An Effective Nondrug Approach to Improve Sleep in Dementia

A multicomponent nonpharmaceutical intervention improves sleep in people with dementia living at home, early results of a new phase 3 randomized controlled trial (RCT) show.

The benefits of the intervention — called DREAMS-START — were sustained at 8 months and extended to caregivers, the study found.

“We’re pleased with our results. We think that we were able to deliver it successfully and to a high rate of fidelity,” said study investigator Penny Rapaport, PhD, Division of Psychiatry, University College London, United Kingdom. 

The findings were presented on July 30 at the Alzheimer’s Association International Conference (AAIC) 2024.

Sustained, Long-Term Effect

Sleep disturbances are very common in dementia. About 26% of people with all types of dementia will experience sleep disturbances, and that rate is higher in certain dementia subtypes, such as dementia with Lewy bodies, said Rapaport.

Such disturbances are distressing for people living with dementia as well as for those supporting them, she added. They’re “often the thing that will lead to people transitioning and moving into a care home.”

Rapaport noted there has not been full RCT evidence that any nonpharmacologic interventions or light-based treatments are effective in improving sleep disturbances.

Medications such as antipsychotics and benzodiazepines aren’t recommended as first-line treatment in people with dementia “because often these can be harmful,” she said.

The study recruited 377 dyads of people living with dementia (mean age, 79.4 years) and their caregivers from 12 national health service sites across England. “We were able to recruit an ethnically diverse sample from a broad socioeconomic background,” said Rapaport.

Researchers allocated the dyads to the intervention or to a treatment as usual group.

About 92% of participants were included in the intention-to-treat analysis at 8 months, which was the primary time point.

The intervention consists of six 1-hour interactive sessions that are “personalized and tailored to individual goals and needs,” said Rapaport. It was delivered by supervised, trained graduates, not clinicians.

The sessions focused on components of sleep hygiene (healthy habits, behaviors, and environments); activity and exercise; a tailored sleep routine; strategies to manage distress; natural and artificial light; and relaxation. A whole session was devoted to supporting sleep of caregivers.

The trial included masked outcome assessments, “so the people collecting the data were blinded to the intervention group,” said Rapaport.

The primary outcome was the Sleep Disorders Inventory (SDI) score. The SDI is a questionnaire about frequency and severity of sleep-disturbed behaviors completed by caregivers; a higher score indicates a worse outcome. The study adjusted for baseline SDI score and study site.

The adjusted mean difference between groups on the SDI was -4.7 points (95% CI, -7.65 to -1.74; P = .002) at 8 months.

The minimal clinically important difference on the SDI is a 4-point change, noted Rapaport.

The adjusted mean difference on the SDI at 4 months (a secondary outcome) was – 4.4 points (95% CI, -7.3 to -1.5; P = .003).

Referring to illustrative graphs, Rapaport said that SDI scores decreased at both 4 and 8 months. “You can see statistically, there’s a significant difference between groups at both time points,” she said.

“We saw a sustained effect, so not just immediately after the intervention, but afterwards at 8 months.”

As for other secondary outcomes, the study found a significant reduction in neuropsychiatric symptoms among people with dementia at 8 months in the intervention arm relative to the control arm.

In addition, sleep and anxiety significantly improved among caregivers after 8 months. This shows “a picture of things getting better for the person with dementia, and the person who’s caring for them,” said Rapaport.

She noted the good adherence rate, with almost 83% of people in the intervention arm completing four or more sessions.

Fidelity to the intervention (ie, the extent to which it is implemented as intended) was also high, “so we feel it was delivered well,” said Rapaport.

Researchers also carried out a health economics analysis and looked at strategies for implementation of the program, but Rapaport did not discuss those results.

Encouraging Findings

Commenting for Medscape Medical News, Alex Bahar-Fuchs, PhD, Faculty of Health, School of Psychology, Deakin University, Victoria, Australia, who co-chaired the session featuring the research,said the findings of this “well-powered” RCT are “encouraging,” both for the primary outcome of sleep quality and for some of the secondary outcomes for the care-partner.

“The study adds to the growing evidence behind several nonpharmacological treatment approaches for cognitive and neuropsychiatric symptoms of people with dementia,” he said. 

The results “offer some hope for the treatment of a common disturbance in people with dementia which is associated with poorer outcomes and increased caregiver burden,” he added. 

An important area for further work would be to incorporate more objective measures of sleep quality, said Bahar-Fuchs.

Because the primary outcome was measured using a self-report questionnaire (the SDI) completed by care-partners, and because the intervention arm could not be blinded, “it remains possible that some detection bias may have affected the study findings,” said Bahar-Fuchs.

He said he would like to see the research extended to include an active control condition “to be able to better ascertain treatment mechanisms.”

The study was supported by the National Institute of Health and Care Research. Rapaport and Bahar-Fuchs reported no relevant conflicts of interest.