Keywords
sleep - cognition - apolipoprotein
Introduction
Obstructive sleep apnea (OSA) is characterized by recurrent episodes of partial (hypopnea)
and complete (apnea) obstruction of the upper airway during sleep, resulting in intermittent
hypoxemia and sleep fragmentation.[1]
[2] It is also a risk factor for other health problems, such as hypertension, sudden
death, stroke,[3]
[4]
[5] and psychiatric conditions like depression, irritability,[6] and cognitive impairment.[7]
The repeated events of hypoxia and reoxygenation during OSA episodes induce an increase
in oxidative stress, inflammatory process,[8] and neurodegeneration.[9]
[10] Evidence suggests that moderate-to-severe OSA (apnea-hypopnea index [AHI] ≥ 15 events/h
of sleep) is associated with increased cognitive decline and brain morphological changes.[11]
[12] However, most studies have failed to reliably establish a relationship between OSA
and neurocognitive changes[13] because of a heterogeneity of factors related to cognitive decline that include
age, sex, obesity, menopause, hypertension, cardiovascular diseases, level of alcohol
consumption, smoking,[14] physical activity,[15] and level of schooling. Also, the apolipoprotein E ε4 polymorphic allele (APOE ε4) confers a high risk of developing cognitive deficits.[16]
The APOE ε4 is produced primarily by astrocytes in the central nervous system as a carrier of
cholesterol and other lipids to support membrane homeostasis, synaptic integrity,
and injury repair. It increases the risk of dementia by initiating and accelerating
amyloid-β accumulation, aggregation, and deposition in the brain. Conversely, APOE ε4 is more frequently present in patients with OSA for reasons that are not fully understood,
but APOE genes have also been proposed as a cause of OSA susceptibility.[17] Despite this evidence, few studies[24]
[27] have included the analysis of APOE ε4 in the assessment of cognitive functioning among patients with OSA.
Therefore, in the present study, we investigated sleep parameters and cognitive function
in patients with different degrees of OSA severity, with particular interest in the
modulation of APOE in the cognitive function in middle-aged adults with moderate-to-severe OSA. We hypothesized
that moderate-to-severe OSA is associated with worse performance in tests for the
attention domain, and when the interaction between OSA and APOE
ε4 occurs, there is a further reduction in performance on the test for the memory domain.
To this end, we studied sedentary patients without other major diseases.
Materials and Methods
Patients
Male and female patients aged between 40 and 65 years underwent nocturnal conventional
polysomnography. The level of physical activity, cognitive function, and APOE ε4 genotyping were collected for every participant. We excluded patients with body mass
index (BMI) > 40 kg/m2, diabetes mellitus, resting blood pressure (BP) > 140/90 mmHg, smoking or alcohol
abuse (2 or more drinks/d), cardiopulmonary disease, chronic renal disease, a history
of major psychiatric disorders, use of medicines that affect sleep and the neurovascular
system, those with < 2 years of schooling, shift workers, and patients with any sleep
apnea treatment. Because hormonal variability during the regular menstrual cycle can
affect cognitive function, all nonmenopausal women were studied between the first
and fifth days after the onset of menstruation. The present study was approved by
the Institutional Committee on Human Research of the Heart Institute at the Teaching
Hospital of the School of Medicine of Universidade de São Paulo (Instituto do Coração,
Hospital das Clínicas, Faculdade de Medicina, Universidade de São Paulo, InCor-HCFMUSP,
in Portuguese; 0833/10), and all subjects provided written informed consent.
Sleep Study
All participants underwent overnight polysomnography (Embla N7000, Medcare Flaga,
Reykjavik, Iceland) in the Sleep Laboratory of InCor-HCFMUSP. Polysomnography was
performed using standardized techniques, with a standard staging system for the sleep
stages used as described in previous studies.[18]
[19] The AHI was determined by the sum of apneas and hypopneas per hour of sleep. We
used a conservative AHI cutoff of ≥ 15 events/hour of sleep, considering the consistent
evidence from brain imaging analysis, highlighting the presence of tissue damage in
the brains of several patients with moderate to severe OSA, as well as a decline in
cognition in several domains.[11]
Cognitive Function
The procedures and descriptions of the cognitive tests used in the present study have
been published elsewhere.[19] The cognitive tests were performed on the same day as the sleep studies. Attention
and executive function were evaluated using the Trail Making Test-Parts A and B, Forward
Digits test, Backward Digits test, Digit Symbol test, Stroop Color Word Test (SCWT),
and Frontal Assessment Battery. Episodic verbal memory and learning abilities were
assessed using the Rey Auditory Verbal Learning Test: RAVLT5–sum of 5 recall trials
of 15 words; and RAVLT late – delayed recall after 30 minutes.
The Trail Making Test-Part A consists of a series of numbers that must be linked with
a pencil by the patient in ascending order. In Part B, the participant alternates
between numbers and letters, with numbers in ascending order and letters in alphabetical
order. The Forward Digits test requires the verbal repetition of digits in the same
order, whereas, in the Backward Digits test, the participant is asked to repeat the
sequence of numbers in the inverse order. In the Digit Symbol test, a visual key consisting
of paired geometric figures and numbers is provided. Participants are asked to apply
a key to supply the proper number that is associated with the specific symbol. The
outcome is the number of correct responses in 90 seconds. The SCWT comprises three
cards containing six lines with four items: part 1 has colored cards (green, pink,
blue, and brown); part 2 consists of neutral words written with the colors of the
tags; and part 3 has the names of colors written in colors that contrast with the
those of the printed words. The participant is asked to verbalize the names of the
colors printed on each card presented as soon as possible. The time starts right after
the instructions, and the time it takes the subject to read each card is recorded.
The Frontal Assessment Battery contains six subtests, which assess conceptualization,
abstraction, lexical fluency, and mental flexibility: motor programming and sensitivity
to interference, including tendency to distraction, inhibitory control, and autonomy.
Each of the subtests is equivalent to a maximum of three points. Together, the 6 subtests
total 18 points, which is the maximum possible score obtained in the Frontal Assessment
Battery. The RAVLT consists of 15 nonrelated words that should be orally repeated
by the patients in 5 consecutive trials testing immediate verbal memory and learning
abilities and remembered after 30 minutes: delayed recall. The Mini-Mental State Examination,
which is a screening neurocognitive tool that covers domains such as orientation,
memory, registration, recall, constructional ability, language, and the ability to
understand and follow commands, was also applied. The intelligence quotient (IQ) was
estimated by using the Wechsler abbreviated scale for intelligence (WASI).
The Short Cognitive Performance Test (Syndrom-Kurztest, SKT),[20] which assesses memory, attention, and related cognitive functions, and the speed
of information processing, was also applied. All cognitive evaluations and data analyses
were conducted by a single investigator, blinded by the study protocol.
DNA Extraction and APOE Genotyping
Genomic DNA was extracted from leukocytes in samples of whole blood. Genomic DNA was
isolated from peripheral leukocytes. Genotypes for APOE ε2/ε3/ε4 were determined by polymerase chain reaction (PCR) followed by restriction fragment
length polymorphism analysis.[21] Briefly, PCR with fluorescent DNA-intercalating SYTO9 (Thermo Fisher Scientific
Inc., Waltham, MA, United States) was performed using the primer sequences 5'-GCCGATGACCTGCAGAAG-3'
and 5'-CACGCGGCCCTGTTCCAC-3' (fragment size 117 pairs base). The individuals were
classified into ε2/ε2, ε2/ε3, ε3/ε3, ε4/ε4, ε3/ε4, and ε2/ε4 genotypes. Individuals with at least one copy of the APOE ε4 allele were considered APOE ε4 carriers.
Level of Physical Activity
The level of physical activity was evaluated using the International Physical Activity
Questionnaire (IPAQ), which estimates the weekly time spent in moderate and intense
physical activity, as well as walking, in the following domains of everyday life:
domestic, work, transportation, leisure, and the time that a person remains seated.[22]
Statistical Analysis
In the present study, the OpenEpi (open source) interface developed for epidemiological
statistics was used for the sample size calculation. A power of 80% and 95% confidence
intervals (95%CIs) were adopted. The sample size calculation was considered based
on our previous study[23] investigating the APOE ε4 allele and cognition. As a result, a value of 36 patients was obtained. However,
considering a possible loss in data collection, we included 55 participants. Data
were expressed as mean ± standard deviation (SD) values. The Chi-squared test (χ2) was used to assess the difference in sex and polymorphism proportions between the
groups. To assess the homogeneity of the sample, we used the Levene test. The normality
of each sample was tested using the Kolmogorov-Smirnov test. The student t-test for non-repeated measures was used in case of homogeneous and Gaussian variables.
For non-homogeneous and/or non-Gaussian variables, the Mann-Whitney test, unpaired
data was used to assess the differences in each group. Differences with p ≤ 0.05 were considered statistically significant. Correlations between the AHI and
the cognitive variables were performed using multiple linear regressions through the
input method. All analyzes were performed using the SPSS Statistics for Windows (IBM
Corp., Armonk, NY, United States) software, version 20.0.
Results
The study included 42 patients with moderate-to-severe OSA (AHI ≥ 15 events/h of sleep)
and 13 patients with no or mild OSA (AHI < 15 events/h of sleep). The baseline characteristics
of the subjects with no or mild OSA and moderate-to-severe OSA were similar regarding
age (49 ± 7 versus 52 ± 7 years respectively; p = 0.30), body mass index (28 ± 1 versus 29 ± 1 kg/m2 respectively; p = 0.17), leisure physical activity (105 versus 450 min/week respectively; p = 0.63). The presence of at least 1 APOE
ε4 allele was found in 1 subject with no or mild OSA and in 9 subjects with moderate-to-severe
OSA.
Regarding sleep apnea patterns, the AHI and arousal index were higher and lower in
the minimum O2 saturation in the moderate-to-severe OSA group than in the no or mild OSA group ([Table 1]). These results remained when excluding carriers of APOE
ε4. (See [Table S1] in the supplemental file for sleep parameters excluding APOE ε4 carriers in the groups.) No significant differences were found in total sleep efficiency,
N1, N2, N3, and rapid eye movement (REM) sleep stages between both groups. In patients
with moderate-to-severe OSA, the comparison between APOE ε4 carriers and noncarriers presented no significant differences (p > 0.05) in sleep parameters ([Table 2]).
Table 1
Sleep parameters of the study participants.
|
Sleep parameters
|
AHI < 15
(n = 13): mean ± SD
|
AHI ≥ 15
(n = 42): mean ± SD
|
p-value
|
|
TST (in minutes)
|
356 ± 63
|
377 ± 54
|
0.25
|
|
N1 (% TST)
|
6 ± 4
|
8 ± 6
|
0.49
|
|
N2 (% TST)
|
57 ± 8
|
58 ± 9
|
0.75
|
|
N3 (% TST)
|
18 ± 7
|
15 ± 10
|
0.31
|
|
REM (%TST)
|
18 ± 6
|
19 ± 7
|
0.71
|
|
Arousal index (events/hour)
|
20 ± 5
|
32 ± 16
|
0.01*
|
|
AHI (events/hour)
|
7 ± 3
|
44 ± 28
|
< 0.001*
|
|
Minimum O2 saturation (%)
|
89 ± 2
|
80 ± 9
|
< 0.001*
|
|
APOE
ε4 carriers (n, %)
|
1 (7.7)
|
9 (21.4)
|
|
Abbreviations: AHI: apnea-hypopnea index; APOE, apolipoprotein E; N1, N2, N3, sleep stages (non-rapid
eye movement); REM: rapid eye movement sleep; SD, standard deviation; TST, total sleep
time.
Note: *p ≤ 0.05.
Table 2
Assessment of sleep parameters among carriers and noncarriers of apolipoprotein E
(APOE) ε4 polymorphic allele with moderate-to-severe sleep apnea (AHI ≥ 15 events/hour).
|
Sleep parameters
|
APOE ε4 (+)
(n = 9): mean ± SD
|
APOE ε4 (-)
(n = 33): mean ± SD
|
p-value
|
|
TST (in minutes)
|
404 ± 30
|
369 ± 57
|
0.08
|
|
N1 (% TST)
|
9 ± 8
|
8 ± 6
|
0.83
|
|
N2 (% TST)
|
60 ± 8
|
57 ± 10
|
0.36
|
|
N3 (% TST)
|
13 ± 8
|
16 ± 10
|
0.44
|
|
REM Sleep (%)
|
18 ± 8
|
19 ± 6
|
0.71
|
|
Arousal index (events/hour)
|
34 ± 23
|
31 ± 14
|
0.72
|
|
AHI (events/hour)
|
35 ± 30
|
46 ± 27
|
0.29
|
|
Minimum O2 Saturation (%)
|
83 ± 10
|
79 ± 9
|
0.21
|
Abbreviations: AHI: apnea-hypopnea index; N1, N2, N3, sleep stages (non-rapid eye
movement); REM: rapid eye movement sleep; SD, standard deviation; TST, total sleep
time.
Cognitive function was worse in terms of processing speed, attention, and inhibitory
control evaluated through the Digit Symbol test and the SCWT-Part 2 in the moderate-to-severe
OSA (p < 0.05; [Table 3]). When excluding APOE ε4 carriers, these results remained worse in the moderate-to-severe OSA compared to
the no or mild OSA group. (See [Table S2] in the supplemental file for the cognitive evaluation excluding APOE ε4 carriers in the groups.) Furthermore, a multiple linear regression analysis among
APOE
ε4 noncarriers found that the AHI explained 61% of the results in the SCWT-Part 2 (p < 0.05). The correlation between AHI and the SCWT-Part 2 was significant (β = 0.610;
95%CI = 0.035–0.214; p = 0.008).
Table 3
Results of the cognitive evaluation in all patients.
|
Cognitive Tests
|
AHI < 15
(n = 13): mean ± SD
|
AHI ≥ 15
(n = 42): mean ± SD
|
p-value
|
|
Years of schooling
|
13 ± 4
|
12 ± 5
|
0.57
|
|
IQ
|
80 ± 16
|
80 ± 13
|
0.93
|
|
MMSE score
|
27 ± 2
|
27 ± 3
|
0.51
|
|
Trail A (in seconds)
|
39 ± 21
|
49 ± 27
|
0.26
|
|
Trail B (in seconds)
|
128 ± 76
|
173 ± 147
|
0.30
|
|
Direct Digits (number of correct answers)
|
6 ± 2
|
5 ± 3
|
0.41
|
|
Indirect Digits (number of correct answers)
|
5 ± 2
|
4 ± 2
|
0.20
|
|
Digit Symbol test (number of correct answers)
|
46 ± 17
|
35 ± 12
|
0.01*
|
|
SCWT- Part 1 (in seconds)
|
17 ± 3
|
22 ± 17
|
0.32
|
|
SCWT- Part 2 (in seconds)
|
20 ± 5
|
26 ± 9
|
0.002*
|
|
SCWT- Part 3 (in seconds)
|
31 ± 8
|
43 ± 23
|
0.07
|
|
Total Frontal Assessment Battery
|
16 ± 2
|
16 ± 2
|
0.85
|
|
SKT Attention
|
2 ± 1
|
3 ± 2
|
0.20
|
|
SKT Memory
|
1 ± 1
|
2 ± 1
|
0.10
|
|
Memory and Learning
|
|
|
|
|
RAVLT A1-A5
|
47 ± 9
|
42 ± 9
|
0.09
|
|
RAVLT late
|
10 ± 2
|
8 ± 3
|
0.17
|
Abbreviations: AHI, apnea-hypopnea index; IQ: intelligence quotient; MMSE: Mini-Mental State Examination;
SCWT, Stroop Color Word Test; SD, standard deviation; SKT, Syndrom-Kurztest; RAVLT,
Rey Auditory Verbal Learning Test (RAVLT A1-A5–sum of 5 recall trials; RAVLT late
– delayed recall after 30 minutes).
Note: *p ≤ 0.05.
In the moderate-to-severe OSA group, the comparison between APOE
ε4 carriers and noncarriers presented significant differences (p = 0.038) in the memory domain evaluated by the RAVLT A1-A5 test ([Table 4]).
Table 4
Results of the cognitive evaluation between apolipoprotein E (APOE) ε4 polymorphic allele carriers (+) and noncarriers (-) among patients with moderate-to-severe
sleep apnea (AHI ≥ 15 events/hour).
|
Cognitive Tests
|
APOE ε4 (-)
(n = 33): mean ± SD
|
APOE ε4 (+)
(n = 9): mean ± SD
|
p-value
|
|
Years of schooling
|
12 ± 4
|
12 ± 8
|
0.37
|
|
MMSE score
|
27 ± 2
|
27 ± 3
|
0.93
|
|
Trail A (in seconds)
|
49 ± 26
|
47 ± 32
|
0.67
|
|
Trail B (in seconds)
|
189 ± 162
|
116 ± 41
|
0.28
|
|
Direct Digits (number of correct answers)
|
5 ± 3
|
5 ± 2
|
0.83
|
|
Indirect Digits (number of correct answers)
|
4 ± 2
|
4 ± 1
|
0.47
|
|
Digit Symbol Test (number of correct answers)
|
35 ± 12
|
32 ± 14
|
0.42
|
|
SCWT – Part 1 (in seconds)
|
20 ± 5
|
30 ± 36
|
0.61
|
|
SCWT – Part 2 (in seconds)
|
26 ± 9
|
27 ± 8
|
0.87
|
|
SCWT – Part 3 (in seconds)
|
40 ± 18
|
53 ± 38
|
0.85
|
|
Total Frontal Assessment Battery
|
16 ± 2
|
17 ± 2
|
0.12
|
|
SKT Attention
|
3 ± 1
|
4 ± 3
|
0.11
|
|
SKT Memory
|
2 ± 1
|
1 ± 1
|
0.24
|
|
Memory and Learning
|
|
|
|
|
RAVLT A1-A5
|
43 ± 9
|
37 ± 8
|
0.038*
|
|
RAVLT late
|
9 ± 3
|
7 ± 3
|
0.29
|
Abbreviations: AHI, apnea-hypopnea index; MMSE: Mini-Mental State Examination; SCWT, Stroop Color
Word Test; SD, standard deviation; SKT, Syndrom-Kurztest; RAVLT, Rey Auditory Verbal
Learning Test (RAVLT A1-A5–sum of 5 recall trials; RAVLT late – delayed recall after
30 minutes).
Note: *p ≤ 0.05.
Discussion
In the present study, we evaluated sleep and cognitive parameters in patients with
and without moderate-to-severe OSA considering the presence of APOE ε4 polymorphism. Important findings emerged from the study. First, in line with previous
studies,[7] patients with moderate-to-severe OSA had lower cognitive performance in terms of
processing speed, divided attention, and inhibitory control (Symbol Digit Test, SCWT-Part
2), then patients with no or mild OSA. In both groups, when excluding APOE ε4 carriers, who tend to be associated with moderate-to-severe OSA, lower cognitive
performance remained for processing speed and attention domains in patients with moderate-to-severe
OSA. Second, among patients with moderate-to-severe OSA, the sleep parameters were
similar in APOE ε4 carriers and noncarriers, but the carriers presented significantly worse memory (RAVLT
A1-A5).
In the present study, the presence of at least 1 APOE ε4 allele was found in 7.7% of the patients with no or mild OSA and in 21.4% of those
with moderate-to-severe OSA. The apparent higher prevalence of the APOE ε4 allele in patients with moderate-to-severe OSA is in line with previous studies[17]
[24] that have indicated that APOE genes are associated with OSA susceptibility. Furthermore, in a previous study[25] in the general population, the APOE ε4 allele was associated with characteristics typical of OSA, including obesity, elevated
total cholesterol levels, low-density lipoprotein cholesterol, and elevated uric acid
levels. However, others studies[26]
[27] found no relationship between APOE ε4 and OSA susceptibility. Importantly, hypertension was identified as a plausible source
of heterogeneity and a confounding factor among studies.[26] Therefore, in the present study, we evaluated the influence of APOE ε4 polymorphism and cognitive functioning in sedentary OSA patients with no other major
comorbidities.
Consistent with the literature, we also found that patients with moderate-to-severe
OSA had lower cognitive performance in the domains of processing speed and attention
than patients with no or mild OSA.[7]
These differences remained even when all APOE ε4 carriers were excluded. Furthermore, by excluding the APOE ε4 carriers from the moderate-to-severe OSA group, the AHI explained 61% of the results
in the attention test of the SCWT-Part 2 (β = 0.610; 95%CI = 0.035-0.214; p = 0.008). The lower cognitive performance may be related to detrimental effects on
the neurocognitive system caused by intermittent hypoxia in brain tissue.[28]
[29]
Studies have shown that structural brain changes have been associated with deficits
in cognitive function in patients with moderate-to-severe OSA,[11]
[30] including the domains of attention, memory, executive function, and information
processing speed.[31]
In line with the studies, we also found lower cognitive performance for processing
speed and attention domains in patients with moderate-to-severe OSA. Except for the
Digit Symbol test and the SCWT-Part 2, none of the other attention tests showed significant
differences between both OSA groups. We speculate that OSA affects attentional performance
differently. The SCWT-Part 2 involves inhibitory control and processing speed, whereas
the Digit Symbol test involves divided attention and visual scanning. These cognitive
functions are specifically associated with the prefrontal region, which is known to
be an area strongly affected by intermittent hypoxia.[32] Areas of structural brain damage are consistent with deficits in attention, processing
speed, and inhibitory functions in OSA. A review study[33] examining neuroimaging in OSA suggested that the causes may be impairment in white-matter
tracts in the frontal region (important for speedy and efficient neural transmission)
and a lack of activation in the dorsolateral prefrontal cortices (crucial brain region
for inhibitory control, an executive function essential for behavioral self-regulation).
Impaired attentional skills, which can occur in patients with more severe OSA, can
also lead to reduced quality of life[34] and increased traffic and work accidents.[35]
[36] Furthermore, impaired inhibitory control may be associated with serious difficulties
and traits of impulsivity when trying to inhibit inappropriate behaviors to control
problems which, in turn, may interfere with sleep quality.
The APOE is a plasma protein that influences lipid metabolism,[25] nervous system growth and repair, synaptic and dendritic remodeling, and amyloid
clearance.[37] Carriers of APOE
ε4 have decreased expression of the APOE protein, making them more susceptible to cerebrovascular disease and Alzheimer disease.[38] The prevalence of APOE
ε4 carriers varies among populations worldwide, with rates of about 20% among Americans
and Europeans.[38]
[39] Among the patients with moderate-to-severe OSA in the present study, the genotype
frequency of the APOE ε4 was of 21.4%, close to a previous study that reported a rate of 23.8%.[23] Considering only patients with moderate-to-severe OSA, our results revealed that
the APOE ε4 carriers had significantly worse memory on the RAVLT test compared with noncarriers.
Our results differ from those of the study by Cosentino et al.,[23] who showed a difference in the memory domain in patients with OSA compared with
controls. However, the authors[23] found no impact of APOE ε4 carriers. We speculate that the difference from our investigation may be explained
by the fact that, in this previous study,[23] factors associated with cognitive decline, including higher BMI, diabetes mellitus,
and smoking,[40]
[41] were significantly higher in the OSA group. A previous study[42] reported dose-response relationship between BMI and the risk of cognitive impairment
and dementia, indicating that the risk is significantly increased when BMI surpasses
29 kg/m2 in middle age. Also, diabetes can influence cognitive function directly (by increasing
fluctuations in blood glucose levels or insulin resistance) or indirectly (leading
to microangiopathy of the brain) increasing the risk of stroke. Furthermore, chronic
nicotine use has been linked to impaired cognitive functioning in middle age.[43]
[44] A meta-analysis[45] identified a cross-sectional association between chronic tobacco smoking and impairments
in several neuropsychological domains, including attention, short- and long-term memory,
cognitive flexibility, cognitive impulsivity, and intelligence. On the other hand,
in the present study, we controlled the groups for a similar level of physical activity,
because physical activity is known to improve cognition and brain function,[46] and it could affect the results of the analysis.
Our results corroborate some findings of the study by Nikodemova et al.,[47] which found that APOE ε4 carriers with OSA had a high risk of developing cognitive deficits in memory and
executive function compared with APOE ε4 noncarriers. In the present study, we added the important information that sleep
parameters were similar in subjects with and without the APOE ε4 polymorphism in the moderate-to-severe OSA group. The presence of moderate-to-severe
OSA itself results in worse cognitive performance in the domains of processing speed,
attention, and inhibitory control. The association between OSA and the presence of
the APOE ε4 polymorphism revealed a decline in episodic memory. Thus, studies on cognition and
OSA may consider the genetic APOE ε4 polymorphism as a factor that may interfere with the results. Future studies including
many patients should also consider the presence of the APOE
ε2 allele, which, in OSA patients, can also cause harmful effects on sleep parameters,
impacting quality of life. Carriers of the APOE ε2 allele had longer sleep latency, lower sleep efficiency, and a higher number of arousals
per hour compared with homozygous carriers of the ε4 and ε3 alleles.[48]
The present study has several potential limitations. First, participants were classified
as having no or mild OSA if the AHI was < 15 events/hour of sleep, and as having moderate-to-severe
OSA if the AHI was ≥ 15 events/hour of sleep. This close cutoff between groups may
have had the effect of decreasing the sensitivity for discovering differences between
the groups. Recent studies indicate that elimination of wastes such as Aβ is improved
when the individual has normal sleep,[49] whereas the concentration or accumulation of Aβ deposition is greater if sleep is
fragmented.[50] However, available evidence suggests that a normal sleep pattern, including fewer
nocturnal awakenings, decreases the negative effects of the ε4 allele,[51] and cognitive impairment may be lower using a cutoff of AHI < 5 events/hour of sleep.
Despite the use of this closed cutoff, we found significant differences between the
groups. Second, the low rate of APOE ε4 allele carriers in the no or mild OSA group does not enable a comparison of cognitive
function regarding the the APOE ε4 polymorphism carriers in the moderate-to-severe group. Third, our analyses were based
on a single measurement of cognitive function without considering intraindividual
variation, so the accuracy of the results may have been affected. Finally, the overall
sample size was relatively small; therefore, the present study may have been insufficient
to detect other cognitive changes. Despite these limitations, but using specific inclusion
criteria, the present study[41]
[42]
[43]
[44] was able to provide a robust conclusion about the influence of APOE ε4 polymorphism and cognitive function in sedentary patients with OSA without other
major comorbidities.
The strengths of the present study are the application of cognitive tests sensitive
to specific domains, the use of nocturnal polysomnography, which is a gold standard
for sleep studies, the fact that the level of schooling and age corresponded in the
two groups, the exclusion of cardiovascular disease, diabetes, and cigarette smoking,
presence of hypertension, medication use (antidepressants, benzodiazepines, or non-benzodiazepine
anxiolytics), and the inclusion only of sedentary patients, because studies[41]
[42]
[43]
[44] have proven that these variables can greatly interfere with cognitive function.
Conclusion
In conclusion, sedentary patients with OSA without other important comorbidities experience
impairments in processing speed, divided attention, and inhibitory control in attentional
tasks. Episodic memory is further impaired by the presence of the APOE ε4 allele.
