Abstract

Introduction

Poor sleep among the general population is prevalent and rising.1,2 Notably, up to 50% of young people rate their sleep quality as poor, and a majority are chronically sleep deprived.3 The negative mental health outcomes of poor sleep have been consistently documented, including links to anxiety, mood disorders, substance use, impulsivity, and aggression,3–6 alongside links to physical health risks such as lower immunity, cancer, cardiovascular disease, and type 2 diabetes.7,8 Current National Institute for Health and Care Excellence (NICE) guidelines recommend face-to-face Cognitive Behavioral Therapy for Insomnia (CBT-I) as first-line treatment for insomnia.9 Yet, this is currently not readily available in the UK, and there is no direct evidence comparing alternative online self-help programs eg Sleepio with face-to-face CBT-I.9 Additionally, CBT-I has a non-response rate of up to 40% and a high remission rate.10 Alternatively, short-term pharmacological treatment for insomnia is advised;11 short-term effects are comparable to CBT-I, and combined, the two treatments have increased efficacy.12 However, pharmacological treatment comes with disadvantages including a variety of side effects and risk of dependency.9,12 Furthermore, insomnia typically returns following withdrawal of pharmacological treatment and a meta-analysis found that the effects of long-term pharmacological treatment may decrease over time.12 Hence, there is growing interest in alternative therapies for reducing sleep problems, such as mindfulness-based interventions (MBIs), which have shown efficacy comparable to CBT-I with high adherence, and no adverse side effects.13,14

Higher trait mindfulness – an individual’s innate tendency to purposefully maintain awareness of the present moment in an open-minded and non-evaluative manner15 – has been linked to a range of positive physical and mental health outcomes including better sleep in a recent meta-analysis.16 There is also evidence that sleep quality serves as a mediator in the relationship between trait mindfulness and better wellbeing, and other research has identified sleep quality as a mediator in the relationship between trait mindfulness and physical health.17,18 These findings highlight the important role of sleep quality in explaining the influence of trait mindfulness on both mental and physical health; however, the relationship between sleep quality and trait mindfulness is not well understood. Cross-sectional19–21 and experimental22 studies investigating the relationship between trait mindfulness and sleep quality have identified psychological distress, eg perceived stress, as a mediating variable. However, this has only been investigated using subjective sleep measures, and not in young adults specifically. There is evidence indicating that subjective and objective sleep measures do not always correlate, perhaps due to underreported sleep duration and overestimated sleep latency.23,24 Given the prevalence and implications of poor sleep among young adults in particular, the relationship between trait mindfulness and sleep quality, and the potential mediating variables, requires further attention.

Factor analyses have led to the conceptualization of trait mindfulness as comprising five facets; mindfulness is the act of observing or attending to internal and external experiences and describing these with awareness of the present moment. This is done with non-judgement of thoughts and feelings and non-reactivity; allowing these to come and go.25 The Five Facets of Mindfulness Questionnaire (FFMQ) was designed to measure mindfulness at the facet level and is widely used.25,26 In adults with insomnia, higher levels of awareness have been linked to less severe symptoms of insomnia, less sleep disturbance, and fewer dysfunctional beliefs about sleep; non-judging was also associated with fewer dysfunctional beliefs about sleep, and both non-judging and non-reactivity were linked to less sleep difficulty.27 Likewise, Xie et al found that awareness, non-judging, and non-reactivity were associated with better sleep health.28 These findings suggest that awareness, non-judging, and non-reactivity play key roles in influencing sleep quality. It is proposed that awareness is necessary to allow modification of thoughts.29 Ong et al’s Meta-cognitive Model of Insomnia posits that primary arousal, involving negative beliefs and attitudes about sleep, directly inhibits sleep.30 The theory describes an iterative process whereby secondary arousal, including awareness of negative beliefs and attitudes, and attachment to them, further precipitates negative cognition, ultimately resulting in further sleep difficulties. Acceptance (non-judging and non-reactivity) deactivates hyperarousal, shifting meta-cognition to a more objective and adaptive state, thus reducing sleep difficulties.

On the other hand, it is well established that high levels of perceived stress negatively impact sleep quality.31–33 Perceived stress refers to the extent to which an individual finds their life unpredictable, uncontrollable, and overloaded.34 Extensive cross-sectional35,36 and longitudinal37 evidence demonstrates a relationship between perceived stress and poor sleep quality, using a variety of self-report and objective measures.38–41 Perceived stress has also been associated with lower levels of trait mindfulness.42 It is thought that mindfulness mitigates stress appraisals and stress reactivity responses, in turn enhancing sleep quality, and other health outcomes.43–45 This may explain why perceived stress appears to mediate the relationship between trait mindfulness and sleep quality,19–21 and provides a rationale for exploring this in more detail, and with inclusion of objective sleep measures.

There is also considerable evidence that rumination, defined by Trapnell & Campbell as “a self-attentiveness motivated by perceived threats, losses or injustices to the self”46 negatively impacts sleep quality due to heightened cognitive and physiological activity prior to sleep onset.47–49 This has been attributed to the negatively toned cognitive arousal associated with rumination, eg repeatedly contemplating the causes and outcomes of a stressful event after it has occurred, with negative self-focus.50,51 Like perceived stress, previous studies have shown that rumination impacts sleep quality based on both subjective,52 and objective measures,53,54 affecting sleep latency,55 disturbance,56 duration,57 and efficiency,58 consequently increasing daytime dysfunction.59 The opposing roles of mindfulness and rumination are not surprising since these are theoretically opposing constructs, which are negatively correlated.60 The Monitor and Acceptance Theory suggests that mindfulness may counteract rumination: attention monitoring (observing), alongside acceptance (non-judging and non-reactivity) can inhibit the negative emotional and cognitive mechanisms linked to rumination.61 Given the interrelations between trait mindfulness, perceived stress, and rumination, this provides a rationale for exploring rumination as another possible mediator in the relationship between trait mindfulness and sleep quality. Rumination has been found to mediate the relationship between mindfulness and poor sleep quality in a previous study.62 The present study is the first to consider both rumination and perceived stress as potential mediators in the relationship between trait mindfulness and sleep quality.

Method

Data Analysis Plan

Data were analyzed using IBM SPSS version 27; a p value of <0.05 was considered statistically significant. There were no outliers within the dataset, with all z-scores falling within the cut-off points of ±3.29.66 Questionnaire measures were completed in full by all participants and for analyses of questionnaire measures the two samples were combined (N = 170). Pearson’s correlations were used to assess relationships between the questionnaire measures as all z-scores for skewness and kurtosis were within the cut-off points of ±3.29.67 From the secondary sample (N = 30), only 23 participants provided seven valid days of actigraphy data, and only 21 participants provided complete sleep diaries; incomplete responses were excluded from analyses. To assess relationships between diary- and actigraphy-based sleep quality and the other measures, both Spearman’s and Pearson’s correlations were used, as diary-based sleep latency, efficiency and self-rated quality, and actigraphy-based efficiency were non-normally distributed (see Supplementary Table 2). Bonferroni correction was used to control for family-wise error in multiple correlations. Mediations were assessed using model four in PROCESS version 3.5 for SPSS. The mediation model is depicted in Figure 1. Assumptions of accuracy, normality, linearity, multicollinearity, homogeneity, and homoscedasticity were met. The indirect effects and 95% confidence intervals were calculated based on 5000 bootstrap samples; effects were considered significant if confidence intervals did not overlap with zero. Mediations based on diary- and actigraphy-based sleep quality could not be assessed due to a lack of significant correlations with trait mindfulness and perceived stress.

Open in a separate window

Figure 1

Hypothesized mediation between trait mindfulness, sleep quality, perceived stress, and rumination.

Results

Mediation Analysis

A mediation analyzing the effect of trait mindfulness on sleep quality, measured by the PSQI, with perceived stress and rumination as mediators and age and gender as covariates was conducted (Figure 1). High levels of mindfulness were negatively associated with perceived stress, b = −0.37, t(166) = −10.50, p < 0.001, and high perceived stress was positively related to poor subjective sleep quality, b = 0.24, t(164) = 5.98, p < 0.001. High levels of mindfulness were also negatively associated with rumination, b = −0.03, t(166) = −8.83, p < 0.001; however, rumination was not significantly associated with subjective sleep quality, b = 0.19, t(164) = 0.46, p = 0.64. The indirect effect through perceived stress was significant, effect size = 0.08, 95% CI [−0.12, −0.06], standardized coefficient β = −0.30, 95% CI [−0.41, −0.21]. However, the indirect effect through rumination was not, effect size = −0.01, 95% CI [−0.03, 0.02], standardized coefficient β = −0.02, 95% CI [−0.11, 0.07]. This indicates that perceived stress fully mediated the relationship between trait mindfulness and subjective sleep quality. Gender was a significant covariate in the relationship between mindfulness and rumination, b = 0.35, t(166) = 3.03, p = 0.0028. Thus, results confirmed the hypothesis that trait mindfulness is indirectly associated with subjective sleep quality through perceived stress but did not indicate any role of rumination in the relationship between trait mindfulness and subjective sleep quality (see Figure 2).

Open in a separate window

Figure 2

The mediating effect of perceived stress in the relationship between trait mindfulness and subjective sleep quality. The negative association between trait mindfulness and sleep quality was fully mediated by perceived stress.

Note: *p < 0.001.

Discussion

This study sought to investigate the interacting effects of trait mindfulness, perceived stress, and rumination in relation to subjective and objective sleep quality. Better sleep quality has been consistently linked to trait mindfulness, however, knowledge around the mediating variables, particularly in young adults, is limited. As expected, high levels of trait mindfulness were associated with low levels of perceived stress, rumination, and better subjective sleep quality (measured by the PSQI). Both rumination and perceived stress were associated with poorer subjective sleep quality on the PSQI. Importantly, perceived stress was found to fully mediate the relationship between trait mindfulness and subjective sleep quality; however, no mediating role was found for rumination. Other than the positive (uncorrected) association between trait mindfulness and diary-based sleep ratings, none of the sleep diary- or actigraphy-based measures were related to any of the other variables, but these results should be regarded as preliminary given the small secondary sample who provided these data. We also explored the effects of specific mindfulness facets.

The negative association between trait mindfulness and rumination is consistent with previous studies and the fact that they are theoretically opposed constructs.60 The Monitor and Acceptance Theory suggests that mindful individuals are more likely to disengage from ruminative thoughts, resulting first from observing, followed by non-judging and non-reactivity, thus reducing stress and stress-related outcomes.19,60,61 While heightened observing of thoughts is a characteristic of high ruminators (a positive association with rumination was found here), the other facets (including non-judging and non-reactivity) were negatively associated with rumination, highlighting the importance of these other factors for counteracting ruminative tendencies. In turn, rumination was negatively related to subjective sleep quality, as is well demonstrated in the literature.47 The heightened cognitive and physiological arousal caused by rumination is thought to underlie this.

Trait mindfulness, and the facets of non-judging, non-reactivity, awareness, and describing, were negatively correlated with perceived stress in the present study. Higher trait mindfulness is consistently associated with lower perceived stress,42,69,70 although fewer studies have explored the effects of individual facets. Here, observing was not associated with perceived stress. This finding is of value as it provides support for previous studies that found likewise. Both Ballantyne et al and McBride et al found that while the other facets were linked to reduced perceived stress, observing was associated with higher stress and physical symptoms of stress.71,72 Again, in relation to Lindsay & Creswell’s Monitor and Acceptance Theory, it seems that attention monitoring alone is insufficient to reduce stress and might in fact be detrimental.61 Without the ability to respond to present moment experience in a non-judgmental/non-reactive way, higher levels of observing may lead to distress and maladaptive coping.73 Other negative effects of observing have been noted, eg the meta-analysis by Sala et al found that higher observing was associated with higher substance use.16

As expected, perceived stress and trait mindfulness were inversely related to subjective sleep quality. Sala et al found that acting with awareness and describing were the most important facets regarding sleep quality while observing was not related, possibly due to higher engagement in health risk behaviors, eg substance use.16 Here, non-judging, non-reactivity, awareness, and describing were seen to be of benefit, while observing had no relationship with sleep quality. In other studies, in undergraduate samples, acting with awareness and non-judgement were associated with higher sleep quality on the PSQI, while observing had negative effects.18,74 Recent work in an undergraduate student sample found that the associations between observing and sleep health were moderated by non-reactivity: observing was associated with worse sleep at lower levels of non-reactivity, but better sleep health at higher levels of non-reactivity.28

Trait mindfulness is known to be a buffer against psychological distress and perceived stress, which in turn impacts health behaviors such as sleep quality.19–21 Simione et al found that, in a general adult sample, mindfulness was negatively related to stress and this effect fully mediated the positive relationship between mindfulness and sleep quality.21 Lau et al and Roberts & Danoff-Burg found a partial mediation effect in undergraduate samples.19,20 The present study sought to clarify this relationship, focusing on young adults, and by including rumination as an additional mediator. This is of value given the well-characterized relationships between mindfulness and rumination, and between rumination and sleep. Thus, reduced rumination is a potential mechanism through which mindfulness could be linked to better sleep. This is the first study to test such a model. Results showed that the relationship between trait mindfulness and subjective sleep quality is fully mediated by perceived stress. However, no mediating effect of rumination was observed.

This confirms that perceived stress plays a key role in the relationship. Thus, mindfulness appears to diminish the negative effects of stress, particularly in high-stress populations such as students, and when there is a direct impact of stress on the health outcome, eg sleep.44 Trait mindfulness and mindfulness-based training have been shown to influence affect regulation, and mindfulness-based training has been shown to alter stress reactivity; this is thought to impact health outcomes by reducing physiological stress responses.44,75,76 These findings provide valuable insight into the mechanisms linking trait mindfulness to sleep quality and are of high relevance to young adults, particularly students who often report poor sleep and high levels of stress;77 stress appears to be the key mechanism and mediating factor (rather than rumination) in the relationship. This finding informs future research and intervention designs: it suggests that interventions should emphasize stress reduction within MBIs, for improving sleep quality among young adults. This is crucial given the high prevalence of sleep issues reported among young adult populations (as seen here and in previous research) and the important consequences of this for young adults’ daytime functioning, as well as longer-term physical and mental health outcomes.

A small secondary sample contributed objective sleep measures, allowing us to conduct exploratory analyses on these. This is of value given that the overwhelming majority of previous work has used subjective sleep measures only. Comparative studies in the sleep literature indicate that participants are often inaccurate when estimating their sleep duration and number of nighttime awakenings,13 with low correlations between self-reported sleep and actigraphy data.78 Here, the objective actigraphy-based measures were not associated with perceived stress, rumination, or trait mindfulness. This accords with a recent meta-analysis by Clancy et al,79 showing that links between rumination and sleep quality are much stronger for subjective, compared to objective measures of sleep quality. High rumination might create a cognitive bias for perceiving and reporting poor sleep quality, as theorized by Harvey et al.80 Similar perception biases could serve to partially explain the relationships between stress, mindfulness, and subjective sleep quality as well. The lack of relationships with the objective sleep measures employed here would support this, but it is important to acknowledge that the non-significant findings may be due to the secondary sample being small and underpowered, potentially resulting in a Type II error. Thus, replication of the present study is required. A recent study in undergraduates derived a composite “sleep health” measure that combined actigraphy data with subjective sleep ratings; this measure correlated with higher mindfulness scores in awareness, non-reactivity, and non-judgment,28 it would be of interest to explore whether this was also the case if only the objective ratings were included.

A weakness of the current study is that we relied on subjective measures of the traits under investigation, making the study susceptible to various biases including social desirability bias, demand characteristics, and order effects. The focus on trait-like characteristics is also a limitation. Pre-sleep rumination has the strongest negative impact on sleep;52 hence, future research should aim to employ measures of state rumination specifically assessing pre-sleep rumination, eg via experience sampling, to confirm the current findings.

Additionally, the study sample was not randomly selected and was recruited from one university, potentially skewing the findings, and limiting their generalizability. The sample was also predominantly female, and the small number of males prevented us from studying potential gender-specific effects. Although a female-dominated sample is not uncommon in the literature, future work should aim to explore the effects in a larger sample that has a more even gender balance.

Finally, a longitudinal design would have allowed inferences regarding causality. In the mediation analysis, trait mindfulness was input as the independent variable, and sleep quality as the dependent variable: this was based on evidence showing reductions in rumination and perceived stress, and improvements in sleep quality, following MBIs.81–84 Still, due to the cross-sectional nature, the findings from this study are strictly correlational and the direction of causality cannot be inferred.

Conclusion

Overall, findings from this study indicate that perceived stress, rather than rumination, is a key mediating variable linking trait mindfulness to better subjective sleep quality in young adults. This adds important mechanistic detail to the literature, and the current findings have practical importance for addressing poor sleep quality, which is highly prevalent among young adults. Results provide a strong rationale for considering MBIs as an alternative to pharmacological treatment, as they have high efficacy and adherence, and no adverse side effects.13 Further, the findings also inform the design of MBIs, indicating that they should focus on reducing perceived stress, to optimize therapeutic outcomes.

References