A Latent Class Analysis of Coping Behaviors in Response to the COVID-19 Pandemic Among Essential Workers of Color

Background and Purpose: Since the onset of the COVID-19 pandemic, essential workers of color were forced to manage daily stressors such as limited protection from the virus and changes in childcare responsibilities while simultaneously delivering critical services. These multilevel stressors had varying consequences for their overall health and wellbeing, including burnout levels and mental health issues. Less is known about the processes by which this population coped with the unexpected changes in daily life and subsequent stressors. To address this research gap, this study sought to understand distinct patterns of coping behaviors among essential workers of color and examine if individuals in these groups differ meaningfully across sociodemographic characteristics, work conditions, and various indicators of mental health.

Methods: Data and Sample. The analytic sample comprised 319 individuals who reported being essential workers during the COVID-19 pandemic. About 50% of participants identified as female, and 92% identified as heterosexual. Nearly 68% of participants identified as Black or African American, 18% identified as Native American or Alaskan Native, 6% identified as Latinx or Hispanic, and 7% identified as Asian or Native Hawaiian. Measures. Respondents were asked if they had engaged in any of 21 possible coping behaviors in response to pandemic-related stress. These 21 behaviors served as observed indicators for the Latent Class Analysis (LCA). Generally, coping behaviors were classified into the following five domains: (a) relaxation/distraction (6 items); (b) substance use (5 items); (c) eating (4 items); (d) physical activity (3 items); and (e) support seeking (3 items). We also assessed differences across latent classes using measures of perceived stress, COVID-19 anxiety, and other socio-demographic and work-related measures. Analysis. LCA was used to explore the presence of unobserved subgroups within the sample that clustered around similar response patterns across the life-stress experiences items. After identifying the optimal or best fitting solution, the three-step procedure outlined by Asparouhov and Muthén (2014) was used to estimate covariate differences across latent classes.

Results: The class enumeration process signaled that a five-class solution was optimal. The five classes were labeled as follows (a) moderately multifaceted coping (9%); (b) highly multifaceted coping (7%); (c) coping through social support, self-care, and distractions (28%); (d) business-as-usual coping (43%); and (e) coping through smoking, drinking, and media use (13%). Levels of perceived stress prior to and during the COVID-19 pandemic were highest among those in the highly multifaceted and moderately multifaceted patterns, although individuals in the smoking, drinking, and media use pattern—disproportionately male (80%)—reported the largest increase in perceived stress over time. Levels of perceived stress and COVID-19 anxiety were lowest among those in the coping through social support, self-care, and distractions pattern.

Conclusions and Implications: To help mitigate suboptimal patterns of coping behaviors related to COVID-19 stress and anxiety, essential workers of color could benefit from efforts aimed at developing and sustaining social support systems and increasing knowledge of and time devoted to implementing healthy self-care practices.