Table 1.
Summary of the previous research findings on sleep disturbances in UHR individuals*.
| First authors | Country | Study design | Study population | Sleep measures | Summary of key findings |
|---|---|---|---|---|---|
| Studies using subjective measures of sleep | |||||
| Goines (31) | Canada, USA | Longitudinal | N = 740 UHR (43% females; aged 18.5 ± 4.26 years) + 280 HC | SOPS sleep disturbance score | • UHR individuals reported higher levels of sleep disturbance than HC (mean sleep disturbance scores of 2.31 ± 1.568 vs. 0.48 ± 0.904, respectively). • In the UHR group: - Baseline sleep disturbances were significantly linked to greater positive symptoms (i.e., paranoia and hallucinations). - No significant differences were found in baseline sleep disturbance between participants who remitted, remained symptomatic, had prodromal progression, or converted to threshold psychosis during follow-up. |
| Grivel (27) | USA | Longitudinal | N = 200 UHR (28% females) | SIPS sleep disturbance score | • UHR individuals with any lifetime trauma (n = 47) had significantly higher sleep disturbance than those with no history of trauma (N = 153) (mean sleep disturbance scores of 3.17 ± 1.539 vs. 2.50 ± 1.727, respectively) |
| Lederman (34) | Australia | Cross-sectional | N = 10 UHR (20% females) + 10 FEP + 10 HC | PSQI | • UHR participants had significantly poorer overall sleep quality than FEP patients and HC (PSQI total scores of 8.0 ± 3.3, 5.5 ± 3.4 and 3.9 ± 1.5, respectively; p = 0.01). Specifically, daytime dysfunction and sleep medication use (psychotropics prescribed with the primary purpose to improve sleep) were significantly greater among UHR participants than both FEP patients and HC. |
| Lindgren (40) | Finland | Longitudinal | N = 54 UHR (81% females; aged 16.7 ± 0.85 years) + 107 non-UHR psychiatric patients | SIPS sleep disturbance score | • Sleep disturbance was significantly associated with current and lifetime suicidality. No association was found between sleep disturbance and intentional self-harm during follow-up (mean sleep disturbance scores of 2.5 ± 1.4 in “No self-harm” group as compared to 2.0 ± 1.2 in “Self-harm” group, p = 0.43). |
| Lunsford-Avery (12) | USA | Cross-sectional | N = 33 UHR (33% females; aged 18.73 ± 1.89 years) + 33 HC | PSQI | • UHR adolescents had significantly higher PSQI total scores, increased latency and greater disturbances compared to HC. • In the UHR group: - More sleep difficulties (increased latency, reduced quality and duration of sleep) were significantly associated with increased negative symptoms. - No association has been found between PSQI variables and positive symptoms. - Bilateral thalamus volume reductions were linked to increased latency, reduced efficiency, and decreased quality of sleep. |
| Lunsford-Avery (41) | USA | Cross-sectional | N = 59 UHR (42% females; aged 18.93 ± 1.67) | PSQI | • A total of 23 UHR participants (33.9%) had poor sleep quality (PSQI>8) • “Poorer sleepers” exhibited lower overall cognitive performance, increased negative symptom severity and similar functioning levels compared to “better sleepers.” • Sleep disturbances (i.e., latency, efficiency and sleep quality) were significantly associated with procedural learning deficits. |
| Michels (36) | Germany | Cross-sectional | N = 14 UHR (36% females; aged 23.29 ± 3.91 years) + 17 patients with schizophrenia + 17 Healthy relatives of patients with schizophrenia + 29 HC | Self-developed Likert-type single items assessing self-reported frequency of dream recall and nightmare during the last 2 months | • UHR participants reported higher nightmare frequencies compared to patients with schizophrenia, first-degree relatives and HC (Means of nightmare frequency of 3.79 ± 1.93, 3.65 ± 2.50, 2.41 ± 2.00, and 1.90 ± 1.92, respectively). • UHR participants had higher dream recall frequencies compared to patients with schizophrenia (while relatives and HC reported lower and similar mean scores) |
| Miller (28) | Canada, USA | RCT | N = 60 UHR (35% females; aged 17.8 ± 4.8 years) | SOPS sleep disturbance score | • Sleep disturbance was reported by 37% of UHR participants (which represents the percent of patients scoring between 3 “moderate” and 6 “extreme” in the SOPS sleep item) |
| Nuzum (19) | UK | Retrospective | N = 795 UHR (44% females; aged 22.72 ± 4.89) | Sleep disturbances reported by clinicians (Any form of insomnia or disturbed sleep that happened more than once and was having an impact on the client's life) | • 59.5%, of UHR individuals experienced sleep problems (22.01% and 58.11% of individuals reported insomnia and disturbed sleep, respectively) |
| Poe (30) | USA | Longitudinal | N = 194 UHR (27% females; aged 20.0 ± 3.8 years) + 66 HC | SIPS sleep disturbance score | • UHR subjects displayed significantly higher sleep disturbance scores than HC. • In the UHR group, sleep disturbance was related to higher positive and negative symptoms and more impaired functioning. |
| Reeve (50) | UK | RCT | N = 160 UHR (39% females; aged 20.9 ± 4.2 years) | Economic Patient Questionnaire Interview | • At baseline, 85% of UHR individuals experienced ‘Bad' night with a mean sleep duration of 4.14 h. • The baseline cross-sectional evaluation revealed that a shorter sleep duration was significantly associated with increased positive symptoms (delusional ideas and hallucinations) and distress. |
| Ruhrmann (39) | England Finland Germany Netherland |
Longitudinal | N = 245 UHR (44% females; aged 23.0 ± 5.2 years) | SIPS sleep disturbance score | • Sleep disturbances score >2 on SIPS helped predict transition to psychosis at 18-month follow-up. |
| Tso (29) | USA | Cross-sectional | N = 203 UHR (43% females; aged 16.8 ± 3.3 years) + 87 individuals with clinically lower risk + 44 very early FEP (<30 days of positive symptoms) | SOPS sleep disturbance score | • UHR participants displayed higher sleep disturbance scores than individuals with clinically lower risk (Scores of sleep disturbance were the highest in FEP patients). |
| Waite (90) | UK | Qualitative | N = 11 UHR (54% females; aged 18.27 ± 1.95 years) | Interviews | • Participants reported delayed sleep phase, lack of routine, circadian rhythm disruption (i.e., day-night reversal). • They also described a complex and reciprocal relationship between sleep disturbance, mental health problems, and daily functioning. |
| Zaks (20) | USA | Longitudinal | N = 478 UHR participants: 67 converters to psychosis (46% females, aged 18.85 ± 4.02 years) and 411 non-converters (45% females; aged 18.30 ± 4.10 years) + 94 HC | PSQI RU-SATED questionnaire |
• All PSQI subscores (i.e., duration, latency, disturbance, efficiency, daytime dysfunction, subjective quality, and medication use) and total score were significantly higher in UHR participants (at a similar extent between converters and non-converters) related to HC; indicating an overall poor sleep quality in UHR groups compared to good sleep quality in HC. • No significant differences were found between UHR and HC individuals in napping frequency or the RU-SATED items timing and regularity. • In UHR individuals, baseline disturbed sleep did not predict conversion to psychosis up to >2 years later. • Sleep disturbance was strongly associated with increased positive symptoms over time. |
| Studies using objective ±subjective measures of sleep | |||||
| Castro (33) | Brazil | Cross-sectional | N = 20 at-risk individuals: 13 UHR and 7 at bipolar risk (35% females; aged 18.3 ± 4.01 years) + 20 HC | Actigraphy PSQI ESS MEQ |
• Participants of the at-risk group had worse sleep quality compared with HC (PSQI total scores of 7.70 ± 3.69 compared to 4.95 ± 2.16, respectively; p = 0.010); whereas no significant differences were noted between the groups regarding sleepiness and chronotype profiles. • The actigraphy data indicated that the at-risk group displayed longer nap duration during waking (44 vs. 23 min), lower autocorrelation functions, lower interdaily stability, higher intradaily variability, lower most active 10 h of the day (M10), and higher beginning of the M10. |
| Lunsford-Avery (35) | USA | Longitudinal | N = 36 UHR (47% females; aged 18.73 ± 1.89 years) + 31 HC |
Actigraphy PSQI |
• The actigraphy data revealed that UHR participants presented increased WASO, decreased efficiency, and increased movements during sleep relative to HC • In the UHR group: increased WASO, decreased efficiency, increased movements, and number of awakenings were longitudinally associated with symptoms over 12-month follow-up. |
| Lunsford-Avery (38) | USA | Longitudinal | N = 34 UHR (56% females; aged 18.79 ± 1.93 years) + 32 HC | Actigraphy | • UHR individuals displayed significantly more fragmented circadian rhythms and later onset of nocturnal rest compared to HC. • In the UHR group: Circadian disturbances were associated with greater psychotic symptoms at baseline, and predicted severity of symptoms and psychosocial dysfunction at 12-months follow-up. |
| Mayeli (21) | USA | Cross-sectional | N = 22 UHR (54% females; aged 20.3 ± 4.6 years) + 20 HC | • hd-EEG • Polysomnography |
• UHR individuals had more WASO and higher NREM sleep gamma EEG power in a large fronto-parieto-occipital area compared to HC. • No significant difference between groups was found regarding arousal index during NREM sleep. • In the UHR group: higher NREM sleep gamma power in medial frontal-anterior frontal and posterior regions was related to worse negative symptoms. |
| Ristanovic (22) | USA | Cross-sectional | N = 57 CHR (aged 18.89 ± 1.82) including 38 participants who had actigraphy data collected + 61 HC | Actigraphy | • Automatic maladaptive responsivity to family stressors (i.e., greater involuntary engagement stress response) was associated with disrupted sleep (i.e., poorer sleep efficiency) in the CHR but not HC group. • Impaired stress tolerance was associated with all objectively assesses sleep parameters (i.e., sleep duration, continuity, and efficiency). |
| Zanini (2015) (32) | Brazil | Cross-sectional | N = 20 UHR (35% females: aged 18.3 ± 3.91) + 20 HC Females of the UHR group=35% | Polysomnography PSQI ESS MEQ |
• UHR individuals reported significantly poorer sleep quality than HC (PSQI total scores of 7.70 (±3.68 vs. 4.95 ± 2.16, respectively; p = 0.007). • No differences found between groups regarding sleepiness and chronotype profiles. • Polysomography findings indicated that the UHR group presented significantly higher sleep latency onset and REMOL than HC. |
*UHR state was evidenced using structured interviews (e.g., CAARMS, Comprehensive Assessment of the At Risk Mental State; SIPS, Structured Interview for Prodromal Symptoms; SOPS, Scale of Prodromal Symptoms).
FEP, First Episode Psychosis; HC, Healthy controls; RCT, Randomized Controlled Trial; PSQI, Pittsburgh 464 Sleep Quality Index; ESS, Epworth Sleepiness Scale; MEQ, Morningness and Eveningness Questionnaire; RU-SATED, Regularity, Satisfaction, Alertness, Timing, Efficiency, Duration; hd-EEG, High Density Electroencephalography; WASO, Wakefulness After Sleep Onset; NREM, Non-Rapid Eye Movement; REMOL, Rapid Eye Movement Sleep Onset Latency.