Attention and non-REM sleep in neuroleptic-naive persons with schizophrenia and control participants

Abstract

The relationship between sleep architecture and attentional performance was evaluated in neuroleptic-naive patients with schizophrenia and healthy controls. Participants were recorded in a sleep laboratory for two consecutive nights after which selective and sustained attention performance was tested. In both groups of participants, Spearman's rho statistics revealed a negative correlation between reaction time on the selective attention task and sleep spindle density. Only control participants showed a negative correlation between reaction time and duration of stage 2 sleep and a positive correlation between reaction time and duration of stage 1 (“light”) sleep. Only persons with schizophrenia showed a negative correlation between reaction time and duration of stage 4 (“deep”) sleep. In the sustained attention task, we found no correlation between reaction time and sleep for control participants while persons with schizophrenia showed a negative correlation between reaction time and duration of stage 4 sleep. It is proposed that EEG sleep spindle activity is associated with automatic attentional processing, while stage 2 sleep continuity in healthy individuals and percentage of stage 4 in patients with schizophrenia are associated with voluntary processes. These results support the existence of a relationship between non-rapid-eye-movement sleep and cognitive performance in healthy individuals as well as in persons with schizophrenia.

Introduction

Attention refers to a number of abilities that include alertness (i.e., the prompt response to a stimulus), selective attention (i.e., the selection of relevant information), sustained attention (i.e., the capacity to focus attention on a task), divided attention (i.e., the capacity to deal with two or more types of information simultaneously), and attention span (i.e., on-line maintenance of information) (Raz, 2004). It is now well recognized that sleep and daytime attentional performance are intrinsically related (for a review, see Forest and Godbout, 2005).

A number of studies using short- and/or long-term total sleep deprivation (TSD) and/or partial sleep deprivation (PSD) paradigms have provided evidence of a relationship between sleep and attention. Using a divided attention task (dual-task interaction), Fisher (1980) found that one night of TSD produces a loss of attentional control in healthy subjects. In another study, after four consecutive nights of TSD, Norton (1970) found a significant breakdown in healthy subjects' capacity to attend selectively to relevant stimuli presented in a card-sorting task. Linde and Bergstrom (1992) showed a lower performance in the Raven progressive matrices task after one night of TSD versus a normal night. According to these authors, such a deficit may reflect impairment in feature-selection functions, possibly related to selective attention problems. Finally, several studies report that healthy participants have an impaired performance on sustained attention tasks after short-term or long-term TSD or PSD (Lovatt and Warr, 1968, Deaton et al., 1971, Taub and Berger, 1973, Webb and Agnew, 1974, Lester et al., 1976, Herscovitch and Broughton, 1981, Smith and Maben, 1993, Blagrove et al., 1995, Martin et al., 1996, Dinges et al., 1997, Drummond et al., 1999).

While studies have classically focused on the role of rapid-eye-movement (REM) sleep in cognitive processes of healthy persons (Maquet et al., 2003; for a review, see Forest and Godbout, 2005), more recent efforts have also been directed toward non-REM sleep and phasic events of sleep (Linde and Bergstrom, 1992, Blagrove et al., 1995, Giuditta et al., 1995, Meier-Koll et al., 1999, Gais et al., 2002). These studies have found, for example, that learning paired words associates is accompanied by increased spindle density in the first part of the following night (Gais et al., 2002). Similarly, Meier-Koll et al. (1999) found an increase in the amount of sleep spindles and stage 2 sleep after learning a spatial memory task (maze). Smith and MacNeill (1994) demonstrated that performance on an implicit/procedural task (pursuit rotor) is dependent upon the amount of non-REM sleep during the night following learning. Walker et al. (2002) obtained similar results with another implicit/procedural task using finger tapping. At the present, however, there are no reports specifically analyzing the relationship between non-REM sleep and attentional performance.

Both sleep and attentional deficits are important features of the symptoms of schizophrenia (Harvey et al., 2001, Chouinard et al., 2005). On the one hand, compared with findings in healthy participants, studies on neuroleptic-naive persons with schizophrenia show longer sleep latency and more time awake, while reduced slow-wave sleep (SWS: sleep stages 3 + 4) and reduced REM sleep latency are still controversial issues (Poulin et al., 2003a, Chouinard et al., 2005). On the other hand, neuroleptic-naive persons with schizophrenia show impaired performance on memory and on selective and sustained attention tasks compared with healthy participants (Lussier and Stip, 2001). Impaired memory performance has been associated with decreased amount of SWS in schizophrenia (Goder et al., 2004). Manoach et al. (2004) also found that chronically medicated patients with schizophrenia do not show sleep-dependent consolidation of procedural learning, either because of reduced amounts of SWS or because of a deficiency in sleep-dependent processes as such. Another study on sleep and attentional performance in schizophrenia showed that poor performance (two or more omission errors) on a Continuous Performance Test (a measure of sustained attention) was associated with low levels of SWS (Orzack et al., 1977). Unfortunately, the fact that the participants were withdrawn for various intervals from unknown doses of neuroleptics (phenothiazine) may have influenced the results. Indeed, neuroleptics have long lasting effects that may outlast withdrawal, as it is the case for sleep organization (Nofzinger et al., 1993). Our recent meta-analysis on sleep in schizophrenia has shown, for example, that sleep disorders are worse for neuroleptic-withdrawal groups than for never-treated groups (Chouinard et al., 2005). The same applies to cognitive performance following neuroleptic withdrawal (Gilbertson and Van Kammen, 1997).

Surprisingly, there are no published reports on the relationship between unmanipulated sleep (i.e., no sleep deprivation or increased presleep cognitive load) and subsequent daytime cognitive performance in both healthy participants and persons with schizophrenia. The present study examines this problem as part of a larger project on the relationship between sleep and cognitive performance in healthy individuals and persons with schizophrenia. It focuses on the putative relationship between sleep and two important dimensions of attentional performance, i.e., sustained and selective attention in patients with schizophrenia and healthy participants. Based on the previous studies discussed above, we hypothesized that low levels of non-REM sleep markers would be associated with lower attentional performance, in both healthy individuals and participants with schizophrenia.