Correlation of attention deficit, rapid eye movement latency and slow wave sleep in schizophrenia patients
Yu-San Chang, MD, MS, Department of Neuropsychiatry, Kai-Suan Psychiatric Hospital, No. 130, Kai-Suan 2 Road, Ling-Ya District, Kaohsiung 802, Taiwan. Email: firstname.lastname@example.org
Aim: Schizophrenia patients present both reduced slow wave sleep (SWS) and shortened rapid eye movement latency (REML) in polysomnographic (PSG) profiles, which have been linked to dopaminergic and muscarinic impairment, respectively. Two main selective attentional systems involve different anatomical structures. The first system is the parietal cortical areas and thalamic areas, which are linked to cholinergic neurotransmission. This is responsible for automatic attention response. The second system is the frontal regions, which are linked to dopaminergic neurotransmission. This is responsible for voluntary control of attentional resources. It was hypothesized that low attentional performance in schizophrenia patients is associated with shortened REML and reduced SWS.
Methods: The PSG profile was correlated with the continuous performance test (CPT) in 15 schizophrenia inpatients under treatment with risperidone. Schizophrenia was diagnosed according to DSM-IV criteria, and clinical symptoms were evaluated on the Brief Psychiatric Rating Scale.
Results: REML was negatively correlated with errors of omission (P < 0.05), reaction time (RT; P < 0.05) and positively correlated with hit rate (HR; P < 0.05). No association was found between SWS and CPT performance.
Conclusions: The significant indicators of CPT represent different attention processes. Errors of omission, which are linked to the problems with automatic attention processing, RT, which represent the speed of automatic processing, and HR, are involved in the integration of autonomic and voluntary attention control. The present results suggest that REML is associated with thalamus-related automatic attention response. Due to study limitations, however, confirmation of these findings in a large-scale controlled study of drug-naïve patients is needed.
THE POLYSOMNOGRAPHIC (PSG) profile of schizophrenia patients remains to be fully delineated. Previous studies have widely agreed, however, that schizophrenia patients generally exhibit disruption of sleep continuity, reduced slow wave sleep (SWS) and reduced rapid eye movement latency (REML) with relative normal REM time and density.1–3 The SWS deficit in schizophrenia patients has been linked to frontal and dopaminergic impairment.4 Although the reason for shortened REML in schizophrenia patients remains unclear, it has been attributed to muscarinic hyperactivity, secondary to increased dopaminergic tone.1,5 Riemann et al. showed that a shortening of REML was achieved by administration of a cholinomimetic agent in a significant number of schizophrenia patients.6 Treatment with first-generation antipsychotics, as well as with anticholinergics has been reported to prolong REM latencies in schizophrenia patients.7–9
Attentional deficits, as measured on the continuous performance test (CPT), are consistently found in patients with schizophrenia and appear to be independent of clinical symptoms.9 There are multiple underlying cognitive processes involved in selective attentional performance. According to Posner and Dehaene, and Posner and Peterson, the two main attentional systems involve different anatomical structures.10,11 The first system consists of the parietal cortical areas and thalamic areas, which are linked to cholinergic neurotransmission. This is responsible for orienting of attention and automatic attention response. The second system consists of the frontal regions, which are linked to dopaminergic neurotransmission. This is responsible for voluntary control of attentional resources.12
Both sleep and attentional deficits are important features of the symptoms of schizophrenia.13 There are few reports on the relationship between sleep and attentional performance in schizophrenia patients. The present study investigated the putative relationship between sleep and CPT performance and tested the hypothesis that low attentional performance is associated with shortened REML and reduced SWS.
This cross-sectional observational study enrolled inpatients in a chronic care ward who met the DSM-IV criteria for schizophrenia14 and who were being treated with risperidone. Patients were excluded if they had substance abuse history or significant medical or neurological problems, including history or clinical evidence of narcolepsy or sleep apnea. The dosage of risperidone was regulated to 4 mg as closely as possible before the assessment of study variables. Any medications that might affect the results of sleep analysis were discontinued except risperidone. PSG examination was performed after the patient's risperidone dosage had been stabilized without adjustment for at least 1 month. Before and after 1 week of PSG study, the severity of mental symptoms and attention performance were evaluated on the Brief Psychiatric Rating Scale (BPRS)15 and CPT, respectively. Written informed consent was obtained from all patients prior to participation in the study, which was conducted with the approval of the Ethics Committee of Kai-Suan Psychiatric Hospital.
Continuous performance test
Patient responses were recorded automatically on a diskette using the Vigil CPT (version 1.2). Numbers between 0 and 9 were randomly presented. A stimulus duration of 85 ms and an inter-stimulus interval of 850 ms were used for all tests. The target stimulus was the number 9 proceeded by the number 1. A total of 480 trials with 36 targets were presented. The following CPT variables were used as indicators: (i) errors of omission (i.e. target presented and the subject did not respond), which is related to the function of selectivity and is likely to involve problems with automatic attention processing, (ii) errors of commission (i.e. subject responded as if the target was presented when in fact no target was presented), which is linked to the function of impulsivity or anticipation and might indicate potential problems with voluntary attention control, (iii) hit rate (HR, i.e. overall accuracy of target discrimination), which, in this priming detection model, indicates ‘be in preparation’ and might be involved in the integration of autonomic and voluntary attention control; (iv) reaction time (RT, i.e. the average time from the onset of each stimulus to the initiation of each response), which represents autonomic/sensory-motor processing speed; and, (v) A′ (i.e. the subject's sensitivity to signals), which represents the sensitivity level of discrimination for signal and its share in attention control regarding HR.16 Lower errors of omission and commission, shortened RT, and higher HR and A′ represent better aspects of selectivity, rapidity of processing and processing capacity.
Daytime napping was not permitted during the sleep study. Full-montage polysomnography was performed that consisted of electroencephalogram (EEG at F3/A2, F4/A1, C4/A1, C3/A2, O2/A1, O1/A2), electro-oculogram (EOG), submental electromyogram (EMG), respiratory monitoring (nasal airflow, abdominal and chest strain gauges), electrocardiogram, finger oximetry, and anterior tibialis EMG. The first night served to exclude sleep apnea and sleep-related periodic leg movements, and was also an adaptation night. PSG data from the second night with EEG, EOG, and submental EMG for sleep staging provided the data for analysis. We scored sleep records visually in 30-s epochs according to the American Academic of Sleep Medicine criteria,17 and the scorers were blinded to the CPT and BPRS data.
The indicators of CPT were partially correlated with sleep variables, controlling for the effects of age, risperidone dosage, education years and BPRS scores. Significance was set at P < 0.05 (two-tailed test).
The study subjects were 15 inpatients with schizophrenia in a chronic care ward (10 male, five female; average age, 39.8 ± 7.8 years; range, 24–49 years; average illness duration, 14.5 ± 7.4 years; average education years, 9.2 ± 2.1 years; range, 6–12 years; average BPRS scores: 11.7 ± 5.7) under treatment with risperidone (average antipsychotic dosage: 4.3 ± 0.5 mg, range: 4–5 mg). The results of PSG parameters and the correlations between CPT and sleep measures are shown in Table 1. REML was negatively correlated with errors of omission (P < 0.05), RT (P < 0.05) and positively correlated with HR (P < 0.05). There was no significant association between SWS and CPT performance.
Table 1. Partial correlations between PSG and CPT variables in 15 schizophrenia patients†
|Sleep efficiency (%)||86.5 ± 6.5||0.02||0.24||−0.04||−0.18||−0.15|
|Sleep latency (min)||25.9 ± 16.5||0.08||−0.51||−0.06||0.42||0.09|
|REM latency (min)||79.1 ± 41.5||−0.64*||0.12||0.64*||−0.69*||0.57|
|Stage 1 (min)||58.9 ± 26.2||−0.57||0.48||0.57||−0.36||0.44|
|Stage 1 (% TST)||11.7 ± 5.0||−0.49||0.30||0.49||−0.24||0.44|
|Stage 2 (min)||241.9 ± 61.5||−0.01||−0.02||−0.01||−0.15||−0.08|
|Stage 2 (% TST)||48.4 ± 11.0||0.12||−0.31||−0.13||0.01||−0.08|
|Slow wave sleep (min)||68.5 ± 47.9||−0.10||0.02||0.10||−0.30||0.06|
|Slow wave sleep (% TST)||14.1 ± 10.0||−0.06||−0.10||0.07||−0.20||0.07|
|REM (min)||130.4 ± 49.2||0.07||0.40||−0.09||0.14||−0.16|
|REM (% TST)||25.8 ± 8.5||0.17||0.27||−0.17||0.28||−0.20|
As shown in Table 1, prolonged REML was associated with fewer errors of omission, shortened RT and higher HR. The significant indicators of CPT represent different attention processes. Errors of omission, which are linked to the problems with automatic attention processing; RT, which represents the speed of automatic processing, and HR are involved in the integration of autonomic and voluntary attention control. The present findings suggest that REML is related to thalamus-related automatic attention processes. It has been suggested that spindle activity blocks sensory input that tries to reach the cortex through the thalamus during sleep. Before and during REM sleep, input from a subpopulation of the mesopontine cholinergic region activates by depolarizing thalamocortical relay neurons, resulting in the disappearance of sleep spindles and K complexes.18 It has been hypothesized that muscarinic hyperactivity results in shortened REML in schizophrenia patients.1,5 During the daytime, muscarinic hyperactivity could interfere with the thalamocortical loop to entertain automatic/sensory motor processes such as those required during selective attention.
In the present study no relationship was found between SWS and CPT performance. Orzack et al. reported that the decrease in SWS in schizophrenia is associated with poor performance on CPT.19 Patients with schizophrenia exhibit frontal and dopaminergic impairment that have been linked to SWS deficits.4 Forest et al. suggested that neuroleptic-naïve schizophrenia patients have negative relationships between SWS and RT in attention tasks, which may be linked to a deficit in voluntary control of attentional resources.20 A previous study reported that risperidone improved increased SWS in schizophrenia patients.21 This unexpected present result may be related to the fact that SWS had been affected by risperidone. Further study is needed to explore the relationships between SWS and attentional performance in schizophrenia.
The most striking result of the present study is the suggestion that a relationship exists between REML and attentional performance in schizophrenia. The small sample size, lack of control group as well as potential confounding issues such as bias of antipsychotic effect, subclinical/clinical mood symptoms and failure to control for circadian variation, however, limit the generalization of our findings. It remains to be established that second-generation antipsychotics significantly affect REML. The present study failed to demonstrate that SWS is associated with voluntary attentional processes in schizophrenia patients treated with risperidone. This study may provide some insight into the possible relationship between REML and attention tasks in schizophrenia. Establishing these relationships will require a further study comparing a large treatment-naïve patient sample with healthy controls.