The COVID-19 pandemic has taken a substantial global toll on health and well-being. As of 12 March 2021, over 117 million people had been infected with the virus with over 2.6 million deaths. These deaths and restrictions to contain the disease have had immense socio-economic implications (Nicola et al., 2020), which together place a burden on overall mental health and well-being (Fisher et al., 2020; Khan et al., 2020). During 2020, daily life changed radically for many, with lockdowns and social-distancing measures being implemented to curb infection rates (Lewnard and Lo, 2020; Warren and Bordoloi, 2020). Individuals and groups at all levels of society have had to manage the consequences of these changes. Health workers have had their duties augmented, with many working to monitor infection spread, care for those infected, and implement rigorous sanitizing and personal protection procedures to minimize their own exposure to the virus (Nienhaus and Hod, 2020). Many retail and entertainment businesses have crumbled, sought temporary government support and/or adapted their business models by offering services online (Bartik et al., 2020; Kim, 2020; Seetharaman, 2020), while other forms of employment have required workers to transfer their duties to home (Bonacini et al., 2021). The creative industries (e.g., film, advertising and fashion, as well as creative occupations such as musicians, dancers, actors, visual artists and designers) have faced immense pressure due to the rapid spread of the virus and physical distancing restrictions. Losses in these industries between 1 April and 31 July 2020 in the United States alone have been estimated at 2.7 million jobs and more than $150 billion USD in sales of goods and services, with the fine and performing arts industries being hardest hit (Florida and Seman, 2020). Many creative and performing arts organizations have also turned to online alternatives (Keller, 2020). The pandemic has thus caused a great deal of stress and fear, even for those who have not been infected (Pfefferbaum and North, 2020).

Alongside these negative impacts, the COVID-19 pandemic has also provided some positive creative opportunities for adaptation (Kapoor and Kaufman, 2020; Kirchner et al., 2021). Emerging evidence suggests that social-distancing measures and stay-at-home orders have encouraged engagement in creative arts activities (Gupta, 2020; Radermecker, 2020). This has included commonplace activities such as baking (Easterbrook-Smith, 2020) and watching films (Mikos, 2020), as well as new and emerging forms of artistic engagement such as singing in online choirs or from balconies (Taylor, 2020; see also www.musicacrossthebalconies.com) and visiting online art galleries (Shehadi, 2020). Creative and artistic activities have served a range of purposes during the COVID-19 pandemic in a variety of cultural contexts, such as health promotion, improving environmental aesthetics and memorializing (de-Graft Aikins, 2020), awareness-raising about the threats posed by the pandemic (Blanc et al., 2020), and emotion regulation (Karwowski et al., 2021). And, while musical activities can serve a range of purposes such as self-expression and social group bonding (Vanstone et al., 2016), musical activities have also facilitated emotion regulation during the pandemic (Martín et al., 2021; Steinberg et al., 2021).

This evidence suggests that creative activities, including specifically artistic creative activities (hereafter ACAs)—defined as creative activities involving the arts specifically (Fancourt et al., 2019)—have supported individuals and groups in regulating their emotions during the pandemic and in supporting their own mental health and well-being. Much of the literature on the emotional and mental health impacts of pandemics and global health crises is either descriptive (e.g., documenting negative effects) and/or prescriptive (e.g., suggesting coping strategies without testing their effectiveness) (Restubog et al., 2020). As such, studies on the role of creative activities in navigating pandemic crises contribute to broader understandings of human psychological responses to epidemic encounters, which are otherwise poorly understood (Curson, 2015).

While research related to the impact of COVID-19 on daily life emerged rapidly from countries with early and restrictive lockdowns (March–June 2020), much less is known about the Australian experience where infection and death rates have been relatively low, partly due to the early and prompt enforcement of restrictions (Pew Research Center, 2020; Hong et al., 2021; World Health Organisation, 2021); by 12 March 2021, Australia had recorded 29,090 total cases and 909 deaths (Australian Government Department of Health, 2021). Much of Australia's success in containing the spread of the virus hinged on two periods of lockdown, with the first lockdown enforced nationwide for a duration of 8 weeks from late March 2020, and the second occurring in the state of Victoria for the 8 weeks following 8 July 2020. These lockdowns in turn raised new questions about the mental health and well-being of the Australian population and the ways these could be effectively supported (Fisher et al., 2020).

There is now a rich body of literature examining the links between creativity and well-being from a range of disciplinary perspectives (Humes, 2011; Gillam, 2018; Garcês et al., 2020; Kiernan et al., 2020). Psychological and philosophical approaches to the study of creativity have often focused on the “Four Ps” of creativity: people, product, process, and press (i.e., environment) (Rhodes, 1961), or have identified different levels of creative behavior, such as the “Four C model,” containing “Big-C” (eminent-level creativity), “Pro-c” (professional level creativity), “little-c” (everyday creativity) and “mini-c” (rudimentary creativity) (Kaufman and Beghetto, 2009). Such approaches have typically viewed creativity as the production of novelty that is both relevant and effective or appropriate in a given context (Cropley, 2011), and as such they usually position creative behavior in opposition to routine or habitual behavior. However, creativity and routine may be interrelated with habits forming a basis for creativity (Dalton, 2004; Chan, 2016). Some studies have also expanded the unit of analysis in creativity research by treating creativity as a socio-cultural act (Glaveanu, 2015), which may be distributed across and between multiple actors and elements (Glaveanu, 2014; Clarke and Doffman, 2017; O'Dair, 2019). Alongside these developments, creativity has also become increasingly relevant to the study of well-being (Basadur and Basadur, 2011; Krippner, 2011; Csikszentmihalyi, 2014; Gillam, 2018; Barker et al., 2019; Kiernan et al., 2020), where emotions themselves may be viewed from a sociological perspective as creative practices (Kiernan, 2020).

The term “well-being,” like “creativity,” also has multiple, domain-specific definitions (Dodge et al., 2012; Oades and Heazlewood, 2017). The hedonic tradition of well-being research has emphasized emotion-related elements such as pleasure attainment and pain avoidance and is sometimes referred to as “subjective well-being” (Kahneman et al., 1991; Diener, 2009), while the eudaimonic tradition has focused on issues such as meaning, self-realization and human development, which are typically associated with “psychological well-being” (Rogers, 1961; Keyes, 2002; Ryff and Singer, 2008; Vittersø, 2016). Well-being research has often intersected with the study of mental health (Gillam, 2018; Barker et al., 2019), partly because anxiety and depression (Galinha and Pais-Ribeiro, 2012) and loneliness (VanderWeele et al., 2012) have been shown to predict subjective well-being. Given this breadth, studies in well-being must clearly delimit which aspect of well-being is being investigated (Dodge et al., 2012). An increasing body of research is also demonstrating clear links between the creative arts, health and well-being (Clift and Camic, 2016; APPG, 2017). Recent studies have proposed that specific emotion-regulation strategies can underpin engagement in ACAs (Fancourt et al., 2019, 2020), while other studies have taken a broader approach to considering the use of ACAs in regulating emotion (in the short-term), mood (in the mid-term) and mental health (in the long-term) (Leckey, 2011; Lith et al., 2013; Fancourt et al., 2016a,b; Vanstone et al., 2016; Garrido, 2017; Leubner and Hinterberger, 2017; Krause et al., 2018; Williams et al., 2018, 2019).

The principal aim of this study was to investigate the role of ACAs in regulating emotion and supporting the mental health and well-being of people in Australia during the COVID-19 pandemic. The secondary aim was to identify which demographic populations were most at risk of poor mental health and well-being. Specifically, we investigated the following five research questions (RQs):

Based on the literature discussed in section “Emotion regulation,” our principal hypothesis was that ACAs would be used during the pandemic to employ avoidance-based emotion regulation strategies rather than approach-based or self-development-based strategies, as defined by Fancourt et al. (2019). Our secondary hypothesis was that music would be used during the pandemic to regulate emotion.

Data were gathered for this study using an online, cross-sectional survey of people residing in Australia.

The survey data were examined according to the ACAs undertaken and ceased during the pandemic and their relationship to well-being and emotion regulation.

This study examined the role of artistic creative activities (ACAs) in regulating emotion and supporting the mental health and well-being of Australians during the COVID-19 pandemic. The findings supported our principal hypothesis that ACAs would be used during the pandemic to employ avoidance-based emotion regulation strategies rather than approach- or self-development-based strategies (see RQ 4 in Aims). Fancourt et al. (2019) argued that emotion regulation strategies were linked with positive “healthy” behavioral outcomes regardless of the strategy being employed, stating, “even for ‘avoidance' behaviors when taking part in artistic creative activities, there are no ‘unhealthy' outcomes from the use of these [emotion regulation strategies]” (p. 22; see also Kashdan et al., 2014). Our findings support this statement to a certain extent, since participants completed the ERS-ACA questionnaire regarding the ACA they had identified as being the most effective at making them “feel better.” This suggests that the avoidance strategy that participants deployed via their nominated ACA provided the pathway toward more positive emotional experiences. This argument is also supported by the fact that participants are not likely to have been aware of the emotion regulation choices that were underpinning their engagement in ACAs. This is important, since deliberate and conscious regulatory choices have been a major focus in the field of emotion regulation, even though many emotion regulation choices are likely to be determined implicitly and without deliberate control (Sheppes et al., 2014). It is thus a strength of the current study that we examined emotion regulation strategies during the COVID-19 pandemic via engagement in ACAs. These findings also support previous research which shows that individuals tend toward reappraisal strategies when faced with low-intensity stimuli and distraction or avoidance strategies when faced with high-intensity stimuli such as a global pandemic (Sheppes et al., 2014).

Regarding our first two research questions (RQ 1 and RQ 2), the findings revealed that watching films or TV shows was the most frequently undertaken ACA by participants, followed by listening to music, cookery/baking and reading literature. However, concerning RQ 3, “Watching films or TV shows” was ranked as the eighteenth most effective ACA at making participants feel better, with “Listening to music,” “Other” (a conglomeration of mostly arts and crafts activities; see Supplementary Table 3), “Singing,” and “Dancing” ranking as the four most effective ACAs for this purpose. These findings build on previous studies which have suggested that people engage in passive leisure activities significantly more often than activities requiring clear rules, higher energy exertion and greater challenge (“flow” activities) even when they know the latter, more active, activities are more likely to make them happy (Schiffer and Roberts, 2018). This is perhaps because people seem to be generally error-prone in predicting future happiness, as studies in affective forecasting have shown (Gilbert et al., 2002), although it may also be a reflection of the marked increase in clinically significant symptoms of depression in Australia during the pandemic (Fisher et al., 2020), which include low energy levels, feelings of hopelessness and lack of interest in doing things (Kroenke et al., 2001). This, in turn, may help to explain our finding that passive or receptive activities (Davies et al., 2012) accounted for three of the four most frequently undertaken ACAs (these being watching films or TV shows, listening to music, and reading novels, stories, poetry or plays; see Supplementary Table 1).

Our findings also show that music listening was both the ACA most likely to make participants feel better (RQ 3), and one that participants could feasibly do during the pandemic. This, in conjunction with the results of the within-subjects ANOVA performed on the MusEQ subscales, supported our secondary hypothesis that music would be used during the pandemic for the purposes of emotion regulation. However, the negative correlation between loneliness (both before and during the pandemic) and the MusEQ “Emotion” subscale suggested that people who were less lonely were less likely to engage with music for the purpose of emotion/mood regulation. Furthermore, while dancing and singing were also among the most effective ACAs at making participants feel better (RQ 3), they also ranked among the activities that participants were most likely to cease during the pandemic (RQ 2), along with rehearsing/performing plays, drama, opera, and musical theater. Notably, three of the top four most effective ACAs at making participants feel better (RQ 3) involved music or dance, whereas listening to music was the only music- or dance-related ACA to feature in the top ten most frequently undertaken ACAs during the pandemic (RQ 1). This is perhaps because dancing, rehearsing plays (opera, musical theater, etc.) and singing often require significant physical space, and for many people this commodity was at a premium during periods of lockdown. The findings of the current study therefore suggest that music listening is a particularly feasible and efficient ACA for supporting well-being during a pandemic, while also suggesting that singing and dancing are ACAs that may warrant more support during future crises of this kind, given their capacity to make people feel better and the difficulty that participants faced in undertaking these activities. From a mental health and well-being perspective, it may be particularly beneficial for future studies to investigate ways of making singing and dancing more possible and feasible during pandemic crises, as some researchers are already doing (Bohn and Hogue, 2021). Future studies of this kind might also explore possible links between the well-being benefits of engaging in ACAs during a pandemic and profession, which, for reasons of space, could not be explored in the present study. It is reasonable to suggest that professional artists and performers, creative arts and music therapists, and amateur or recreational artists and musicians may engage in ACAs in different ways and for different reasons during periods of enforced lockdown associated with public health crises.

Given that music listening is typically considered a passive or receptive activity (Davies et al., 2012), it is interesting that music listening was the only activity of this type to feature among the ten most effective ACAs at making participants feel better (see Supplementary Table 5), except for “Other,” which itself only includes one passive/receptive activity (see Supplementary Table 3). This suggests that music listening differs in some important well-being-related respects from other passive or receptive ACAs such as reading or watching films. This can perhaps be explained by theories of music listening that frame the activity as active and imaginative rather than as a passive act of perception, such as persona theory (Robinson, 2005; Cochrane, 2010; Peters, 2015). However, this well-being benefit may also be explained by the fact that music listening can be integrated into, or accompany, daily activities such as cooking or doing house chores much better than some other receptive activities such as watching TV/films or reading (DeNora, 2000; Vanstone et al., 2016), which in turn suggests that the well-being benefit may originate in the combination of aesthetic and practical or task-oriented elements. Since music listening ranked as the most effective ACA at making participants feel better during the pandemic, further investigation of the processes by which music listening facilitates well-being benefit is warranted.

Concerning RQ 5, the findings also showed that anxious and depressed Australians during the COVID-19 pandemic seem to be turning to music as a coping mechanism or emotional crutch significantly more than people who were not experiencing these mental health issues, or those who were experiencing these issues to a lesser degree. This was also supported by two other findings: a) significant negative correlations between well-being scales and age, indicating that the pandemic had a significantly greater negative impact on the anxiety, depression and loneliness of younger participants than older participants; and b) significant negative correlations between the MusEQ subscales and age, indicating that younger participants showed stronger engagement with music than older participants. This supports prior suggestions that music plays a particularly crucial role in adolescents' development of identity (Tarrant et al., 2002). It also supports the work of Groarke and Hogan (2016), who examined how the adaptive functions of music listening co-function within an enhancement system that supports well-being. These authors observed age differences in the ways that functions of music listening were considered adaptive, with younger participants emphasizing affect regulation and social connection, while older participants emphasized eudaimonic functions such as transcendence and personal growth (p. 769). The finding that younger people were more negatively impacted by the pandemic in terms of anxiety, depression and loneliness than older people also warrants consideration. This could perhaps be a reflection of the destabilizing impact of the pandemic on younger people as they attempt to plan for their future (gaining education and employment, starting families, and so on), although the prevalence of generalized anxiety disorder is expected to decrease with age, partly because emotional control typically increases and anxiety and worry are expressed differently in older adults (Nilsson et al., 2019).

Also concerning RQ 5, this study reported significant positive relationships between participants who reported exercising more during the pandemic, and high scores (i.e. feeling better) on three of the four well-being scales. These three scales were specifically for anxiety, depression, and loneliness “since the pandemic.” Furthermore, the fact that this relationship was not present for loneliness “before the pandemic” suggests that the role and importance of exercise on mental health may have increased during the pandemic. However, the causality of these relationships cannot be established. While it is possible that increasing the amount of exercise undertaken during the pandemic had the effect of reducing symptoms of anxiety, depression and loneliness, it is also possible that those who were less impacted by these factors simply felt more motivated to exercise. Indeed, a mixture of both interactions could be present. Similarly, as negative relationships were observed between all four well-being scales and Activity hours, it is possible that an increase in ACA engagement tended to produce small decreases in well-being, or that respondents who felt more anxious, depressed, and lonely were less likely to spend as many hours engaging with ACAs overall; as above, it is not possible to determine the causality of this observed relationship. While no significant relationship was observed between household size and loneliness ratings “since the pandemic,” household produced a significant positive correlation with loneliness ratings “before the pandemic.” This finding suggests that prior to the COVID-19 pandemic, those living in larger households tended to feel less-lonely than those living in smaller households, although this difference has since disappeared. Future studies should examine this relationship in additional detail to pinpoint if certain household sizes provide an optimal boost to well-being during a pandemic, and similarly which household sizes can be expected as most at-risk. Future studies might also explore whether associations exist between demographic data, engagement in ACAs, and other factors that have become associated with life during periods of lockdown such as working from home, home schooling and childcare arrangements.

We conclude that artistic creative activities, including musical activities, have served as an important resource for facilitating emotion regulation and supporting the mental health and well-being of Australians during the COVID-19 pandemic. Listening to music was the most effective ACA for making participants in this study “feel better,” and while dancing and singing were also ranked as effective, these two ACAs were among the most frequently ceased during the pandemic. We suggest that, where possible, specific ACAs should be prioritized during a pandemic for well-being purposes, and that further research is needed into developing pandemic-safe approaches to dancing and singing, which showed strong emotion-regulatory potential, but which participants found difficult to pursue. Additionally, given its prevalence, further research into what distinguishes music listening from other passive or receptive creative activities (Davies et al., 2012) in terms of well-being benefits may prove beneficial for future crises on a large scale. We also note that younger Australians and those exercising less appear to be the most at risk of experiencing symptoms of anxiety, depression and loneliness. This indicates that specific care needs to be taken in circumstances such as the COVID-19 pandemic to support these at-risk populations. Finally, we argue that these findings indicate that ACAs can form an important and efficient part of Australia's mental health response to pandemic crises.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the Human Research Ethics Committee of The University of Melbourne. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

FK, JD, MH, and SG developed the research questions and study design and obtained ethics approval. The analyses were undertaken by AC and the findings were discussed by all authors. The first draft of the article was produced by FK and AC with guiding input from JD. Subsequent drafts and revisions also included further substantial intellectual input from MH and SG. All authors helped circulate the survey and recruit participants. The final draft was approved by all authors.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.