| dc.description.abstract | Background Sleep disorders are frequent in patients with dementia. Sleep disturbances can negatively influence the disease burden of dementia and accelerate the rate of cognitive decline. It is hypothesized that sleep pathology, such as REM sleep behavior disorder (RBD), can occur many years before symptomatic disease and that it can be one of the early manifestation of neurodegeneration in alpha-synucleinopathies, such as dementia with Lewy bodies (DLB). DLB is the third most frequent type of dementia, after Alzheimer’s dementia (AD) and vascular dementia (VaD). However, there are few validation studies of the DLB consensus criteria from 2005, and the criteria appear to still have low sensitivity. Objectives The main goals of this thesis were: 1) to investigate the frequency of sleep disturbances in dementia and compare sleep disturbance profiles in a group of patients with AD and DLB; 2) to assess the potential influence of probable RBD (pRBD) on the rate of dementia progression; and 3) to compare neuropathological diagnoses with clinical diagnoses based on DLB consensus criteria from 2005. In the first publication, we investigated the prevalence of different sleep disturbances (SD) in mild AD and mild DLB dementia, in order to characterize the profile of sleep disorders in these two dementias. We studied the relationship between SD, RBD, and neuropsychological symptoms to find out if they could be associated with a higher incidence of psychopathological symptoms, i.e. depression, anxiety, hallucinations. In the second publication, we tested the hypothesis that RBD could be a factor for a more aggressive dementia course in light of previous reports that idiopathic RBD (iRBD) is a risk factor for mild cognitive impairment (MCI) and dementia. RBD seems also to worsen the course of Parkinson’s disease (PD) and accelerate conversion from MCI to Parkinson’s disease dementia (PDD). We examined the association of pRBD with global cognitive decline and decline in the cognitive subdomains in a 4-years longitudinal study. In the third publication, we analyzed reliability between clinical and neuropathological diagnoses in the selection of 56 autopsied patients to characterize sensitivity and specificity of consensus criteria from 2005. Moreover, we discussed potential reasons for misdiagnosis. Subjects and methods The thesis was based on data from the prospective DEMVEST study which started in 2005 and was conducted until 2013. Patients with dementia and their caregivers were recruited from clinics of old age psychiatry and geriatrics in Western Norway. The inclusion criterion was mild dementia with Mini-Mental State Examination (MMSE) ≥20. Dementia diagnoses were made according to established criteria, and pathologically confirmed in a subsample of 56 cases. Patients with acute cognitive disturbances and known chronic psychiatric illness were excluded. MMSE was used for cognitive screening. A battery of neuropsychological tests was used for measuring cognition: Stroop test, Controlled Oral Word Associations Test, semantic fluency (COWAT), Boston Naming Test 15 items (BNT), Trail Making Test A and B (TMT A, TMT B), the California Verbal Learning Test-II (CVLT-II), Silhouettes and Cubes on the Visual Object and Space Perception Battery (VOSP). Standard statistical analyses were performed for demographical and clinical comparisons. Longitudinal analyses were conducted with generalized linear mixed models (GLMM). Results A total of 266 patients were included in the DEMVEST study, and our cohort consisted of 139 with AD, 82 with DLB, 17 with PDD, 13 with VaD and 5 with FTD, six with other diagnoses (normal ageing, mixed AD and VaD, alcoholic dementia, unspecified dementia). Numbers of participants varied in the three papers according to purposes of the studies. Paper 1: 56% of patients with dementia, represented by AD and DLB, had at least one sleep disorder, with predominance of insomni (34.8%). Sleep problems were significantly more frequent in the group of patients with DLB than with AD (73.2% vs. 45.7%, p<0,001). DLB diagnoses also increased the risk of having three or more sleep disorders. Having at least one sleep-related problem was associated with significantly higher rates of psychopathological symptoms compared to patients without any sleep disorders. The second most common sleep problem in the DLB group after insomni was pRBD, which was present in 40% of patients. Paper 2: Of the total 246 patients included at baseline we identified 47 (19.1%) with pRBD. Patients with pRBD had the same progression rate of dementia as patients without RBD, both in global cognition measured by MMSE and in different cognitive domains, during 4 years of follow-up. However, the results showed that patients with RBD had significantly lower performance in figure copying in MMSE, VOSP Cube, TMT A and Stroop Words compared to RBD-negative patients throughout the observation period. Paper 3: In a cohort of 56 patients who came to autopsy we found 31 cases with neuropathologically verified AD, 20 with DLB and PDD and five with other diagnoses. Results showed that for overall DLB/ PDD diagnosis there was a relative good consistency between the clinical and pathological diagnoses, with a sensitivity (SN) and specificity (SP) of 80% and 92% and that the positive predictive value (PPV) and negative predictive value (NPV) were 84% and 89% respectively. For possible and probable DLB alone (without PDD), the values of 73% (SN), 93% (SP), 79% (PPV) and 90% (NPV) were found. Of the 56 patients, seven were misdiagnosed; three patients with false positive DLB diagnosis and four with false negative DLB diagnosis. For AD diagnosis SN, SP, as well as PPV and NPV were 81%, 88%, 89% and 79%, respectively. We also discussed factors that could affect the accuracy of clinical diagnosis. Conclusions Sleep disturbances are common in dementia, but patients with DLB have more sleep problems and a more complicated pattern of sleep pathology than those with AD. Occurrence of at least one sleep disturbance seems to be related to a more severe course of psychiatric symptoms. We could not demonstrate that pRBD is a risk factor for faster progression of global cognition and cognition in various cognitive domains of dementia. Sensitivity, specificity and accuracy of clinical diagnoses were similar to results from previous studies which applied DLB consensus criteria from 2005 and showed that there is still a need to improve sensitivity. | en_US |
