Around half of people with dementia have trouble sleeping, wake often and wander during the night. This can greatly affect them, their carers, and other family members or care home residents.
Hypnotic medication is often given to people with dementia to help them to sleep. However, we do not have evidence from clinical trials about the effects of these drugs. Many hypnotics have side effects or hangover effects of daytime sedation, and in the long run can become addictive.
Z-drugs (zolpidem, zopiclone or zaleplon) in particular are often used, but it is not known how safe or effective they are for people with dementia.
Using data from Clinical Practice Research Datalink (CPRD) and existing clinical trials and cohort studies, we will estimate the effect of Z-drug in people with dementia.
- What are the benefits and harms of Z-drugs when used for sleep disturbance in people with dementia?
- Is there any difference, in terms of benefits and harms, between Z-drugs and other approaches to management of sleep disturbance in people with dementia?
1. The project team will perform an inception cohort study using data extracted from Clinical Practice Research Database. Target population in this stage are people with dementia (PwD) in the UK with a first sleep disturbance reported to primary care, not restricted by age or dementia subtype, and includes those living in the community and in residential care. Patients will be followed from first recorded diagnosis of sleep disorder and the rates of adverse events such as falls and fractures, mortality, inception, and stroke will be analysed. Potential confounders in the analysis will be taken into account, such as proxies for sleep disturbance severity, lifestyle factors and cardiovascular health, and dementia subtype. Cox regression models will be used to estimate the hazard ratio for the effect of sleep medication class compared with no treatment on each binary outcome.
2. Data from clinical research studies (RCT and cohort study) will be incorporated, including trials of other interventions. The target population in this stage includes all PwD and their informal carers if living at home. Some outcomes of interests are quality of life of both PwD and their carers, cognitive functions and sleep disturbance. Mixed effect multilevel regression models will be used to estimate the effect of Z-drugs on each patient and carer outcome, and a further separate model will be developed to take into account the nature of each outcome.
This study will provide evidence base for sleep management in dementia, in particular with respect to the safety and the effectiveness of pharmacological interventions, particularly the Z-drugs.
This study will contribute to academic and policy debates about the use of Z-drugs and will help the carers and clinicians to balance potential risks versus benefits, and support treatment decisions.
PrincipaI Investigator: Dr George Savva
Patient and public involvement:
This study is funded by the National Institute for Health Research (NIHR) Heath Technology Assessment Programme.
The total cost is £298,000, for 18 months (from June 2016 to Nov 2017).