Association Between Self-Reported Sleep and Cognitive Function in Patients With Mild Cognitive Impairment
Article information
Abstract
Objectives
This study aimed to investigate the association between self-reported sleep and cognitive function in individuals with mild cognitive impairment (MCI) to understand potential implications for Alzheimer’s disease prevention.
Methods
This retrospective cohort study included 80 patients with MCI and 70 controls. Participants completed standardized questionnaires to assess self-reported sleep quality (Korean version of the Pittsburgh Sleep Quality Index [PSQI-K]), daytime sleepiness (Epworth Sleepiness Scale [ESS]), and insomnia severity (Insomnia Severity Index [ISI]). Cognitive function was evaluated using the Seoul Neuropsychological Test and the Korean version of the Mini-Mental State Examination. The Korea Instrumental Activities of Daily Living Scale was used to assess instrumental activities of daily living. Correlation analyses examined the relationship between sleep-related parameters and cognitive function.
Results
The results indicated no significant differences in PSQI-K and ESS scores between patients with MCI and the control group. Correlation analyses revealed that poorer sleep quality was associated with reduced frontal and executive functions in patients with MCI, particularly in tests such as Controlled Oral Word Association Test (supermarket, -0.311, p<0.001) and Trail Making Test (TMT) B (0.232, p<0.001). Additionally, daytime dysfunction was associated with poorer cognitive performance across language and executive domains (e.g., Korean Boston Naming Test: -0.290, p<0.001; TMTA: 0.248, p<0.001). In both groups, ISI scores were linked to cognitive functions, particularly in attention, phonemic fluency, and executive function (e.g., digit span, backward: -0.225, p<0.01; TMTA: 0.327, p<0.01).
Conclusions
These findings suggest that sleep disturbances significantly impact cognitive function and daily living abilities in patients with MCI.
INTRODUCTION
The increasing prevalence of Alzheimer’s disease (AD) underscores the critical need for effective preventive measures. Mild cognitive impairment (MCI) is considered a transitional stage to dementia, with an annual conversion rate from MCI to AD estimated at 10%–15% [1,2]. Among the various factors influencing cognitive decline, sleep disorders have been associated with cognitive impairment. Given that memory consolidation occurs during sleep, sleep disorders are acknowledged as modifiable risk factors for dementia [3].
A previous studies indicated that sleep disturbances, as observed through polysomnography, can manifest during the MCI stage [4]. Individuals with MCI experience sleep disorders at a higher rate, ranging from 14% to 49% more than normal individuals [5]. However, many patients with MCI may not perceive significant changes in their sleep patterns, underscoring the necessity for sleep assessments. Subjective sleep quality reports have been validated as useful tools for distinguishing between individuals with and without sleep disorders [6]. Additionally, research using sleep questionnaires has found associations between cognitive impairment and sleep integrity [7].
Detecting the transition from MCI to early AD sensitively involves assessing Instrumental Activities of Daily Living Scale (IADL) and cognitive function test [8]. Although research on the relationship between cognitive function and sleep has yielded limited findings, some studies suggest that frontal lobe dysfunction influences sleep quality [9,10]. While good sleep quality in older adults has been associated with independent daily functioning, the relationship between IADL and sleep disturbances in patients with MCI remains poorly understood [11]. Therefore, we aim to explore the relationship between cognitive impairment components and sleep quality, daytime sleepiness, and insomnia scales in patients with MCI, as well as investigate the relationship between IADL and subjective sleep reports.
METHODS
Participants
This retrospective cohort study analyzed data from the Department of Neurology at Samsung Medical Center, Seoul, Korea. The study included 286 patients who presented with cognitive impairment and sleep disorders as their primary complaints, and who completed neuropsychological tests and sleep questionnaires between November 2023 and January 2024. Patients with insufficient questionnaire content (n=21), structural abnormalities that may affect cognitive function in brain imaging (n=45), accompanying neurological diseases such as AD or Parkinson’s disease (n=56), or head trauma (n=14) were excluded because sleep and cognitive function evaluation could be affected. The diagnosis of MCI was based on participants with subjective memory complaints having a Korean IADL (K-IADL) [12] score of <0.44, and cognitive decline substantiated by a neuropsychological test. Among 150 participants, 80 were patients with MCI, and 70 were control participant (Fig. 1).
Enrollment log of the study. We enrolled 150 participants in 2023–2024 and patients were classified as MCI group and healthy control group. SNSB, the Seoul Neuropsychological Screening Battery; AD, Alzheimer’s disease; PD, Parkinson disease; MCI, mild cognitive impairment.
The study received approval from the institutional review board of the Samsung Medical Center, and the requirement for informed consent was waived owing to the non-interventional observational nature of the study (approval number: 2019-01-051).
Sleep-related measures
Subjective sleep quality was measured for all participants using the Korean version of the Pittsburgh Sleep Quality Index (PSQI-K) [13]. The PSQI-K is a comprehensive tool that evaluates the quality and quantity of sleep over the past month and consists of 18 questions and 7 sub-items. Each subdomain is scored from 0 to 3, giving a total score of 0 to 21, with a high score indicating poor sleep quality. This study classified poor sleep quality based on the PSQI-K total score of 5 point above. Daytime sleepiness was assessed using the Epworth Sleepiness Scale (ESS) [14]. This scale evaluates the likelihood of drowsiness in eight daily situations on a scale of 0–3 (0=never drowsy; 1=slightly drowsy; 2=moderately drowsy; and 3=highly likely to be drowsy). The total ESS score ranges from 0 to 24, and a score of 10 or higher indicates excessive daytime sleepiness. Additionally, the Insomnia Severity Index (ISI) [15] was used to assess insomnia symptoms, on a scale of 0–4, based on frequency: never (0/month); rarely (1/month or less); sometimes (2–4/month); often (5–15/month); or almost always (16–30/month). The integrated insomnia symptom score was calculated by summing the minimum score of 0 and the maximum score of 20, and insomnia presence or absence was classified based on 8 points.
Cognitive measures
The neuropsychological evaluation tool utilized in this study was the Seoul Neuropsychological Screening Battery (SNSB), incorporating the Korean versions of the Mini-Mental State Examination (K-MMSE) and the Clinical Dementia Rating (CDR) scale [16]. Developed domestically by Kang and Na in 2003, the SNSB comprehensively assesses various cognitive functions for dementia evaluation [16]. The SNSB administered in the present study encompassed standardized evaluation items including attention concentration (digit span); language abilities (spontaneous speech, comprehension, repetition, reading, writing, and Korean Boston Naming Test [K-BNT]); spatiotemporal perception and composition abilities clock drawing test [CDT]; memory; and frontal lobe function (Seoul Verbal Learning Test [SVLT], digit symbol coding, Controlled Oral Word Association Test [COWAT], and Trail Making Test [TMT]).
This study compared the raw scores obtained from each SNSB test. Additionally, the K-IADL, a standardized version of the IADL questions developed by Kang et al. [12], evaluates complex functions like phone usage, shopping, eating, financial management, household chores, transportation use, hobbies, and leisure activities. The K-IADL assesses performance across 11 IADL items over the past 4 weeks. Each item is scored on a scale of 0–3. The final score is calculated by summing the scores of the applicable items and dividing by the total number of items.
Statistical analyses
The baseline characteristics are presented as numbers and proportions for categorical variables and mean with standard deviations for continuous variables. The correlation between the neuropsychological test results and the sleep questionnaire was calculated using the Pearson correlation test to estimate the correlation between each cognitive function factor and K-IADL and the sleep questionnaire. Additionally, the PSQI-K, ESS, and ISI scores were divided into two groups based on the total points: subclinical insomnia (5, 10, and 8 points) and insomnia (15 points). A chi-squared test or an independent t-test was performed to compare the results of the neuropsychological test.
RESULTS
Participants
The clinical characteristics of the participants are summarized in Table 1. In the MCI group, 70% were male (n=56) and 30% were female (n=24), compared to 60% male (n=42) and 40% female (n=28) in the control group (p=0.266). The mean age was 75.7±4.9 years in the MCI group and 73.8±3.1 years in the control group, with no significant difference between the two groups (p=0.635). The K-MMSE scores were 25.9±2.40 in the MCI group and 28.29±1.46 in the control group (p<0.001), and K-IADL scores were 0.14±0.15 and 0.06±0.12 in the MCI and control groups, respectively (p<0.001). Overall, sleep questionnaire scores tended to be higher in the control group, with the ISI score notably averaging 7.51±6.18 in the MCI group and 10.81±7.07 in the control group, indicating a significant difference (p=0.002).
Characteristics of participants (n=150)
Correlations between neuropsychological factors with sleep quality, daytime sleepiness, and insomnia severity
The relationships between neuropsychological factors and sleep evaluation for each group are presented in Tables 2 and 3. Consistent with the control group results (Table 2), there were no differences in PSQI-K scores and each neuropsychological test item. However, ISI scores were associated with verbal memory, specifically the SVLT recall (SVLT recall: -0.241, p=0.03). Regarding the ESS, it was observed that higher levels of daytime sleepiness were associated with lower K-MMSE scores.
Correlation between sleep parameters and cognition in healthy control
Correlation between sleep parameters and cognition in MCI
A significant association was observed between high PSQI-K scores and poor performance on tests assessing frontal lobe and executive function scores (COWAT, phonemic: -0.325, p=0.005; TMTA: 0.274, p=0.005) (Table 3). Based on a PSQI-K score of 5 points, the group with poor sleep quality exhibited a significant decrease in scores in digit span, backward, COWAT, and phonemic (Fig. 2). The ESS scores were correlated with not only frontal lobe and executive function scores (COWAT, phonemic: -0.237, p=0.012; TMTA: 0.283, p=0.005; TMTB: 0.227, p=0.005) but also with K-BNT, which assessed language ability (K-BNT: -0.28, p=0.005). The group with excessive daytime sleepiness was confirmed to have significantly lower scores in TMTA, SVLT recall, COWAT phonemic, and K-BNT (Fig. 3). The ISI total score was related to concentration (digit span, backward: -0.225, p=0.005), frontal and executive function (COWAT, phonemic: -0.237, p=0.012; TMTA: 0.283, p=0.005; TMTB: 0.227, p=0.005), visual memory scores (CDT, total, p=0.005), and verbal test (K-BNT: -0.271, p=0.005). The normal group with ISI scores below 8 points had significantly higher scores than that of the group with insomnia in attention, COWAT, and phonemic. When divided based on a criterion of 15 points, a significant relationship was observed only in TMTA (Fig. 4). Multiple regression analyses adjusted for age, sex, education, hypertension, and diabetes mellitus, revealed that a decrease in COWAT phonemic scores and an increase in TMTA completion time were significantly associated with higher PSQI-K and ISI scores (Table 4).
Neuropsychological test scores of patients with and without clinically relevant poor sleep quality at the PSQI-K. Poor sleep quality showed a significant decrease in scores in digit span, backward, COWAT phonemic. *p<0.05 statistically significant. PSQI-K, Korean version of the Pittsburgh Sleep Quality Index; COWAT, Controlled Oral Word Association Test; TMTA, Trail Making Test A; TMTB, Trail Making Test B; SVLT, Seoul Verbal Learning Test.
Neuropsychological test scores of patients with and without clinically relevant daytime sleepiness at the ESS. Excessive daytime sleepiness group was confirmed to have significantly lower scores in TMTA, SVLT recall, COWAT phonemic, and K-BNT. *p<0.05 statistically significant. ESS, Epworth Sleepiness Scale; COWAT, Controlled Oral Word Association Test; TMTA, Trail Making Test A; TMTB, Trail Making Test B; SVLT, Seoul Verbal Learning Test; CDT, clock drawing test; K-BNT, Korean Boston Naming Test; K-MMSE, Korean version of the Mini-Mental State Examination.
Neuropsychological test scores of patients with and without clinically relevant insomnia at the ISI. The normal group with an ISI of less than 8 points had significantly higher scores than the group with insomnia in attention, COWAT, phonemic. When divided based on a criterion of 15 points, a significant relationship was observed only in TMTA. *p<0.05 statistically significant. ISI, Insomnia Severity Index; COWAT, Controlled Oral Word Association Test; TMTA, Trail Making Test A; TMTB, Trail Making Test B; SVLT, Seoul Verbal Learning Test; CDT, clock drawing test; K-BNT, Korean Boston Naming Test; K-MMSE, Korean version of the Mini-Mental State Examination.
Multiple regression analysis of the sleep assessment and cognition score in MCI
Association between K-IADL and sleep disorders
Patients with MCI exhibited significant associations with all sleep assessments, whereas the healthy control group did not show significant relationships between K-IADL and sleep assessment scores. Higher K-IADL scores, indicating greater disability in daily life, were associated with higher PSQI-K, ESS, and ISI scores (Table 4).
DISCUSSION
This study suggests that self-reported sleep in patients with MCI is associated with frontal lobe dysfunction and that excessive daytime sleepiness and insomnia are linked to declines in attention and verbal abilities. Additionally, IADL scores were correlated with poor sleep quality, increased daytime sleepiness, and the severity of insomnia.
Self-reported sleep assessments in patients with MCI did not significantly differ from those in the control group. Sleep plays a crucial role in promoting the structural plasticity of the hippocampus, thus influencing memory formation and cognitive decline in the older adult population [17]. However, the impact of subjective sleep quality on cognitive function has shown inconsistent results. A community-based cross-sectional study of 1,443 individuals found an association between MCI and sleep disorders [18]. Conversely, a different study dividing 459 older adult participants into good and poor sleepers found no significant differences in Alzheimer’s Disease Neuropsychological Battery scores between the groups [19]. Another study comparing sleep disorders between 117 patients with AD and normal individuals reported higher Pittsburgh Sleep Quality Index (PSQI) scores in the normal group, likely owing to reduced accuracy in self-reporting among cognitively impaired patients [9]. Furthermore, a study on 59 older adults with MCI revealed a 61% discrepancy between subjective and objective sleep measures [20]. A systematic review of 71 studies found that objective measures indicated poor sleep quality in 92.8% of pathological older adult participants, whereas subjective measures showed differences in only 40.9% of the studies [21]. This indicates that memory-impaired older individuals may underreport their symptoms.
We examined the relationship between cognitive domains and sleep within each group. The MCI group showed an association of poor sleep quality and insomnia with reduced frontal executive function. The severity of insomnia was also linked to declines in visual memory and verbal abilities. Among healthy older adults, insufficient sleep is known to be correlated with a faster rate of cognitive decline [22]. A meta-analysis examining the relationship between sleep duration and various cognitive domains found that shorter sleep increased the risk for deficits in executive function, verbal memory, and working memory with odds ratios ranging from 1.33 to 1.35 [23]. However, studies focusing on the relationship between specific cognitive domains and sleep in cognitively impaired patients are limited.
A study involving 63 patients with AD and 54 healthy individuals found that sleep latency, duration, and efficiency were all associated with executive function in patients with AD [9]. Our results indicate that in the MCI group, sleep quality (as measured using PSQI-K), daytime dysfunction, and ISI scores were related to frontal executive function, suggesting that sleep deficiency affects executive abilities. Research on the relationship between sleep quality and brain atrophy found that poor sleep quality was associated with atrophy in widespread frontal, temporal, and parietal regions [24]. Additionally, sleep quality was associated with the orbitofrontal and prefrontal cortices, correlating with increased severity of insomnia symptoms [25].
This mechanism can be explained by the role of pathological proteins and neuroinflammation. Sleep deprivation induces the accumulation of beta-amyloid peptides and tau proteins in the cerebrospinal fluid [26]. Chronic poor sleep can cause neural damage and inflammation due to prolonged high cortisol levels, which contribute to cognitive impairment [27].
In our study, excessive daytime sleepiness was correlated with both executive function and language abilities in patients with MCI. A previous study found that excessive daytime sleepiness in older adults is associated with lower Mini-Mental State Examination scores [28]. Moreover, excessive daytime sleepiness may result from early damage to cholinergic neurons in the brainstem, which play a critical role in the arousal system during the early stages of neurodegeneration [29].
Our results suggest that poor sleep quality and excessive daytime sleepiness can influence cognitive decline. Furthermore, we found that IADL were associated with sleep quality, insomnia, and daytime sleepiness. Several studies have reported links between poor sleep quality and Activities of Daily Living (ADL) in older adults. Studies examining the relationship between ADL and sleep found that lower ADL scores were associated with poorer sleep quality [30]. In addition, older individuals with poor sleep quality had significantly impaired ADL [31]. Furthermore, a cohort study based on the Chinese Hainan centenarians cohort study confirmed that poor sleep quality increased the risk of severe ADL impairment regardless of sleep duration [32]. Evaluating independent ADL is crucial for assessing pathological cognitive decline, particularly in patients with MCI, where early detection of deterioration is challenging.
A previous similar study classified patients with subjective memory impairment and MCI into good sleepers and poor sleepers based on sleep quality, and examined the differences in clinical features and neuropsychological test results between the two groups [10]. While the study differentiated cognitive test scores by PSQI within MCI and control groups, our current study has additional significance as it also considered indicators of daytime sleepiness and insomnia. Moreover, through correlation analysis, we identified trends between sleep indicators and frontal executive functions. Based on these findings, our study provides a comprehensive clinical evaluation of various sleep assessments and neuropsychological functions in patients with MCI.
There are some limitations to our study. First, sleep was assessed using self-report questionnaires rather than multifaceted tools. The lack of polysomnography results prevented us from identifying the presence of sleep disorders such as sleep apnea. Second, the participants were individuals who visited a sleep clinic, possibly including those with existing sleep disorders, which may limit generalizability. Third, as a cross-sectional study, it is challenging to determine the precedence of cognitive impairment and sleep disorders. Further longitudinal studies are needed to evaluate changes in sleep with cognitive decline in patients with MCI. Lastly, our study relied solely on self-reports from individuals with MCI. The inclusion of caregiver questionnaires could further enhance the assessment modalities in future research.
Notes
Eun Yeon Joo, the Editor-in-Chief of the Journal of Sleep Medicine, was not involved in the editorial evaluation or decision to publish this article. All remaining authors have declared no conflicts of interest.
Author Contributions
Conceptualization: Eun Yeon Joo, Heewon Bae. Data curation: all authors. Formal analysis: Eun Yeon Joo, Heewon Bae. Funding acquisition: Eun Yeon Joo. Investigation: all authors. Methodology: Eun Yeon Joo, Heewon Bae. Project administration: Eun Yeon Joo. Supervision: Eun Yeon Joo. Validation: Seonjeong Kim, Yi-Seul Choo. Visualization: all authors. Writing—original draft: Eun Yeon Joo, Heewon Bae. Writing—review and editing: Eun Yeon Joo, Heewon Bae.
Funding Statement
This research was supported by Samsung Medical Center Grant (OTC 1190671).
Acknowledgements
Thank you to the patients at the Veterans Health Service Medical Center who generously participated in this study.