The effect of bright light on sleep in nursing home patients with dementia

Background: Up to 70% of nursing home patients with dementia suffer from disrupted sleep, often characterized by multiple awakenings at night and excessive daytime sleep. Sleep disruption may have negative effects on the cognition, mood, behaviour, and well-being of nursing home patients, while also representing a challenge for nursing home staff. However, few sleep scales are developed and validated specifically for the nursing home setting. Sleep problems among nursing home patients are frequently treated by medications, which are associated with severe side effects, including daytime sleepiness, and an increased risk of falls. Thus, there is a need for non-pharmacological interventions to improve sleep in this population. Bright light treatment (BLT) may represent such an intervention, providing increased light exposure aiming to impact sleep, circadian rhythmicity, mood, and/or behaviour. Light is the most important zeitgeber to the circadian system, and consequently has a significant impact on sleep-wake behaviour. Unfortunately, studies have reported low indoor light levels in nursing homes, which in combination with dementia-related neuropathology and age-related reductions in light sensitivity, are likely to contribute to sleep problems in this population. The aim of this thesis was to investigate whether increasing daytime light exposure, by means of BLT, can improve sleep in nursing home patients with dementia, and also to address methodological challenges in this field of research. Methods: Paper 1 is a systematic review of the literature, focusing on the methodological features of the included studies, in addition to their findings. Paper 2 and 3 are based on data from the DEM.LIGHT trial; a cluster-randomized placebo-controlled trial conducted in Norwegian nursing homes, including 69 patients. The intervention comprised a diurnal cycle of ambient light with a maximum of 1,000 lux and 6,000 Kelvin (K) from 10:00-15:00, administered using light emitting diode (LED) light. Before and after this interval, the light levels gradually increased/decreased in lux and K. In the placebo condition, standard light levels were maintained at 150-300 lux and approximately 3,000 K throughout the day. The intervention and placebo lights were installed in the common rooms of the included nursing home units. Outcomes were measured at baseline and at follow-up at week 8, 16, and 24. Paper 2 was a validation study of a proxy-rated sleep scale, using the baseline data from the DEM. LIGHT trial. Actigraphy was used as the reference standard. Paper 3 reported on the sleep outcomes of the trial, which were the primary outcomes. Results: Paper 1 found that there are promising, though inconsistent, results regarding the effect of BLT on sleep and circadian rhythmicity in dementia. Large heterogeneity in terms of interventions, study designs, population characteristics, and outcome measurement tools may explain some of the inconsistencies of results across studies. Paper 2 showed that the proxy-rated Sleep Disorder Inventory (SDI) had satisfactory internal consistency and convergent validity. Using actigraphy as the reference standard, the SDI was termed clinically useful, and we suggested a cut-off score of five or more as defining disrupted sleep in nursing home patients with dementia. These results should be interpreted keeping in mind that actigraphy have some important weaknesses, such as underestimating wake time. Paper 3 evaluated the effects of the BLT on sleep and found an improvement in sleep according to the SDI scores in the intervention group, as compared to the control group, from baseline to week 16 and baseline to week 24. There was no effect in terms of sleep measured by actigraphy. Conclusion: In summary, this thesis found that the evidence for an effect of BLT on sleep in nursing home patients with dementia is promising, but equivocal. Importantly, the research field faces some important methodological challenges, such as accurately measuring sleep. The SDI may represent a valid tool to measure sleep in the nursing home setting, which may be used both by researchers and by practitioners. Although the results of this thesis are not conclusive regarding the effect of BLT on sleep in nursing home patients with dementia, it may represent a step forward in understanding the potential value of BLT in this population, and may lay the ground for further investigation. The lack of an improvement on the SDI at week 8 indicates that the effect of BLT may take a long time to manifest in this population