The relationship between sleep EEG-based brain age and cognitive function in OSA patients.
Poster No:
2569
Submission Type:
Abstract Submission
Authors:
Soonhyun Yook1, Hae Ree Park2, Eun Yeon Joo3, Hosung Kim4
Institutions:
1University of Southern California, San Gabriel, CA, 2Department of Neurology, Inje University College of Medicine, Ilsan Paik Hospital, Ilsan, Gyeonggi-do, 3Department of Neurology, Neuroscience Center, Samsung Medical Center, Samsung Biomedical Research In, Seoul, Seoul, 4University of Southern California, Los Angeles, CA
First Author:
Co-Author(s):
Hae Ree Park
Department of Neurology, Inje University College of Medicine, Ilsan Paik Hospital
Ilsan, Gyeonggi-do
Department of Neurology, Inje University College of Medicine, Ilsan Paik Hospital
Ilsan, Gyeonggi-do
Eun Yeon Joo
Department of Neurology, Neuroscience Center, Samsung Medical Center, Samsung Biomedical Research In
Seoul, Seoul
Department of Neurology, Neuroscience Center, Samsung Medical Center, Samsung Biomedical Research In
Seoul, Seoul
Introduction:
Obstructive sleep apnea (OSA) is one of the most common sleep disorders, characterized by recurring breathing interruptions during sleep due to airway obstruction, resulting in intermittent hypoxemia and fragmented sleep.1,2 These events contribute to oxidative stress, systemic inflammation, and adverse effects on brain structure and cognition in individuals with OSA.3,4
Recent research from our group introduced a novel metric, the brain age index (BAI), derived from electroencephalogram (EEG) signals during sleep. BAI revealed an association with cortical thickness, suggesting a link between neuroelectrophysiological aging (BAI measured on EEG) and brain structural aging (cortical atrophy) in OSA.
However, it remains unexplored whether faster aging observed on EEG is associated with declined cognitive function. This study thus attempts to answer this novel question by examining whether sleep EEG-BAI correlates with cognitive function, providing insights into reported cognitive impairments in OSA populations.
Recent research from our group introduced a novel metric, the brain age index (BAI), derived from electroencephalogram (EEG) signals during sleep. BAI revealed an association with cortical thickness, suggesting a link between neuroelectrophysiological aging (BAI measured on EEG) and brain structural aging (cortical atrophy) in OSA.
However, it remains unexplored whether faster aging observed on EEG is associated with declined cognitive function. This study thus attempts to answer this novel question by examining whether sleep EEG-BAI correlates with cognitive function, providing insights into reported cognitive impairments in OSA populations.
Methods:
1. Participants
We studied individuals from Samsung Medical Center, including 78 OSA patients with an AHI > 15 and 77 normal subjects with an AHI < 5, all aged 18 and above. Normal subjects had an average age of 37.9 ± 11.0, while OSA patients had an average age of 47.8 ± 10.3.
2. Brain age prediction
We utilized sleep EEG data, comprising six channels (F3, F4, C3, C4, O1, O2). The sleep EEG data were transformed into scalograms and fed into a 3D deep learning-based brain age prediction model.5 BAI was calculated by subtracting chronological age from predicted age.
3. Cognitive score
Neuropsychological tests were conducted on 155 subjects, which we derived five composite score: 1) Attention and Executive Function, 2) Verbal Fluency, 3) Verbal Memory, 4) Visual Memory, and 5) Visuospatial Ability.
4. statistical analysis
We conducted linear regression and Student's t-tests for group comparisons between the normal and OSA groups while correcting for possible confounding effects of age and body-mass index (BMI). Subsequently, we correlated cognitive composite scores with BAI within each normal and OSA group. All statistical results were adjusted using false discovery rate correction.
We studied individuals from Samsung Medical Center, including 78 OSA patients with an AHI > 15 and 77 normal subjects with an AHI < 5, all aged 18 and above. Normal subjects had an average age of 37.9 ± 11.0, while OSA patients had an average age of 47.8 ± 10.3.
2. Brain age prediction
We utilized sleep EEG data, comprising six channels (F3, F4, C3, C4, O1, O2). The sleep EEG data were transformed into scalograms and fed into a 3D deep learning-based brain age prediction model.5 BAI was calculated by subtracting chronological age from predicted age.
3. Cognitive score
Neuropsychological tests were conducted on 155 subjects, which we derived five composite score: 1) Attention and Executive Function, 2) Verbal Fluency, 3) Verbal Memory, 4) Visual Memory, and 5) Visuospatial Ability.
4. statistical analysis
We conducted linear regression and Student's t-tests for group comparisons between the normal and OSA groups while correcting for possible confounding effects of age and body-mass index (BMI). Subsequently, we correlated cognitive composite scores with BAI within each normal and OSA group. All statistical results were adjusted using false discovery rate correction.
Results:
1. Comparison between normal and OSA groups
The OSA group was on average 9.9 years older than the normal group. BAI in the OSA group was significantly higher at 2.5±7.8 years compared to -0.08±6.95 in the normal group. Furthermore, the normal group exhibited superior verbal fluency compared to the OSA group (t=3.14, p=0.01, Table 1).
2. Association between BAI and cognitive function
In the normal group, none of the five cognitive scores exhibited a significant correlation with BAI. In contrast, within the OSA group, attention and executive function negatively correlated with BAI (ß=-3.52, p=0.01, Table 2).
The OSA group was on average 9.9 years older than the normal group. BAI in the OSA group was significantly higher at 2.5±7.8 years compared to -0.08±6.95 in the normal group. Furthermore, the normal group exhibited superior verbal fluency compared to the OSA group (t=3.14, p=0.01, Table 1).
2. Association between BAI and cognitive function
In the normal group, none of the five cognitive scores exhibited a significant correlation with BAI. In contrast, within the OSA group, attention and executive function negatively correlated with BAI (ß=-3.52, p=0.01, Table 2).
·Table 1. Comparison between normal and OSA groups
·Table 2. Association between cognitive function and BAI
Conclusions:
In this study, we compared cognitive function between normal and OSA groups and explored the relationship between BAI and cognitive function. The OSA group had higher BAI and lower scores in verbal fluency. These differences may be linked to hypoxemia6 and sleep fragmentation7 that are main characteristics of OSA, which lead to neuroinflammation 8 and glymphatic dysfunction9. This may subsequently affect brain structural integrity and normal neural activity, thereby reducing cognitive function.3
An interesting finding is that within the OSA group, attention and executive function correlated with BAI. Previous research has consistently noted impaired function in these domains among older adults with sleep apnea.10 In the current study, a limitation is acknowledged in the notably younger healthy subjects compared to the OSA group. Further investigations focusing on old adults with OSA are required to elucidate whether brain aging mediates the cognitive decline associated with sleep apnea. In conclusion, our study proposes that BAI could be a valuable biomarker for assessing cognitive function, particularly in the context of OSA.
An interesting finding is that within the OSA group, attention and executive function correlated with BAI. Previous research has consistently noted impaired function in these domains among older adults with sleep apnea.10 In the current study, a limitation is acknowledged in the notably younger healthy subjects compared to the OSA group. Further investigations focusing on old adults with OSA are required to elucidate whether brain aging mediates the cognitive decline associated with sleep apnea. In conclusion, our study proposes that BAI could be a valuable biomarker for assessing cognitive function, particularly in the context of OSA.
Modeling and Analysis Methods:
EEG/MEG Modeling and Analysis
Novel Imaging Acquisition Methods:
EEG 2
Perception, Attention and Motor Behavior:
Sleep and Wakefulness 1
Keywords:
Cognition
Electroencephaolography (EEG)
ELECTROPHYSIOLOGY
Sleep
1|2Indicates the priority used for review
Provide references using author date format
2. Jordan, A. S. (2014) Adult obstructive sleep apnoea. The Lancet, 383, 736-747.
3. Lim, D. C. (2014) Obstructive sleep apnea and cognitive impairment: Addressing the blood–brain barrier. Sleep Medicine Reviews, 18, 35-48.
4. Desseilles M. (2008) Neuroimaging Insights into the Pathophysiology of Sleep Disorders. Sleep, 31, 777-794.
5. Yook, S. (2022) Novel neuroelectrophysiological age index associated with imaging features of brain aging and sleep disorders. NeuroImage, 264, 119753.
6. Berr, C. (2000) Cognitive decline is associated with systemic oxidative stress: the EVA study. Journal of the American Geriatrics Society, 48, 1285-1291.
7. Benkirane, O. (2022) Impact of sleep fragmentation on cognition and fatigue. International journal of environmental research and public health, 19, 15485.
8. Daulatzai, M. A. (2015) Evidence of neurodegeneration in obstructive sleep apnea: Relationship between obstructive sleep apnea and cognitive dysfunction in the elderly. J Neurosci Res, 93, 1778-94.
9. Lee, H. (2022) Glymphatic system dysfunction in obstructive sleep apnea evidenced by DTI-ALPS. Sleep Medicine, 89, 176-181.
10. Alchanatis, M. (2008) Comparison of cognitive performance among different age groups in patients with obstructive sleep apnea. Sleep and breathing, 12, 17-24.