Insomnia in older adults Consequences, assessment and treatment

Objective: Insomnia is one of the most common conditions in older adults, affecting almost one third of adults aged 55 years or older. Insomnia has been linked to a number of individual adverse medical and psychological consequences, in addition to having large socioeconomic consequences in terms of lost productivity and risk of accidents. Still, only ten percent of chronic insomniacs receive treatment for their condition, of which the vast majority is treated pharmacologically. Psychological treatment options have previously been successfully applied in younger adults, but only a few studies have focused on nonpharmacological treatment in the elderly. In the first paper of this thesis, we investigated empirically the contribution of insomnia to the award of disability pensions. The aim of the study was to estimate the independent effect of insomnia on permanent work disability by statistically controlling for a range of possible confounders. In the second paper we examined both the short- and long-term treatment effect of cognitivebehavior therapy (CBT) and pharmacological treatment in older adults suffering from insomnia. In the third paper, we compared both the psychometric properties and clinical utility of two objective measures of sleep (polysomnography and actigraphy) in the same study sample as in paper II. Methods: In the first paper we used a historical cohort design to estimate the effect of DSM-IV defined insomnia on permanent work disability. Baseline data comprised 37 308 working age people from a large population based Norwegian health study. The outcome was subsequent award of a disability pension 18 to 48 months after baseline assessment, as registered in the National Insurance Administration. In order to estimate the independent contribution of insomnia, we statistically controlled for a range of possible confounders. These included both physical and mental symptoms and diagnoses, as well as lifestyle behaviors, shift work and other potential socio-demographic factors. The two papers from the treatment study are based on a double-blinded randomized controlled trial of 46 patients with chronic primary insomnia who received either cognitivebehavior therapy (CBT, n=18), sleep medication (7.5 mg zopiclone daily, n=16), or placebo (n=12). All active treatments lasted 6 weeks with follow-ups conducted at 6 months. Ambulant clinical polysomnography (PSG), actigraphy and sleep diaries were used on all three assessment points. In paper II we primarily focused on the treatment effect of the two active interventions, while we in paper III explored the ability of actigraphy to correctly classify sleep/wake patterns. Results: From the first paper it is concluded that insomnia is a robust and independent predictor of subsequent permanent work disability. Socio-demographic and shift work characteristics barely affected the association, which also remained significant after adjusting for both psychiatric and physical morbidity and health-related behaviors. In the treatment study (paper II) we found that CBT produced significantly better short- andlong-term treatment effect than zopiclone. We found no significant outcome differences between zopiclone and placebo. Patients receiving CBT spent significantly less time awake and had higher sleep efficiency than patients treated pharmacologically. Patients in the CBT condition spent significantly more time in slow wave sleep (stages 3/4) than either of the two other conditions. In paper III we found that actigraphy performed poorly in detecting wakefulness, and consequently overestimated the patients’ total sleep time and sleep efficiency. Compared with PSG, actigraphy captured only part of the treatment effects on total wake time and sleep onset latency, and failed to detect significant changes in sleep efficiency. Conclusions: Even though a diagnosis of insomnia is legally insufficient for the award of a disability pension, we found that insomnia is an independent predictor of subsequent work disability. Considering the direct costs of disability expenditures, in addition to the indirect costs from lost productivity, sleep medications and sleep-related accidents, we believe that early detection and interventions for insomnia should receive increased focus. Based on the findings from the treatment study we conclude that interventions based on CBT are superior to zopiclone treatment, both in terms of short- and long-term management of insomnia in older adults. We also conclude that the clinical utility of actigraphy is still suboptimal in older insomniacs, and should hence not be used in a clinical setting without parallel use of additional assessment tools