The effects of recovery sleep on pain perception: A systematic review

Highlights

• Possible pain normalization after recovery sleep (RS) has yet been understudied.

• Data from 29 animal and human studies on pain-related RS-effects are reviewed.

• RS can reset hyperalgesic changes produced by sleep disruption.

• Extended periods of RS are necessary to counterbalance profound hyperalgesia.

• RS-effects have yet been studied only after prior experimental sleep deprivation.

Abstract

Experimental studies highlight profound effects of sleep disruptions on pain, showing that sleep deprivation (SD) leads to hyperalgesic pain changes. On the other hand, given that sleep helps normalizing bodily functions, a crucial role of restorative sleep in the overnight restoration of the pain system seems likely. Thus, a systematic review of experimental studies on effects of recovery sleep (RS; subsequently to SD) on pain was performed with the aim to check whether RS resets hyperalgesic pain changes occurring due to SD. Empirical animal and human studies including SD-paradigms, RS and pain assessments were searched in three databases (PubMed, Web of Science, PsycINFO) using a predefined algorithm. 29 studies were included in this review. Most results indicated a reset of enhanced pain sensitivity and vulnerability following RS, especially when total SD was implemented and pressure pain or painful symptoms (human studies) were assessed. Further research should focus on whether and how recovery is altered in chronic pain patients, as this yields implications for pain treatment by enhancing or stabilizing RS.

Introduction

Reciprocal associations between sleep and pain have been proven since chronic pain conditions are often accompanied by sleep disturbances (Finan et al., 2013; McBeth et al., 2015; Sivertsen et al., 2015; Smith & Haythornthwaite, 2004; Tang et al., 2012) and since non-restorative sleep itself is predictive for the development of pain and the exacerbation of existing pain (Gupta et al., 2007; Mork & Nilsen, 2012). Additionally, experimental sleep deprivation (SD) studies have been conducted to assess effects of sleep alterations on pain in well-controlled laboratory settings, consistently showing hyperalgesic changes following SD (e.g. lower pain thresholds, enhanced pain complaints) (Karmann et al., 2014; Kundermann et al., 2004a; Lautenbacher et al., 2006). On the other hand, several findings may lead to believe that undisturbed (restorative) sleep might in turn help normalizing the pain system overnight. In accord, studies improving sleep quality/quantity in sleep-deprived healthy subjects tended to show reduced pain sensitivity thereafter (Roehrs et al., 2012). Further, improvements in sleep after cognitive behavioral therapy in patients with chronic pain (osteoarthritis) were associated with reduced pain (Smith et al., 2015; Vitiello et al., 2014).

During the past decades a major focus in sleep and pain research was on examining the deleterious effects of sleep deprivation on pain, which are meanwhile well known and repeatedly reviewed. But in line with the recent shift towards perspectives of resilience in pain research, an emphasis on positive outcomes and possible mechanisms of resilience, as these can be directly targeted in interventions (Goubert & Trompetter, 2017), begins to emerge. Therefore, the question whether recovery sleep is able to reset the deleterious effects of SD on pain has gained special interest. However, a systematic evaluation of effects of recovery sleep (RS) on pain is still missing, leaving unanswered whether undisturbed sleep following SD helps to reset occurring pain changes.

Also, mechanisms linking sleep and pain are not conclusively clarified, and thus, how restorative sleep might contribute to a pain normalization. For instance, slow wave sleep (SWS) rebound during RS was emphasized to be involved in a normalization of pain processing (Azevedo et al., 2011; Onen et al., 2001; Smith et al., 2007). Further, since pain is modulated by sleep-dependent (e.g. by homeostatic sleep drive) and circadian processes, chronobiological contributions seem likely (Hagenauer et al., 2017). Other perspectives focus on the opioidergic system (Ukponmwan et al., 1984), central serotonergic transmission (Foo & Mason, 2003), stress (Alexander et al., 2009; Araujo et al., 2011; McEwen & Karatsoreos, 2015) as well as central (Krause et al., 2019) and autonomic nervous system activity (Meerlo et al., 2008) since these factors are related to both pain perception and sleep modulation.

Taking these considerations into account, the main objectives of this review are (1) to systematically summarize effects of RS on pain by retrieving all relevant publications using predefined criteria and (2) to elaborate possible mechanisms of action. For this purpose, empirical animal and human studies conducting experimental SD and subsequent RS were reviewed. Pain had to be measured as the dependent variable and compared between the RS treatment and either an untreated control group or a baseline-condition.

Additionally, we aimed at identifying variables possibly moderating associations between RS and pain, therefore accounting for divergent results. Age and sex of subjects were considered as moderating variables, since both are associated with sleep (Luca et al., 2015) and pain (Lautenbacher et al., 2005; Mogil, 2012; Vendruscolo et al., 2004). In animal (rodent) studies, strains of rodents were considered (Mogil, 2009). Due to enhanced statistical power in larger samples, we included sample size as moderating variable. To reveal if results differ according to methods of pain induction or measures of pain perception, these variables were included as well, as done by previous reviews on SD-effects on pain (Karmann et al., 2014; Kundermann et al., 2004a; Lautenbacher et al., 2006). Since design and duration of sleep manipulations are of important influence, SD design and length (Karmann et al., 2014; Kundermann et al., 2004a; Lautenbacher et al., 2006) as well as length of RS (Jay et al., 2007; Lamond et al., 2007) were considered. The described moderators could be chosen since publications retrieved for this review allowed for their consideration.

Based on our theoretical assumptions, we hypothesized that RS (subsequently to SD) normalizes the pain system overnight. To our knowledge, this is the first review to systematically examine effects of RS on pain, therefore adding to and extending previous findings in this field of research.