We conducted an EMA that involves repeated sampling of individuals’ current behaviors and experiences in real time and in their natural environments [18] during a postlockdown period (from August 8, 2020, to November 1, 2020) in Germany, when restrictions were lenient (eg, no private or public meeting restrictions, reopening of most leisure facilities, bars, and catering facilities; see Multimedia Appendix 1). EMA aims to minimize recall bias, maximize ecological validity, and approximate temporal causality (ie, Granger causality) and allows researchers to study microprocesses that influence behavior in real-world contexts [19]. Participants were recruited via online advertisements on universities’ websites, Twitter, and eBay classifieds. Participants had to fill in an online prequestionnaire on the Siuvo Intelligent Psychological Assessment Platform. After an initial contact via phone or email, we sent participants our study information, informed consent, and a QR code (to install a smartphone app) by mail.

We targeted vulnerable individuals who reported at least mild psychological distress and sometimes felt lonely amid the COVID-19 pandemic. We used the COVID-19 Peritraumatic Distress Index (CPDI [20]; cutoff score=28, indicating mild distress) questionnaire and the short-form version of the UCLA Loneliness Scale (ULS-8 [21]; cutoff score=16, indicating mild loneliness), respectively. Other inclusion criteria were being at least 18 years of age, not working night shifts, not currently infected with COVID-19, using an Android smartphone, and speaking fluent German. The CPDI was designed to evaluate changes in mental health status, cognitive skills, avoidance and compulsive behavior, physical symptoms, and loss of social functioning due to the COVID-19 pandemic. The questionnaire has been previously validated in a sample in Germany [20].

We used a smartphone app called “movisensXS” (movisens GmbH), which was developed for research purposes. The app is compliant with the General Data Protection Regulation (European Union) and Berlin Data Protection Act (Berliner Datenschutzgesetz – BlnDSG). Participants completed a 20-minute baseline assessment, followed by 7 consecutive days in which they received 8 randomized prompts between 8 AM and 10 PM. The study procedure was approved by the Ethics Committees of Charité – Universitätsmedizin Berlin (ref: EA2/143/20) and Freie Universität Berlin (ref: 030/2020).

To quantify COVID-19–related distress, we measured worries about the COVID-19 pandemic, perceived restrictions due to the COVID-19 pandemic, COVID-19 information exposure, and feelings of loneliness. Finally, we measured respondents’ momentary negative mood (anxiety, depression, fatigue, stress, and unhappiness). All questions were measured on a visual analogue scale ranging from 0 (not at all) to 100 (very much). To account for the steady increase in active COVID-19 cases in Germany during the time of measurement [22], we included daily COVID-19 cases as a predictor in our analysis. Our smartphone study consisted of a sociodemographic assessment (ie, age, gender, years of education) and the EMA. The exact EMA items can be found in Multimedia Appendix 1 and online at [23].

All statistical analyses were conducted in R (version 3.5.3; R Foundation for Statistical Computing [24]). To consider the hierarchical data structure and autoregressive parameters, we performed model selection using autoregressive (AR) multilevel models with the dependent variable negative mood. We followed the approach by Haan-Rietdijk et al [25]; details about the model selection procedure can be found in Multimedia Appendix 1 and online at [23].

We found that negative effects of the COVID-19 pandemic on mental health outlast lockdown measures. In line with findings from the first COVID-19 wave [8,26-30], we found that loneliness, worrying about COVID-19, and perceived restriction increased negative mood during a postlockdown period. Similar to the Ebola pandemic [31], possible reasons for the lasting effect of the COVID-19 pandemic might be worries about the negative economic consequences, concern about resurgence of the virus, struggles to rebuild social networks, and/or deliberately withdrawing from social contacts to avoid infection.

Furthermore, we found a negative carryover effect of mood between days (mood inertia), indicating dysfunctional mood regulation. Restrictive policies during the COVID-19 pandemic can impact mental health, possibly due to impaired mood homeostasis (ie, failure to positively regulate mood via mood-modifying activities) [7]. Importantly, our results show that even when the acute threat and restrictive measures are less pronounced, negative daily mood inertia remains.

Neither COVID-19 information seeking nor active COVID-19 cases increased negative mood. This contrasts with previous findings from lockdown periods [6,32]. For example, an EMA study during the first lockdown in Germany and Austria reported increased perceived COVID-19–related restrictions that were positively associated with increased daily news consumption, especially in individuals living alone [32]. In addition, an EMA study conducted in New Jersey in the United States between April 24 and May 26, 2020, showed that undergraduates felt more anxious about COVID-19 on days when the number of new cases and deaths due to COVID-19 were higher [6]. Our opposing finding might be caused by the relatively low domestic case numbers and associated news during the postlockdown period. Moreover, negative COVID-19 news might have less impact on mood over the course of the pandemic as people get accustomed to it.

We did not make explicit comparisons to participant status before the COVID-19 outbreak or to a control group, which limits generalizability to other populations. Furthermore, we did not measure adaptability, which has been associated with positive mood (eg, optimism and satisfaction) [33]. Finally, we did not assess the nature of COVID-19 worries and reasons for feeling restricted.

Even if cases are low and lockdown policies are lenient, mental health is still challenged by COVID-19–related stressors. Although information exposure to COVID-19 and daily COVID-19 cases had no impact on mood, we found a day-to-day carryover effect of negative mood. Moreover, COVID-19–related restriction, worry about COVID-19, and loneliness increased negative mood. Thus, the negative impact of the COVID-19 pandemic on mental health outlasts lockdown measures and mental health challenges will likely continue after the pandemic.