Introduction

The use of social media has become widespread during later years, and many–first and foremost younger people–spend considerable time engaging with them [1,2]. The opportunities that were introduced with social media were manifold. It became possible to engage with people and events across the world only by a keystroke, and news, entertainment, and a plethora of special interest groups were instantly accessible. This led to optimism, particularly due to the prospects of being able to utilize such new technologies to support interpersonal connection and wellbeing, as some studies showed [3,4].

Over the years, however, researchers have inquired about a range of possible negative consequences of social media, often including poorer mental health, or factors assumed to be in the pathway between social media use and poorer mental health. Theoretical models focusing on motivation, such as the uses and gratifications theory [5], posit that people use social media for the perceived gratifications they can provide. However, if the person’s needs (e.g., for belonging and support) exceed the perceived gratification of these needs, using social media may become excessive based on the hope that more intense use will result in the desired needs gratification [6]. Excessive social media use may develop into social media addiction, generally referring a maladaptive pychological dependency of online social networks to the extent that behavioral addiction symptoms occur [6]. In turn, studies suggest that such behaviors may impede normal functioning in central areas of life such as relationships, studies and work, and health and well-being [7]. With a view to mental health, studies have shown associations between higher social media use and factors such as stress [8,9], anxiety and depression symptoms [10–13], and loneliness [14–16]. In addition, motives for social media use, and not least the personal experience from using them, may be even more relevant for health outcomes than the actual time spent using them, as suggested from recent studies [13,17–19].

Health promoting behaviors, such as physical activity and adequate sleep, have also been studied in relation to social media use. Studies have shown lower levels of physical activity [20], and lower levels of adequate sleep [21,22], among people with higher levels and intensity of social media use. Sleep quality and sleep problems may also be related to the time of the day when social media are used, as late night use has been shown to be related to poorer sleep quality among university students while using them earlier in the evening was related to better sleep [23]. In addition, physical activity and sleep quality have been shown to partially mediate relationships between social media use and mental health and wellbeing [24,25]. In other words, the relationship between social media use and poorer mental health and wellbeing may in part be accounted for by the negative impact of social media use on health-promoting lifestyle behaviors. Considering the potential of social media to interfere with sleep [26,27], the higher psychological distress levels among those with sleep problems [28], and the increasing attention to social media addiction across the world [29], further investigations into the role of social media for sleep problems are warranted.

The pandemic that besieged the world in early 2020, referred to as the COVID-19 pandemic, did not spare any country its devastating impacts [30]. Contracting the disease had multiple negative effects, most notably difficulty breathing, and could be fatal, in particular for old people and people with chronic health problems. In fact, by January 2022 more than six million people worldwide had died from the disease [30]. Although the number of cases and deaths varied substantially between countries, the pandemic had a range of other consequences, such as schools, shops and workplaces closing, leading to a number of challenges for individuals and families such as home schooling, working remotely, or even losing one’s job. Thus, many people’s daily routines changed radically during the pandemic and a general sense of crisis emerged, with inherent risks of inducing sleep problems. Indeed, several studies surfaced that focused on sleep problems during the pandemic. A number of systematic reviews and meta-analyses of such studies reported a global prevalence of sleep problems in the general population. For example, Alimoradi and co-workers reported a prevalence of 18% [31], whereas Jahrami and co-workers reported a prevalence of 32% [32]. Despite diverging prevalence rates across these meta-analyses, they substantiate that sleep problems was a major public health issue at the time. Moreover, while the psychological consequences of the pandemic were widespread in the population, those who had personal experience with the disease, or had experienced that someone in the family had been ill, were reported to have comparably poorer mental health and more worry about themselves and their close ones [33], factors which in turn have been closely linked with sleep problems [34].

Due to restrictions during the pandemic, individuals leaned more on social media and other electronic methods of communication with family and friends, but also more generally has the use of social media increased over the past years. Between 2015 and 2023 the average daily time spent using social media increased by approximately 40 minutes, with the recent average being reported at 151 minutes [35]. Considering the evidence, the increased use of social media may give rise to more sleep problems among people. However, while the literature on associations between social media use, sleep problems, and poorer mental health is substantial, many of these studies have focused on children and adolescents [e.g., 26,27,36,37]. Moreover, these associations may be different in a pandemic context where a range of other factors and events can give cause for sleep problems. The gap in the literature concerning these associations among the general adult population, and during the extraordinary context of the COVID-19 pandemic, constitutes the rationale for this study. The aim of the study was to answer the question: What is the association between daily hours of social media use and self-reported sleep problems in a cross-national adult population in the late COVID-19 pandemic phase?

Materials and methods

Results

Sample characteristics in relation to sleep problems

Of the 858 (61.1%) participants who reported sleep problems during the past weeks, a substantial proportion (n = 353, 41.1%) related their sleep problems to their experience with COVID-19. Women reported sleep problems more often than men (64% versus 49%, p<0.001). Compared to their counterparts, participants with sleep problems had higher mean GHQ rating (16.9 versus 11.7, p<0.001), indicating higher levels of psychological distress, and reported more time spent on social media (mean rating 4.4 versus 4.1, p<0.001). No other variables showed a significant relationship to sleep problems in the initial analyses. A detailed overview of the sample characteristics as related to sleep problems is provided in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Sample characteristics by sleep problems.

https://doi.org/10.1371/journal.pone.0318507.t001

Adjusted associations with sleep problems

Based on the single variable logistic regression analyses, female gender, higher GHQ ratings, and more time spent on social media were significantly associated with sleep problems (all p<0.001). Age and employment were carried over to the multivariable analysis due to associations with sleep problems showing p≤0.30.

In the multivariable (adjusted) analysis, sleep problems were associated with higher age (OR: 1.13, p = 0.01), female gender (OR: 1.69, p<0.001), having employment (OR: 1.34, p = 0.04), and higher GHQ ratings (OR: 1.20, p<0.001). Time spent on social media was no longer significantly associated with sleep problems after adjustment (OR: 1.06, ns). Results from the single and multivariable logistic regression analyses are displayed in Table 2.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Unadjusted and adjusted logistic regression analyses showing associations with sleep problems (n = 1405).

https://doi.org/10.1371/journal.pone.0318507.t002

Discussion

Although social media has often been blamed for disrupting the sleep and subsequent psychosocial wellbeing of individuals [26,27], the findings from this study suggest that in the pandemic context, they may be falsely blamed. The association between time spent on social media and sleep problems was no longer statistically significant in the multivariable analysis adjusting for other variables. This suggests that while there was an initial bivariate relationship between social media usage and sleep problems, this relationship was likely the result of one or more variables being related to both social media use and sleep problems (confounding). Female gender, higher psychological distress, higher age, and having employment were significantly associated with sleep problems after adjustment.

Noteworthy associations with sleep problems were shown for female gender and psychological distress, and there is reason to suggest that these factors might also explain the results for social media use. The higher levels of psychological distress and sleep problems among women has been well documented, both prior to and during the COVID-19 pandemic [50–52]. Studies have also pointed to girls and women as more frequent users of social media, compared to boys and men [53–55], and also that the association between higher social media use and lower wellbeing is stronger for girls [54]. Thus, if adult women use social media more than men and also respond to their social media experiences with more distress, this may explain why the association between social media use and sleep problems was weakened and no longer statistically significant after controlling for gender and psychological distress.

While female gender and higher psychological distress are linked, the multivariate model showed that both of these variables had a unique relationship with sleep problems. Thus, more sleep problems among women than men may partially, but not fully, be explained with reference to higher psychological distress levels among women. The processes driving sleep problems may therefore concern a range of other factors than psychological distress. Previous research has suggested biological factors such as differences in sex steroids [56] and women’s menstruational cycle [57], but women may also be psychologically more prone to lie awake pondering about things, without necessarily feeling distressed. As we do not have data to explore these possibilities further, we suggest that future research may address the processes driving sleep problems in a gender perspective.

The association between age and sleep problems was suppressed in the unadjusted analysis, but became statistically significant when cancelling out the effects of the other included variables. During the pandemic crisis, people of younger age have reported higher levels of psychological distress, compared to older people [58,59], despite older people being more prone to become seriously ill from contracting COVID. Thus, confounding from psychological distress may explain the initial results for the association between age and sleep problems. However, after adjustment, higher age was associated with sleep problems,which is in line with results from previous studies demonstrating that sleep problems is common among people of older age [60,61].

Having employment was significantly associated with sleep problems in the multivariable analysis. This was a surprise, considering that not having a job would commonly be related to more economic concerns, which in turn may affect sleep negatively [62–64]. However, having a job may also be the cause of daily concerns, both directly related to the work tasks and the work environment [65], but also indirectly, considering for example commuting and balancing work-life commitments. Moreover, being employed in the COVID-19 era may be particularly stressful, given the demands for adapting to new ways of organizing work, including more digitalization, more working from home–and therefore also more role overload and family distraction [66]. Possibly, these and other aspects of having employment during the COVID-19 pandemic might translate into more sleep problems.

No association was shown between having had the COVID disease and sleep problems. While previous studies have shown considerably more sleep problems among COVID patients compared to the general population (75% vs. 32%) [32] one should consider the risks as perceived in different phases of the pandemic. In the early phase, the number of cases and deaths was rocketing and there were still no vaccines available, both of which contributed to incrasing people’s psychological distress [67]. In later phases of the pandemic, people had had time to adapt to the situation, vaccines had become available, and many had already been infected with COVID and had recovered. All of these factors would indicate that at the time of the data collection for this study, approximately two years after the pandemic outbreak, sleep problems would not be significantly different for people who had been infected with COVID compared to those who had not.

Limitations

The study had a cross-sectional design, thus the results cannot be used to imply cause and effect relationships. We do not have information about the specific social media people used, and we do not know how, why, or when they were used, since this study only collected data on time spent using social media. As the study is concerned with associations with sleep problems, the lacking information about these aspects of social media use represent a limitation. The recruitment strategy, relying on dissemination on social media, required that all participants had to use a computer, tablet, or smartphone to be able to access and respond to the survey. Thus, the population should be specified as ‘users of social media’ as a subset of the general population. In the dataset we noticed more participants in the younger age groups and assume this may have been a result of the data collection strategy. The study may have limited transferability to the general population, given the skewed sample distributions on age, gender and education. Moreover, while all countries are considered representative of Western culture, there may still be cultural differences between the countries. Given that larger proportions of younger participants came from the USA and UK, and larger proportions of older participants came from Australia, it is possible that the age-related results are influenced by culture. Limitations also include that the results are solely based on the participants’ self-report, and the use of measurement methods–including the sleep problems measure–being suboptimal and with unknown psychometric properties. Suggestions for the methodological improvement of future studies in this area thereore include longitudinal data collections, sampling methods that secure the sample’s representativity of the selected population, use of well-established and good quality measurements, and a more detailed inquiry into the various aspects of social media use that goes beyond time use.

Conclusions

To improve the health and wellness of individuals globally, knowledge about their influencing factors are vital. This study showed, in contrast to expectations, that time spent using social media was unrelated to people’s sleep problems during the late phase of the COVID-19 pandemic. Higher psychological distress levels, on the other hand, were associated with sleep problems, and women were more inclined to report sleep problems than men. While the results suggest that time on social media in and of itself may not play a role for people’s sleep quality, other aspects of social media use (eg, time of the day, content of interactions, associated stress experience) may be more relevant for understanding sleep problems. Investigating the complexities of social media use in relation to sleep patterns and problems appears to be a fruitful way forward. In addition, as the study showed that characteristics such as age, gender, psychological distress, and employment contribute to predicting sleep disturbances, it underscores that these factors should continue to be used for adjustment in the future exploration of other predictors. Increasing the understanding of the complexity of factors related to sleep problems can support the development of prevention and intervention strategies to improve sleep and overall psychosocial wellness.

Supporting information

Acknowledgments

The authors are grateful for the time and efforts made by the survey participants, and thank Hilde Thygesen (University of South-Eastern Norway, Norway), Mary Ruffolo (University of Michigan, USA) and Janni Leung (The University of Queensland, Australia) for their contributions to the research group in general and for their contributions to the data collection in particular.