Sleep Disorders in Early Psychosis: Incidence, Severity, and Association With Clinical Symptoms

Sarah Reeve; Bryony Sheaves; Daniel Freeman


Schizophr Bull. 2019;45(2):287-295. 

In This Article



Sixty participants were recruited from 4 NHS Trusts: Oxford Health NHS Foundation Trust, Berkshire Healthcare NHS Foundation Trust, Central and North West London NHS Foundation Trust, and Northamptonshire Healthcare NHS Foundation Trust. The inclusion criteria were a diagnosis of nonaffective psychotic disorder; outpatient status; and aged between 18 and 30. Exclusion criteria were primary affective disorder; primary substance abuse disorder; organic or neurological disorder; and nonfluency in English. Diagnoses and current medication were taken from clinical notes. All participants provided written informed consent to take part in the study. The study received approval from an NHS Research Ethics Committee.


Diagnostic Interview for Sleep Disorders. The Diagnostic Interview for Sleep Patterns and Disorders (DISP[20]) was the primary sleep disorder assessment. This structured interview assesses a range of sleep disorders. Symptoms are rated according to diagnostic criteria, and severity was rated as mild, moderate, or severe based on the number and frequency of symptoms, duration of disorder, and the distress and impairment reported (see Supplementary Appendix 1 for the scoring algorithm). Diagnostic requirements were made as conservative as possible by including criteria from the DSM-5,[19] ICSD-2,[21] and ICSD-3[22] classification systems, and requiring that all criteria must be satisfied for a positive diagnosis to be made. Dual ratings for diagnosis were carried out for 5 interviews, and inter-rater reliability calculated for a further 20 assessments. For some disorders (period limb movement syndrome, sleepwalking, REM sleep behavior disorder, narcolepsy, and hypersomnia) polysomnographic monitoring is required to confirm the diagnosis, therefore these are treated as positive screens rather than diagnoses. For sleep apnea, the number of symptom indicators endorsed is reported. For all other sleep disorders listed in Table 1 self-reported symptoms, in the context of the interview, are sufficient to confirm diagnosis.

If participants met criteria for diagnosis or positive screen for a particular disorder, they were then asked (a) if they have discussed the sleep problem with a medical professional and (b) if they are receiving treatment. Any treatment was then scored according to whether it was a recommended treatment or a nonrecommended treatment according to NICE Guidance or other published clinical recommendations—eg, for insomnia[23] or nightmare disorder.[24]

Sleep Recording.Sleep Diary: Participants completed a Consensus Sleep Diary[25] for 7 days. Average sleep duration, sleep onset and offset, and sleep efficiency (the percentage of time in bed spent asleep) were calculated from the sleep diaries. Where less than 3 days of the diary were filled then the data were excluded from analysis.

Actigraphy: Participants wore a wrist-based activity monitoring device (CamNTech Motionwatch 8) for the same 7 days to allow objective recording of activity, which provides estimates of sleep variables. Actigraphic data were analyzed within the proprietary software package (Motionware V1.1.25, CamNTech). Where less than 3 days of actigraphy data were available then the data were excluded from analysis. Actigraphic data were also checked against the diary, and where actigraphic data deviated significantly from the diary the actigraphic data were excluded from all further analyses. Each sleep period was analyzed by entering the time to bed and time out of bed from the sleep diary. If there were no diary data for a particular day visual inspection was applied to find the most likely sleep window—this process was used for a maximum of 3 nights per participant. In both cases, the software algorithm uses movement data to estimate the time spent asleep within the window, from which it calculates the sleep onset and offset, sleep duration, and sleep efficiency.

Sleep-50. A subgroup (n = 29) completed the Sleep-50.[26] These data were used to validate DISP outcomes (Supplementary Appendix 2). The Sleep-50 is a self-report measure comprised of a number of subscales assessing particular sleep disorders as according to DSM-IV (TR).[27] Agreement was assessed by testing if the total score in the relevant Sleep-50 subscale was higher in those receiving the DISP diagnosis vs those without the DISP diagnosis for that particular sleep disorder. This was testable for insomnia, nightmare disorder, restless leg syndrome (RLS), circadian disorder, and sleep walking. For insomnia and nightmare disorder, the relevant Sleep-50 subscale score was significantly higher for those receiving a diagnosis from the DISP (P = .003 and P= .001). The Sleep-50 subscale score was not significantly higher for RLS, circadian disorder or sleep walking in those receiving the respective DISP diagnoses. For RLS, this does appear to be a genuine lack of concordance between these measures (possibly due to the Sleep-50 indexing PLMS in the same subscale as RLS), but in the latter 2 cases the diagnostic groups were likely too small to detect differences (both n= 2).

Psychiatric Symptoms. Psychotic experiences were assessed with the Specific Psychotic Experiences Questionnaire[28] (SPEQ), a self-report measure assessing the past month. The subscales for paranoia, hallucinations, cognitive disorganization, and grandiosity were used in the current study. The paranoia subscale is formed of 15 statements rated by the participant for frequency of the thought from 0 ("not at all") up to 5 ("nearly all the time"), with a maximum possible score of 75. The subscale for hallucinations is comprised of 9 items rated on the same scale to comprise a maximum score of 45. For cognitive disorganization, 11 statements are marked as "yes" (this does apply to me) or "no" (this doesn't apply to me), scored as 1 and 0, respectively. The grandiosity subscale is formed of 8 statements rated for agreement by the participant on a 0 ("not at all") to 3 ("completely") scale, with a maximum score of 24. Higher scores indicate more severe psychotic experiences.

Depression and anxiety were assessed using 2 subscales from the Depression Anxiety and Stress Scale (DASS-21),[29] a self-report measure assessing the past month. The 7 items for each subscale are rated from 0 (did not apply to at all) to 3 (applied to me very much). Higher scores indicate higher levels of anxiety and depression.

Psychological Wellbeing. Health-related quality of life was assessed with the EQ-5D-5L.[30] This self-report questionnaire assesses 5 health-related domains (mobility, self-care, ability to do usual activities, pain or discomfort, and anxiety and depression). Participants rate on a 1–5 scale their extent of issues in each domain (where 1 is no problems, 5 is extreme problems). This score is then entered into an algorithm which weights the responses by domain and population norms to give a decimal score (between 0 and 1), where higher values indicate higher health-related quality of life.

Fatigue was measured via the Multi-Dimensional Fatigue Symptom Inventory (MFSI) short form.[31] This scale comprises 30 items in 5 subscales (general, emotional, physical, mental, vigor). Each item is rated from 0 ("not at all") to 4 ("extremely") by the participant. The total score for the questionnaire is the sum of the general, emotional, physical, and mental subscale scores minus the vigor subscale score. The maximum score is 100, with a greater score indicating greater fatigue.

Analysis. All data were analyzed with SPSS 23.[32] Descriptive variables from the diagnostic interview were reported. These are supplemented by statistics from the diary and actigraphic recordings (sleep duration, onset/offset, and efficiency) where recording is required in the diagnostic criteria, ie, circadian disorders and hypersomnia. Differences in sleep recording data were also analyzed for those with and without insomnia to support the interview-reported symptoms. Differences in psychotic experiences, mood, and wellbeing between those with and without sleep disorders were assessed via independent samples t-tests. Secondary analyses tested medication differences between those with and without sleep disorders. All hypothesis testing was two-tailed.