Vascular dementia (VaD), the second most common cause of dementia after Alzheimer dementia (AD), represents not less than 20% of cases and is expected to increase in coming years . The pervasiveness of both VaD and AD rises dynamically with age, with the risk of VaD doubling every 5.3 years . Vascular-type pathology and mixed pathology are respectively two and three times more likely in demented patients . In Egypt, the dementia prevalence ranged from 2.01% to 5.07% and dementia increased with age, with the rapid increase among those aging >80 .
Because of their dependency and negative financial impact upon their families and healthcare providers, VaD is a major public health issue and its prevention playing the crucial role .
Sleep is important to normal cognitive functioning, particularly for the formation and consolidation of new memories. Circadian rhythm and sleep complaints are common in the aging population, particularly in those with dementia . In vascular dementia, neuropsychiatric manifestations are very common (90%), and the most common presenting symptoms are sleep disturbances (61%), depression (46%), and apathy (44%). Sleep disturbance is one of the most important domains of behavioral and psychological symptoms in patients with dementia. Past research has demonstrated that sleep disturbance is related to reduce executive function and an increased mortality rate in older adults .
Accurately diagnosing sleep disorders in dementia patients can be quite tricky, due to an abundance of underlying causes, mitigating factors and common causal symptoms. Given the rising prevalence of vascular dementia with aging, the development of more sensitive diagnostic and prognostic tools will be critical for elucidating and modifying VaD pathophysiology to alleviate the personal and socioeconomic impact of cognitive decline in the elderly . Also any delay of early detection or corrective treatment of vascular cognitive impairment (VCI) causes the condition to progress to dementia . Also, patients with mixed pathologies have nearly twice the incremental risk of dementia compared with patients with only Alzheimer-type lesions. Consequently, many cases of dementia could be prevented or delayed by targeting the vascular component .
In the current study, we assessed the disturbance in sleep pattern by both the subjective sleep assessment; Epworth Sleepiness Scale (ESS) & Pittsburgh Sleep Quality Index (PSQI) and objective sleep measurements by polysomnography (PSG) in patients with VaD, and compared it to healthy cognitive elderly individuals.
2. Subject and Methods
2.1. Study Subject
Twenty patients with VaD patients according to the criteria of Vascular Dementia in DSM-V  (10/10 M/F, mean age: 66.05 ± 2.04) and 20 healthy subjects (10/10 M/F, mean age: 65.25 ± 2.22 y) age and sex matched, were enrolled in the study from the outpatient clinics of Geriatrics, Neurology and Psychiatry in Ain Shams University Hospitals, from December 2017 till December 2018, after the patients or their caregivers signed an informed consent. Sample size calculation was based on a one-way ANOVA study, the total sample of 40 subjects achieves 80% power to detect differences among the means versus the alternative of equal means using an F test with a 0.05 significance level. Experimental procedures were previously approved by the Ethical Committee for Human Research at the faculty of medicine Ain Shams University.
Both healthy subjects and VaD patients underwent a full medical, neurological, psychiatric history and clinical examinations. Brain magnetic resonance imaging (MRI) was also performed on all candidates. Exclusion criteria were; other types of dementias, psychiatric disorders affecting sleep, and/or major medical illness affecting sleep, the use of medication affecting the sleep-wake cycle. Patients were at mild to moderate stages of VaD, as assessed using the Clinical Dementia Rating (CDR)  and the Mini-Mental State Examination. Cognitive assessment included: Mini Mental State Examination (MMSE) was used to provide a simple global measure of cognitive functioning . Mood assessment by Cornell Scale for Depression in Dementia  and Taylor Anxiety Scale were used to exclude patients with depression or anxiety . Subjective sleep assessment was assessed using psychometric sleep assessment questionnaire (applied to the caregivers); an Arabic version for sleep evaluation . This instrument includes validated Arabic translations of other sleep assessment scales; 1) The PSQI  and 2) Subjective daytime sleepiness assessment using the ESS . Objective sleep assessment was done using PSG. PSG record includes the following parameters; the electroencephalogram (EEG, leads C3-A2 and 01-A2), electrooculogram (EOG), electromyogram of the chin (chin EMG), electrocardiogram (ECG), the abdominal & thoracic respiratory movements, oral & nasal airflows, and arterial oxygen saturation is also measured. The PSG recordings were taken from 22:00 to 7:00 of the following day. The PSG parameters were scored by visual assessment using the standard criteria . The following data were analyzed: a) Sleep efficiency and continuity indices; sleep latency (SL, the time span between “lights out” and onset of the first stage of sleep), sleep efficiency (SE, the percentages of the sleep time for the recording time) and arousal index (AI, a tendency for more wake after sleep onset arousal). b) Hypnogram indices: the percentage of stage I, II, III, IV, Slow wave sleep (SWS) for total sleep time and the percentage of REM sleep for total sleep time (REM %). c) REM sleep indices, the following were analyzed: REM; duration of REM time in total sleep time, 1st REMD; first REM period duration and REML; REM latency (the time span between sleep onset and onset of the first REM sleep). d) Sleep fragmentation parameters: total sleep time (TST); the time from sleep onset to the end of the final sleep epoch minus time awake, time spent in wakefulness after sleep onset (WASO); the time spent awake between sleep onset and end of sleep and the Sleep fragmentation index (SFI); any sleep stage shift and the total number of awakenings, divided by TST/h. e) Respiratory events includes APNEA, HYPONE, RDI (Respiratory Distress Index) & Oxygen Saturation (SPO2). f) PLMS; periodic limb movements.
2.2. Statistical Analysis
Data were processed by standard analytical procedures  to determine the awake time after sleep onset, sleep efficiency, percentage of non-REM sleep in Stages 1 to 3 of sleep, percentage of REM sleep, and REM sleep latency. Results are expressed as mean, standard deviation (SD) or number (%). Comparison between categorical data (number (%)) was performed using Chi square test. Comparison between values of different variables in the two studied groups was performed using either unpaired t test or Mann-Whitney test whenever it was appropriate. Correlation between sleep parameters and different parameters in patient group was performed using either Pearson or Spearman’s Rank correlation coefficient. SPSS computer program (version 19 windows) was used for data analysis. P value ≤ 0.05 was considered significant.
3.1. Demographic Data
VaD patients and control group showed similar demographic profiles (Table 1). There were with no statistical significant difference as regards age or sex between both groups. But there was statistical significant difference between both groups regarding body mass index (BMI) with high BMI in the VaD group.
Table 1. Demographic features in the two studied groups.
There was statistical significant difference between VaD group and control group regarding hypertension and hyperlipidemia. Sleep disorders were present in 20 patients (100%) in VaD group, but only in 5 (25%) patients in control group. The sleep disorders were insomnia in 3 (15%) patients, excessive daytime sleepiness EDS in 13 (65%) patients, sleep disordered breathing (SDB) in 10 (50%) patients, REM behavioral disordered (RBD) in one (5%) patient and restless leg syndrome RLS & PLM in 3 (15%) patients.
As regards activities of Daily Living (ADL) there was statistical significant difference between VaD group and control group, most of VaD patients had low ADL. There was statistical significant difference between VaD group and control group regarding CDR (Table 2).
There was a statistical significant difference between VaD group and control group regarding MMSE (Table 3).
For the MRI brain findings in VaD group; there were large cortical infarction in 10 (50%) patients, subcortical infarction in 6 (30%) patients, lacunar infarcts in 9 (45%) patients, diffuse white matter hyperintensities in 12 (60%) patients and brain atrophy in 20 (100%) patients.
3.2. Subjective Sleep Assessment
There was a statistical significant difference between VaD group and control group regarding ESS and PSQI between both groups with higher results among patient group (Table 4).
There was a statistical significant correlation between CDR & MMSE and PSQI (the more severe the cognitive impairment, the poorer sleep quality). But, there was no statistical significant correlation between CDR & MMSE and ESS (no relation between degree of cognitive impairment and excessive daytime sleepiness) (Table 5).
3.3. Polysomnography Results
As regards sleep efficiency and continuity, the results showed poorer sleep efficiency,
Table 2. Clinical dementia rating scale (CDR) in the studied groups.
Table 3. MMSE in the studied groups.
Table 4. ESS & PSQI in the two studied groups.
Table 5. Correlation between severity of the disease and subjective sleep assessment.
prolonged SL, and a tendency for more wake after sleep onset (AI) in patient group. As regards Hypnogram, the results showed that VaD patients spent more time in stage I and II sleep, but had lower percentage of SWS and REM sleep compared to control group. As regards REM parameters the results showed VaD patients had lower percentage of REM sleep, delayed REML and decreased 1st REMD compared to control group. As regards Sleep fragmentation parameters, the results showed that VaD patients had short TST, more WASO (more time sent awake) and high SFI (more sleep stage shifts) compared to control group. As regards respiratory events; VaD patients had more Apnea, Hypopnea, RDI and high SpO2 desaturation compared to control group. Finally, as regards PLMS there was a statistical significant increase of PLMS with RLS in VaD group in comparison to control groups (Table 6).
As regard Sleep parameters in the VaD group in correlation to ADL scores, there were statistical significant negative correlation between AI & RDI and ADL ; high AI & RDI is associated with low ADL (more functional impairment), but there was no statistical significance for Sleep efficiency (Table 7).
There were statistical significant difference for Apnea, RDI & SPO2 and BMI score (more Apnea/H and RDI with high BMI index and more average SpO2 desaturation with high BMI), but not for SE and AI (Table 8).
There was no statistical significant difference as regards sleep parameters (SE, SL, SWS and AI), grade of dementia (MMSE and CDR) and subjective sleep assessment (PSQI) in patient group (Table 9).
There was statistical significant difference for SE, AI, Apnea, RDI & SPO2 with the ESS score severity. Also, there was statistical significant difference for BMI in comparison to ESS score severity; severe ESS score with high BMI (Table 10).
Table 6. Polysomnography results in the two studied groups.
Table 7. Correlation between sleep parameters in the patient group and ADL.
Table 8. Sleep parameters in the patient group classified according to BMI subgroups.
There were no significant correlation between sleep parameters and MRI findings; either subjective sleep assessment; ESS & PSQI or objective PSG parameters; SE, SL, AI, RDI & SPO2, WASO and SFI. The MRI findings were; the presence of cortical or subcortical infarction, frontal, thalamic or lacunar infarction, different types of diffuse white matter hyperintensities, degree of brain atrophy and side of infarction; left, right or bilateral (Table 11 & Table 12).
Table 9. Correlation between sleep parameters and grade of dementia (MMSE and CDR) and PSQI in patient group.
Table 10. Correlation between sleep parameters and ESS score severity (normal, mild, moderate, severe).
Table 11. Correlation between sleep parameters and site of lesion.
Table 12. Correlation between sleep parameters and type of diffuse white matter hyperintensities, degree of brain atrophy and side of the lesion.
Sleep is considered to be important for cognitive function, cognitive deficits and sleep disorders are influenced by one another . Given the rising prevalence of cerebrovascular diseases and dementia with aging, the development of more sensitive neuropsychological and neuroimaging diagnostic and prognostic tools, will be crucial for elucidating and modifying VCI pathophysiology to develop new modes of intervention for disease prevention and treatment.
The present study revealed significant correlation for Apnea, RDI & SPO2 with BMI; more Apnea and RDI with high BMI and more average SpO2 desaturation with high BMI.
Hypertension was present in 80% of VaD patients in this study. These findings denote the importance of hypertension as risk factors for vascular dementia. This fact was supported by the report of Hebert et al. (2000) . Also, hyperlipidemia was present in 80% and cardiac diseases in 60% of VaD patients. Diabetes mellitus (DM) was present in 50% of VaD patients in this study. The association between dementia and DM can be explained through the effect of DM on cerebral blood flow especially with regards to reactivity and autoregulation or serving as a risk factor for clinically overt and silent brain insults.
In this study, the most common sleep complaints among VaD patients were; EDS (65%), SDB (50%), insomnia (15%) and RLS & PLMS (15%) and RBD (5%) respectively. These results were in accordance with that of Ramirez-Santos et al. (2015) who stated that in mild cognitive impairment and dementia in general (regardless of type) are associated with a marked increase in sleep-related complaints . The global prevalence reported for any sleep disorder was 82%, for insomnia 37.1%, for hypersomnia 47.8%, for parasomnia 21.4% and apnea 54.5%. Also, Guarnieri et al. (2012) found that VaD patients had a higher frequency of sleep disturbance (insomnia, EDS, SDB, RBD & RLS) than that observed in AD patients . The highest risk increase with respect to AD was for SDB with the high frequency of obstructive sleep apnea (OSA) reported in stroke patients. Both the vascular damage and SDB have some common risk factors such as excessive BMI and hypertension .
In this study, there was a statistical significant correlation between CDR & MMSE and PSQI; the more severe the cognitive impairment, the poorer sleep quality which explain the high prevalence of SDB, insomnia, RLS & PLMS and RBD in VaD patients. Similar findings were reported by Mondal et al. (2013) who found that an abnormal ESS were more likely to have an abnormal PSQI score . But, the results of this study showed no statistical significant correlation between CDR & MMSE and ESS; no relation between grade of cognitive impairment and degree of excessive daytime sleepiness which means that EDS can be found in the earliest stage of vascular dementia. Similar findings were reported by Miu, Szeto (2012) who found no correlation between MMSE and ESS among 105 of mild to moderate dementia patients . Furthermore, Merilino et al. (2010) who showed that, although insomnia represented the most common sleep disturbance in 750 subjects aged 65 years or older, 86 of them were diagnosed as demented, and this was not associated with cognitive impairment . However, the severity of SPO2 desaturation, was associated with high ESS; excessive daytime sleepiness and also those with an abnormal ESS had higher BMI and higher AHI.
Taken together, the data suggests that SE and continuity was poorer in patients with VaD in this study, as the sleep-wake cycle is regulated by a complex interplay of mechanisms located mainly in the brainstem, hypothalamus, and thalamus . And any lesion such as an acute stroke, which directly affects the thalamocortical network function has the potential to disrupt the sleep-wake cycle and lead to sleep disturbances . As lesions of the cortex might compromise this process with abnormal deactivation of frontal and thalamic areas from presleep wakefulness to non-REM sleep and hence primarily affect sleep continuity .
In the present study there was a significant negative correlation between AI & RDI and ADL and significant correlation for SE, AI & RDI with the ESS score severity. Similar results were reported by Jiang et al. (2013) as they found that patients with vascular cognitive impairment-no dementia had higher PSQI scores compared with controls . Compared with controls, patients had reduced TST, decreased SWS and REM sleep, longer SL, lower SE, and increased AI and PLMS index.
The present study was guided by the idea that REM sleep parameters may reflect different pathophysiological mechanisms between AD and VaD.
As, cholinergic neurons are important determinants of REM sleep, with cholinergic activity low during SWS and high during REM sleep . Stroke causes a central imbalance of neurotransmitters, such as acetylcholine, serotonin, and melatonin, causing sleep structure abnormalities . A similar REM parameter dysfunction was present in AD; reduced REM duration, and increased first REML episode . But, this can be explained because of the dependence of REM sleep on the integrity of cholinergic neurotransmission and the widespread deterioration of cholinergic systems throughout the basal forebrain in AD .
The present study revealed a positive significant correlation between ESS, Apnea, RDI & SPO2 and BMI score. Similar findings were reported by Bassetti et al. (2006), as they found a significant correlation between apnea-hypopnea index and BMI . Also, Erkinjuntti et al. (1987) reported that patients with multi-infarct dementia tended to have more apneas/hypopneas than those with AD, and apneas/hypopneas tended to increase in direct proportion to the severity of dementia . However, Karaca (2016) found no correlation between BMI and sleep quality (PSQI) . This discrepancy may result from differences in evaluation methods; subjective and/or PSG, and differences in the patients’ age range.
Circadian rhythm disturbances are common in patients with dementia and affect more than 80 % of those over age 65, resulting in insomnia, excessive daytime sleepiness, and day/night reversal . Similar to the present study, Guarnieri et al. (2012) found that VD patients had disrupted sleep-wake cycles associated with shorter sleep periods and lower sleep quality which were more significant than those in AD patients . Both the degradation of sleep quality and the disintegration of the sleep-wake cycle in VaD may reflect the disruptive effects of the vascular lesions on the neural network dedicated to sleep regulation. Since most of the lacunes in VaD are located predominately in the internal capsule, basal ganglia, and the periventricular white matter. These lesions could disconnect the pathways leading to and from the suprachiasmatic nucleus, which might be involved in the regulation of the circadian sleep-wake cycle. Furthermore, stroke lesions may alter other circadian functions such as sleep-related secretion of growth hormone and melatonin .
Post stroke hypersomnia or EDS is due to reduced arousal because of lesions involving the ascending arousal pathways. This occurs in patients with bilateral lesions of the thalamus, subthalamic and hypothalamic area, tegmental midbrain, and pons, where fibers of the ascending arousal pathways can be severely injured even by single small lesions . White matter hyperintensities (WMHs) severity was significantly associated with sleep disturbance, with most symptoms related to daytime hypersomnolence and restless sleep. This finding might be explained by disruption of the frontal-subcortical neuronal circuits and basal ganglia . Baillet et al. (2017) found that a higher sleep fragmentation was associated with a reduction in white matter integrity due to WMHs . Also, greater sleep fragmentation was associated with more severe arteriolosclerosis and subcortical infarcts in brain autopsies in community-dwelling older people .
Strokes related insomnia are associated with caudate or subcortical, thalamic and brainstem (thalamo-mesencephalic, pontomesencephalic, and pontine tegmentum) lesions. Patients may presented by inversion of the sleep-wake cycle with insomnia, night-time agitation, and daytime hypersomnia .
Researchers have been unable to link SDB frequency, type, or severity to the location of the stroke, however autonomic networks responsible for respiratory control may be disrupted with lesions in forebrain structures that control respiration as part of integrated behaviors such as speech or exercise . Hemispheric strokes in the frontal cortex, basal ganglia, or internal capsule may cause respiratory apraxia, with impaired voluntary modulation of breathing amplitude and frequency, leaving patients unable to take a deep breath or hold the breath . Also, the medulla may be less responsive to rising Pco2 levels during sleep . SDB includes; OSA, central and mixed apnea; is linked with white matter disease on magnetic resonance imaging and silent strokes . SDB patients are at a higher risk of developing cognitive impairment or incident AD  .
OSA is a common feature of vascular dementia, and leads to fragmented sleep, increased nocturnal confusion, and excessive daytime sleepiness . In the acute post-stroke period, there is a high prevalence of central apneas, which typically resolve . Central sleep apnea and Cheyne-Stokes respiration is caused by motor dysfunction that lead to destabilization of the upper airways due to involvement of pyramidal-related musculature without affecting swallowing and cause a higher instability of the upper respiratory tract during the night .
Post stroke PLMS may be of primary type; is not always associated with sleep disturbance, and may be due to unilateral hemispheric, pontine base or tegmentum and spinal strokes . But, patients with bilateral RLS had lesions in both the corona radiata and basal ganglia, whereas patients with lesions only in the corona radiata had either contralateral or bilateral RLS . Most commonly, RLS was accompanied by PLMS in sleep . A common mechanism behind both PLM and RLS might be due to dysfunction of the dopaminergic system, possibly on the level of either pre- and/or post-synaptic striatal and/or spinal dopamine receptors . Evidence supporting an association between wandering in dementia and RLS and/or PLMS; as neuroimaging studies have suggested reduced dopamine reuptake in the caudate and putamen among AD patients who wander relative to those who did not .
Cases of stroke have been described in association with RBD include lesions in the pontine tegmentum, midbrain, or paramedical thalamus which may trigger visual hallucinations, especially at sleep onset . Thalamus, temporal, parietal, and occipital lobes strokes may lead to increased dreaming and nightmares and/or a syndrome of dream-reality confusion . Patients with strokes in the pons, midbrain, or paramedian thalamus may experience peduncular hallucinosis which characterized by complex, colorful, dreamlike visual hallucinations, especially in the evening and at sleep onset. Peduncular hallucinosis may represent a release of REM sleep mentation and may be associated with insomnia .
In this study there were no significant correlation between MRI findings (the presence of cortical or subcortical infarction, frontal, thalamic or lacunar infarction, different types of diffuse white matter hyperintensities, degree of brain atrophy and side of infarction; left, right or bilateral) and sleep parameters (subjective sleep assessment; ESS & PSQI or PSG parameters; SE, SL, AI, RDI & SPO2, WASO and SFI) and this might be due to associated widespread injury and dysfunction throughout the brain in cases of stroke . Similar findings were reported by Karaca (2016) who reported that there was no significant correlation between right-left cerebral involvement and sleep quality . Lutsey et al. (2016) found that neither OSA nor abnormal sleep duration were statistically significantly associated with cerebral infarcts, white matter brain volumes or regional brain volumes by MRI imaging in VaD and AD patients .
However, some studies suggest that right-sided strokes decrease REM and REM density, while left-sided strokes decrease NREM stages. Körner et al.  found that slow-wave sleep was decreased in infarctions of the left hemisphere stroke patients. While Pasic et al. (2011) and Da Rocha et al. (2013) reported that the involvement of the right cerebral hemisphere is more frequent in patients with post stroke sleep disorder; insomnia, fragmented sleep, difficulty in falling asleep, greater sleep latency and worse subjective sleep quality  . Also, Wu et al. (2016) found that patients with minor thalamic lesions are at increased risk for sleep disturbance, sleep-related breathing disorders, and memory deficits . In this respect, the results are contradictory, and further studies with large number are needed.
The available data confirms that sleep is significantly impaired in patients with VaD at both subjective level and the objective by PSG recordings which may be used as a surrogate marker of vascular dementia.
Much evidence suggests that daytime sleepiness, sleep disturbances, or SDB are associated with an increased risk for vascular dementia, ischemic stroke, hypertension, and heart disease . As sleep disorders are frequent and can have serious consequences on patient’s health and quality of life and some sleep disorders are more challenging to treat, most can be easily managed with adequate interventions.
Simple questions of the patient or bed-partner for the symptoms and signs of the OSA, such as loud snoring, observed apneas, and daytime sleepiness, would help identify those in need of further diagnostic evaluation as OSA which is a treatable disorder . A careful clinical evaluation of sleep disorders should be performed routinely in the clinical setting of persons with cognitive decline. Instrumental assessment; PSG should be used in selected patients.
5. Limitations and Strengths
This study has some limitations which have to be taken into consideration. First, the sample size was relatively small, resulting in low statistical power for detecting significant differences between groups. Second, the difficulty of doing second PSG to confirm results and any day to day variation of sleep disturbances. Third, the sample lack different stages of VaD, as severe cases were excluded from the study.
Despite these limitations this work is unique in that: 1) used an extensive clinical evaluation including a combination of neurological examinations and a detailed neuropsychological tests in order to define the cognitive status of each participant; 2) assessment of sleep disorders by means of subjective and objective instruments as, PSG is essential for a correct diagnosis of some sleep disturbances like SDB and measures sleep architecture; 3) state-of-the-art brain MRI assessment was used and; 4) provide insights into sleep disturbances in VaD patients and highlights the importance of this frequently missed aspect in the care of dementia patients and their caregivers.
The information provided in this study helps provide insights into the importance of sleep disturbance in vascular dementia patients and highlights the importance of this frequently missed aspect in the care of dementia patients and their caregivers. Sleep disorders can be detected in early stages of VaD.
The most common sleep complaints among VaD patients were EDS, SDB, insomnia and RLS & PLMS and RBD respectively. The more severe the cognitive impairment (MMSE & CDR), the poorer the sleep quality (PSQI) which explains the high prevalence of SDB, insomnia, RLS & PLMS and REM behavioral disordered in VaD patients.
PSG study of VaD patients showed poorer SE, prolonged SL and high AI. Also, VaD patients spend more time in stage I, less time in SWS and REM sleep, together with delayed REML, and less 1st REML. Another PSG findings, increased sleep fragmentation (more WASO & SFI) and increased AI and PLMS index. VaD patients had significant high Apnea, Hypopnea & RDI score with high average SPO2 desaturation. Early detection and treatment of SDB in these cases may be of utmost importance and may prevent further cognitive decline.