REM sleep fragmentation associated with depressive symptoms and genetic risk for depression in a community-based sample of adolescents

Highlights

• REM fragmentations are short arousals during REM sleep inducing discontinuity of sleep.

• Depressive symptoms and polygenic risk score (PRS) for somatic complaints are independently associated with more fragmented REM sleep.

• Fragmented REM sleep may be associated with less efficient regulation of negative affect.

• REM fragmentation is a conceptually distinct phenomenon from nighttime awakenings.

• REM fragmentation and REM density may be distinct mechanisms lowering the quality of sleep.

Abstract

Introduction

Fragmented REM sleep may impede overnight resolution of distress and increase depressive symptoms. Furthermore, both fragmented REM and depressive symptoms may share a common genetic factor. We explored the associations between REM sleep fragmentation, depressive symptoms, and a polygenic risk score (PRS) for depression among adolescents.

Methods

About 161 adolescents (mean age 16.9 ± 0.1 years) from a birth cohort underwent a sleep EEG and completed the Beck Depression Inventory-II the same day. We calculated PRSes for depressive symptoms with PRSice 1.25 software using weights from a recent genome-wide association study for dimensions of depressive symptoms (negative emotion, lack of positive emotion and somatic complaints). REM fragmentation in relation to entire REM duration was manually calculated from all REM epochs. REM latency and density were derived using SomnoMedics DOMINO software.

Results

PRSes for somatic complaints and lack of positive emotions were associated with higher REM fragmentation percent. A higher level of depressive symptoms was associated with increased percent of REM fragmentation and higher REM density, independently of the genetic risks. Belonging to the highest decile in depressive symptoms was associated with a 2.9- and 7.6-fold risk of belonging to the highest tertile in REM fragmentation and density. In addition, higher PRS for somatic complaints had an independent, additive effect on increased REM fragmentation.

Limitation

A single night's sleep EEG was measured, thus the night-to-night stability of the REM fragmentation-depressive symptom link is unclear.

Conclusion

Depressive symptoms and genetic risk score for somatic complaints are independently associated with more fragmented REM sleep. This offers new insights on the quality of sleep and its relation to adolescents’ mood.

Introduction

The duration and quality of Rapid Eye Movement (REM) sleep has been proposed to be related to affective and cognitive processes (Walker and van der Helm, 2009). REM sleep is associated with procedural and emotional memory consolidation (Groch and Wilhelm, 2013, Ackermann and Rasch, 2014), although the precise function of REM sleep remains partly unresolved (Ackermann and Rasch, 2014, Peever and Fuller, 2017, Tempesta and Socci, 2018). Increasing evidence points to the role of REM sleep in emotion regulation. For example, in depressive individuals, REM sleep in specific may contribute to enhanced consolidation of negative memories (Harrington et al., 2018). REM sleep can also contribute to maintaining optimal emotional homeostasis by decreasing accumulation of negative affectivity (Gujar and McDonald, 2011, van der Helm and Yao, 2011). REM sleep deprivation, in turn, is associated with increased emotional reactivity compared to NREM sleep deprivation (Rosales-Lagarde et al., 2012). At the theoretical level, it has been suggested that REM sleep promotes both reactivation of previously acquired affective experiences in the limbic system and their integration to semantic memory, leading to reduced amygdala activity and affective memory trace over time (Walker and van der Helm, 2009). In support of this, it has been shown how rapid eye movements in REM sleep cause transient, time-locked activation of the amygdala (Corsi-Cabrera et al., 2016), confirming the role of limbic system in reprocessing and consolidation of emotional experiences during REM sleep.

As reviewed by Palagini (Palagini et al., 2013), REM sleep alterations, such as decrease in the interval from sleep onset to the first REM episode (REM latency), increase of total REM sleep duration, and increased REM density (the frequency of rapid eye movements per REM period) have been frequently observed in people diagnosed with major depressive disorder. It has been proposed that these changes might be prodromal and residual properties with respect to depressive episodes. For instance, shortened REM latency can persist beyond the depressive episode and increase the likelihood of relapse (Paykel, 2008). Increased REM density, in turn, has been associated with increased memory consolidation of negatively-valenced emotional content (Gilson et al., 2015). Thus, these alterations may be trait markers of depression vulnerability that are non-specific, as not all depressed individuals display REM alterations (Riemann et al., 2001). In addition, other psychiatric disorders (e.g., PTSD) are associated with different REM dysfunctions (Habukawa et al., 2018). It has also been suggested that there are underlying genetic drivers that may cause both REM sleep changes and vulnerability for depression (Palagini et al., 2013). These may equally relate to the individual vulnerability for stress.

Expanding upon traditional ways of looking at REM alterations, recent studies have examined REM sleep fragmentation (i.e., the number and duration of short arousals that disrupt the continuity of the REM period). People with chronic PTSD have more fragmented sleep, even when compared to individuals with major depressive disorder (MDD) (Habukawa et al., 2018), and this was related to the subjective experience of trauma-related nightmare intensity in those with PTSD. Thus, REM sleep continuity may function to de-potentiate emotional load; a function that is disrupted when REM is fragmented by brief arousals. Indeed, in an experimental fMRI study by van der Helm et al. (van der Helm et al., 2011) it was shown that when REM sleep was characterized by lower gamma power at the prefrontal region there was a greater overnight de-potentiation of neural and behavioral responsivity to affective stimuli. Insomnia severity and an experience of distress lasting over the night have also been associated with nocturnal hyperarousal and REM discontinuity (Wassing et al., 2016). It was suggested that hyperarousal and related REM discontinuity plays a significant role in the regulation of emotions in insomnia, depression, and PTSD, by weakening the de-potentiation of negative distress.

The current study explores the question of REM fragmentation, density and latency in relation to depressive symptoms among a community cohort of individuals born in 1998, who underwent a sleep EEG at the age of 17 years. We hypothesized to find associations between higher level of depressive symptoms and lower REM sleep quality. Given many young people experience poor sleep (Bartel and Gradisar, 2015, Crowley and Wolfson, 2018) and subthreshold depression during adolescence (Bertha and Balazs, 2013, Wesselhoeft and Sorensen, 2013) this sample in this developmental period represents a good ‘model’ to test the proposed associations.

Finally, there is evidence for a genetic role linking depression and various aspects of REM sleep (see Palagini et al., 2013, for review). In the present study, we explored a new aspect of REM sleep dysregulation. We also examined whether the effect of depressive symptoms on REM sleep would be enhanced by genetic vulnerability to depression. Towards this aim, we built a polygenic risk score (PRS) derived from a recent genome-wide association study (GWAS) for depressive symptoms (Demirkan et al., 2016). The asset in polygenic scores is that they combine together many genetic variants obtained from large, already existing genome-wide association studies. Polygenic score is thus a weighted genetic index score for a risk concerning a specific trait or disease. In relation to a candidate gene approach, PRS has the advantage of representing hundreds, or even thousands of SNPs related to a complex trait in one score. This is a novel approach in studies on REM sleep, as the previous candidate gene (Chang et al., 2016) or twin studies (Markovic et al., 2018) on REM have not directly addressed the question of REM sleep quality. In sum, our study thus seeks new understanding on the quality of REM sleep in relation to individual resilience and mental health from both genetic and self-experienced perspectives.