Actigraphic-measured sleep disturbance predicts increased positive symptoms in adolescents at ultra high-risk for psychosis: A longitudinal study

Abstract

Background

Sleep disturbance is prevalent among patients with psychosis, yet little is known about sleep health during the ultra high-risk (UHR) period. This study used actigraphy to evaluate sleep in healthy control (HC) and UHR adolescents to examine the relationship between sleep disturbance and psychosis symptoms at baseline and 12-month follow-up, as well as comparisons between objective and subjective measurements of sleep functioning in UHR youth.

Method

Thirty-six UHR and 31 HC youth participated in a baseline evaluation including 5 nights of actigraphy, subjective measurement of sleep health (Pittsburgh Sleep Quality Index; PSQI), and clinical interviews. Clinical measures were repeated with UHR youth (N = 23) at a 12-month follow-up.

Results

The actigraphy data indicated that UHR youth displayed increased wake time after onset (WASO), increased movements during sleep, and decreased efficiency compared to HC, and several markers of sleep disturbance including decreased efficiency, increased WASO, number of awakenings, and increased movements were associated with symptomatology in the UHR group. Interestingly, there were associations between actigraph and self-report indices of sleep duration and efficiency (at the trend level) but not awakenings. Several objective measures of sleep disturbance and one self-reported measure (disrupted continuity) predicted the longitudinal course of symptoms over 12 months in the UHR group.

Conclusions

Taken together, the results suggest a potential role for sleep problems in the etiology of schizophrenia, and highlight sleep health as a possible target for prevention/intervention efforts. Additionally, actigraphy represents an inexpensive, sensitive measurement providing unique information not captured by self-report, and may be an informative adjunct to UHR assessments.

Introduction

Sleep disturbance is a prevalent symptom among individuals with psychosis, including deficits in duration, efficiency, and continuity. Among individuals with schizophrenia, sleep problems are associated with greater distress and reduced quality of life (Cohrs, 2008, Waters and Manoach, 2012) and occur regardless of medication status (Chouinard et al., 2004) or mood state (Wulff et al., 2012). Additionally, sleep disturbance is associated with increased illness severity (Cohrs, 2008), often precedes relapse (Benson, 2008), and when targeted in treatment, improves sleep and psychosis symptoms (Kantrowitz et al., 2010). Taken together, converging evidence suggests that poor sleep may play an important role in schizophrenia pathophysiology (Keshavan and Tandon, 1993, Monti and Monti, 2005).

Despite the pervasiveness of sleep disturbance in psychosis populations, few studies have investigated its prevalence before schizophrenia onset or how it relates to the emergence and course of symptoms over time among at-risk adolescents. At present, it is unclear whether objectively-measured sleep disturbance (e.g., deficits in duration, efficiency, continuity) precedes psychosis onset, emerges secondary to exposure to schizophrenia, or reflects comorbid mood disorders. To the extent that sleep problems precede illness onset and predict increased symptoms over time, it may represent an important target for early identification, prevention, and treatment strategies for at-risk youth.

Although few studies have examined sleep in the prodromal period, retrospective data from schizophrenia samples suggest that impairments in sleep duration and continuity precede psychosis onset (Yung and McGorry, 1996, Lunsford-Avery and Mittal, 2013). Three studies have investigated sleep in at-risk populations. Our recent investigation of self-reported sleep disturbance in the prodrome examined relationships between sleep deficits and psychosis symptoms at a single time point (Lunsford-Avery et al., 2013). Results revealed increased sleep latency and nocturnal awakenings among UHR youth compared to healthy controls (HC), with significant associations between poor sleep quality and increased negative symptoms. Two additional studies have observed sleep physiology abnormalities in at-risk samples (Keshavan et al., 2004, Castro et al., 2012). Overall, these results suggest sleep difficulty in the prodromal period and point to its possible role in the etiology and pathophysiology of psychosis.

Although existing data have increased the understanding of sleep disturbance in at-risk samples, they are limited for several reasons. First, cross-sectional studies have not permitted determination of the temporal order of increases in sleep disturbance versus symptom severity. Second, prior studies have utilized either self-report or polysomnography. Self-reports are vulnerable to recall biases, particularly among adolescents at risk for psychosis (Granö et al., 2011). Whole-night polysomnography is widely viewed as the gold-standard of sleep measurement. It is, however, expensive, infeasible in many UHR youth-based clinics, and often implemented in laboratory settings, which may not accurately reflect sleep in natural settings (Meltzer et al., 2012).

In the current study, we extended our initial investigation (Lunsford-Avery et al., 2013) by including actigraphic assessment of sleep and a follow-up clinical time point. Our first aim evaluated whether objectively-measured sleep (duration, efficiency, wake time after onset (WASO), total movements) is disrupted in UHR youth compared to HC using a novel measurement approach (actigraphy), and if present, how sleep deficits relate to symptom severity in UHR adolescents. Actigraphy has established reliability and validity, is relatively inexpensive, provides an objective measure of sleep in the natural environment, and supplies data over longer periods of time than polysomnographic studies (Sadeh, 2011, Meltzer et al., 2012).

A second aim evaluated the relationship between actigraphic measures of sleep duration, efficiency, and continuity with self-reports of corresponding constructs from the Pittsburgh Sleep Quality Index (PSQI; Buysse et al., 1989). To the degree that actigraphy provides important information over and above subjective assessments of sleep, it may represent an informative supplement to current UHR assessments. Finally, the current study employed a longitudinal design to assess the extent to which sleep health at clinical intake predicts increased symptom severity over a 12-month period. If specific areas of disturbed sleep predict worsened illness over time, elucidating these deficits may improve our understanding of the etiology of psychosis and highlight potential biomarkers.