If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The COVID-19 virus has resulted not only in high rates of morbidity and mortality across the globe, but in widespread mental health problems and sleep disruption, likely as a result of pandemic-related stressors. The current study examines associations among COVID-related stress, sleep quality, and mental health.
Cross-sectional data were collected via online surveys in May 2020.
were 2541 community adults ages 18-70 from Israel (N = 1969) and the U.S. (N = 572).
Participants completed measures of COVID-related stress, sleep quality, and symptoms of anxiety, depression and adjustment disorder.
Participants reported high rates of depression and anxiety symptoms, adjustment difficulties, and poor sleep quality. In both countries, COVID-related stressors were associated with both anxiety and depression, and these associations were mediated by sleep disturbances.
These results support the role of sleep in mental health difficulties. Widespread, accessible, evidence-based interventions are urgently needed to improve health and mental health and to promote resilience in preparation for future global crises.
The majority of these initial studies collected self-report data from adults in a single country. Successful management of the current and future pandemics requires the international research community to examine the wide-ranging and complex sleep health and mental health consequences of COVID-19.
A large body of research suggests that trauma and/or significant stressors are associated with sleep difficulties, including difficulty falling asleep or maintaining sleep, poor sleep quality, and/or irregular sleep patterns.
Globally, researchers are examining the effects of COVID-19 on sleep health. In a Taiwanese survey of adults recruited via Facebook, just over 50% reported sleep disturbances, which were associated with worries about COVID-19, lower social support, work difficulties, and poorer physical health.
investigated social and biological rhythms as well as sleep during a 6-week period (March-April 2020) in Austria, Germany, and Switzerland, and found that COVID-19 restrictions reduced social jetlag, with an overall increase in sleep duration, yet a slight decrease in sleep quality. Similarly, a survey study conducted in Italy during lockdown and resulting home confinement found that sleep timing became more delayed with people going to bed and waking up later, and spending more time in bed, but paradoxically, also reporting poorer sleep quality. Studies in Italy and Morocco reported increased sleep difficulties and decreased sleep quality with greater sleep difficulties associated with higher levels of depression, anxiety, and stress.
Although sleep disturbances have traditionally been viewed as a secondary consequence of mental health problems, increasingly sleep quality, sleep problems, and disrupted sleep have been examined for their role in causing, mediating, and/or maintaining mental health problems.
An early review paper on mental health during COVID-19 suggests that symptoms of anxiety and depression and stress are common psychological reactions to the COVID-19 pandemic, and associated with disturbed sleep.
Studies of adults during the initial phase of COVID-19 lockdown support this view. For example, among Spanish adults a variety of factors were associated with higher risk of mental health symptoms and/or low psychological well-being including increased time on the internet, and, in particular, higher prevalence of prior sleep difficulties.
Likewise, an online survey of adults in Australia revealed that negative changes in physical activity, sleep, smoking, and alcohol intake were associated with higher depression, anxiety and stress symptoms.
Most recently, a team in Brazil found that the association of physical inactivity and sedentary behaviors with mental health factors is partly mediated by worsening sleep quality during the COVID-19 pandemic quarantine.
In addition to sleep health challenges, evidence from a range of countries and cultures suggests that adults are experiencing multiple COVID-related stressors including financial and health-related stress, social isolation, and difficulty obtaining needed supplies, as well as persistent worries about one's self or family members becoming ill.
The prevalence of anxiety and depressive symptoms were 3 to 4 times higher than in the prior year. In a another large sample of US adults with no reported prior history of mental health problems, more than 25% reported psychological distress in the early weeks of COVID-19 (March 2020).
as a preoccupation with an identifiable stressor or failure to adapt to that stressor within the first 6 months of experiencing it), research has not yet documented the prevalence or correlates of adjustment disorders during the pandemic.
In summary, sleep health behaviors are clearly associated with both physical and mental health yet may be disrupted by social isolation and changes in daily routine prompted by the COVID-19 pandemic. The current study combined data from 2 large, diverse samples of adults collected in the early stages of the pandemic in the United States and Israel to examine the following hypotheses: 1) Participants with more COVID-related stress will report significantly more adjustment difficulty, depression, anxiety, and poorer sleep quality; and 2) Sleep quality will mediate the associations between COVID-related stress and depression and anxiety symptoms.
Participants included 2541 adults ages 18 to 75 years (M = 39.6 years) recruited on-line using methods described below. As displayed in Table 1, the study included 1969 participants from Israel (they could complete the survey in either Hebrew or Arabic) and 572 from the United States (all completed the survey in English). Inclusion criteria included being between 18 and 75 years of age and speaking the language in which the survey was administered. No exclusion criteria were applied. Fifty-two percent of the combined sample were female; the majority were employed (78.8%), married (65.9%), with children at home (56.0%), and had at least a college degree (61%).
Table 1Demographic Characteristics of Participants
American sample % (n)
Israeli sample % (n)
Effect size (Hedges’ g/φ)
36.94 ± 11.51
40.4 ± 13.76
Gender (% female)
Employment status (% employed)
Marital status (% married)
% with any children at home
65.84 ( 376)
Education (% with college degree or higher)
Notes: N = 2541 for full sample, n = 572 for US sample, n = 1969 for Israeli sample. Employment status = not employed (include furloughed, unpaid, & not looking for work) vs. employed (paid). Marital status = Married vs. not married (Single/ Separated/ Divorced/ Widowed). Children at home = none at home (includes adult children) vs. 1+ children at home. Education = with vs. without a college degree or higher.
Demographics. All participants provided their demographic information including age, sex, education level, employment status, income, and presence of children under the age of 18. (Race and ethnicity were obtained for both samples, but categories were not comparable across the 2 countries.)
COVID-related stress. Participants rated how stressful each of 13 COVID-related stressors had been in the past 6 months from 1 (Not at all stressful) to 4 (Very stressful).
Stressors included financial problems, inability to spend time with friends or family, changes to normal routines, cancellation of travel plans, challenges at home, trouble obtaining needed supplies or services, hearing distressing news reports, uncertainty about self or others getting COVID-19, difficulty completing work or educational responsibilities, increased work or family responsibilities, and uncertainty about the future. A total COVD-related stressors score was computed for each participant by summing the scores from all 13 items, such that higher scores indicate greater stress. This measure was created for the current study, thus prior reliability and validity statistics are unavailable. Internal consistency (Cronbach's alpha) was .87 for both the American and Israeli samples.
Changes in sleep habits. Using a measure created for this study, participants were asked if they had experienced 6 specific changes to sleep habits since the pandemic started: go to bed later, wake up later, sleep more hours, sleep fewer hours, sleep more often in front of a screen (eg, TV, tablet, smart phone), and use a screen more often in the bedroom. These items were endorsed as Yes or No.
Sleep quality. The Pittsburgh Sleep Quality Index (PSQI)
is the most widely used self-report sleep quality assessment tool for clinical and non-clinical populations. The PSQI consists of 18 items: 14 are rated as 0–3, while 4 open-ended questions are used to evaluate perceived sleep quality. These items are used to generate categorical scores representing the PSQI's 7 subscales, and a total score reflecting the sum of the subscales. The total score provides an efficient summary of the participants’ sleep experience and quality over the previous 2 weeks, with higher scores indicating poorer sleep quality. A total score higher than 5 implies the existence of sleep disturbances. In the current study, the internal reliability of the 7 subscale scores (Cronbach's alpha) was .73 (U.S) and .72 (Israel).
Depression and anxiety symptoms. Depression and anxiety were measured with 2 subscales from the self-report DASS-21, 21-item short form derived from the Depression Anxiety Stress Scales (DASS).
Seven items measuring depression, and 7 measuring anxiety, were endorsed on a Likert-type scale from 0 (did not apply to me) to 3 (applied to me very much); scores within each domain are summed such that higher scores indicate more depression or anxiety. The DASS-21 demonstrates good reliability and validity
consists of 4 items measuring adjustment to a specific stressor (in this case, the COVID-19 pandemic) using a 4-point Likert scale ranging from 1 (Never) to 4 (Often). Two items measure success in adapting to the stressful event, and are reverse scored, while 2 items measure failure to adapt. Responses are summed, such that scores range from 4 to 16 with higher scores indicating greater difficulty adjusting; a score higher than 8.5 indicates the existence of adjustment difficulty.
In the current study, the internal reliability (Cronbach's alpha) was .75 (U.S.) and .84 (Israel).
All study procedures were approved by the Institutional Review Boards at both sites. Data are cross-sectional and were collected separately via convenience sampling in 2 counties. For the Israeli sample, all questionnaires were translated from English to Hebrew and Arabic by one person and then back translated to English by a different person. In Israel, participants were recruited in May 2020 through iPanel (https://www.ipanel.co.il), a large Israeli online platform, with over 100,000 panel members. iPanel participants complete tasks for points, which can be converted into gift certificates. In the United States, participants were recruited in May 2020 through Amazon Mechanical Turk (MTurk), an online platform in which people complete tasks for a small, monetary sum ($1 for this study). Several studies support the utility of collecting data via MTurk, including for obtaining a diverse sample and obtaining information on mental health problems.
In the United States sample, data were collected via MTurk from 578 people, but 5 were excluded because they failed attention checks, 3 were excluded because they completed the measures in less than 10 minutes, and 4 were excluded because their responses to sleep quality questions were implausible (eg, they reported getting up before going to bed each day or reported getting more hours of sleep than they spent in bed each night). In the Israeli sample, sleep quality data of 56 participants (out of 1969) were excluded because of similar implausible responses.
All analyses were conducted using IBM SPSS version 27. Demographic data were analyzed via independent-samples t tests or Pearson's chi-squared tests. Because of site differences in all demographic variables (see Table 1), site was controlled (when appropriate) in all multivariate analyses, as was age, gender, marital status and employment. Site pairwise comparisons were conducted using Pearson's chi-squared for differences in sleep habits since COVID-19, and t-test with the 95% bootstrapped confidence intervals for PSQI subscales. Analyses examining mediated mediation (model 8) used the PROCESS (v 3.5) macro for SPSS.
Significance of effects in the mediation analyses were determined by lack of zero in the 95% bootstrapped confidence intervals (5000 samples). Effect size estimators were Hedges' g for t-tests, phi (φ) for chi-squared tests, and adjusted R2 for multiple regression. In addition, analyses for site differences and for mediation were re-run using randomly selected, equal size groups (n = 540 in each group), and results (not shown) were nearly identical to those presented for the full sample.
As shown in Table 1, participants from the 2 countries differed demographically, such that those from Israel were significantly older; more likely to be female; and less likely to be employed, married, have children at home, or have a college degree or more education. Therefore, all multivariate analyses controlled for site. Table 2 displays means, standard deviations, and inter-correlations among key study variables. Key variables showed a general pattern of positive correlations that were medium to large in magnitude.
A substantial proportion of the sample reported mental health problems, including 15% reporting moderate to severe levels of depression symptoms, 20% reporting moderate to severe levels of anxiety symptoms, and more than half (54.5%) reporting significant difficulty adjusting to the pandemic. Furthermore, 56% of the sample were poor sleepers, as indicated by PSQI total scores of 5 or greater. Participants also reported experiencing on average several COVID-related stressors; those related to social isolation (ie, being unable to spend time with close friends or family and being unable to participate in normal routines or activities) were rated as the most stressful.
Table 2Descriptive Information and Correlations Among Stress, Mental Health, and Sleep Quality
Full sample Mean ± SD
American sample (n = 572)
Israeli sample (n = 1969)
Adjustment disorder total score
DASS-21 depression scale
DASS-21 anxiety scale
13 items COVID stressors
PSQI total score
Adjustment disorder total score
9.09 ± 3.12
10.67 ± 2.74
8.64 ± 3.07
DASS-21 depression scale
5.57 ± 5.7
9.56 ± 6.13
4.41 ± 5.00
DASS-21 anxiety scale
4.15 ± 5.36
9.00 ± 6.15
2.74 ± 4.15
13 item COVID stressors
32.94 ± 8.22
33.2 ± 8.22
32.86 ± 8.22
PSQI total score
9 ± 3.89
4.91 ± 3.27
Note: All correlations survived the Bonferroni correction (significant at P < .005). ***P < .001.
A large majority of the sample (88.9%) reported that their sleeping habits had changed since the pandemic started in at least one of the 6 ways assessed. Almost 70% of the participants reported they were going to bed later, and almost 50% reported waking up later in the morning (see Table 3). American participants were more likely to report each of the 6 changes in sleep habits than the Israeli sample.
Table 3Sleep Habits Changes Since COVID-19
Full sample % (n)
American sample % (n)
Israeli sample % (n)
Effect size (φ)
Go to bed later
Wake up later in the morning
Sleep more hours
Sleep fewer hours
Sleep more often in front of screen
Use screen in bedroom more than usual
Note: N = 572 for US sample, 1969 for Israeli sample. All comparisons survived the Bonferroni correction (significant at P < .007). Results (percentage and effect size) remained virtually the same when analysis was conducted on randomly selected, equal size groups (n = 540 in each group), except for "Go to bed later", which narrowly missed the cutoff of the Bonferroni correction.
Table 4 displays the PSQI total and component scores for the full sample as well as site comparisons. American participants reported poorer sleep quality than Israeli participants on all 7 PSQI components, with small to large effect sizes, as well as on total PSQI scores as shown in Table 2. In addition, Spearmen correlations between PSQI subscales and symptoms of depression and anxiety revealed that the strongest correlations were between use of sleeping medication and both depression (rs = .55, P < .001) and anxiety (rs = .57, P < .001), and between sleep disturbances and both depression (rs =.52, P < .001) and anxiety (rs=.49, P < .001).
Table 4Sleep Quality: PSQI Subscales
American sample M ± SD
Israeli sample M ± SD
t (df = 2451)
95% Bootstrapped- CI
Subjective sleep quality
1.27 ± 0.84
1.15 ± 0.75
1.37 ± 0.89
1.14 ± 1.02
0.71 ± 0.88
0.48 ± 0.76
0.95 ± 1.20
0.11 ± 0.44
1.78 ± 0.78
1.04 ± 0.68
Use of sleeping medication
1.24 ± 1.05
0.27 ± 0.72
1.64 ± 1.00
0.69 ± 0.79
Note: n = 540 US sample, n = 1913 Israeli sample. PSQI subscales scored 0 to 3 with higher scores indicating poorer quality sleep. Comparisons with a p-value <.007 are considered statistically significant (Bonferroni correction). Bootstrap (5000 samples, BCa confidence intervals) was used, as the normality and equality of variances assumptions were violated. Effect sizes are Hedges' g. PSQI = Pittsburgh Sleep Quality Index BCa = Bias-corrected and accelerated
Multiple regression was used to examine COVID-related stress in relation to adjustment problems, depression, anxiety, and poorer sleep quality, controlling for site, age, gender, marital status, and employment status. As hypothesized, participants who reported more COVID-related stress reported significantly poorer adjustment, more depression, anxiety, and poorer sleep quality, with moderate effect sizes (see Table 5).
Table 5COVID-Related Stress in Relation to Psychological Outcomes and Sleep
Note: Each row represents a unique model. Regression coefficients indicate the association between COVID stress and the outcomes, controlling for site, age, gender, marital status and unemployment. R2 indicates the total variance in the dependent variables explained by COVID stress + control variables.
Finally, we examined whether sleep quality mediated the above-reported associations between COVID-related stress and both depression and anxiety, and whether these associations were moderated by site (US/Israel). The model for depression (F[7,1113]= 238.61, P < .001), as well as the moderation effect (b = −.12, bootstrapped 95% CI [−.19, −.05]) were statistically significant. The mediation (indirect effect) of sleep quality on depression was significant for both samples but was stronger for the American sample (b = .16, bootstrapped 95% CI [.13, .19]), compared to the Israeli sample (b = .09, bootstrapped 95% CI [.06, .11]). As with the model for depression, the model for anxiety (F[7,1113]= 330.32, P < .001), as well as the moderation effect (b = −.21, bootstrapped 95% CI [−.27, −.14]) were statistically significant. The mediation (indirect effect) of sleep quality on anxiety was significant for both samples but was stronger for the American sample (b = .14, bootstrapped 95% CI [.12, .17]), compared to the Israeli sample (b = .08, bootstrapped 95% CI [.05, .10]). In other words, a significant portion of the association between stress and psychological symptoms was accounted for by sleep quality in both samples (see Fig. 1 for depression and Fig. 2 for anxiety).
The global COVID-19 pandemic has resulted not only in unprecedented morbidity and mortality, but also in widespread financial strain and social isolation with devastating effects on mental health and well-being.
This study used data from 2 large community samples of adults collected in the United States and Israel during the early stages of the pandemic (May 2020) to examine sleep and mental health in association with COVID-19 stressors. Although demographically different, these 2 samples completed the same measures, allowing for an examination of these variables among a large, diverse international sample. We found that substantial numbers of adults reported mental health problems during the early stage of the pandemic, including high levels of anxiety and depression symptoms and significant difficulty adjusting, as well as a number of COVID-related stressors, with social isolation and home confinement perceived as particularly stressful. Furthermore, more than half the sample were poor sleepers, and 89% reported a change in their sleeping habits since the pandemic started. These high rates of distress and sleep difficulties are consistent with other community samples from a variety of different countries surveyed during the COVID-19 pandemic and indicate the need for interventions to reduce stress and symptom levels and improve healthy behaviors.
Additional, longitudinal research is also needed to examine whether these high rates of distress and sleep difficulties represent an immediate response to the pandemic that attenuates over time, or a more chronic response requiring intervention. Gruber et al.,
and others have argued that mental health interventions must be widely available, developmentally appropriate, and accessible (eg, via telehealth) to meet the changing needs of the population; these guidelines can be applied to behavioral health interventions as well.
Although these analyses focused on sleep quality, it should also be noted that some participants reported none of these difficulties. During crises such as the COVID-19 pandemic, individuals, communities and cultures exhibit diverse characteristics and response styles which determine how well they can resist negative effects, and instead display growth or resilience.
and to promote resilience at the individual and community level before the next global crisis.
Our significant finding demonstrating that sleep quality mediated associations between COVID-related stressors and mental health outcomes speaks to the central role that sleep plays in promoting resilience in the face of stress or adversity and in regulating the immune response.
found that adults with insomnia reported higher sleep reactivity to stress and lower psychological resilience than good sleepers. Going forward, it would be helpful to better understand what aspects of sleep quality, sleep patterns, and other sleep health variables are significant in resilience. Nonetheless, taken together, these findings point to reciprocal relationships among sleep quality, emotional regulation, and the ability to adaptively overcome stress.
There were unexpected differences in mental health and sleep characteristics of the 2 samples. Although participants in both samples reported similar COVID-19 related stressors, the U.S. sample reported higher rates of anxiety and depression, consistent with Barzilay et al.,
Alternatively, Israeli adults’ historical experience with stress and trauma differs from U.S. adults’ experience and may promote greater resilience in Israeli culture. Despite these differences, the finding that in both samples the association between the variables is similar indicates that findings have global and not just local validity.
The current study has several limitations including that all data were derived from self-report at a single point in time, early in the COVID-19 pandemic. Thus, we could not assess changes in stress, sleep or mental health over time; nor could we test causal relations among these variables. Reports of health behaviors and sleep quality may not have reflected a change from pre-COVID-19 and were not validated by more objective measures such as a medical exam. Additionally, the PSQI assesses self-reported sleep quality at one point in time and, therefore, to better understand the role of sleep in resilience, more objective estimates of sleep over work and non-workdays would be advantageous. Furthermore, there were differences in demographic characteristics of the 2 samples, suggesting the need to obtain nationally representative samples from multiple countries.
Despite these limitations, these findings have a number of implications. First, by presenting data from 2 different countries (the U.S. and Israel), both of which are heterogeneous with respect to ethnicity, age, and family composition, this study underscores the truly international impact of the pandemic, including the resulting widespread disruptions in sleep and mental health. Second, these data reflect known linkages between stress and mental health, and in particular point to the importance of sleep quality as a mediator. As the pandemic enters its second year, interventions to improve health behaviors and mental health, and strengthen resilience, are urgently needed. Public health campaigns and behavioral health interventions designed to promote healthy sleep during the pandemic have the potential not only to improve mental health, but to boost immunity to the virus.