Daily associations between sleep and physical activity: A systematic review and meta-analysis

Summary

The day-to-day variations of sleep and physical activity are associated with various health outcomes in adults, and previous studies suggested a bidirectional association between these behaviors. The daily associations between sleep and physical activity have been examined in observational or interventional contexts. The primary goal of the current systematic review and meta-analysis was to summarize existing evidence about daily associations between sleep and physical activity outcomes at inter- and intra-individual level in adults. A systematic search of records in eight databases from inception to July 2019 identified 33 peer-reviewed empirical publications that examined daily sleep–physical activity association in adults. The qualitative and quantitative analyses of included studies did not support a bidirectional daily association between sleep outcomes and physical activity. Multilevel meta-analyses showed that three sleep parameters were associated with physical activity the following day: sleep quality, sleep efficiency, and wake after sleep onset. However, the associations were small, and varied in terms of direction and level of variability (e.g., inter- or intra-individual). Daytime physical activity was associated with lower total sleep time the following night at an inter-person level with a small effect size. From a clinical perspective, care providers should monitor the effects of better sleep promotion on physical activity behaviors in their patients. Future studies should examine sleep and physical activity during a longer period and perform additional sophisticated statistical analyses.

Systematic review registration

https://osf.io/w6uy5/.

Introduction

Sleep and physical activity are both important health behaviors in the general population. It has been well-established that low physical activity levels and poor sleep are related to a wide range of medical problems and chronic health conditions, including cardiovascular disease, overweight and obesity, type 2 diabetes, depression, and anxiety disorders [1,2]. Furthermore, these behaviors are likely to interact with each other. Chennaoui et al. [3] presented potential biological pathways explaining possible reciprocal interactions between sleep and physical activity in the short- and long-term. They suggested that acute or repeated physical activity increased Total Sleep Time (TST) through a decrease of insulin resistance and inflammation markers concentration, better regulation of circadian rhythm, and release of brain-derived neurotrophic factor. In return, sleep deprivation could alter the physical performance or facilitate fatigue during physical activity [4] by increasing cortisol concentration, decreasing growth hormone, and prolactin concentration and stimulating inflammation markers [5,6].

Long-term bidirectional associations between sleep and physical activity have been hypothesized in narrative reviews [3,7,8] and examined in prospective studies. Reciprocal relationships between better sleep quality and high physical activity were found over 3 y in samples of students [9] and adults aged 42–72 y [10] in longitudinal studies. Another 2-y longitudinal study found a unidirectional, positive, and significant association between sleep quality and physical activity among community-dwelling older adults [11].

Short-term associations between sleep and physical activity have also been experimentally tested. Previous investigations examined the temporal consequences of a sleep-deprived night on physical activity. An induced sleep restriction was associated with a significantly lower time spent in objectively measured physical activity the following day, as well as the lower intensity of activity [4,12]. However, a contradictory finding was found in another investigation showing that increased energy expenditure was observed the day following a short night [13]. The other side of this association whether physical activity is associated with sleep the following night; have been tested in a recent meta-analysis of experimental studies. This meta-analysis concluded that a session of physical exercise was associated with higher Sleep Efficiency (SE), TST, lower Sleep of Latency (SOL), and Wake After Sleep Onset (WASO) the following night [14]. These short-term associations were characterized by a small-to-moderate effect size. These experimental studies thus suggested a possible bidirectional association between sleep and physical activity outcomes. However, these findings were limited in terms of ecological validity (i.e., the behaviors were not measured in participants' daily life and over multiple days).

In this perspective, previous studies investigated sleep – physical activity associations using ambulatory and daily measures and indicated that these two behaviors exhibited day-to-day variations [[15], [16], [17]]. Furthermore, repeated measures provide data to differentiate inter- and intra-individual variations (i.e., daily deviations from an individual’s usual behavior). Consequently, a better understanding of inter- and intra-individual associations between sleep and physical activity could be developed.

The daily associations between sleep and physical activity have been examined in observational or interventional contexts with self-report or objective measures (e.g., sleep device, pedometer) including a large spectrum of participants (e.g., older people with insomnia symptoms, athletes). An observational study used 7-d accelerometer data to examine the bidirectional association between sleep and physical activity in women with breast cancer [18] and found a unidirectional association. The lower WASO, Total Wake Time (TWT), and TST (at intra-individual level) on the previous night was significantly related to a higher daily physical activity the next day within-participants. However, no significant relationships were found at the inter-individual level (e.g., participants with a higher level of physical activity did not report significantly different sleep parameters than participants with lower physical activity level, [18]). Another observational study [19] asked depressed and pair-matched non-depressed participants to rate their sleep quality and duration and to wear an accelerometer for 30-d. Only one directional association was significant. The higher sleep quality was related to higher physical activity the following day within participants and between participants.

Other studies with repeated daily measures examined these associations with different sleep–physical activity measures or characterization and tested unidirectional or bidirectional associations. For instance, Bittner et al. [17] analyzed the unidirectional association between sleep parameters and physical activity the following day. Gabriel et al. [15] explored bi-directional associations between self-reported sleep parameters and several physical activity outcomes (e.g., moderate and vigorous physical activity, light physical activity). These methodological features do not facilitate a clear understanding of daily associations between sleep and physical activity. Moreover, a different pattern of associations was observed at the inter- and intra-individual level. Despite the evidence supporting the short-term associations between these two behaviors, it is currently unclear if these associations are empirically validated in an ecological context. A better understanding of short associations between sleep and physical activity could provide useful information for the development of future interventions and theoretical refinement [16].

The primary goal of the current systematic review and meta-analysis was to summarize existing evidence about daily associations between sleep and physical activity outcomes at inter- and intra-individual level in adults.