Public mental health during and after the SARS-CoV-2 pandemic: Opportunities for intervention via emotional self-efficacy and resilience

Abstract

Importance: During the pandemic, the number of United States adults reporting clinically significant symptoms of anxiety and depression sky-rocketed, up from 11% in 2020 to more than 40% in 2021. Our current mental healthcare system cannot adequately accommodate the current crisis; it is therefore important to identify opportunities for public mental health interventions.

Objective: Assess whether modifiable emotional factors may offer a point of intervention for the mental health crisis.

Design setting and participants: From January 13 to 15, 2022, adults living in the United States were recruited via Amazon Mechanical Turk to complete an anonymous survey.

Main outcomes and measures: Linear regressions tested whether the primary outcomes during the SARS-CoV-2 pandemic (depressive and anxiety symptoms, burnout) were associated with hypothesized modifiable risk factors (loneliness and need for closure) and hypothesized modifiable protective factors (the ability to perceive emotions and connect with others emotionally; emotion-regulation efficacy; and resilience, or the ability to "bounce back" after negative events).

Results: The sample included 1,323 adults (mean [SD] age 41.42 [12.52] years; 636 women [48%]), almost half of whom reported clinically significant depressive (29%) and/or anxiety (15%) symptoms. Approximately 90% of participants indicated feeling burned out at least once a year and nearly half of participants (45%) felt burned out once a week or more. In separate analyses, depressive symptoms (Model A), anxiety symptoms (Model B), and burnout (Model C) were statistically significantly associated with loneliness (βModel A, 0.38; 95% CI, 0.33-0.43; βModel B, 0.30; 95% CI, 0.26-0.36; βModel C, 0.34; 95% CI, 0.28-0.41), need for closure (βModel A, 0.09; 95% CI, 1.03-1.06; βModel B, 0.13; 95% CI, 0.97-0.17; βModel C, 0.11; 95% CI, 0.07-0.16), recent stressful life events (βModel A, 0.14; 95% CI, 0.10-0.17; βModel B, 0.14; 95% CI, 0.11-0.18; βModel C, 0.10; 95% CI, 0.06-0.15), and resilience (βModel A, -0.10; 95% CI, -0.15 to -0.05; βModel B, -0.18; 95% CI, -0.23 to -0.13; βModel C, -0.11; 95% CI, -0.17 to -0.05). In addition, depressive and anxiety symptoms were associated with emotional self-efficacy (βModel A, -0.17; 95% CI, -0.22 to -0.12; βModel B, -0.11; 95% CI, -0.17 to -0.06), and beliefs about the malleability of emotions (βModel A, -0.08; 95% CI, -0.12 to -0.03; βModel B, -0.09; 95% CI, -0.13 to -0.04). Associations between loneliness and symptoms were weaker among those with more emotional self-efficacy, more endorsement of emotion malleability beliefs, and greater resilience, in separate models. Analyses controlled for recent stressful life events, optimism, and social desirability.

Conclusion and relevance: Public mental health interventions that teach resilience in response to negative events, emotional self-efficacy, and emotion-regulation efficacy may protect against the development of depressive symptoms, anxiety, and burnout, particularly in the context of a collective trauma. Emotional self-efficacy and regulation efficacy may mitigate the association between loneliness and mental health, but loneliness prevention research is also needed to address the current mental health crisis.

Keywords: COVID-19; burnout; depression; emotional self-efficacy; loneliness; public mental health; resilience.

Figure 1

Correlations between modifiable protective/risk factors and mental health symptoms and burnout. Zero-order correlations between modifiable protective (solid black)/risk factors (dotted) and (A) depressive symptoms, (B) anxiety symptoms, and (C) burnout. All correlations are greater than zero at p < 0.001. N = 1,334.

Figure 2

Emotional self-efficacy moderates the association between loneliness and mental health symptoms. The association between loneliness (possible score range: 8–32) and mental health symptoms [(A) depression possible score range: 0–27; (B) anxiety possible score range: 0–12] is moderated by emotional self-efficacy in separate hierarchical regression models with all specified covariates included. The simple slopes for 16th (3.38), 50th (4.13), and 84th (5.75) percentiles of emotional self-efficacy scores are displayed (all p < 0.001). Models included the following covariates: binary gender, sex, race, ethnicity, whether in a relationship, better or worse off financially due to the pandemic (reference group: no impact of pandemic), less than or more than 4-year college degree attainment (reference group: 4-year college degree), social desirability, and dispositional optimism. Analyses only included participants with binary gender identity because a sample size for non-binary gender meant the study was underpowered to test for any statistically significant differences for this groups. N = 1,323.