on a weekly or a bi-weekly basis in partnership with seven US federal statistical agencies, namely: Bureau of Labor Statistics, National Center for Health Statistics, United States Department of Agriculture's Economic Research Service, National Center for Education Statistics, Department of Housing and Urban Development, Social Security Administration, Bureau of Transportation Statistics.^22,23

Measurement of Job-Related Income Loss, Health Outcomes, Job Sector, and Covariates

To examine temporal trends, we selected independent HPS samples for April (Phase 1, Week 1), August (Phase 2, Week 13), and December (Phase 3, Week 21). We restricted the study sample to those aged 18-64 years given our primary aim of examining inequalities in job losses and resultant health impact by job sector. The seven binary outcome variables were job-related income loss, expected job/income loss, and five health measures: self-assessed fair/poor health, serious depression, serious anxiety, serious worry, and serious lack of interest. Job-related income loss was based on the question: "Have you, or has anyone in your household experienced a loss of employment income since March 13, 2020?” Expected household job loss was based on the question, "Do you expect that you or anyone in your household will experience a loss of employment income in the next four weeks because of the coronavirus pandemic?”

Self-assessed general health status was based on the question, "Would you say your health in general, is excellent, very good, good, fair, or poor?” The binary health status was defined with two categories: fair/ poor health vs. excellent/very good/good health. The depression variable was derived from the question, "Over the last seven days, how often have you been bothered by... feeling down, depressed, or hopeless? Would you say not at all, several days, more than half the days, or nearly every day?” Respondents who reported feeling down, depressed, or hopeless nearly every day of the past week were defined as having serious depression. The “anxiety” variable was derived from the question, "Over the last seven days, how often have you been bothered by the following problems... Feeling nervous, anxious, or on edge? Would you say not at all, several days, more than half the days, or nearly every day?” Respondents who reported feeling nervous, anxious, or on edge nearly every day of the past week were defined as having serious anxiety. The “worry” variable was derived from the question, "Over the last seven days, how often have you been bothered by the following problems... Not being able to stop or control worrying? Would you say not at all, several days, more than half the days, or nearly every day?” Respondents who reported not being able to stop or control worrying nearly every day of the past week were defined as having serious worry. The “lack of interest” variable was derived from the question, "Over the last seven days, how often have you been bothered by the following problems... having little interest or pleasure in doing things? Would you say not at all, several days, more than half the days, or nearly every day?” Respondents who reported having little interest or pleasure in doing things nearly every day of the past week were defined as having a serious lack of interest.

The sample size for the December 2020 HPS was 69,944 for all ages and 52,441 for the 18-64 age group. There were 941 (1.79%) observations with missing data for job sector and job-loss variables, excluding which resulted in an effective sample size of 51,500 for analysis. The sample size for all ages, for the 18-64 age group, and effective sample size for analysis were 109,051, 83,643, and 82,173 for the August HPS and 74,413, 57,086, and 56,156 for the April HPS, respectively.

The primary independent variable of interest was the job sector which consisted of 6 categories: government, private company, non-profit organization including tax-exempt and charitable organizations, self-employed, working in a family business, and unemployed. Based on the social determinants of health framework and previous research, the following covariates of job loss, self-assessed health, and mental health were considered: age, gender, race/ethnicity, marital status, region of residence, educational attainment, household income, housing tenure, and actual or expected job loss (for health outcomes only).^16,24 Age was categorized into 5 groups: 18-24, 25-34, 35-44, 45-54, and 55-64. Race/ethnicity consisted of non-Hispanic Whites, non- Hispanic Blacks, Asians, the other/multiple-race group, and Hispanics. The other/multiple-race group consisted of American Indians and Alaska Natives (AIANs), Native Hawaiians and Pacific Islanders, and mixed-race individuals. Marital status was grouped into currently married, widowed, divorced/ separated, and never married. Education consisted of 5 groups: <high school, high school, some college, bachelor's degree, and master's degree or higher. Household income was measured as of 2019 and included 9 categories: <$25,000; $25,000-$34,999; $35,000-$49,999; $50,000-$74,999; $75,000-$99,999; $100,000-$ 149,999; $150,000-$ 199,999; $200,000+; and missing. Housing tenure consisted of homeowners and renters. Region of residence included Northeast, Midwest, South, and West.

Statistical Methods

Multivariable logistic regression was used to examine inequalities in actual job-related income losses since the start of the pandemic and expected job/income losses in the four weeks post-survey. Logic regression was also used to model inequalities in self-assessed health and four mental health measures by job sector in each time period after controlling for selected socioeconomic and demographic characteristics. The Chi-square statistic was used to test the overall association between the job sector and prevalence of job-related income losses and 5 health measures, whereas the two-sample t-test was used to test the difference in prevalence between any two groups or time periods. To account for the complex sample design of the HPS, SUDAAN software was used to conduct all statistical analyses, including the logistic modeling procedure RLOGIST.²⁵

Trends and Disparities in Job-Related income Losses Due to the Pandemic by Job Sector

In December, 51.3% of private-sector workers, 64.0% of self-employed workers, and 66.3% of unemployed adults reported that they or someone in their household experienced a loss of employment income since the start of the pandemic in March 2020 (Table 1 and Figure 1). This percentage was the lowest among those working in the public sector (35.2%) and the non-profit sector (45.0%). For workers in the private and non-profit sectors, household job/income losses increased by 26% between April 2020 and December 2020 (Figure 1).

Close

Figure 1:

Increases in Job-Related Income Losses Since the Pandemic and Expected Job/Income Losses in the Next Four Weeks among Adults Aged 18-64 Years by Job Sector, United States, April, August, and December 2020

Source: Data derived from April, August, and December 2020 Household Pulse Survey. Notes: Chi-square statistics for testing the overall association between job sector and job losses were statistically significant at P<0.001 for each time period. Increases or changes in job losses by job sector between April and December were statistically significant at P<0.05.

Export to PPT

Compared with the government-sector workers, self-employed, unemployed, private sector, non-profit sector, or family business workers had, respectively, 223%, 207%, 84%, 59%, and 45% higher adjusted odds of experiencing job/income losses (Table 1). Disparities in prevalence and adjusted odds of expected job/income losses according to job sector were similar to those in actual job/income losses (Table 1 and Figure 1).

Trends and Disparities in Physical and Mental Health Problems During the Pandemic by Job Sector

Between April and December 2020, the prevalence increased by 42.8% for fair/poor health; 55.3% for serious depression; 33.6% for serious anxiety; 42.1% for serious worry; and 55.1% for serious lack of interest across all job sectors (Table 2). Unemployed adults and workers in each sector reported marked increases in poor overall health and serious mental health problems during April-December 2020. In December 2020, the prevalence of fair/poor health ranged from 12.3% for workers in the non-profit sector and 14.7% in the government sector to 17.1% for those in the family business, 17.9% for the self- employed, and 33.6% for unemployed adults (Table 2). Controlling for covariates, the odds of reporting fair/ poor health were 74% higher for the unemployed than government-sector workers (Table 2).

Prevalence of serious depression varied from a low of 9.5% for workers in the government sector and 10.3% for workers in the non-profit sector to a high of 14.6% for the self-employed, 15.3% for those in the family business, and 22.8% for unemployed adults. Controlling for covariates, the odds of serious depression were 45% higher for the self-employed and 118% higher for the unemployed, compared to government-sector workers (Table 2). Prevalence and odds of serious anxiety, serious worry, and serious lack of interest were generally lowest among workers in the government and non-profit sectors and highest among the self-employed, those in the family business, and the unemployed (Table 2).

impact of Household Job/income Losses on Physical and Mental Health Problems by Job Sector During the Pandemic

In each sector, workers with job/income losses had substantially higher risks of poor physical and mental health than their counterparts with no job/ income losses (Table 3). In December, the prevalence of fair/poor health varied from 7.0% for the self- employed and 9.3% for workers in the non-profit sector without job/income losses to 23.7% for the self-employed with job/income losses and 34.8% for unemployed adults with job/income losses (Table 3).

Compared to workers in the government sector with no job/income losses, workers in the private and government sectors, self-employed, and unemployed adults with job/income losses had 1.6 to 2.5 times higher adjusted odds of experiencing fair/poor health.

Prevalence of serious depression varied from 7.0% for workers in the government sector and 7.1% for workers in the non-profit sector without job/income losses to 18.5% for those in the family business with job/income losses and 26.9% for unemployed adults with job/income losses. Compared to workers in the government sector with no job/income losses, workers from all other sectors with job/income losses experienced 2.0 to 3.6 times higher odds of serious depression (Table 3). Patterns of disparities in serious anxiety, serious worry, and serious lack of interest by job sector and job/income losses were similar to those for poor self-assessed health and serious depression, with workers with no job/losses in each sector experiencing almost twice the risk of serious mental health problems than their counterparts with job/income losses (Table 3).

Limitations

This study has some limitations. HPS does not collect information on many factors that could help explain the rising prevalence of and disparities in physical and mental health problems among workers in various job sectors during the pandemic, including job insecurity, work-related stress, autonomy, control, and decision making in the workplace, social and physical isolation, lack of social support or connectedness, loss of self-identity and sense of control, unhealthy diet, lack of physical inactivity, smoking, alcohol, and other substance use.^4,29 Nevertheless, the analysis of temporal data indirectly sheds light on the significance of these factors which are central to understanding how certain groups of workers have remained vulnerable to an increased likelihood of experiencing poor health and psychological distress as the pandemic has continued to disproportionately affect their social and economic well-being.

Second, the job-sector variable in HPS does not provide details about different industries such as agricultural, non-agricultural, manufacturing, construction, and service sectors, and types of public-sector workers such as federal, state, or local government employees, and armed forces. Third, HPS is a cross-sectional survey; causality cannot be inferred especially for the association of the job sector with self-assessed health and mental health. However, reverse causality is less of an issue, since employment in different job sectors is less likely to change during the course of the pandemic, and self- assessed health and mental health were measured at the time of or the week preceding the survey. Fourth, the respondents in HSP are more likely to be women and non-Hispanic Whites and have higher education, compared with the American Community Survey.³⁴ This might have resulted in an underrepresentation of workers in more disadvantaged job sectors such as self-employment, family business, and the unemployed. However, we addressed disproportionate sampling of demographic characteristics by using survey weights, which rakes the demographics of the interviewed persons to education attainment/sex/age distributions and ethnicity/race/sex/age population distributions.²²

Compliance with Ethical Standards