Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Self-quarantining, social distancing, and mental health during the COVID-19 pandemic: A multi wave, longitudinal investigation

Abstract

Social isolation and disconnectedness increase the risk of worse mental health, which might suggest that preventive health measures (i.e., self-quarantining, social distancing) negatively affect mental health. This longitudinal study examined relations of self-quarantining and social distancing with mental health during the COVID-19 pandemic. A U.S. national sample (N = 1,011) completed eight weekly online surveys from March 20, 2020 to May 17, 2020. Surveys assessed self-quarantining, social distancing, anxiety, and depression. Fixed-effect autoregressive cross-lagged models provided a good fit to the data, allowing for disaggregation of between-person and within-person effects. Significant between-person effects suggested those who engaged in more self-quarantining and social distancing had higher anxiety and depression compared to those who engaged in less social distancing and quarantining. Significant within-person effects indicated those who engaged in greater social distancing for a given week experienced higher anxiety and depression that week. However, there was no support for self-quarantining or social distancing as prospective predictors of mental health, or vice versa. Findings suggest a relationship between mental health and both self-quarantining and social distancing, but further longitudinal research is required to understand the prospective nature of this relationship and identify third variables that may explain these associations.

Citation: Lee J, Wilson J, Oosterhoff B, Shook NJ (2024) Self-quarantining, social distancing, and mental health during the COVID-19 pandemic: A multi wave, longitudinal investigation. PLoS ONE 19(2): e0298461. https://doi.org/10.1371/journal.pone.0298461

Editor: Lakshit Jain, University of Connecticut Health Center: UConn Health, UNITED STATES

Received: October 20, 2023; Accepted: January 24, 2024; Published: February 26, 2024

This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.

Data Availability: Preregistration information and all research materials are available from the Open Science Framework (https://osf.io/56fcv?view_only=002af22544b547c387624f726486311b) (https://osf.io/s9uxv/?view_only=290032cc1ec94381981fda8858ee1180).

Funding: This work was supported by a RAPID grant from the National Science Foundation (https://www.nsf.gov/) [BCS-2027027] awarded to NS. The sponsor/funder did not play any role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

The Coronavirus disease 2019 (COVID-19) pandemic has negatively impacted both physical and mental health. Although effective containment of highly communicable diseases, such as COVID-19, largely depends on limiting social contact [1], there are concerns that such preventive health measures (i.e., self-quarantining, social distancing) may negatively affect mental health, increasing anxiety and depression [2–4]. Though these measures are effective at reducing transmission and illness from COVID-19 [5], self-quarantining and social distancing may prompt feelings of disconnectedness and social isolation [2]. Furthermore, physically distancing from one another may increase opportunities for anxious rumination [6]. Social connectedness benefits mental health [7], whereas social isolation negatively affects health [8]. However, little empirical research has longitudinally examined the proposed relations of self-quarantining and social distancing with mental health. Thus, the goal of the present study was to utilize a multi-wave, longitudinal study design to determine the directional associations of self-quarantining and social distancing with symptoms of anxiety and depression during the COVID-19 pandemic.

Self-quarantining

Humans are social animals [9], and decades of research link social isolation with worse mental health outcomes [10]. Thus, it may not be surprising that public health measures encouraging social isolation may threaten mental well-being. Self-quarantining policies encourage individuals who have contracted COVID-19 or who may have been exposed to the virus to self-isolate and stay away from other people for several days to monitor symptoms and/or recover [11]. Indeed, some research lends support for an association between self-quarantining and poorer mental health. A review of studies examining the effects of isolation on mental health among individuals quarantined due to infectious diseases (e.g., Middle East respiratory syndrome, Ebola) found positive associations between quarantining and depression, anxiety, irritability, and post-traumatic stress symptoms [12].

In the context of the COVID-19 pandemic, mass-level quarantining (e.g., quarantining enforced at the national level) has been associated with symptoms of anxiety and depression. A meta-analysis of cross-sectional, longitudinal, and experimental studies demonstrated a small positive effect for the association between mass-level quarantine duration and symptoms of anxiety and depression [13]. Further, a meta-analysis of cross-sectional and longitudinal studies examining the impact of national lockdowns (i.e., requirement of citizens to stay home) concluded that the negative effect of lockdown measures on anxiety and depression in the general population is small in magnitude [14]. The results of both meta-analyses challenge claims that COVID-19 lockdowns may have had a considerable negative effect on population mental health. However, both meta-analyses focused on mass quarantine measures (i.e., mass-level quarantine duration, assessment of mental health before and after lockdown orders, comparing regions with and without lockdown orders), which limits conclusions regarding the individual-level self-quarantining (i.e., isolating oneself from others to recover from and/or avoid spreading COVID-19).

Regarding self-quarantining at the individual level, findings are mixed and few. Cross-sectional research conducted in various countries (e.g., Albania, Argentina, China, Hong Kong) demonstrated that self-quarantining was positively associated with anxiety and/or depressive symptoms [15–17]. However, other cross-sectional studies conducted in the U.S. suggest self-quarantining is not significantly associated with anxiety and depressive symptoms [18, 19]. Further, in a recent two-wave longitudinal study that assessed three independent U.S. samples, self-quarantining was not associated with changes in anxiety and depressive symptoms from pre-pandemic to during the pandemic [20]. To date, literature regarding associations between self-quarantining during COVID-19 and mental health is limited to mixed findings from cross-sectional and two-wave longitudinal studies.

Social distancing

Social distancing involves maintaining a distance of 6 feet from others and avoiding crowded and poorly ventilated establishments [21]. Social distancing may disrupt social processes that are often beneficial to mental health, such as day-to-day interactions and the availability of social support [22]. Although quarantining has been studied in the context of past pandemics, social distancing policies are unique to the COVID-19 pandemic, and existing evidence regarding the link between social distancing and mental health is mixed. Cross-sectional studies demonstrated that greater social distancing was associated with worse anxiety and/or depressive symptoms in samples recruited from Germany [23] and the U.S. [22, 24, 25]. Yet, other cross-sectional research indicates that greater social distancing was associated with less anxiety and depressive symptoms in a sample from Hong Kong [17]. Moreover, some cross-sectional [18, 19] and two-wave longitudinal [20] studies demonstrated that social distancing is not significantly associated with anxiety and depressive symptoms in U.S. adults. Such findings are inconsistent with predictions that physically distancing from one another would be associated with poorer mental health [4, 26].

Methodological limitations of existing research

There are several methodological limitations associated with existing studies examining the association between self-quarantining, social distancing, and mental health. First, the extant literature is primarily limited to cross-sectional (single time point) study designs. Cross-sectional research can provide between-person data that speak to associations across a set of individuals; however, conclusions about fluctuations in variables within an individual (e.g., are changes in self-quarantining associated with changes in anxiety within a person) cannot be made [27].

Second, two-wave longitudinal designs are prone to confounding variables and little can be concluded about how change occurs [28] or the directionality of associations. From two time points, it cannot be discerned which variable changes first or whether there is a bidirectional association. Indeed, it is possible that self-quarantining or social distancing may have a bidirectional relationship with mental health, such that those with higher levels of anxiety and/or depression are more likely to engage in social distancing and self-quarantining. Theoretically, avoidance is a key component of anxiety [29]. Thus, highly anxious individuals who are concerned about contracting COVID-19 may be more likely to engage in preventive behaviors. Consistent with this notion, Velikonja and colleagues [30] found that anxiety was associated with a broad measure of preventive measures that included social distancing. Furthermore, theories of depression suggest that behavioral withdrawal is a core component of depression that may serve to regulate mood [31] or provide short-term relief from potentially aversive experiences [32]. Thus, people experiencing high levels of depression may be more prone to engaging in behavioral withdrawal, such as self-quarantining or social distancing. No study to date has examined the potentially reciprocal prospective effects between self-quarantining, social distancing, and mental health.

Third, existing research largely focused on mass-level quarantining rather than individual-level engagement, which limits understanding of how self-quarantining and mental health vary within an individual. To address these limitations, repeated measurement from individuals over at least three time points is recommended [33]. Thus, given the inconsistencies in the literature and the infrequent nature of pandemics, more multi-wave longitudinal studies are necessary to provide a better understanding of how self-quarantining and social distancing may be associated with mental health to better inform interventions for future pandemics.

Present study

On March 11, 2020, the World Health Organization declared COVID-19 a pandemic [34], prompting many countries, including the United States, to implement self-quarantining and social distancing policies to reduce the spread of the disease. On March 15–16, 2020, states across the U.S. shut down schools and other public areas (e.g., restaurants, bars) and the White House issued social distancing guidelines to prevent the spread of COVID-19 [34, 35]. On April 24, 2020, some U.S. states began to partially reopen [34]. The goal of the present study was to assess fluctuations in variables between and within individuals, and longitudinally assess the extent to which self-quarantining and social distancing prospectively predicted mental health (i.e., anxiety and depression), and vice versa, over eight weeks in a national sample of U.S. adults from March 20, 2020 to May 17, 2020. This timeframe is uniquely marked by the introduction of self-quarantining and social distancing policies in March 2020 and the loosening of these policies by May 2020. Thus, we would expect this novel and transitional period (as opposed to a period where policies have been in place for longer) to be associated with changes in self-quarantining, social distancing, and psychological adjustment.

We expected greater days of quarantining per week to be associated with greater anxiety and depressive symptoms. Although there are mixed findings regarding social distancing, we hypothesized that greater social distancing per week would be associated with worse mental health due to the shared feature of social disconnection present in both self-quarantining and social distancing. Furthermore, given avoidance and behavioral-based theories of anxiety and depression, respectively, we hypothesized that greater anxiety and depression would predict greater engagement in self-quarantining and social distancing.

Materials and methods

Participants and procedure

Participants were recruited through the panel provider Qualtrics for a larger 29-wave longitudinal study about the effects of COVID-19. For the purposes of the present study, only the first eight waves were examined. A minimum sample of 500 was estimated using Monte Carlo simulations (N  =  10,000) for the larger study to provide sufficient power (>95%) to detect anticipated effects (β = .15 to .20) assuming α = .05. To account for missing or unusable data, at least 1,000 U.S. individuals were recruited. Data were collected every week for eight weeks from a national sample of 1,213 U.S. residents from March 20, 2020 to May 17, 2020. Two hundred two participants were excluded due to atypical response patterns (e.g., open-ended responses unrelated to the prompt, straightline responses) or failure to complete at least 60% of either the first survey or primary study variables. It was not possible to meaningfully examine group differences between participants that were included and excluded in the present analyses because all of the 202 excluded participants discontinued the survey before providing demographic variables. Additionally, primary study variable data were only available for 1 to 51 participants. The final sample comprised 1,011 participants (51% women, Mage = 46.6 years, SDage = 16.5, range: 18–85 years; 76.1% White, 7.1% Asian, 6.2% Black, 3.7% Latinx, 3.4% more than one race, 0.6% “Other,” 0.4% Native American; Mdneducation = College graduate; Mdnincome = $70,000-$79,999).

Institutional Review Board (IRB) approval was obtained from the first author’s institution (Protocol #L20-0018). As the study was conducted online and anonymously, written consent was not feasible and consent was requested electronically. The IRB approved use of electronic consent. Participants read and acknowledged an online consent form and indicated their consent by clicking on an "Agree" radio button. Those who clicked "Disagree" were automatically exited out of the Qualtrics survey. Participants completed primary study measures and other questionnaires in a random order, with one exception. Demographics were assessed at the end of the Wave 1 survey. Self-quarantining was not assessed at Wave 1 but was included in the Wave 2 through 8 surveys. Surveys were sent out on a weekly basis, and each online survey took approximately 25 minutes. Participants were provided monetary compensation in an amount established by the panel provider. Preregistration information and research materials (e.g., data, code) are publicly available (https://osf.io/56fcv?view_only=002af22544b547c387624f726486311b and https://osf.io/s9uxv/?view_only=290032cc1ec94381981fda8858ee1180).

Measures

A list of measures that were included as part of the larger study is in S1 Appendix. The present study focused on the measures below.

Self-quarantining.

Starting with the second wave, participants indicated (“Yes” or “No”) whether they self-quarantined (i.e., stay at home to avoid contracting or spreading COVID-19) over the past week [20]. For individuals who responded “Yes,” they were asked to indicate how many days in the past week they had self-quarantined (1 to 7 days). A continuous variable was created by coding “No” responses as 0 and utilizing the number of days self-quarantined in the past week for “Yes” responses.

Social distancing.

Six items assessed social distancing [36]. Participants indicated the extent to which they engaged in five items related to avoiding contact with others (e.g., ‘Avoid going to school/job,’ ‘Avoid hugging someone for greeting’) within the past week on a scale ranging from 1 (Not at all) to 6 (Multiple times a day). Additionally, participants reported the extent to which they engaged in social distancing in the past week on a scale ranging from 1 (Not at all) to 5 (A great deal). This measure of social distancing has demonstrated acceptable reliability in prior research (α = .76) [36]. All items were standardized and averaged to create a composite score. Higher scores indicate more social distancing. This measure demonstrated acceptable reliability across all eight waves within our sample (αs = .75 –.78).

Anxiety symptoms.

The Generalized Anxiety Disorder screener (GAD-7) [37] is a 7-item measure of anxiety. Participants indicated their experience of anxiety symptoms (e.g., ‘Feeling nervous, anxious, or on edge’) over the past two weeks for the first wave and over the past week for subsequent waves using a scale ranging from 0 (Not at all) to 3 (Nearly every day). Scores are summed and range from 0–21, with the following severity ratings: minimal to no anxiety symptoms (0–4), mild (5–9), moderate (10–14), and severe (15+). Scores of ≥10 suggest presence of GAD. The GAD-7 has demonstrated good reliability in clinical (α = .92) [37] and community-based samples (α = .89) [38]. Prorated scores were computed to account for individuals who missed two or fewer items [39]. The GAD-7 demonstrated strong reliability across all eight waves within our sample (αs = .95 –.96).

Depressive symptoms.

The abbreviated 8-item version of the Patient Health Questionnaire (PHQ-9) [40] assessed depressive symptoms. One item about suicidal ideation was excluded from this study. Participants indicated their experience of depressive symptoms (e.g., ‘Feeling down, depressed, or hopeless’) over the past two weeks for the first wave and over the past week for subsequent waves using a scale ranging from 0 (Not at all) to 3 (Nearly every day). Scores are summed and range from 0–24, with the following severity ratings: minimal (0–4), mild (5–9), moderate (10–14), moderately severe (15–19), and severe (20–24). Scores of ≥10 suggest presence of major depression. The PHQ-9 has demonstrated strong internal reliability (α = .89) in prior research [40]. Prorated scores were computed to account for individuals who missed two or fewer items [41]. The abbreviated PHQ-9 demonstrated strong reliability across all eight waves within our sample (αs = .93 –.94).

Data analytic plan

Fixed-effect autoregressive cross-lagged (FACL) structural equation models were used to examine the extent to which self-quarantining, social distancing, anxiety, and depression covary within individuals over time. FACL models can isolate within-person effects from between-person effects with cross-lagged data [27]. Between-person effects provide information regarding how variables may differ between individuals (e.g., Do people who engage in more self-quarantining report greater depressive symptoms relative to those who engaged in less self-quarantining?). Within-person effects provide information regarding how variables change within an individual over time (e.g., Is engaging in more self-quarantining one week compared to one’s average level of self-quarantining associated with greater anxiety during the same week compared to one’s average level of anxiety?). Lastly, cross-lagged effects provide information regarding the directional effect of two variables across time (e.g., Does self-quarantining prospectively predict greater depressive symptoms or vice versa?).

Latent intercepts were fixed for each construct, which specified mean scores for week-level self-quarantining, social distancing, and mental health. Autoregressive and cross-lagged paths were specified among the residuals of the weekly indicators. Consistent with prior research [27], the variances of the residuals were constrained to be equal. For parsimony, we also constrained each autoregressive and cross-lagged effect to be equal given that we did not expect differences in these estimates across weeks. Covariances were specified across constructs for the same week and were constrained to be equal to indicate weekly effects. A covariance was also specified across intercepts for the latent variables and was used to indicate an overall between-person effect.

Attrition varied across waves, as individuals could complete surveys even if they missed a particular week. Attrition (n) at each wave was as follows: Wave 1 (0), Wave 2 (190), Wave 3 (237), Wave 4 (262), Wave 5 (287), Wave 6 (317), Wave 7 (360), and Wave 8 (351). Little’s test of Missing Completely at Random (MCAR) for each of the eight waves was not significant, ps = .16 - .78, indicating that there was no evidence supporting the hypothesis that the data were not MCAR. Thus, missing data were assumed to be missing at random and estimated using multiple imputation [42]. Analyses were conducted with imputed and non-imputed data. All non-imputed data results are in S2, S3 and S5–S7 Tables.

Analyses were conducted using the lavaan package [43] in RStudio. Given that chi-square tests often inflate Type I error rates in large samples [44], acceptable model fit was evaluated with the Comparative Fit Index (CFI) ≥ .95, Tucker-Lewis Index (TLI) ≥ .95, Root Mean Square Error of Approximation (RMSEA) ≤ .06, and Standardized Root Mean Square Residual (SRMR) < .08 [45].

Results

Means and standard deviations are presented in Table 1. Within this sample, clinical cutoffs for generalized anxiety and major depression were calculated. At Wave 1, 27.2% of participants reported clinically significant anxiety, followed by 23.6%, 23.3%, 24.1%, 20.6%, 19.1%, 16.9%, and 18.5% for subsequent waves. Additionally, 25.6% of participants reported clinically significant depression at Wave 1, followed by 25.0%, 25.7%, 21.4%, 22.0%, 22.5%, 19.8%, and 19.7% for subsequent waves. Data regarding anxiety and depression severity are provided in S1 and S2 Tables.

thumbnail
Download:
Table 1. Means (standard deviations) for repeated-measures ANOVAs.

https://doi.org/10.1371/journal.pone.0298461.t001

Changes over time

To examine changes in self-quarantining, social distancing, anxiety, and depressive symptoms over time, four repeated-measure ANOVAs were conducted (Fig 1 and Table 1). For each ANOVA, time (seven or eight weekly time points) was the within-subjects variable. Repeated measures ANOVA indicated a main effect of time for self-quarantining (F(5.43, 5483.19) = 14.43, p < 0.001, ηp2 = 0.014), social distancing (F(6.45, 6509.84) = 2.84, p = 0.008, ηp2 = 0.003), anxiety symptoms (F(5.85, 5907.07) = 47.28, p < 0.001, ηp2 = 0.045), and depressive symptoms (F(6.19, 6249.17) = 19.77, p < 0.001, ηp2 = 0.019). Self-quarantining, anxiety symptoms, and depressive symptoms generally decreased over time. Levels of social distancing remained relatively consistent across the eight waves, except for Wave 7, where social distancing was relatively lower than other waves. Patterns of results were similar for the non-imputed data, except for social distancing such that the main effect of time for social distancing was not significant, F(6.36, 2703.26) = 1.13, p = 0.34, ηp2 = 0.003.

thumbnail
Download:
Fig 1. Visual representation of repeated measures ANOVA assessing changes in self-quarantining, social distancing, anxiety symptoms, and depressive symptoms.

Note. Y-axes reflect average scores. The social distancing composite reflects six items which were standardized and averaged to have a mean of 0 and standard deviation of 1.

https://doi.org/10.1371/journal.pone.0298461.g001

Correlations

Bivariate correlations assessed associations among key variables across all waves (Tables 2 and 3). Correlations for all variables are in S4 and S5 Tables. Greater self-quarantining was generally significantly associated with greater anxiety and depressive symptoms within and across all waves. Greater social distancing was significantly associated with greater anxiety symptoms and greater depressive symptoms within and across all waves. For non-imputed data, patterns were comparable to patterns resulting from imputed data except self-quarantining was not consistently significantly associated with anxiety and depressive symptoms.

thumbnail
Download:
Table 2. Bivariate correlations between preventive measures and symptoms of anxiety.

https://doi.org/10.1371/journal.pone.0298461.t002

thumbnail
Download:
Table 3. Bivariate correlations between preventive measures and symptoms of depression.

https://doi.org/10.1371/journal.pone.0298461.t003

Self-quarantining and anxiety symptoms

The model testing links between self-quarantining and anxiety symptoms provided a good fit to the data (Table 4). Table 5 provides model estimates. There were significant positive between-person associations, such that those who engaged in greater self-quarantining reported higher anxiety symptoms compared to those who engaged in less self-quarantining. There were no significant weekly within-person effects among self-quarantining and anxiety symptoms, indicating that those who engaged in greater self-quarantining for a given week relative to their own average across the study did not report greater anxiety symptoms that week relative to their own average across the study. There was also no evidence of significant cross-lagged effects for engagement in self-quarantining predicting anxiety symptoms the subsequent week, nor anxiety symptoms predicting engagement in self-quarantining the subsequent week. For non-imputed data, between-person, within-person, or cross-lagged effects were not significant for associations between self-quarantining and anxiety symptom.

thumbnail
Download:
Table 4. Fixed-effect autoregressive cross-lagged model fit statistics.

https://doi.org/10.1371/journal.pone.0298461.t004

thumbnail
Download:
Table 5. Standardized and unstandardized estimates for models.

https://doi.org/10.1371/journal.pone.0298461.t005

Self-quarantining and depressive symptoms

The model testing links between self-quarantining and depressive symptoms provided a good fit to the data (Table 4). Table 5 provides model estimates. There were significant positive between-person associations, such that those who engaged in greater self-quarantining reported higher depressive symptoms compared to those who engaged in less self-quarantining. There were no significant weekly within-person effects among self-quarantining and depressive symptoms, suggesting that those who engaged in greater self-quarantining for a given week relative to their own average across the study did not report greater depressive symptoms that week relative to their own average across the study. There was also no evidence of significant cross-lagged effects for self-quarantining predicting next week depressive symptoms. However, there was a significant cross-lagged effect for depressive symptoms predicting next week self-quarantining. Those who had higher depressive symptoms one week relative to their eight-week average were more likely to engage in self-quarantining the next week relative to their eight-week average. For example, if Person A had depressive symptoms in Week 3 that was higher than their average level of depressive symptoms across the eight-week study, then they were likely to engage in more self-quarantining in Week 4 than they usually would. For non-imputed data, between-person, within-person, or cross-lagged effects were not significant for associations between self-quarantining and depressive symptoms.

Social distancing and anxiety symptoms

The model testing links between social distancing and anxiety symptoms provided a good fit to the data (Table 4). Table 5 provides model estimates. There was a significant positive between-person effect, such that those who engaged in greater social distancing reported greater anxiety symptoms compared to those who engaged in less social distancing. There was a significant positive weekly within-person effect between social distancing and anxiety symptoms, indicating that those who engaged in greater social distancing for a given week relative to their own average across the study also experienced greater anxiety that week relative to their own average across the study. For example, if Person B social distanced more in Week 3 compared to their average amount of social distancing across the eight-week study, then they also experienced more anxiety symptoms in Week 3 compared to their average level of anxiety across the eight-week study. There was no evidence of significant cross-lagged effects for social distancing predicting anxiety symptoms the subsequent week or anxiety symptoms predicting social distancing the subsequent week. Patterns of results did not differ for non-imputed data.

Social distancing and depressive symptoms

The model testing links between social distancing and depressive symptoms provided a good fit to the data (Table 4). Table 5 provides model estimates. There were significant positive between-person associations, suggesting that those who engaged in greater social distancing reported greater depressive symptoms compared to those who engaged in less social distancing. There was a significant weekly within-person effect between social distancing and depressive symptoms, indicating that those who engaged in greater social distancing for a given week relative to their own average across the study also experienced greater depressive symptoms that same week relative to their own average across the study. There were no significant cross-lagged effects for engagement in social distancing predicting depressive symptoms the subsequent week, nor depressive symptoms predicting engagement in social distancing the subsequent week. For non-imputed data, there was a significant positive between-person effect for social distancing and depressive symptoms, and within-person or cross-lagged effects were not significant.

Discussion

The current study assessed fluctuations in self-quarantining, social distancing, and mental health symptoms (i.e., anxiety and depression) between and within individuals, as well as prospectively, over eight weeks in a national sample of U.S. adults from March 20, 2020 to May 17, 2020. We found between person effects indicating that those who engaged in greater self-quarantining or social distancing reported higher anxiety and depressive symptoms relative to those who engaged in less self-quarantining or social distancing. Further, we found weekly within-person effects for associations between social distancing, but not self-quarantining, and mental health symptoms. That is, people who engaged in greater social distancing for a given week relative to their own average across the study reported greater symptoms of anxiety and depression that week relative to their own average across the study. Finally, contrary to expectations, we generally did not find evidence for prospective associations between self-quarantining or social distancing with mental health symptoms.

Consistent with concerns and empirical evidence suggesting increases in psychological maladjustment pre- to during the pandemic [20, 46], the present sample reported striking levels of anxiety and depressive symptoms at Wave 1. Within this sample, 27.2% reported clinically significant anxiety symptoms and 25.6% reported clinically significant depressive symptoms. Of note, in a national U.S. adult sample surveyed in 2019, 6.1% reported clinically significant anxiety symptoms [47], and 7.0% reported clinically significant depressive symptoms [48]. These findings speak to the high levels of psychological distress experienced by U.S. adults during the initial weeks of the COVID-19 pandemic. However, results revealed that average levels of anxiety and depressive symptoms significantly decreased from Wave 1 to Wave 8, suggesting that psychological distress waned eight weeks into the pandemic. Effect sizes indicate that this decrease was slightly stronger for anxiety symptoms (ηp2 = 0.045) than for depressive symptoms (ηp2 = 0.019) based on Cohen’s benchmarks of small (ηp2 = .0099), medium (ηp2 = .0588), and large (ηp2 = .1379) effects [49, 50].

Findings also provide insight into engagement with self-quarantining and social distancing behaviors at the individual level among U.S. adults. During Wave 2, individuals reported an average of 4.70 days of self-quarantining over the prior week. At Wave 8, participants reported an average of 4.14 days of self-quarantining over the prior week. Levels of engagement in social distancing at Wave 1 and Wave 8 were comparable. ANOVAs indicated that self-quarantining generally decreased over time whereas there was a stark reduction in social distancing at Wave 7. Of note, the effect size for self-quarantining was small (ηp2 = 0.014), whereas the effect size for social distancing was minimal (ηp2 = 0.003). Potentially, with the loosening of COVID-19 preventive policies, people may have viewed self-quarantining as less of a necessary action, whereas they may have become more accustomed to engaging in habitual social distancing. Indeed, reductions in self-quarantining, anxiety symptoms, and depressive symptoms from Waves 1 and 2 to Wave 8 are striking given how early this was in the pandemic.

Bivariate correlations provide cross-sectional evidence that greater engagement in both self-quarantining and social distancing is associated with greater anxiety and depressive symptoms. As discussed previously, prior research provides mixed findings regarding the associations of self-quarantining and social distancing with symptoms of anxiety and depression, with some researchers finding significant positive associations [15, 24] and others finding non-significant associations [17, 18]. Inconsistencies could be due to several factors such as variability in sample sizes, sample demographics, and time of data collection. Thus, the significant correlations should be interpreted in the context of the present sample (a large national sample of 1,011 U.S. adults) and time period (March 20 to May 17, 2020).

The present findings overcome methodological limitations of prior cross-sectional and two-point longitudinal research by isolating between- and within-person effects, allowing conclusions about associations of self-quarantining and social distancing with symptoms of anxiety and depression at the group and individual level. Broadly, the present findings suggest that those who self-quarantine or social distance more also report greater levels of anxiety and depression than those who self-quarantine or social distance less (βs = .14 –.30). Effect sizes indicate that these between-person associations are large, based on guidelines for small (β = .03), medium (β = .07), and large (β = .12) effects for standardized coefficients in cross-lagged models [51]. However, associations at the within-person level indicate that more social distancing, but not self-quarantining, for a given week relative to an individual’s own average across the study was associated with greater symptoms of anxiety and depression that week relative to their own average across the study. Though the between-person associations were stronger than the within-person associations, the within-person results revealed notable small-medium effect sizes (βs = .05). Perhaps at the individual level, something more complex may be occurring when one self-quarantines. Although social isolation is generally associated with negative consequences, it is also keeping one safe from being sick or infecting others. Thus, the detrimental effects of self-quarantining may be buffered by the positive consequences (e.g., limiting exposure) of quarantine. It is also possible that individuals may not have felt fully isolated when self-quarantining given technologies that allow for social connection [52].

Although findings highlight several between-person and weekly within-person associations between social distancing and mental health symptoms, as well as between-person associations between self-quarantining and mental health, there was little evidence of weekly cross-lagged associations. These findings suggest that mental health consequences of self-quarantining and social distancing may be more transient, impacting individuals in the moment but not necessarily carrying over to have effects on subsequent weeks of functioning. The lack of evidence for prospective, predictive associations may indicate that potential third variables are at play. Other pandemic stressors may be more salient when thinking about mental health, such as job insecurity and economic concerns [18] or workplace characteristics (e.g., frontline worker position) [53]. Alternatively, factors related to the innate human capacity for adaptability and resilience may be at play [54]. Indeed, cross-sectional research has identified resiliency factors, such as hope [55] and coping flexibility [56], to have protective effects on psychological functioning in the context of the COVID-19 pandemic. Utilizing longitudinal analyses to identify pandemic and individual-related variables associated with the link between self-quarantining, social distancing, and mental health will inform interventions targeting anxiety and depression throughout the COVID-19 or future pandemic.

Limitations

These findings should be considered in the context of some limitations. Although the design of this study allowed for dynamic longitudinal analyses of within-person, between-person, and cross lagged effects, causal claims cannot be made and there may be underlying effects of third variables (e.g., loneliness, online social connection) [57, 58]. While experimental designs can help elucidate causality and third variables, it would be unethical to implement an experimental design that involves enforced isolation and social disconnection. Further, despite recruitment of a large national sample, the sample was limited to those involved with the panel provider and was not nationally representative. Lastly, the present study’s weekly assessment structure limits the ability to test temporal effects on a smaller scale (e.g., daily basis or within-day fluctuations). Ecological momentary assessments could provide insight into smaller scale temporal fluctuations between self-quarantining, social distancing, and mental health.

Conclusion

This study is a novel longitudinal analysis of associations between self-quarantining, social distancing, and symptoms of anxiety and depression among a U.S. sample during an 8-week period around the introduction and loosening of preventive COVID-19 policies. Previous correlational research suggests associations between self-quarantining, social distancing, and poor mental health at the group level (e.g., people who self-quarantine more report greater depression). Findings confirm between-person (group) associations of self-quarantining and social distancing with symptoms of anxiety and depression. Results also provide evidence for within-person effects indicating that when an individual engaged in greater social distancing compared to their own average, they also reported greater anxiety and depression relative to their own average. Future research using longitudinal, within-subject methods is needed to understand other factors that may affect associations between self-quarantining, social distancing, anxiety, and depression. Such information would help guide policies and interventions related to behavioral and mental health throughout the COVID-19 pandemic and future pandemics.

Supporting information

S1 Appendix. List of all study measures.

https://doi.org/10.1371/journal.pone.0298461.s001

(DOCX)

S1 Table. Severity of anxiety and depression using imputed data.

https://doi.org/10.1371/journal.pone.0298461.s002

(DOCX)

S2 Table. Severity of anxiety and depression using non-imputed data.

https://doi.org/10.1371/journal.pone.0298461.s003

(DOCX)

S3 Table. Means (standard deviations) for repeated-measures ANOVA using non-imputed data.

https://doi.org/10.1371/journal.pone.0298461.s004

(DOCX)

S4 Table. Bivariate correlations between study measures using imputed data.

https://doi.org/10.1371/journal.pone.0298461.s005

(XLSX)

S5 Table. Bivariate correlations between study measures using non-imputed data.

https://doi.org/10.1371/journal.pone.0298461.s006

(XLSX)

S6 Table. Fixed-effect autoregressive cross-lagged model fit statistics using non-imputed data.

https://doi.org/10.1371/journal.pone.0298461.s007

(DOCX)

S7 Table. Standardized and unstandardized estimates for models using non-imputed data.

https://doi.org/10.1371/journal.pone.0298461.s008

(DOCX)

References

  1. 1. Wilder-Smith A, Freedman DO. Isolation, quarantine, social distancing and community containment: pivotal role for old-style public health measures in the novel coronavirus (2019-nCoV) outbreak. J Travel Med. 2020;27: taaa020. pmid:32052841
  2. 2. Aleman A, Sommer I. The silent danger of social distancing. Psychol Med. 2022;52: 789–790. pmid:32624043
  3. 3. Hossain MM, Sultana A, Purohit N. Mental health outcomes of quarantine and isolation for infection prevention: A systematic umbrella review of the global evidence. Epidemiol Health. 2020; e2020038. pmid:32512661
  4. 4. Venkatesh A, Edirappuli S. Social distancing in covid-19: what are the mental health implications? BMJ. 2020; m1379. pmid:32253182
  5. 5. Courtemanche C, Garuccio J, Le A, Pinkston J, Yelowitz A. Strong social distancing measures in the United States reduced the COVID-19 growth rate: study evaluates the impact of social distancing measures on the growth rate of confirmed COVID-19 cases across the United States. Health Aff. 2020;39: 1237–1246. pmid:32407171
  6. 6. Unützer J, Kimmel RJ, Snowden M. Psychiatry in the age of COVID ‐19. World Psychiatry. 2020;19: 130–131. pmid:32394549
  7. 7. Saeri AK, Cruwys T, Barlow FK, Stronge S, Sibley CG. Social connectedness improves public mental health: Investigating bidirectional relationships in the New Zealand attitudes and values survey. Aust N Z J Psychiatry. 2018;52: 365–374. pmid:28803484
  8. 8. Umberson D, Karas Montez J. Social relationships and health: a flashpoint for health policy. J Health Soc Behav. 2010;51: S54–S66. pmid:20943583
  9. 9. Frith U, Frith C. The social brain: allowing humans to boldly go where no other species has been. Phil Trans R Soc B. 2010;365: 165–176. pmid:20008394
  10. 10. Leigh-Hunt N, Bagguley D, Bash K, Turner V, Turnbull S, Valtorta N, et al. An overview of systematic reviews on the public health consequences of social isolation and loneliness. Public Health. 2017;152: 157–171. pmid:28915435
  11. 11. Centers for Disease Control and Prevention. Isolation and precautions for people with COVID-19. [cited 10 Oct 2023]. Available: https://www.cdc.gov/coronavirus/2019-ncov/your-health/isolation.html
  12. 12. Brooks SK, Webster RK, Smith LE, Woodland L, Wessely S, Greenberg N, et al. The psychological impact of quarantine and how to reduce it: rapid review of the evidence. Lancet. 2020;395: 912–920. pmid:32112714
  13. 13. Jin Y, Sun T, Zheng P, An J. Mass quarantine and mental health during COVID-19: a meta-analysis. J Affect Disord. 2021;295: 1335–1346. pmid:34706447
  14. 14. Prati G, Mancini AD. The psychological impact of COVID-19 pandemic lockdowns: a review and meta-analysis of longitudinal studies and natural experiments. Psychol Med. 2021;51: 201–211. pmid:33436130
  15. 15. Mechili EA, Saliaj A, Kamberi F, Girvalaki C, Peto E, Patelarou AE, et al. Is the mental health of young students and their family members affected during the quarantine period? evidence from the COVID‐19 pandemic in Albania. Psychiatric Ment Health Nurs. 2021;28: 317–325. pmid:32657469
  16. 16. Wang W, Huang W, Liu X, Hennessy DA. Psychological impact of mandatory COVID-19 quarantine on small business owners and self-employed in China. Curr Psychol. 2023;42: 17291–17303. pmid:34155428
  17. 17. Zhao SZ, Wong JYH, Wu Y, Choi EPH, Wang MP, Lam TH. Social distancing compliance under COVID-19 pandemic and mental health impacts: a population-based study. Int J Environ Res Public Health. 2020;17: 6692. pmid:32937929
  18. 18. Wilson JM, Lee J, Fitzgerald HN, Oosterhoff B, Sevi B, Shook NJ. Job insecurity and financial concern during the COVID-19 pandemic are associated with worse mental health. J Occup Environ Med. 2020;62: 686–691. pmid:32890205
  19. 19. Wilson JM, Lee J, Shook NJ. COVID-19 worries and mental health: the moderating effect of age. Aging Ment Health. 2021;25: 1289–1296. pmid:33291948
  20. 20. Haliwa I, Wilson J, Lee J, Shook NJ. Predictors of change in mental health during the COVID-19 pandemic. J Affect Disord. 2021;291: 331–337. pmid:34087628
  21. 21. Centers for Disease Control and Prevention. How to protect yourself and others. [cited 10 Oct 2023]. Available: https://www.cdc.gov/coronavirus/2019-ncov/prevent-getting-sick/prevention.html
  22. 22. Marroquín B, Vine V, Morgan R. Mental health during the COVID-19 pandemic: effects of stay-at-home policies, social distancing behavior, and social resources. Psychiatry Res. 2020;293: 113419. pmid:32861098
  23. 23. Benke C, Autenrieth LK, Asselmann E, Pané-Farré CA. Lockdown, quarantine measures, and social distancing: associations with depression, anxiety and distress at the beginning of the COVID-19 pandemic among adults from Germany. Psychiatry Res. 2020;293: 113462. pmid:32987222
  24. 24. Baiden P, LaBrenz CA, Findley E. Social distancing and anxiety among female caregivers of children ages zero-to-five during coronavirus disease (COVID-19) lockdown in the United States. J Affect Disord Rep. 2021;5: 100154. pmid:34642679
  25. 25. Kämpfen F, Kohler IV, Ciancio A, Bruine de Bruin W, Maurer J, Kohler H-P. Predictors of mental health during the Covid-19 pandemic in the US: role of economic concerns, health worries and social distancing. PLoS ONE. 2020;15: e0241895. pmid:33175894
  26. 26. Carvalho Aguiar Melo M, de Sousa Soares D. Impact of social distancing on mental health during the COVID-19 pandemic: an urgent discussion. Int J Soc Psychiatry. 2020;66: 625–626. pmid:32438880
  27. 27. Berry D, Willoughby MT. On the practical interpretability of cross‐lagged panel models: rethinking a developmental workhorse. Child Dev. 2017;88: 1186–1206. pmid:27878996
  28. 28. Rogosa D, Brandt D, Zimowski M. A growth curve approach to the measurement of change. Psychol Bull. 1982;92: 726–748.
  29. 29. Behar E, DiMarco ID, Hekler EB, Mohlman J, Staples AM. Current theoretical models of generalized anxiety disorder (GAD): Conceptual review and treatment implications. J Anxiety Disord. 2009;23: 1011–1023. pmid:19700258
  30. 30. Velikonja NK, Erjavec K, Verdenik I, Hussein M, Velikonja VG. Association between preventive behaviour and anxiety at the start of the COVID-19 pandemic in Slovenia. Slov J Public Health. 2020;60: 17–24. pmid:33488818
  31. 31. Sloman L. Evolved mechanisms in depression: the role and interaction of attachment and social rank in depression. J Affect Disord. 2003;74: 107–121. pmid:12706512
  32. 32. Jacobson NS, Martell CR, Dimidjian S. Behavioral activation treatment for depression: returning to contextual roots. Clin Psychol. 2001;8: 255–270.
  33. 33. Chan D. Data analysis and modeling longitudinal processes. Group Organ Manag. 2003;28: 341–365.
  34. 34. Centers for Disease Control and Prevention. CDC museum COVID-19 timeline. [cited 10 Oct 2023]. Available: https://www.cdc.gov/museum/timeline/covid19.html
  35. 35. U.S. Department of Defense. Coronavirus timeline. [cited 10 Oct 2023]. Available: https://www.defense.gov/Spotlights/Coronavirus-DOD-Response/Timeline/
  36. 36. Shook NJ, Sevi B, Lee J, Oosterhoff B, Fitzgerald HN. Disease avoidance in the time of COVID-19: The behavioral immune system is associated with concern and preventative health behaviors. PLoS ONE. 2020;15: e0238015. pmid:32817714
  37. 37. Spitzer RL, Kroenke K, Williams JBW, Löwe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med. 2006;166: 1092. pmid:16717171
  38. 38. Löwe B, Decker O, Müller S, Brähler E, Schellberg D, Herzog W, et al. Validation and standardization of the generalized anxiety disorder screener (GAD-7) in the general population. Medical Care. 2008;46: 266–274. pmid:18388841
  39. 39. Cha DS, Carmona NE, Rodrigues NB, Mansur RB, Lee Y, Subramaniapillai M, et al. Cognitive impairment as measured by the THINC-integrated tool (THINC-it): the association with self-reported anxiety in Major Depressive Disorder. J Affect Disord. 2018;238: 228–232. pmid:29886204
  40. 40. Kroenke K, Spitzer RL, Williams JBW. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16: 606–613. pmid:11556941
  41. 41. American Psychiatric Association. Severity measure for depression—adult (adapted from the Patient Health Questionnaire–9 [PHQ-9]). 2021 [cited 10 Oct 2023]. Available: https://www.psychiatry.org/File%20Library/Psychiatrists/Practice/DSM/APA_DSM5_Severity-Measure-For-Depression-Adult.pdf
  42. 42. Wayman JC. Multiple imputation for missing data: what is it and how can I use it. Chicago, IL; 2003. Available: https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=14e7b7ccb82a7802555a598fe32bb0897fe692fd
  43. 43. Rosseel Y. lavaan: An R package for structural equation modeling. J Stat Soft. 2012;48.
  44. 44. Shi D, Lee T, Maydeu-Olivares A. Understanding the model size effect on SEM fit indices. Educ Psychol Meas. 2019;79: 310–334. pmid:30911195
  45. 45. Hu L, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Struct Equ Modeling. 1999;6: 1–55.
  46. 46. Pfefferbaum B, North CS. Mental health and the Covid-19 pandemic. N Engl J Med. 2020;383: 510–512. pmid:32283003
  47. 47. Terlizzi EP, Villarroel MA. Symptoms of generalized anxiety disorder among adults: United States, 2019. Centers for Disease Control and Prevention; 2020. Report No.: 378.
  48. 48. Villarroel MA, Terlizzi EP. Symptoms of depression among adults: United States, 2019. Centers for Disease Control and Prevention; 2020. Report No.: 379.
  49. 49. Cohen J. Statistical power analysis for the behavioural sciences. New York: Academic Press; 1969.
    • 50. Richardson JTE. Eta squared and partial eta squared as measures of effect size in educational research. Educ Res Rev. 2011;6: 135–147.
    • 51. Orth U, Meier LL, Bühler JL, Dapp LC, Krauss S, Messerli D, et al. Effect size guidelines for cross-lagged effects. Psychol Methods. 2022. pmid:35737548
    • 52. Canale N, Marino C, Lenzi M, Vieno A, Griffiths MD, Gaboardi M, et al. How communication technology fosters individual and social wellbeing during the Covid-19 pandemic: preliminary support for a digital interaction model. J Happiness Stud. 2022;23: 727–745. pmid:34177363
    • 53. Rossi R, Socci V, Pacitti F, Di Lorenzo G, Di Marco A, Siracusano A, et al. Mental health outcomes among frontline and second-line health care workers during the Coronavirus Disease 2019 (COVID-19) pandemic in Italy. JAMA Netw Open. 2020;3: e2010185. pmid:32463467
    • 54. PeConga EK, Gauthier GM, Holloway A, Walker RSW, Rosencrans PL, Zoellner LA, et al. Resilience is spreading: mental health within the COVID-19 pandemic. Psychol Trauma. 2020;12: S47–S48. pmid:32496106
    • 55. Braun-Lewensohn O, Abu-Kaf S, Kalagy T. Hope and resilience during a pandemic among three cultural groups in Israel: the second wave of Covid-19. Front Psychol. 2021;12: 637349. pmid:33679564
    • 56. Cheng C, Wang H, Ebrahimi OV. Adjustment to a “New Normal:” coping flexibility and mental health issues during the COVID-19 pandemic. Front Psychiatry. 2021;12: 626197. pmid:33815166
    • 57. Antonelli-Salgado T, Monteiro GMC, Marcon G, Roza TH, Zimerman A, Hoffmann MS, et al. Loneliness, but not social distancing, is associated with the incidence of suicidal ideation during the COVID-19 outbreak: a longitudinal study. J Affect Disord. 2021;290: 52–60. pmid:33991946
    • 58. Stuart J, O’Donnell K, O’Donnell A, Scott R, Barber B. Online social connection as a buffer of health anxiety and isolation during COVID-19. Cyberpsychol Behav Soc Netw. 2021;24: 521–525. pmid:33601944