Prevalence, associated factors and treatment of sleep problems in adults with intellectual disability: a systematic review

In people with intellectual disability (ID), impaired sleep is common. Life expectancy has increased in this group, and it is known that in general population sleep deteriorates with aging. Therefore the aims of this systematic review were to examine how sleep problems are defined in research among adults and older people with ID, and to collect information on the prevalence, associated factors and treatment of sleep problems in this population. PubMed, EMBase, PsycINFO and Web of Science were searched for studies published between January 1990 and August 2011. All empirical studies covering sleep problems in adults with ID were included, and assessed on quality (level of evidence), using a slightly modified version of the SIGN-50 methodology checklist for cohort studies. Of 50 studies that were included for systematic review, one was of high quality, 14 were well conducted, 14 were well conducted but with a high risk of bias, and 21 were non-analytical. The reported estimated prevalence rates of sleep problems in adults with ID ranged from 8.5% to 34.1%. A prevalence of 9.2% was reported for significant sleep problems. Sleep problems were associated with the following factors: challenging behavior; respiratory disease; visual impairment; psychiatric conditions; and using psychotropic, antiepileptic and/or antidepressant medication. Little information was found on older people specifically. Two studies reported treatment effects on sleep problems in larger populations; their findings suggest that non-pharmaceutical interventions are beneficial. Research on the prevalence, associated factors and treatment of sleep problems in adults and older people with ID has mainly focused on subjectively derived data. The definitions used to describe a sleep problem are not uniform, and associations are mainly described as correlations. In order to give recommendations for clinical practice further research is needed, involving objective measurements and multivariate analysis.