Older schizophrenia patients have more disrupted sleep and circadian rhythms than age-matched comparison subjects

Abstract

Patient reports and laboratory studies suggest schizophrenia patients have disrupted sleep across age groups. Studies have not compared overall sleep/wake patterns or circadian (24-h) activity rhythms of older community dwelling schizophrenia patients to matched comparison subjects. This study examined whether older schizophrenia patients had more disrupted sleep/wake patterns and circadian activity rhythms than age- and gender-matched normal comparison subjects (NCS). Twenty-eight older schizophrenia patients and 28 age- and gender-matched NCS were studied with three days of continuous wrist actigraphy. Nighttime and daytime actigraphically estimated sleep and wake, circadian activity rhythms and light exposure patterns were compared with and without years of education as a covariate. Patients spent longer in bed, had more disrupted nighttime sleep, slept more during the day, and had less robust circadian rhythms of activity and light exposure compared to NCS. Differences persisted in education-adjusted analyses. Within patients, working was associated with improved sleep and circadian rhythms. Findings suggest the sleep and circadian rhythm disruption of older schizophrenia patients was more extensive than that of matched NCS suggesting patients’ sleep disruption was above and beyond what is attributable to advanced age alone. A need exists to develop multicomponent interventions to address sleep difficulties specific to older schizophrenia patients.

Introduction

Individuals with schizophrenia commonly complain of disturbed nighttime sleep both in inpatient (Morgan and Drew, 1970, Rotenberg et al., 2000) and outpatient settings (Haffmans et al., 1994, Sweetwood et al., 1976). Laboratory polysomnographic studies confirm that nighttime sleep of patients with schizophrenia is highly fragmented (Tandon et al., 1992, Tandon et al., 1988, Nofzinger et al., 1993, Neylan et al., 1992, Goldman et al., 1996, Lauer et al., 1997, Keshavan and Tandon, 1993). In addition, REM and non-REM sleep abnormalities have been described in both medicated and unmedicated patients (Lauer et al., 1997, Jus et al., 1973, Ganguli et al., 1987, Hudson et al., 1993, Kempenaers et al., 1988, Stern et al., 1969). Studies suggest that these differences in nighttime sleep architecture are related to important clinical outcomes, medications, and specific psychiatric symptoms (Benson and Zarcone, 1994, Zarcone, 1979, Keshavan et al., 1994, Tandon et al., 1988, Goldman et al., 1996, Kajimura et al., 1996, Hoffmann et al., 2000).

In addition to subjective and objective nighttime sleep disruption, some studies suggest circadian rhythms of hormones and activity patterns are also disrupted in schizophrenia patients (Rao et al., 1994, Wirz-Justice et al., 2001, Wirz-Justice et al., 1997). Based upon analysis with the traditional cosine fitting method, a descriptive study from our laboratory showed that older patients with schizophrenia have disturbed circadian activity rhythms, disrupted nighttime sleep and excessive daytime sleep (Martin et al., 2001). Like other studies in the field, this original descriptive study was limited by the absence of sleep and circadian rhythm data from comparison subjects and by the method used to analyze circadian activity rhythms. Previous studies have not compared sleep/wake recordings or circadian rhythms of older community dwelling schizophrenia patients to matched comparison subjects, so it has not been possible to determine if the disruption in sleep/wake patterns was a result of the psychiatric disorder or a result of aging independent of the psychiatric disorder. Since the publication of our original report, we studied a group of age- and gender-matched normal comparison subjects (NCS) to address this limitation. We also employed a newly developed sophisticated method for circadian rhythm modeling, which enabled examination of additional circadian rhythm characteristics (Martin et al., 2000, Marler et al., 2004).

The current study examined differences between the older patients with schizophrenia we originally studied (Martin et al., 2001) and a group of matched NCS in daytime and nighttime actigraphically estimated sleep/wake and circadian rhythms of activity. We hypothesized that: (a) patients would sleep more at night, but would have poorer nighttime sleep quality than NCS; (b) during the day, patients would be sleepier than NCS; (c) patients would have less robust circadian activity rhythms than NCS; and (d) worse nighttime sleep quality and less robust circadian rhythms would be associated with lower levels of light exposure during the daytime hours. Secondary analyses were undertaken to examine the impact of education, working and living situation on sleep within the schizophrenia patients as well.