The Effect of Systematic Light Exposure on Sleep in a Mixed Group of Fatigued Cancer Survivors

Abstract

Study objectives: Sleep disturbances are commonly reported by cancer survivors. Systematic light exposure using bright light has been used to improve sleep in other populations. In this secondary data analysis, the effect of morning administration of bright light on sleep and sleep quality was examined in a mixed group of fatigued cancer survivors.

Methods: Forty-four cancer survivors screened for cancer-related fatigue were randomized to either a bright white light or a comparison dim red light condition. Participants were instructed to use a light box every morning for 30 minutes for 4 weeks. Wrist actigraphy and the Pittsburgh Sleep Quality Index were administered at 4 time points: prior to light treatment (baseline), 2 weeks into the intervention, during the last week of the intervention, and 3 weeks postintervention. Thirty-seven participants completed the end-of-intervention assessment.

Results: Repeated-measures linear mixed models indicated a statistically significant time × treatment group interaction effect with sleep efficiency improving more in the bright light condition over time compared with the dim light condition (F3,42 = 5.55; P = .003) with a large effect size (partial η2 = 0.28). By the end of the intervention and 3 weeks postintervention, mean sleep efficiency in the bright light group was in the normal range. Medium to large effect sizes were also seen in sleep quality, total sleep time, and wake after sleep onset for participants favoring the bright light condition.

Conclusions: The results suggest that systematic bright light exposure in the morning may have beneficial effects on sleep in fatigued cancer survivors. Larger scale efficacy trials are warranted.

Clinical trial registration: Registry: ClinicalTrials.gov, Title: Treating Cancer-Related Fatigue Through Systematic Light Exposure, Identifier: NCT01873794, URL: https://ichgcp.net/clinical-trials-registry/NCT01873794.

Keywords: actigraphy; cancer; fatigue; light therapy; sleep disturbance; sleep efficiency; sleep quality.

© 2018 American Academy of Sleep Medicine

Figures

Figure 1. Consort diagram.

Figure 2. Least squares mean and standard error bars for subjective sleep quality over time.

After controlling for baseline levels of sleep quality, the repeated-measures linear mixed model showed no effects for time (P = .31), treatment condition (P = .12), nor time by treatment condition (P = .11). PSQI = Pittsburgh Sleep Quality Index.

Figure 3. Least squares mean and standard error bars for total sleep time over time.

The repeated-measures linear mixed model showed that neither the main effect for time (P = .93) nor treatment condition (P = .74) were statistically significant. Time × treatment was also not significant (P = .06).

Figure 4. Least squares mean and standard error bars for sleep efficiency over time with mean levels of time in bed (in hours) included for reference.

The repeated-measures linear mixed model showed that the main effect for time was not significant (P = .49). The main effect for treatment condition was also not statistically significant (P = .07). However, the time × treatment condition was statistically significant (P = .003). tib = time in bed in hours.

Figure 5. Least squares mean and standard error bars for WASO over time.

The repeated-measures linear mixed model showed that the main effects for time (P = .66) and for treatment condition (P = .07) were not statistically significant. The time × treatment condition was also not statistically significant (P = .06). WASO = wake after sleep onset.