Differential effects of COVID-19 stressor types on mental health and moderating roles of coping strategies

Abstract

Stress exposure often leads to poor mental health, but whether certain types of stressors predict differential levels of mental health symptoms and whether coping differentially moderates these associations remain untested. This study examined whether COVID-19 stressor types (financial-, activity-, and infection-related) differentially predicted subsequent mental health (general distress, posttraumatic stress symptoms (PTSS), and positive states of mind) and whether different types of coping moderated these relationships. In April, 2020, questionnaires assessing the impact of COVID-19 on daily life were administered to an adult sample across the US (N = 1,546); 861 (55.7%) of whom completed questionnaires one month later. Nearly all stressor types independently predicted each aspect of mental health. Behavioral disengagement moderated links between both activity- and infection-related stressors and mental health. Seeking emotional social support moderated links between infection-related stressors and both general distress and positive states of mind as well as activity-related stressors and general distress. Positive reappraisal moderated links between financial-related stressors and positive states of mind. Acceptance coping predicted better mental health but was not a significant moderator. Stressor types and coping strategies are important determinants of effects of COVID-19 on mental health; our findings suggest possible screening and intervention implications and directions for future research.

The unprecedented stress of the COVID-19 pandemic has led to substantial social, economic, physical, and mental health impacts worldwide. The prolonged nature of the pandemic differentiates it from other stressors and may magnify its effects on mental health, including not only increased negative symptoms such as general distress (anxiety, depression, and stress) and posttraumatic stress symptoms (PTSS), but also reduced positive experiences, such as in positive states of mind (PSOM). For instance, the likelihood of US adults screening positive for anxiety and depressive disorders quadrupled from 2019 to 2020 (Twenge et al., 2021) and it has been estimated that 26% of the general population had PTSS or posttraumatic stress disorder (PTSD) as a result of the pandemic (Czeisler et al., 2020). Additionally, studies have reported that positive aspects of mental health including life satisfaction and positive affect significantly declined in the early months of the pandemic (Zacher & Rudolph, 2021).

Although associations of high stress exposure and negative mental health outcomes are well-established, relatively little is known about how different types of stressors stemming from a commonly-experienced situation such as the pandemic differentially affect mental health. The pandemic spans many different types of stressors, from financial concerns (e.g., loss of employment or income; Galea & Abdalla, 2020) to activity/routine-related concerns (e.g., changes to work or school routine; Shaw et al., 2020) to infection-related concerns (e.g., risk of becoming infected, stigma related to high risk jobs/certain age groups; Quadros et al., 2021). Greater understanding of how these different types of stressors predict mental health symptoms could help improve both preventative screening and treatment of mental health symptoms.

In addition, although some theories (e.g., transactional theory of stress and coping; Lazarus & Folkman, 1987) suggest that the different strategies individuals implement to deal with stressors (i.e., to cope) might differentially buffer these links between types of stress exposure and aspects of mental health, research examining this notion is rare. The transactional theory of stress and coping posits that stress responses are highly dependent on one’s appraisals of the situation and interactions with the environment, which in turn, impacts one’s coping strategy and resulting mental health responses (Lazarus & Folkman, 1987).

The current study aimed to determine whether different types of stressors all independently or differentially predict different types of mental health symptoms, and whether different types of coping differentially buffer these relationships. Investigating these novel questions contribute to better understanding of risk and resilience factors related to stress and coping trajectories and may have downstream implications for preventive assessments and treatments for mental health symptoms.

Types of Stressors and Mental Health Symptoms

Prior studies have demonstrated links between overall stress and mental health (e.g., Shields & Slavich, 2017). However, very few have dismantled the global construct of “stress” to examine the roles of domain-specific stressors such as the financial-, activity-disruption-, and infection-related stressors of the pandemic and whether each differentially and/or independently predicts mental health. Some preliminary evidence suggests that stressor types may predict different mental health outcomes. For example, one study found adolescents’ responses to stress in different domains differentially predicted their depressive symptoms; depressive symptoms were predicted by rumination about dependent and social events rather than independent or non-social events (Nicolai et al., 2013). Moreover, some stressors (e.g., interpersonal relationship struggles, financial difficulties, loss of loved ones) appear more likely to provoke depressive symptoms while others (e.g., sickness or injury) are more closely associated with anxiety symptoms while yet others (e.g., COVID-19-related discrimination) may be more likely to provoke PTSS (Hahm et al., 2021; Rosendal et al., 2011; Shin et al., 2017).

Another possibility, albeit with less support, is that stressor types independently predict mental health. One study of Arab adults included financial-, activity-, and infection-related COVID-19 stressors simultaneously in a larger model and found all independently predicted mental health symptoms (Kira et al., 2021). Given the importance of the contextual factors (e.g., geography, current number of cases, cultural considerations) in assessing impact of different stressors pandemic, these questions need to be further investigated. Although findings are mixed, prior literature more consistently supports the notion that specific types of stressors predict differential mental health symptoms. Additionally, to our knowledge, no study to date has explored how coping strategies may impact the mental health effects of different types of stressors, which has important potential implications for identifying modifiable treatment targets.

The Role of Coping in Different Types of Stress Exposure

Importantly, while stress exposure is generally associated with a wide range of negative outcomes, individuals vary a great deal in the extent to which they develop such negative outcomes. This variation suggests that factors may buffer the relationship between stress exposure and mental health (Troy & Mauss, 2011). Coping strategies are increasingly recognized as one such factor in individuals’ resilience to stressors (Labrague et al., 2021). However, findings on the effects of coping on mental health have been mixed (Main et al., 2011; Rettie & Daniels, 2021). Many studies have identified maladaptive coping strategies (e.g., behavioral disengagement) that adversely influence mental health, but literature on protective coping strategies that improve mental health is less abundant or consistent (e.g., Freyhofer et al., 2021; Mahmoud et al., 2012; Main et al., 2011). For instance, one study of adults in the United Kingdom found that during the COVID-19 pandemic, maladaptive coping (e.g., denial, behavioral disengagement) was significantly associated with poorer mental health while coping generally considered adaptive (e.g., acceptance, active coping) was not (Rettie & Daniels, 2021). Even in studies in which adaptive coping was associated with better mental health, effect sizes were consistently larger for maladaptive coping strategies (Lehavot, 2012).

No coping strategy is efficacious across all situations (Troy & Mauss, 2013) but rather efficacy of any given strategy depends on the type of stressor and the demands it places on the individual (e.g., Mattlin et al., 1990; Rottweiler et al., 2018). However, the relative effectiveness of different coping strategies under different circumstances is poorly understood (Eckenrode, 2013). Research on situational determinants of coping effects has largely focused on stress appraisals (e.g., controllability; Finkelstein-Fox & Park, 2019) rather than objective characteristics (e.g., domain) of the stressor. Very little is known about how the effects of different types of exposure emanating from the same situation might be more or less buffered by different types of coping. Coping strategies offer a modifiable treatment target; thus if certain coping strategies are more effective in specific contexts, this research could be clinically impactful.

The types of stressors individuals faced in the pandemic offers an opportunity to examine how different coping strategies might differentially buffer the effects of different types of stressors. For example, activity-related stressors (including changes to personal, work, and school routines) may result in social isolation, and thus, seeking emotional support may be an effective coping strategy. This hypothesis is based on results of studies that have shown the importance of social support during life transitions that require adjustments in personal, social, and academic domains, which is also applicable in the context of COVID-19 (Urquhart & Pooley, 2007). On the other hand, seeking emotional support may be less effective (or not possible due to social distancing constraints) for exposure to infection-related stressors while acceptance may be more useful due to the uncertainty of the disease. Although social support is often positive, it may also have negative implications, such as when the goals of the support-seeker and support-giver misalign as well as the potential costs of burdening others (e.g., Brashers et al., 2004). Due to the exploratory nature of this study, specific hypotheses about coping strategy effectiveness are necessarily speculative because of the limited research to date.

Aims and Hypotheses

This novel study aims to determine if exposure to three different types of COVID-19 stressors results in differential mental health symptoms and whether different coping strategies moderate these relationships. While coping clearly plays a role in the impact of stressors on individuals’ mental health, the specifics of this role are less clear. Some prior studies have conceptualized coping as a mediator between stress and mental health outcomes (e.g., Lee & Song, 2021), while others have examined coping as a moderator of that relationship (e.g., Main et al., 2011). The current study aims to understand whether coping strategies are differentially effective for different types of stressor exposure (as opposed to investigating coping as a mechanism explaining how stress exposure leads to distress); thus, coping was conceptualized as a moderator. Mental health was assessed broadly to include two aspects of distress, general and COVID-19-specific.In addition, because mental health also includes positive aspects that are not merely the inverse of negative symptoms (Park et al., 2022), we included a measure of positive well-being, the ability to achieve positive states of mind (PSOM). The coping strategies selected for this study included behavioral disengagement, positive reappraisal, seeking emotional support, and acceptance to cover both emotion-focused and meaning-focused coping strategies shown to vary in effectiveness in previous studies (Guo et al., 2013; Rettie & Daniels, 2021). Problem-focused coping strategies (e.g., active coping) were not included due to the low controllability of stressors early in the pandemic. The hypotheses are as follows:

Hypothesis 1: Specific types of stressors will differentially, rather than independently, predict both positive (i.e., PSOM) and negative (i.e., general distress and PTSS) aspects of mental health.

For instance, based on the integrated specificity model of stress (i.e., different types of stressful conditions elicit different psychobiological impacts; Kemeny, 2003), financial-related stressors will be more strongly associated with increased general distress (e.g., depression) symptoms, while activity-related stressors (e.g., cancellation of celebrations and vacations) will be more strongly associated with decreased PSOM. Due to the nature of infection related stressors and the large number of COVID-related deaths early in the pandemic (Koh et al., 2021), these types of stressors will be more likely associated with increased PTSS. Due to the very limited amount of prior research on this topic, this hypothesis is exploratory and further research must be conducted to inform more specific hypotheses for each stressor type.

Hypothesis 2: Coping strategies will have each have a significant main effect on both positive (i.e., PSOM) and negative (i.e., general distress and PTSS).

Seeking emotional support, positive reappraisal, and acceptance will generally be adaptive coping strategies (i.e., associated with lower general distress and PTSS and higher PSOM), while behavioral disengagement will generally be maladaptive (i.e., associated with higher general distress and PTSS and lower PSOM).

Hypothesis 3: Coping strategies will differentially moderate the linkage between certain stressor types and both positive (i.e., PSOM) and negative (i.e., general distress and PTSS); thus, certain coping strategies will strengthen or weaken the relationship between stress and mental health depending on the stressor type.

This hypothesis is based on prior research supporting that the efficacy of coping strategies depends on contextual factors (e.g., Mattlin et al., 1990; Rottweiler et al., 2018). For example, activity-related stressors may result in social isolation, and thus, seeking emotional support may be an effective coping strategy and attenuate the relationship between activity-related stressors and poor mental health. On the other hand, seeking emotional support may be less effective (or not possible due to social distancing constraints) for exposure to infection-related stressors and may amplify this relationship between infection-related stressors and poor mental health, while acceptance may be more useful due to the uncertainty of the disease. Due to the exploratory nature of this study, specific hypotheses about effectiveness of each coping strategy mentioned above are necessarily speculative because of the limited research to date.

Methods

Study design and participants

Participants were recruited from Amazon’s Mechanical Turk (MTurk) online worker pool. MTurk workers are more diverse than typical student or online forum samples and fairly representative of larger populations, and prior research has shown the MTurk data to be high quality, replicable, and valid (Sheehan, 2018). Longitudinal data were collected at six timepoints through April 25^th, 2021. Inclusion criteria were as follows: participants needed to be 18 years or older, living in the United States, and English-speaking. After consenting, participants completed questionnaires on the impact of COVID-19 on their daily lives. Participants received $2 for each timepoint completed. The data presented focus on the survey results from the first (T1) and second (T2) timepoints, which were collected between April 8 – 25, 2020 and between May 15 – 29, 2020, respectively. At T1, there were 1,546 participants; 861 (55.69%) of whom completed the T2 assessments. All study materials were approved by the University of Connecticut IRB (X20–0057) as an exempt protocol.

Measures

Measure reliability was calculated using McDonald’s omega (ω; maximum likelihood) using Hayes’ OMEGA macro (Hayes & Coutts, 2020), in accordance with recent research supporting it as a more optimal measurement of reliability compared to other more traditional estimations (Hayes & Coutts, 2020). Reliability for scales that contained fewer than three items was calculated using Cronbach’s alpha (α).

Covariates

Demographics included gender, age, race, and income. Participants were permitted to check all applicable categories when reporting their race. Race and gender were measured at T1, while age and income were measured at T2.

COVID-19 stressors

A measure of COVID-19 stressors (Park et al., 2020) was administered. Stressor exposure was assessed by asking whether individuals experienced each of 23 stressors in the past week (“yes” = 1 or “no” = 0). The scale has demonstrated convergent and divergent validity and a unidimensional factor structure (Tambling et al., 2021). The scale was divided into 3 subcategories: financial-related stressors (e.g., loss of current job security or income), activity-related stressors (e.g., changes to daily work routines), and infection-related stressors (e.g., risk of becoming infected). The summed stress exposure score in each subcategory was used in the analysis. In this sample, McDonald’s omega was .67 for infection-related stressors, .71 for activity-related stressors, and .56 for financial-related stressors.

Mental health

General distress.

The 21-item Depression Anxiety Stress Scales (DASS-21; Lovibond & Lovibond, 1995) was used to assess general distress in the past week. The DASS-21 includes 7 items each to assess anxiety, depression, and stress. Items are rated from 0 (“did not apply to me at all”) to 3 (“applied to me very much” or “most of the time”). The scores on each subscale are summed and then multiplied by two to calculate the total score comparable to previous studies using the longer format of the measure (Henry & Crawford, 2005). The DASS has been shown to be reliable and valid in both non-clinical and clinical samples (Henry & Crawford, 2005). The DASS-21 was administered at both T1 and T2 and McDonald’s omega was .96 at both timepoints.

Peritraumatic distress and posttraumatic stress symptoms (PTSS).

Two measures were used to assess COVID-19 specific distress. At T1, the Peritraumatic Distress Inventory (PDI) was administered, which is a 13-item self-report measure to assess emotional and physiological distress experienced during or immediately after a traumatic event (i.e., COVID-19 pandemic). The PDI has been shown to be predictive of the development of PTSS or PTSD (Guardia et al., 2013). The pandemic was just beginning at T1, so the PDI was used to assess distress whereas the Impact of Events Scale-Revised (IES-R) was introduced at T2 to measure PTSS. The IES-R is a 22-item self-report instrument assessing how distressing each of 22 symptoms has been in the past week in response to a specific traumatic event, adapted here to ask with respect to the COVID-19 pandemic. In prior studies, the IES-R demonstrated high internal consistency (α = 0.94) and criteria validity (Chen et al., 2011). Items are rated from 0 (“not at all”) to 4 (“extremely”) and the total score is summed. Higher scores indicate greater concern for a PTSD diagnosis. McDonald’s omega was .90 for the PDI at T1 and .95 for the IES-R at T2.

Positive states of mind.

The PSOM scale was introduced at T2 to assess positive emotional and cognitive experiences. This 7-item measure has been found to be inversely related to anxiety and stress (Adler et al., 1998). In prior studies, the PSOM has shown satisfactory internal consistency (Cronbach’s α = .72; Adler et al., 1998). Items are rated from 1 (“not at all”) to 5 (“very much”). McDonald’s omega was .87 in this sample at T2.

Coping

From the Brief COPE (Carver, 1997), the 2-item behavioral disengagement (“I’ve given up trying to deal with it” and “I’ve been giving up the attempt to cope”; Cronbach’s α = .79) and seeking emotional support (“I’ve been getting emotional support from others” and “I’ve been getting comfort and understanding from someone”; Cronbach’s α = .85) subscales were administered. The questionnaire asked how individuals had “coped with COVID-19 related stressors in the last week”. Items are rated from 1 (“I haven’t been doing this at all” to 4 (“I’ve been doing this a lot”). Additionally, the acceptance (“I think that I have to accept that this has happened” and “I think that I have to accept the situation”; Cronbach’s α = .78) and positive reappraisal (“I think I can learn something from the situation” and “I think that I can become a stronger person as a result of what has happened”; Cronbach’s α = .83) subscales of the Cognitive Emotion Regulation Questionnaire-short form (CERQ; Garnefski & Kraaij, 2006) were administered. Items are rated from 1 (“almost never”) to 5 (“almost always”). Scores for each subscale are then summed.

Data analysis plan

Descriptive statistics and bivariate correlations among study variables are reported. Attrition analyses were conducted to determine if any of the variables were significantly different for participants who completed both T1 and T2 from those who dropped out after T1. A series of four (one for each coping type) 5-step hierarchical regressions with bias-corrected 95% confidence intervals (n=1,000) were conducted to examine the effects of stressor exposure, coping, and their interactions on mental health. The stressor types and coping strategies were centered at their means prior to calculating the interactions to reduce multicollinearity. SPSS (version 26) was used for data cleaning and SPSS (version 28) was used for the analyses. Hayes’ PROCESS Macro (Hayes, 2017; Model 1) was used to plot and probe (using the Johnson-Neyman technique; Johnson & Fay, 1950), the interactions for the significant moderators at mean, one standard deviation above, and one standard deviation below mean (if lower than the range, the minimum score was used).

Results

Sample description

Sample characteristics are reported in Table 1. The sample included 668 males (43.21%), 850 females (54.98%), and 32 participants (2.07%) endorsing nonbinary, transgender, self-described or prefer not to say. Due to the small percentage of this latter group, gender was dichotomized (men/women) and the rest were eliminated from the analysis. Average age was 39.3 (SD = 14.1). Most participants identified as White (n = 1,249; 80.79%); others identified as Black/African American (n = 209; 13.52%), Asian/Asian American (n = 208, 13.45%), Native Hawaiian (n = 89, 5.76%), and American Indian/Alaska Native (n = 119, 7.78%). As the numbers of each non-White group were too small to analyze comparatively, race was collapsed into a dichotomous variable (White/non-White). Average income was $59,550 (SD = $21,112.5).

Table 1.

Descriptive Statistics and Pearson Correlations for Study Variables

	n (%)/M (SD)	Range	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
1. Gender (T1)		N/A
Male	668 (43.2%)
Female	850 (54.8%)
2. Race (T1)		N/A	−.01
White	1,132 (72.3%)
Non-White	433 (27.7%)
3. Age (T2)	39.26 (14.142)	20–89	−.03	.18**
4. Income (T2)	$59,550 ($21,112.5)	< $5000 – > $200,000	−.02	.00	.01
5. Financial Stress (T2)	1.3 (1.2)	0–5	−.05	.04	−.08*	−.04
6. Activity Stress (T2)	4.2 (2.5)	0–10	−.10**	−.02	−.15**	.13**	.50**
7. Infection Stress (T2)	3.5 (1.7)	0–8	−.09**	.05	−.11**	.01	.35**	.40**
8. Behavioral Disengagement Coping (T2)	1.4 (.6)	1–4	.02	−.04	−.27**	−.11**	.18**	.14**	.11**
9. Positive Reappraisal Coping (T2)	5.8 (2.4)	2–10	−.04	−.06	−.04	.11**	.06	19**	.09*	−.01
10. Emotional Support Coping (T2)	2.2 (.9)	1–4	−.09**	.01	−.04	.08*	.14**	.22**	.23**	.07	.33**
11. Acceptance Coping (T2)	7.0 (2.4)	2–10	−.05	.02	.14**	.02	.02	.19	.12**	−.01	.39**	.22**
12. General Distress (DASS) (T1)	29.3 (.7)	0–126	−.10**	−.001	−.29**	−.16**	.27**	.26**	.29**	.50**	.01	.09*	−.04
13. General Distress (DASS) (T2)	22.3 (25.4)	0–124	−.04	−.05	−.28**	−.13**	.33**	.28**	.31**	.57**	.01	.14**	−.01	.85**
14. Peritraumatic Distress (PDI) (T1)	10.6 (.2)	0–48	−.18**	.00	−.22**	−.07	.31**	.34**	.33**	.32**	.06	.25**	.01	.68**	.61**
15. Posttraumatic Stress Symptoms (IES-R) (T2)	17.1 (16.1)	0–8	−.04	−.08*	−.24**	−.06	.38**	.41**	.35**	.55**	.11**	.26**	.07	.69**	.80**	.64**
16. Positive States of Mind (PSOM) (T2)	22.5 (4.8)	7–28	.08*	.05	.23**	.23**	−.22**	−.15**	−.18**	−.40**	.22**	.11**	.15**	−.59**	−.63**	−.42**	−.49**

Note. T1 = Time 1; T2 = Time 2; DASS = Depression Anxiety Stress Scales; IES-R = Impact of Events Scale-Revised; PSOM = Positive States of Mind. Race (0 = Non-white; 1 = White) and gender (0 = Female; 1 = Male) were analyzed as dichotomous variables.

^*p<.05

^**p<.01

Attrition analyses

Responses were obtained from 1,546 participants at T1 and 861 at T2. Completers and dropouts differed significantly, respectively, on DASS (M = 23.9, SD = 24.8 vs. M = 35.7, SD = 28.4, t(1407.81) = −8.61, p<.001); all three types of stressors (Financial-related: M = 1.6, SD = 1.2 vs. M = 2.0, SD = 1.3, t(1425.53) = −3.95, p<.001; Activity-related: M = 5.2, SD = 2.2 vs. M = 5.6, SD = 2.2, t(1508.64) = −3.15, p<.001; Infection-related: M = 3.9, SD = 1.6 vs. M = 4.3, SD = 1.7, t(1466.73) = −5.003, p<.001); behavioral disengagement coping (M = 1.4, SD = 0.7 vs. M = 1.6, SD = .73, t(1426.15) = −4.383, p<.001); and acceptance coping (M = 7.0, SD = 2.1 vs. M = 6.7, SD = 2.0, t(1520.01) = 2.486, p = .013). Completers and dropouts did not significantly differ on positive reappraisal coping or emotional social support coping. IES-R and PSOM measures were introduced at T2, so attrition analyses were not conducted for these scores.

Intercorrelations among study variables

Bivariate correlational analysis is shown on Table 1. Stressor types were moderately correlated with each other (rs = .35 to .50). Coping intercorrelation sizes ranged from low to moderate (rs = −.01 to .39), with the weakest between acceptance and behavioral disengagement and the strongest between positive reappraisal and acceptance. Most coping types significantly correlated with stressor types (rs = .11 to .23) with the exception of positive reappraisal and financial-related stressors and acceptance and both financial- and activity-related stressors. All three stressor types were significantly correlated with both distress (rs = .28 to .41) and PSOM (rs = −.15 to −.22). Most coping types were significantly correlated with each mental health variable (rs = .11 to .57). Mental health variables correlated moderately to highly with each other (rs = −.49 to .80).

Models of stress exposures as predicting mental health with coping as a moderator

Five-step hierarchical multiple regression analyses examined the effects of stress exposures predicting mental health and the potential moderating effects of coping strategies on the relationship between each COVID-19 stressor type and mental health variable. Analysis of tolerance statistics confirmed no violations of multicollinearity, as all tolerance values were above 0.10; the minimum tolerance value was 0.622. No Variance Inflation Factors exceeded 10; the largest was 1.607. The first step included demographic items. Step 2 added each COVID-19 stressor type (financial, activity, and infection-related) to the regression model. The first two steps were identical across all types of coping and thus are shown only in Table 2. Coping strategy was added at the third step. Interactions between each stressor type and coping were added at the fourth step. Lastly, both types of baseline distress were added at the fifth step for DASS and PTSS scores to predict residual change. PSOM scores were not collected at baseline and thus were not included in this step. Coefficients for the last two steps (i.e., the primary outcomes) can be found in Tables 2–5. Complete tables including coefficients and standard errors for all steps can be found in the Supplement (Tables S1–S4).

Table 2:

Hierarchical regressions with behavioral disengagement coping predicting mental health

	Behavioral disngement coping
	DASS					PTSS					PSOM
Steps of the hierarchical regression models	B	β	R2	ΔR2	F Statistics	B	β	R2	ΔR2	F Statistics	B	β	R2	ΔR2	F Statistics
Step 1	See Table S1	See Table S1	.09	.09	F(4, 816) = 20.68***	See Table S1	See Table S1	.07	.07	F(4, 778) = 13.86***	See Table S1	See Table S1	.10	.10	F(4,632) = 18.22***
(Demographics)
Step 2	See Table S1	See Table S1	.22	.13	F(7, 813) = 32.45***	See Table S1	See Table S1	.27	.20	F(7, 775) = 41.06***	See Table S1	See Table S1	.16	.05	F(7, 629) = 16.45***
(Stressor Types)
Step 3	See Table S1	See Table S1	.42	.20	F(8, 812) = 73.54***	See Table S1	See Table S1	.46	.19	F(8, 774) = 83.02***	See Table S1	See Table S1	.23	.07	F(8, 628) = 23.28***
(Coping Strategy)
Step 4			.43	.01	F(11, 809) = 54.30***			.47	.01	F(11,771) = 61.71***			.25	.02	F(11,625) = 19.08***
Gender	−1.32	−0.03				−0.55	−0.02				0.65	0.07
Race	−2.05	−0.04				−2.45*	−0.07				0.26	0.02
Age	−0.18***	−1.00				−0.06	−0.05				0.04**	0.11
Income	−0.70**	−0.08				−0.13	−0.02				0.35***	0.20
Fin. Stress	2.63***	0.12				1.73***	0.13				−0.48**	−0.12
Act. Stress	0.83*	0.08				1.34***	0.20				−0.04	−0.02
Inf. Stress	2.42***	0.16				1.64***	0.17				−0.29**	−0.10
BD	18.50***	0.47				11.61***	0.47				−2.63***	−0.35
Fin. × BD	−0.23	−0.01				−0.67	−0.04				0.40	0.08
Act. × BD	1.21*	0.08				0.87**	0.09				0.14	0.05
Inf. × BD	−0.45	−0.02				−0.85*	−0.06				0.30	0.08
Step 5			.76	.33	F(12, 808) = 207.27***			.61	.14	F(12,770) = 99.71***	N/A	N/A	N/A	N/A	N/A
Gender	0.01	0.00				1.27	0.04
Race	−2.12*	−0.04				−2.32**	−0.06
Age	0.01	0.00				0.00	0.00
Income	0.09	0.01				0.02	0.00
Fin. Stress	1.31**	0.06				0.91**	0.07
Act. Stress	0.16	0.02				0.88***	0.13
Inf. Stress	0.63*	0.04				0.90***	0.09
BD	6.63***	0.170				9.08***	0.37
Fin. × BD	0.94	0.04				−0.22	−0.01
Act. × BD	0.81*	0.05				0.82**	0.09
Inf. × BD	−0.34	−0.02				−0.53	−0.04
BL DASS/PDI	0.74***	0.72				0.87***	0.44

Note. DASS = Depression Anxiety Stress Scales; PTSS = Posttraumatic Stress Symptoms; PSOM = Positive States of Mind; BD = Behavioral Disengagement; Fin. = Financial-Related Stressors; Act. = Activity-Related Stressors; Inf. = Infection-Related Stressors; BL = Baseline; PDI = Peritraumatic Distress. Race (0 = Non-white; 1 = White) and gender (0 = Female; 1 = Male) were analyzed as dichotomous variables. Age, income, stressors, and coping were centered at their means. Complete table including coefficients and standard errors for all steps can be found in the Supplement (Table S1).

^*p < .05.

^**p < .01.

^***p <.001.

Table 5:

Hierarchical regressions with acceptance coping predicting mental health

	Acceptance Coping
	DASS					PTSS					PSOM
Steps of the hierarchical regression models	B	β	R2	ΔR2	F Statistics	B	β	R2	ΔR2	F Statistics	B	β	R2	ΔR2	F Statistics
Steps 1 & 2
See Table 2
Step 3(Coping Strategy)	See Table S1	See Table S1	.22	.00	28.37***	See Table S1	See Table S1	.27	.00	36.44***	See Table S1	See Table S1	.18	.02	16.88***
Step 4			.22	.00	20.90***			.28	.00	26.59***			.18	.01	12.66***
Gender	−0.41	−0.01				0.23	0.01				0.67	0.07
Race	−1.85	−0.03				−2.45*	−0.07				0.36	0.03
Age	−0.39***	−0.22				−0.20***	−0.17				0.05***	0.15
Income	−1.12***	−0.13				−0.39*	−0.07				0.39***	0.23
Fin. Stress	4.05***	0.19				2.56***	0.19				−0.53**	−0.13
Act. Stress	1.06**	0.10				1.49***	0.23				−0.11	−0.06
Inf. Stress	2.65***	0.18				1.62***	0.17				−0.36**	−0.13
Acc	−0.13	−0.01				0.37	0.06				0.29***	0.15
Fin. × Acc	0.32	0.03				0.11	0.02				0.09	0.05
Act. × Acc	0.01	0.00				0.01	0.00				−0.03	−0.04
Inf. × Acc	−0.39	−0.06				−0.14	−0.04				−0.06	−0.05
Step 5			.73	.51	179.60***			.49	.22	61.75***	N/A	N/A	N/A	N/A	N/A
Gender	0.55	0.01				2.16*	0.07
Race	−1.92	−0.03				−2.23*	−0.06
Age	−0.05	−0.03				−0.10**	−0.09
Income	0.02	0.00				−0.15	−0.03
Fin. Stress	1.81***	0.08				1.48***	0.11
Act. Stress	0.13	0.01				0.89***	0.14
Inf. Stress	0.48	0.03				0.73**	0.08
ES	0.27	0.03				0.36*	0.05
Fin. × ES	0.08	0.01				0.01	0.00
Act. × ES	0.04	0.01				0.07	0.03
Inf. × ES	−0.20	−0.03				−0.09	−0.02
BL DASS/PDI	0.82***	0.80				1.05***	0.53

Note. DASS = Depression Anxiety Stress Scales; PTSS = Posttraumatic Stress Symptoms; PSOM = Positive States of Mind; Acc = Acceptance; Fin. = Financial-Related Stressors; Act. = Activity-Related Stressors; Inf. = Infection-Related Stressors; BL = Baseline; PDI = Peritraumatic Distress. Race (0 = Non-white; 1 = White) and gender (0 = Female; 1 = Male) were analyzed as dichotomous variables. Age, income, stressors, and coping were centered at their means. Complete table including coefficients and standard errors for all steps can be found in the Supplement (Table S4).

^*p < .05.

^**p < .01.

^***p <.001.

Demographics and stressor types as predictors (steps 1 and 2):

Demographic variables (age, gender, race, and income) accounted for a significant amount of variance for each of the mental health variables (DASS: 9.20%; PTSS: 6.65%; PSOM: 10.34%, ps <.001). Younger age and lower income significantly predicted higher DASS and PTSS scores, while older age and higher income significantly predicted higher PSOM score. Race and gender were not significantly predictive of any of the mental health variables with the exception of non-White race and higher PTSS score. When stressor types were introduced in the second model, the total variance accounted for in the overall models remained significant (DASS: 21.84%; PTSS: 27.06%; PSOM: 15.48%, ps <.001). All three stressor types significantly predicted each of the mental health variables, with the exception of activity-related stressors and PSOM score (Table 2).

Coping strategy, interactions of stressor types and coping strategy, and baseline distress as predictors (steps 3–5):

Behavioral disengagement coping.

In the third model, behavioral disengagement coping was introduced; the total variance explained by the model as a whole was significant for each of the mental health variables (DASS: 42.01%; PTSS: 46.18%; PSOM: 22.87%, p <.001). Behavioral disengagement coping significantly positively predicted DASS and PTSS and negatively predicted PSOM. In the fourth model, interactions between behavioral disengagement coping and stressor types were introduced and the total variance accounted for in the overall models remained significant (DASS: 42.47%; PTSS: 46.82%; PSOM: 25.14%, ps <.001).

There was a significant interaction between activity-related stressors and behavioral disengagement coping on both DASS and PTSS scores (Figures 1 and 2). At high levels of activity-related stressors, the PTSS and general distress scores across the levels of behavioral disengagement differed the most, and those with high levels of behavioral disengagement reported high mental distress scores. Johnson-Neyman results indicate that activity-related stressors were related to general distress when behavioral disengagement coping was at mean (p = .02) and one standard deviation above mean (p < .001), but not one standard deviation below the mean (p = .33). Infection-related stressors were related to PTSS at all levels of behavioral disengagement (p <.001). A significant interaction was also found between infection-related stressors and behavioral disengagement coping on PTSS score (Figure 3). At low levels of infection-related stressors, PTSS scores across levels of behavioral disengagement differed the most, and those with high levels of behavioral disengagement reported high PTSS scores. Johnson-Neyman results indicate infection-related stressors were related to PTSS at all levels of behavioral disengagement (p <.003) The interaction of financial-related stressors and behavioral disengagement coping was not significant for any of the variables. Lastly, when baseline distress scores were added into the fifth model, the total variance explained for the overall models was significant (DASS: 75.47%; PTSS: 60.74%, ps <.001) and the interaction between activity-related stressors and behavioral disengagement coping remained significant for DASS and PTSS (Table 2).

Figure 1.

Interaction of Activity-Related Stressors x Behavioral Disengagement Coping on DASS Score

Figure 2.

Interaction of Activity-Related Stressors x Behavioral Disengagement Coping on PTSS Score

Figure 3.

Interaction of Infection-Related Stressors x Behavioral Disengagement Coping on PTSS Score

Positive reappraisal coping.

In the third model, positive reappraisal coping was introduced and the total variance explained by the model as a whole was significant for each of the mental health variables (DASS: 21.98%; PTSS: 27.26%; PSOM: 21.14%, ps <.001). In the fourth model, the interactions between positive reappraisal coping and stressor types were introduced and the total variance accounted for in the overall models remained significant (DASS: 22.17%; PTSS: 27.62%; PSOM: 22.23%, ps <.001). None of the interactions were significant, with the exception of financial-related stressors and positive reappraisal coping on PSOM score (Figure 4). High positive reappraisal coping was most effective at increasing PSOM at high levels of financial-related stressors. Johnson-Neyman results indicate financial-related stressors were associated with PSOM when positive reappraisal coping was one standard deviation below mean (p <.001) and at mean (p = .01) but not one standard deviation above the mean (p = .94). Lastly, when baseline distress scores were added into the fifth model, the total variance for the overall models was significant (DASS: 72.73%; PTSS: 49.27%, ps <.001) and the interaction of activity-related stressors and positive reappraisal on PTSS score became significant (Table 3).

Figure 4.

Interaction of Financial-Related Stressors x Positive Reappraisal Coping on PSOM Score

Table 3:

Hierarchical regressions with positive reappraisal coping predicting mental health

	Positive Reappraisal Coping
	DASS					PTSS					PSOM
Steps of the hierarchical regression models	B	β	R2	ΔR2	F Statistics	B	β	R2	ΔR2	F Statistics	B	β	R2	ΔR2	F Statistics
Steps 1 & 2
See Table 2
Step 3 (Coping Strategy)	See Table S1	See Table S1	.22	.00	25.59***	See Table S1	See Table S1	.27	.00	32.26***	See Table S1	See Table S1	.21	.06	21.04***
Step 4			.22	.00	20.94***			.28	.00	F26.75***			.22	.01	285.19***
Gender	−0.54	−0.01				0.18	0.01				0.64	0.07
Race	−2.05	−0.04				−2.51*	−0.07				0.35	0.03
Age	−0.39***	−0.22				−0.19***	−0.16				0.07***	0.19
Income	−1.12***	−0.13				−0.41*	−0.07				0.37***	0.22
Fin. Stress	3.98***	0.18				2.52***	0.18				−0.48**	−0.12
Act. Stress	1.13**	0.11				1.42***	0.22				−0.19*	−0.10
Inf. Stress	2.69***	0.18				1.72***	0.18				−0.33**	−0.12
PR	−0.42	−0.04				0.30	0.05				0.49***	0.25
Fin. × PR	0.27	0.03				0.29	0.05				0.19**	0.11
Act. × PR	0.08	0.02				0.06	0.02				−0.06	−0.08
Inf. × PR	−0.25	−0.04				−0.14	−0.04				0.01	0.01
Step 5			.73	.51	179.55***	2.12*	0.07	.49	.22	62.33***	N/A	N/A	N/A	N/A	N/A
Gender	0.48	0.01				−2.23*	−0.06
Race	−2.03	−0.04				−0.09**	−0.08
Age	−0.04	−0.02				−0.16	−0.03
Income	0.02	0.00				1.41***	0.10
Fin. Stress	1.73***	0.08				0.83***	0.13
Act. Stress	0.11	0.01				0.81**	0.09
Inf. Stress	0.57	0.04				0.28	0.04
PR	0.05	0.01				0.09	0.02
Fin. × PR	0.21	0.02				0.17*	0.06
Act. × PR	0.09	0.02				−0.12	−0.03
Inf. × PR	−0.20	−0.03				1.05***	0.53
BL DASS/PDI	0.82***	0.80

Note. DASS = Depression Anxiety Stress Scales; PTSS = Posttraumatic Stress Symptoms; PSOM = Positive States of Mind; PR = Positive Reappraisal; Fin. = Financial-Related Stressors; Act. = Activity-Related Stressors; Inf. = Infection-Related Stressors; BL = Baseline; PDI = Peritraumatic Distress. Race (0 = Non-white; 1 = White) and gender (0 = Female; 1 = Male) were analyzed as dichotomous variables. Age, income, stressors, and coping were centered at their means. Complete table including coefficients and standard errors for all steps can be found in the Supplement (Table S2).

^*p < .05.

^**p < .01.

^***p <.001.

Seeking emotional social support coping.

In the third model, seeking emotional support coping was introduced and the total variance was significant for each of the mental health variables (DASS: 22.03%; PTSS: 28.84%; PSOM: 17.50%, ps <.001). In the fourth model, interactions between seeking emotional support coping and stressor types were introduced and the total variance accounted for in the overall models remained significant (DASS: 22.93%; PTSS: 29.29%; PSOM 18.83%, ps <.001). The interaction of infection-related stressors and seeking emotional support significantly predicted DASS scores (Figure 5). At low levels of infection-related stressors, high emotional support predicted higher general distress scores, while at high levels of infection-related stressors, high emotional support predicted lower general distress scores. Johnson-Neyman results indicated that infection-related stressors are associated with general distress when emotional support coping is one standard deviation below mean (p <.001) and at mean (p <.001), but not one standard deviation above mean (p = .11). The interaction of activity-related stressors and seeking emotional support was significant on both DASS and PSOM scores (Figures 6 and 7). At low levels of activity-related stressors, high emotional support predicted lower general distress scores, while at high levels of activity-related stressors, high emotional support predicted increased general distress score. Johnson-Neyman results indicated that activity-related stressors are associated with general distress when emotional support coping is at mean (p = .01) and one standard deviation above mean (p <.001), but not one standard deviation below mean (p = .92). Emotional support coping was most effective at increasing PSOM when individuals reported low activity-related stressors. Johnson-Neyman results indicated that activity-related stressors are associated with PSOM when emotional support coping one standard deviation above mean (p = .002), but not at mean (p = .13) or one standard deviation below mean (p = .38). The interactions of financial-related stressors and seeking emotional support coping were not statistically significant. Lastly, when baseline distress scores were added into the fifth model, the total variance explained for the overall models were significant (DASS: 72.81%; PTSS: 49.20%, ps <.001), but no interactions remained significant (Table 4).

Figure 5.

Interaction of Activity-Related Stressors x Seeking Emotional Support Coping on DASS Score

Figure 6.

Interaction of Infection-Related Stressors x Seeking Emotional Support Coping on DASS Score

Figure 7.

Interaction of Activity-Related Stressors x Seeking Emotional Support Coping on PSOM Score

Table 4:

Hierarchical regressions with seeking emotional social support coping predicting mental health

	Emotional Social Support Coping
	DASS					PTSS					PSOM
Steps of the hierarchical regression models	B	β	R2	ΔR2	F Statistics	B	β	R2	ΔR2	F Statistics	B	β	R2	ΔR2	F Statistics
Steps 1 & 2
See Table 2
Step 3 (Coping Strategy)	See Table S1	See Table S1	.22	.00	28.68***	See Table S1	See Table S1	.29	.02	39.21***	See Table S1	See Table S1	.18	.02	16.62***
Step 4			.23	.01	21.88***			.29	.01	29.04***		0.07	.19	.01	13.18***
Gender	−0.89	−0.02				0.22	0.01				0.70*	0.02
Race	−1.80	−0.03				−2.41*	−0.07				0.23	0.18
Age	−0.40***	−0.22				−0.19***	−0.16				0.06***	0.22
Income	−1.18***	−0.13				−0.43*	−0.08				0.38***	−0.13
Fin. Stress	4.08***	0.19				2.49***	0.18				−0.54**	−0.06
Act. Stress	0.97*	0.09				1.37***	0.21				−0.12	−0.13
Inf. Stress	2.50***	0.17				1.48***	0.16				−0.36**	0.14
ES	1.25	0.04				2.43***	0.14				0.77***	0.07
Fin. × ES	−0.58	−0.02				0.02	0.00				0.32	−0.13
Act. × ES	1.02*	0.09				0.53	0.07				−0.27**	0.07
Inf. × ES	−1.46*	−0.08				−0.46	−0.04				0.24
Step 5			.73	.50	180.32***			.49	.20	62.14***	N/A	N/A	N/A	N/A	N/A
Gender	0.44	0.01				2.10*	0.07
Race	−2.00	−0.03				−2.23*	−0.06
Age	−0.04	−0.02				−0.09**	−0.08
Income	−0.00	0.00				−0.17	−0.03
Fin. Stress	1.87***	0.09				1.49***	0.11
Act. Stress	0.07	0.01				0.85***	0.13
Inf. Stress	0.40	0.03				0.71*	0.07
ES	1.14*	0.04				1.11*	0.06
Fin. × ES	−0.65	−0.03				−0.33	−0.02
Act. × ES	0.31	0.03				0.35	0.05
Inf. × ES	−0.33	−0.02				−0.11	−0.01
BL DASS/PDI	0.82***	0.80				1.02***	0.52

Note. DASS = Depression Anxiety Stress Scales; PTSS = Posttraumatic Stress Symptoms; PSOM = Positive States of Mind; ES = Emotional Support; Fin. = Financial-Related Stressors; Act. = Activity-Related Stressors; Inf. = Infection-Related Stressors; BL = Baseline; PDI = Peritraumatic Distress. Race (0 = Non-white; 1 = White) and gender (0 = Female; 1 = Male) were analyzed as dichotomous variables. Age, income, stressors, and coping were centered at their means. Complete table including coefficients and standard errors for all steps can be found in the Supplement (Table S3).

^*p < .05.

^**p < .01.

^***p <.001.

Acceptance coping.

In the third model, acceptance coping was introduced and the total variance was significant for each of the mental health variables (DASS: 21.85%; PTSS: 27.36%; PSOM: 17.70%, ps < .001). The fourth model introduced the interactions between acceptance coping and stressor types and the total variance accounted for in the overall models remained statistically significant (DASS: 22.20%; PTSS: 27.48%; PSOM: 18.22%, ps <.001). None of the interactions were significant. Lastly, when baseline distress scores were added into the fifth model, the total variance for the overall models were significant (DASS: 72.73%; PTSS: 49.04%, ps <.001), but interactions remained non-significant (Table 5).

Discussion

This study advances our understanding of the ways that different types of stressors may influence mental health following protracted community-wide traumas and the roles that different types of coping play within them. To our knowledge, this is the first study to examine the impact of different stressors simultaneously arising from a pervasive stressor (i.e., the pandemic) on mental health in the US; thus, these results contribute to a fuller understanding of the interplay of stressors and coping and mental health symptoms. Given the dearth of research on these topics, our hypotheses were made tentatively; results generally supported our expectations.

The first hypothesis, that differential mental health symptom presentations will emerge based on exposure to specific types of stressors, was not supported. Nearly all stressor types significantly predicted each dimension of mental health. Even though the stressor types are related, they each carried weight in predicting each aspect of mental health, which is consistent with some prior research (e.g., Kira et al., 2021). However, other previous studies have reported differing results, indicating that some types of stressors are more likely to negatively impact mental health than are others (e.g., financial stressors were not related to mental health in college students; Hubbard et al., 2018). Our results indicate that all three stressor types independently predicted each aspect of mental health, reflecting that exposure types each carry their own weight and have a cumulative impact. This finding has clinical implications for screening and assessment as it highlights the importance of assessing different domains of stressors to get a fuller sense of stressors to predict approximate risk or resilience in developing mental health symptoms.

The second hypothesis, that coping strategies will have significant main effects on positive and negative aspects of mental health was partially supported. As expected, behavioral disengagement was inversely associated with all three measured aspects of mental health, consistent with prior studies (Lehavot, 2012; Turner-Cobb et al., 2002). Perhaps more surprisingly, seeking emotional support was associated with both higher PTSS and PSOM. Additionally, positive reappraisal and acceptance coping predicted PSOM, but not distress.

The third hypothesis, that coping strategies will differentially moderate links between specific stressor types and mental health, was generally supported. The results are in line with theories of transactional stress and coping, suggesting that one’s responses to a stressful event depend on interactions among environmental, contextual, and individual factors (Lazarus & Folkman, 1987). The overall pattern of results indicated that although some types of coping (e.g., acceptance) were not shown to be effective in any situation, no one coping strategy is effective across all situations, and, rather, the effects of any coping strategy depends on situational determinants of stress. Generally, when faced with stressors that may be unprecedented or feel more uncontrollable (i.e., activity- and infection-related stressors as opposed to financial-related stressors), behavioral disengagement coping tended to exacerbate distress while seeking emotional support largely was a protective coping strategy (not only in reducing distress but also in increasing PSOM). Positive reappraisal and acceptance largely did not seem to be effective coping strategies when faced with any type of stressor. None of the coping strategies examined reduced distress under financial-related stressors, indicating that when faced with those types of stressors, perhaps a more active form of coping is warranted. Detailed results for each type of coping are reported below.

Behavioral Disengagement

Though relationships were small, behavioral disengagement significantly buffered links between activity-related stressors and distress. Since activity-related stressors (e.g., disruptions in daily routines) likely decrease individuals’ engagement in their typical responsibilities, it is logical that coping by disengaging further increases the likelihood of heightened distress. Behavioral disengagement exacerbated the adverse effects of stressors, although for general distress, this exacerbation occurred only at average or higher levels of behavioral disengagement. Taken together, these findings indicate even a small amount of behavioral disengagement coping can have a negative impact. Behavioral disengagement not only was associated distress cross-sectionally, but also predicted increased distress from T1 to T2, warranting future research to explore these questions over a greater timespan to understand long-term trajectories and changes in mental health symptoms. Behavioral disengagement (regardless of the severity) also exacerbated the association between infection-related stressors and PTSS. Though the relationship size was small, this finding makes sense given that there are few active ways to cope with the uncertainty of the disease and utilizing behavioral disengagement (i.e., giving up attempts to cope) can result in COVID-19-specific distress (Kalaitzaki et al., 2022). Due to the uncertainty of the disease, perhaps high enough levels of infection-related stressors were predictive of PTSS on their own, while at lower levels, behavioral disengagement coping would have less of a relative influence.

Seeking Emotional Support

Seeking emotional support predicted higher PTSS and PSOM. That seeking social support was positively predictive of PTSS is perhaps surprising since some other studies have shown seeking social support to be a protective factor for PTSS (e.g., Dirkzwager et al., 2003). However, coping through seeking social support has been shown to have mixed efficacy (e.g., Main et al., 2011). Individuals who experienced more PTSS may have more actively sought social support more than did those with fewer symptoms and/or may not have received adequate support (Okafor et al., 2022). Seeking emotional support buffered links between infection-related stressors and general distress. At low levels of infection-related stressors, high emotional support-seeking predicted higher general distress while at high levels, it predicted lower general distress. Perhaps when exposed to low levels of infection-related stressors, discussing the uncontrollable aspects of the disease with others can increase general distress (Palant & Himmel, 2019), while at high levels of infection-related stressors, anxiety surrounding the uncertainty of the disease is already very high, so seeking emotional support from others may reduce general distress and provide comfort. Average levels of seeking emotional support were found to buffer general distress, but once seeking emotional support reached a certain level, it was no longer protective. This finding indicates that some emotional support is helpful, but above-average social support seeking may not be effective at reducing general distress when faced with infection-related stressors. Seeking emotional support also buffered links between activity-related stressors and both general distress and PSOM. Similarly, seeking emotional support is effective as a buffer only to a certain extent – as activity-related stressor exposure increased, seeking emotional support was not as effective. This finding was surprising because as individuals lose the ability to participate in activities (such as missing celebrations and trips), seeking emotional social support in other ways might be expected to be beneficial to combat social isolation (Xu et al., 2020). However, we found just the opposite. Perhaps at high levels of activity-related stressors, it is not possible to participate in social activities due to pandemic limitations, which can cause frustration, and thus seeking emotional support is less satisfying and increases general distress. It is also possible that those seeking emotional support did not receive adequate support since others are experiencing their own pandemic-related stress, which could, in turn, increase distress in the support seeker (Okafor et al., 2022). These findings are only significant when levels of seeking emotional support were average or above average, indicating that seeking emotional support needs to reach a certain threshold in order to be an effective buffer against lower levels of activity-related stressors. Additionally, only above average levels of seeking emotional support significantly buffered the relationship between activity-related stressors and PSOM, indicating that emotional support needs to be sought to a greater extent in order to effectively facilitate PSOM. Taken together, seeking emotional support generally has a positive impact on distress, but specific effects varied by amount of social support seeking.

Positive Reappraisal and Acceptance

Positive reappraisal and acceptance coping predicted PSOM, but not distress, indicating that these types of coping may be more useful at increasing positive wellbeing. Prior studies have found mixed results on the efficacy of positive reappraisal and acceptance coping (Kraaij et al., 2002), with some studies reporting results consistent with our findings (e.g., Nowlan et al., 2016) while, other studies reporting that acceptance and positive reappraisal were effective at reducing distress (Charlton & Thompson, 1996; Saxon et al., 2017) as well as increasing positive emotion (Folkman & Moskowitz, 2000).

In a time of high uncertainty where available coping strategies are limited (i.e., active coping is sometimes not possible when faced with uncontrollable stressors; Ding et al., 2021), it would make sense that accepting and reappraising the situation could be beneficial strategies. However, these types of coping did not buffer effects of stressors on distress. The magnitude of the pandemic may have been so great that accepting the situation or gaining a positive view was not always achievable.

Implications

The results indicate that when faced with certain COVID-19 related stressors, it may be more effective to seek emotional social support and avoid behavioral disengagement, as positive reappraisal and acceptance were not shown to reduce poor mental health outcomes. Seeking emotional support during a pandemic can be challenging, so it is important to find creative ways to engage with others (e.g., socially distanced walks, connecting electronically; Freyhofer et al., 2021). Ways to avoid behavioral disengagement could perhaps be targeted by methods such as behavioral activation (i.e., increasing engagement with pleasurable activities; Carona et al., 2014), which may make coping by giving up less likely. Results indicate that coping strategies generally considered adaptive (e.g., acceptance and positive reappraisal) are not always effective at reducing distress, findings consistent with prior studies that have found mixed effects of adaptative coping (Freyhofer et al., 2021; Mahmoud et al., 2012). In terms of clinical implications, these findings suggest that clinicians may want to focus on facilitating certain coping strategies (e.g., seeking emotional support and behavioral engagement) rather than others (e.g., acceptance and reappraisal) with patients. These coping strategies may be more or less effective based on the type of stressors experienced and thus, clinicians may explore the different domains of stress origin to determine appropriate modifiable treatment targets.

Strengths and Limitations

One limitation of this study is that it focused largely on cross-sectional data (at the second time point). Though baseline distress was included in the final steps of the regressions for general distress scores, no baseline scores were available for the PSOM regressions since that measure was introduced to the survey at T2. Additionally, PDI scores were used in the PTSS analyses because the PTSS measure was also introduced to the survey at T2. Asking similar questions longitudinally using the same measures at all timepoints would be valuable. Another limitation is that participants were not asked what types of coping they used with each type of stressor individually, but rather what types of coping they used overall. Additionally, controlling for baseline pre-pandemic levels of stress would have been ideal, but was not possible as data collection began after the start of the pandemic. Lastly, the internal reliability of the financial stressor subscale was relatively low, and thus, those results should be interpreted cautiously.

Despite these limitations, these results provide directions for future research. One avenue is to ask how individuals are coping with each specific stressor type to assess the efficacy of specific coping strategies on stressor types. Another is to ask similar questions in the context of other broad stressors (e.g., disasters, chronic illness) that encompass many types of smaller stressors (Lock et al., 2012). Future research could also examine other types of stressors (e.g., political stressors) as well as other dimensions of mental health (e.g., examining anxiety, depression, and stress separately rather than combining them as general distress).

Conclusions

The present study provides a foundation for this future research by advancing understanding of stress processes. First, the results of this novel study support the notion that different stressor types independently predict each mental health outcome, underscoring the importance of continuing to research the role of different types of stress exposure rather than simply overall exposure to a major stressor to better evaluate cumulative impact. Second, coping strategies differentially moderate the relationship between stressor type and mental health, indicating that the effectiveness of certain types of coping strategies depend on the type of stress exposure. This finding has treatment implications as targeting promotion and avoidance of certain types of coping through intervention should perhaps be tailored based on the type of stress exposure experienced. Overall, these findings provide preliminary evidence that stressor types and coping strategies are key constructs to consider when evaluating mental health outcomes and resilience factors, thus providing promising directions for future studies.