Abstract
Introduction:
We evaluated whether insomnia symptom severity was associated with cognitive function, and whether this relationship was modified by biomarkers associated with Alzheimer’s disease risk.
Methods:
We examined insomnia symptoms and neuropsychological performance 3.4 years later in 511 dementia-free Framingham Heart Study participants (62.65 ± 8.7 years, 50.9% male). Additionally, we explored insomnia symptoms combined with self-reported short habitual sleep duration and effect modification by apolipoprotein E (APOE) ε4 allele status.
Results:
More severe insomnia symptoms were associated with lower performance on global cognition, and immediate and delayed Logical Memory recall, especially when insomnia symptoms were combined with short sleep duration. The association between insomnia symptoms and poorer memory recall was more pronounced in APOE ε4 allele carriers.
Discussion:
Insomnia symptom severity was associated with worse subsequent global cognitive and memory performance, which was especially apparent in APOE ε4 allele carriers, suggesting that poor sleep might be particularly detrimental when the brain is already vulnerable to neurodegeneration.
Keywords: aging, Alzheimer’s disease, apolipoprotein E ε4, cognition, cognitive aging, C-reactive protein, dementia, inflammation, insomnia, memory, neuropsychology, sleep
1 |. BACKGROUND
Insomnia is one of the most prevalent sleep disorders, especially in older adults.1,2 It is characterized by frequent and chronic difficulties in falling or maintaining sleep, or by early awakenings, which are associated with daytime impairments (e.g., fatigue, cognitive complaints, mood disturbances, impaired daily activities, behavioral problems).3 Although cognitive dysfunction in insomnia might be attributable to insufficient and fragmented sleep, subsequent sleepiness, fatigue, and lower vigilance from the night before, chronic insomnia might impair the brain in a more profound and long-standing way.
Insomnia is a disorder of hyperarousal, chronic stress, and elevated hypothalamic-pituitary-adrenal axis activity, which has been associated in some studies with elevated amyloid burden4 and impaired metabolic and inflammatory responses,5–7 all of which have the potential to affect brain health and cognitive function. Therefore, in older adults with preexisting risk factors for cognitive decline, insomnia might be an additional insult to the brain and result in lower cognitive performance. Although recent reports described insomnia as a risk factor for incident dementia and Alzheimer’s disease (AD),8–11 the existing literature is inconsistent as to how insomnia associates with cognitive dysfunction in middle-aged and older individuals, with heterogeneity across studies regarding the presence or nature of findings potentially due to variability in the operationalization of the insomnia diagnosis and cognitive testing.12 Further inconsistency may stem from different study populations with varying levels of vulnerability to cognitive dysfunctions, perhaps due to differences in genetic susceptibility to AD or systemic inflammation.
In this study, we sought to evaluate whether the severity of insomnia symptoms was associated with subsequent cognitive performance in a large population-based cohort of middle-aged and elderly individuals. Second, we evaluated whether the association between insomnia symptom severity and cognition varied according to two important dementia risk factors: the apolipoprotein E (APOE4) ε4 allele, the strongest genetic risk factor for AD, and systemic inflammation measured by C-reactive protein (CRP) levels, an active participant and signaling protein in inflammatory processes.13–15 We hypothesized that more severe insomnia symptoms would be associated with poorer cognitive performance and these associations would be more pronounced in persons with a higher risk of dementia (i.e., APOE ε4 carriers or higher CRP levels) or when combined with short sleep duration.
2 |. METHODS
2.1 |. Sample
This study included a subset of 511 participants from the Framingham Heart Study (FHS) Offspring cohort16 who participated in the Sleep Heart Health Study at the sixth FHS examination cycle (from 1995 to 1998) and completed a neuropsychological assessment at the seventh FHS examination cycle (from 1999 to 2002). This final sample of 511 participants was achieved by excluding participants aged < 40 years (as more significant age-related cognitive decline occurs after early adulthood, n = 6);17 participants with incomplete sleep and insomnia symptoms assessment (n = 44); and participants with neurological conditions such as dementia, stroke, and other neurological diseases (n = 56). All participants gave their written informed consent before the beginning of the study. The institutional review board at Boston University Medical Center approved the study.
2.2. |. Sleep assessment: insomnia symptoms and other sleep characteristics
Participants were evaluated with self-reported sleep questions during the evening, followed by an overnight in-home polysomnography. For the present study, we used data from five Likert-type questions representing insomnia symptoms. These questions represent the insomnia symptoms according to both the Diagnostic and Statistical Manual of Mental Disorders V (DSM-5) and the International Classification of Sleep Disorders-Third Edition (ICSD-3).18,19 The first three of these questions are very similar to questions contained within the Insomnia Severity Index.20
The questions were: “How often do you have trouble falling asleep?”; “How often do you wake up during the night and have difficulty getting back to sleep?”; “How often do you wake up too early in the morning and are unable to get back to sleep?”; “How often do you feel unrested during the day, no matter how many hours of sleep you had?”; “How often do you not get enough sleep?” Each question had the following response options: never (0/month); rarely (1/month or less); sometimes (2–4/month); often (5–15/month); or almost always (16–30/month). These answers were given a score of zero to four, four representing “almost always.” We calculated a combined insomnia symptom score by summing the answers of all five questions for a minimum score of 0 and a maximum score of 20, with higher scores representing more severe insomnia symptoms.
Because the first three questions (trouble falling asleep, difficulty resuming sleep, and waking up too early) are more closely related to insomnia than the other two questions (feeling unrested, not getting enough sleep) that might represent other sleep disorders as well, we performed additional sensitivity analyses with an alternative score including only the first three questions representing the core insomnia symptoms, for a minimum score of 0 and a maximum score of 12.
To better describe our sample, we also evaluated other self-reported measures, such as habitual sleep duration, the Epworth Sleepiness Scale, and naps per week (proportion of nappers and nap frequency). We also report objective sleep parameters measured with polysomnography closely associated with insomnia,21,22 including total sleep time, sleep latency, sleep efficiency, wake after sleep onset, awakenings by hour of sleep, as well as sleep stages. For descriptive purposes, we also reported variables representing obstructive sleep apnea (OSA): apnea-hypopnea index (AHI, hypopnea with an oxygen desaturation ≥4%), sleep time spent with SpO2 < 90%, mild to severe OSA (AHI > 5), moderate to severe OSA (AHI > 15). The montage, scoring criteria, and procedures of the polysomnography were published previously.23–25 Of note, polysomnography is not recommended nor required in the evaluation of insomnia symptoms,26 and was only included for descriptive purposes and to account for other sleep features such as OSA.
2.3 |. Cognitive assessment
Participants underwent a neuropsychological assessment, which was described previously.27 Using the complete neuropsychological assessment, a composite global measure of cognitive function was derived using principal component analysis forcing a single factor solution, as published by our group previously.28 We also investigated individual cognitive tests including Logical Memory immediate and delayed recall assessing verbal episodic memory; Trail Making Test Part A and Part B minus A to measure attention and executive functions; Similarities to measure abstract reasoning; and Hooper Visual Organization Test to assess visuoperceptual organization. All scores were coded such that higher scores represent better cognitive performance. These tests were selected from the full neuropsychological assessment to cover many different cognitive domains while also being somewhat parsimonious in selecting a limited number of tests measuring any single domain.
2.4 |. Biomarkers as modifiers
We evaluated the moderating effect of APOE ε4 allele carrier status and CRP levels. The genotyping of the two polymorphic sites of the APOE gene was made from whole blood, amplified by polymerase chain reaction for 35 cycles (DNA Thermal Cycler, PTC-100, MJ Research), and separated by electrophoresis. Participants were classified according to their APOE ε4 allele carrier status, that is, at least one ε4 allele for carriers (ε4/ε4, ε3/ε4, ε2/ε4) and no ε4 allele for non-carriers (ε3/ε3, ε3/ε2, ε2/ε2). High-sensitivity CRP levels were measured in serum during the sixth FHS examination (concomitantly with the sleep assessment) and assessed with a reagent containing monoclonal antibodies specific to human CRP (Dade Behring high-sensitivity CRP reagent). Participants were classified by CRP levels into the following categories according to American Heart Association guidelines: low inflammation < 1 mg/L, average inflammation 1 to 3 mg/L, and high inflammation > 3 mg/L.29
2.5 |. Clinical covariates
All covariates were assessed during the sixth FHS clinic examination (concomitantly with the sleep assessment). Body mass index (BMI) was calculated from each participant’s height and weight. The Revised Framingham Stroke Risk Profile (FSRP) score, a clinical 10-year risk prediction tool for incident stroke, was assessed using clinical information, including age, smoking status (self-report, ≥1 cigarette per day within the year), prevalent cardiovascular diseases (coronary heart disease, peripheral arterial disease, and/or heart failure), atrial fibrillation, diabetes mellitus (fasting blood glucose ≥ 126 mg/dL, or use of oral hypoglycemic agents or insulin), and hypertension (systolic ≥140 mmHg, diastolic ≥90 mmHg, or use of antihypertensive medications).30 The presence of depressive symptoms was defined as Center for Epidemiologic Studies Depression Scale ≥16,31 self-reported depression diagnosis, and/or current antidepressant usage. The use of sleeping pills was self-reported, with regular usage defined as at least one day per week.
2.6 |. Statistical analyses: primary and secondary analyses
Statistical analyses were performed using SAS software V9.4 (SAS Institute). For descriptive purposes, we performed Pearson and Spearman correlations between insomnia symptom severity scores and objective sleep parameters measured with polysomnography.
For the primary analysis exploring the associations between insomnia symptom severity and cognitive function, we used multivariate linear regression models. In initial models, age, age2 (because of the non-linear association between age and cognition in our cohort), sex, education, and the time interval between the sleep and neuropsychological assessments were included as covariates (Model 1). Subsequent models included additional adjustments for APOE ε4 allele carrier status, BMI, regular usage of sleeping pills, depressive symptoms, and the FSRP score (Model 2). These covariates were selected because of their association with either cognitive function and/or sleep. Insomnia symptom severity scores were standardized. To normalize distributions and to reduce the impact of outliers, natural log transformations were applied to Trail Making and Hooper Visual Organization Test scores and BMI. None of these variables had zero values. Participants with missing data were removed on an analysis-by-analysis basis. Primary results were considered significant at P < .05 adjusted for false discovery rate (FDR).
For our secondary hypothesis exploring whether biomarkers of dementia moderate the relationship between insomnia symptoms and cognition, an interaction term was added to evaluate the moderating effect of the APOE ε4 allele (dichotomous) and CRP levels (three levels: low, average, high) on the primary associations that showed significance. These analyses, and stratified regressions when interactions were significant, were adjusted for Model 1 covariates. All results were considered significant if P < .05, except for tests of interaction that were considered significant at P < .10 because they are generally less powerful than main effects.
2.7 |. Sensitivity analyses
To evaluate whether our primary findings were accounted for by other sleep features and OSA, we performed a series of sensitivity analyses adjusted for Model 2 covariates in addition to sleep parameters entered individually in a separate model. These variables represented OSA (AHI, sleep time with SpO2 < 90%, AHI > 5, AHI > 15), naps (frequency and nappers),and slow-wave sleep (N3 sleep, in percentage and minutes). The AHI was natural log transformed, and a very small value (“0.00001”) was added because a few participants had values equal to zero.
To evaluate whether our findings were the same when defining insomnia symptoms based on the core symptoms, we repeated our primary and secondary analyses with the insomnia symptom severity score computed with three core questions (trouble falling asleep, difficulty resuming sleep, waking up too early) rather than five questions.
Because clinical diagnoses of insomnia were not rendered in our sample, we performed a sensitivity analysis including participants that are most likely to present with clinical insomnia because they displayed high frequency of a core insomnia symptom, that is, participants that answered “almost always” to either one of these symptoms: trouble falling asleep, difficulty resuming sleep, and waking up too early. Because “almost always” corresponds to 16 to 30 times a month, and because the diagnosis criteria for insomnia states that the symptoms are present ≥3 times per week (i.e., ≥12–15 times a month), participants who answered “almost always” were most likely to fulfill a clinical diagnosis of insomnia. These participants were compared to those that answered “never” or “rarely” to all three questions.
To evaluate whether participants with high frequency of a core insomnia symptom who reported short sleep duration showed a similar pattern of cognitive performance than our primary findings, we evaluated participants that answered “almost always” to one of the three core insomnia questions and that reported sleeping < 6 hours per night compared to participants that answered “never” or “rarely” to all three questions, adjusted for Model 2 covariates.
To evaluate whether the association between insomnia symptom severity scores and cognitive performance varied depending on habitual sleep duration, we explored interactions by habitual sleep duration reported as < 6 hours per night compared to ≥6 hours per night for significant primary findings, adjusted for Model 1 covariates.
3 |. RESULTS
3.1 |. Sample characteristics
The sample included 511 participants aged between 44 and 84 years at the time of their sleep assessment. The distribution of participants’ answers to each of the five questions regarding insomnia symptoms as well as the insomnia symptom severity score are presented in Figure 1. Sample characteristics are described in Table 1. On average, cognitive testing was performed 3.4 years after the sleep assessment and blood tests. Participants with higher insomnia symptom severity comprised more women, reported shorter habitual sleep duration, slightly more sleepiness, and more of them used sleeping pills regularly (see Table 1). Both insomnia symptom severity scores did not correlate with nap frequency (r = 0.046, P = .301; r = 0.031, P = .481). A higher Epworth Sleepiness Scale score correlated with more severe insomnia symptoms calculated on five questions (r = 0.115, P = .011) but not three questions (r = 0.043, P = .343).
FIGURE 1.
Distribution of each insomnia symptom and combined score of insomnia symptom severity. Each question (A-E) had the following response options, which were given a score of zero to four: never (0/month); rarely (1/month or less); sometimes (2–4/month); often (5–15/month); or almost always (16–30/month). The combined insomnia symptom score (F) was obtained by summing the answers of all five questions, with higher scores representing more severe insomnia symptoms
TABLE 1.
Demographic, clinical, and sleep characteristics of the complete sample by low and high insomnia severity symptom score on five questions
| Mean (standard deviation) or n (%) |
|||
|---|---|---|---|
| Characteristics | Complete sample (n = 511) | Low insomnia symptom severity (< median, n = 222) | High insomnia symptom severity (≥ median, n = 289) |
| Demographic | |||
| Age at sleep evaluation (years) | 58.6 (8.8) | 59.2 (8.6) | 58.2 (8.8) |
| Age at neuropsychological assessment (years) | 62.5 (8.7) | 63.0 (8.5) | 62.1 (8.8) |
| Time between sleep and cognitive testing (years) | 3.4 (1.0) | 3.3 (1.0) | 3.4 (1.0) |
| Sex, men, n (%)‡ | 251 (49.1) | 133 (59.9) | 118 (40.8) |
| Education, n(%)* | |||
| No high school degree | 25 (4.9) | 12 (5.4) | 13 (4.5) |
| High school degree | 155 (30.3) | 57 (25.7) | 98 (33.9) |
| Some college education | 160 (31.3) | 65 (29.3) | 95 (32.9) |
| College graduate | 171 (33.5) | 88 (39.6) | 83 (28.7) |
|
| |||
| Biomarkers and clinical covariates | |||
| Apolipoprotein E ε4 allele carrier, n (%) | 107 (21.4) | 41 (19.0) | 66 (23.2) |
| C-reactive protein levels (mg/L) | |||
| Low inflammation < 1 mg/L, n(%) | 147 (29.5) | 60 (27.5) | 87 (31.0) |
| Average inflammation 1–3 mg/L, n(%) | 155 (31.1) | 73 (33.5) | 82 (29.2) |
| High inflammation > 3 mg/L, n(%) | 197 (39.5) | 85 (39.0) | 112 (39.9) |
| Depression symptoms, n (%)‡ | 64 (12.5) | 14 (6.3) | 50 (17.3) |
| Systolic blood pressure (mmHg) | 126 (17.0) | 125 (16.4) | 127 (17.8) |
| Treated for hypertension, n (%) | 118 (23.1) | 52 (23.5) | 66 (22.8) |
| Framingham Stroke Risk Profile score, score units | 0.03 (0.04) | 0.04 (0.04) | 0.03 (0.04) |
| Body mass index (kg/m2) | 28.1 (5.0) | 28.2 (4.8) | 28.0 (5.2) |
| Diabetes mellitus, n (%) | 46 (9.0) | 21 (9.5) | 25 (8.7) |
| Smoking, n (%) | 72 (14.1) | 32 (14.4) | 40 (13.8) |
| Prevalent atrial fibrillation, n (%) | 11 (2.2) | 7 (3.2) | 4 (1.4) |
| Prevalent cardiovascular disease, n (%) | 32 (6.3) | 12 (5.4) | 20 (6.9) |
|
| |||
| Insomnia symptom severity score | |||
| Score computed on five questions, score units on 20‡ | 7.7 (3.9) | 4.3 (1.6) | 10.3 (3.1) |
| Score computed on three questions, score units on 12‡ | 4.6 (2.7) | 2.5 (1.3) | 6.2 (2.3) |
|
| |||
| Self-reported sleep parameters | |||
| Self-reported habitual sleep duration (h)‡ | 7.1 (1.2) | 7.3 (1.0) | 6.8 (1.3) |
| Epworth Sleepiness Scale, score units* | 6.9 (4.1) | 6.4 (4.1) | 7.3 (4.1) |
| Regular usage of sleeping pills, n (%)‡ | 87 (17.0) | 17 (7.7) | 70 (24.2) |
| Nap frequency, #/week | 2.0 (3.4) | 1.8 (2.5) | 2.2 (4.0) |
| Nappers, n (%) | 243 (48.1) | 105 (47.5) | 138 (48.6) |
|
| |||
| Sleep parameters at the polysomnography | |||
| Total sleep time, h | 6.3 (0.9) | 6.3 (0.8) | 6.2 (1.0) |
| Sleep latency, min | 22.0 (12.2) | 20.8 (20.4) | 22.9 (25.1) |
| Sleep efficiency, % | 83.2 (9.7) | 84.3 (8.6) | 82.4 (10.4) |
| Wake after sleep onset (min) | 55.8 (40.0) | 54.0 (37.0) | 57.1 (42.2) |
| Number of awakenings per hour of sleep | 3.7 (1.6) | 3.8 (1.7) | 3.7 (1.6) |
| Apnea-hypopnea index, events/hour | 8.7 (12.6) | 8.6 (11.1) | 8.8 (13.6) |
| Mild to severe obstructive sleep apnea, n (%) | 221 (46.2) | 97 (47.1) | 124 (45.6) |
| Moderate to severe obstructive sleep apnea, n (%) | 85 (17.8) | 38 (18.5) | 47 (17.3) |
| Sleep time with SpO2 < 90%, % of total sleep time | 2.7 (8.3) | 2.7 (9.6) | 2.7 (7.1) |
| N1 sleep, % of total sleep time | 4.9 (3.2) | 5.0 (3.3) | 4.9 (3.2) |
| N1 sleep, min | 18.6 (12.2) | 18.9 (12.9) | 18.3 (11.7) |
| N2 sleep, % of total sleep time | 55.8 (11.1) | 55.7 (11.2) | 55.9 (11.0) |
| N2 sleep (min) | 211.2 (49.6) | 211.9 (48.1) | 210.6 (50.8) |
| N3 sleep, % of total sleep time | 18.4 (10.6) | 17.8 (10.8) | 18.8 (10.5) |
| N3 sleep (min) | 69.9 (41.6) | 68.9 (44.0) | 70.7 (39.6) |
| REM sleep, % of total sleep time* | 20.9 (5.9) | 21.5 (5.3) | 20.4 (6.3) |
| REM sleep (min) | 80.1 (27.0) | 82.6 (24.2) | 78.2 (28.8) |
|
| |||
| Neuropsychological assessment | |||
| Composite global cognitive score | −0.11 (0.95) | −0.07 (0.96) | −0.14 (0.95) |
| Logical Memory immediate recall, score units | 11.1 (3.3) | 11.4 (3.2) | 11.0 (3.3) |
| Logical Memory delayed recall, score units | 10.2 (3.4) | 10.4 (3.4) | 10.0 (3.5) |
| Trail Making Test Part A (min) | 0.56 (0.23) | 0.58 (0.28) | 0.55 (0.18) |
| Trail Making Test Part B minus A (min) | 0.87 (0.79) | 0.85 (0.79) | 0.89 (0.78) |
| Similarities, score units | 16.3 (3.7) | 16.5 (3.4) | 16.1 (3.9) |
| Hooper Visual Organization Test, score units | 24.9 (3.2) | 24.8 (3.4) | 25.1 (3.1) |
Notes: Univariate t tests were performed for continuous variables and Chi-square tests were performed for categorical variables between groups under and over the insomnia symptom severity score median.
P < .05;
P < .01;
P < .001.
Table 2 shows how insomnia symptom severity scores correlate with objective sleep measurement on the night of the polysomnography. More severe insomnia symptoms (five questions) correlated with lower sleep efficiency and a shorter time in REM sleep. More severe insomnia symptoms (three core questions) correlated with lower sleep efficiency, longer wake after sleep onset, and a shorter time in N2 and REM sleep. Interestingly, both scores did not associate with the AHI.
TABLE 2.
Correlation matrix between insomnia symptom severity and objective sleep parameters at the polysomnography
| [1] | [2] | [3] | [4] | [5] | [6] | [7] | [8] | [9] | [10] | [11] | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| [1] Insomnia symptoms (5) | – | ||||||||||
| [2] Insomnia symptoms (3) | 0.92‡ | – | |||||||||
| [3] Total sleep time | −0.06 | −0.09 | – | ||||||||
| [4] Sleep latency | 0.05 | 0.03 | −0.39‡ | – | |||||||
| [5] Sleep efficiency | −0.15† | −0.21‡ | 0.71‡ | −0.56‡ | – | ||||||
| [6] Wake after sleep onset | 0.05 | 0.15‡ | −0.35‡ | 0.15† | −0.83‡ | – | |||||
| [7] # of Awakenings | −0.06 | −0.02 | −0.17† | 0.01 | −0.28‡ | 0.38‡ | – | ||||
| [8] Apnea-hypopnea index | −0.04 | −0.05 | −0.19‡ | 0.07 | −0.16† | 0.15† | 0.17‡ | – | |||
| [9] N1 min | −0.05 | −0.01 | 0.15† | 0.03 | −0.08 | 0.26 | 0.37‡ | 0.02 | – | ||
| [10] N2 min | −0.07 | −0.12† | 0.55‡ | −0.17† | 0.35‡ | −0.10* | 0.05 | 0.01 | 0.16‡ | – | |
| [11] N3 min | 0.06 | 0.07 | 0.27‡ | −0.19‡ | 0.29‡ | −0.20‡ | −0.24‡ | −0.18† | −0.32‡ | −0.50‡ | – |
| [12] REM min | −0.09* | −0.11* | 0.58‡ | −0.19‡ | 0.46‡ | −0.28‡ | −0.07 | −0.23‡ | 0.08 | 0.05 | −0.01 |
Notes: R values are represented. Spearman correlations were used when variables were not distributed normally or symmetrically, and Pearson correlations were used otherwise. Insomnia symptoms (5), insomnia symptoms severity score calculated on five questions; Insomnia symptoms (3), insomnia symptoms severity score calculated on three questions.
P < .05;
P < .01;
P < .001.
3.2 |. Associations between the severity of insomnia symptoms and cognitive performance
More severe insomnia symptoms were associated with lower global cognitive functioning and episodic memory as measured by the Logical Memory test, immediate and delayed recall for both statistical models (see Table 3). No associations were observed for Similarities, Trail Making Test, or Hooper Visual Organization Test.
TABLE 3.
Linear regressions models between the severity of insomnia symptoms and cognitive performance
| Cognitive performance | Insomnia symptom severity, five questions |
Insomnia symptom severity, three questions |
High frequency of a core insomnia symptom |
|||
|---|---|---|---|---|---|---|
| β (standard error) | P * | β (standard error) | P * | β (standard error) | P | |
| Composite global cognitive score | ||||||
| Model 1 | −0.098 (0.035) | .020 | −0.110 (0.035) | .005 | −0.396 (0.139) | .005 |
| Model 2 | −0.098 (0.038) | .023 | −0.114 (0.037) | .005 | −0.359 (0.148) | .016 |
|
| ||||||
| Logical Memory, immediate recall | ||||||
| Model 1 | −0.382 (0.140) | .020 | −0.435 (0.140) | .005 | −0.931 (0.488) | .058 |
| Model 2 | −0.447 (0.150) | .014 | −0.503 (0.147) | .003 | −0.905 (0.523) | .085 |
|
| ||||||
| Logical Memory, delayed recall | ||||||
| Model 1 | −0.452 (0.146) | .014 | −0.539 (0.145) | .001 | −1.109 (0.516) | .033 |
| Model 2 | −0.521 (0.156) | .013 | −0.607 (0.152) | .001 | −1.028 (0.554) | .065 |
|
| ||||||
| Trail Making Test, Part A (inverted) | ||||||
| Model 1 | −0.006 (0.014) | .812 | 0.010 (0.014) | .552 | 0.056 (0.057) | .333 |
| Model 2 | 0.005 (0.015) | .812 | 0.019 (0.015) | .274 | 0.088 (0.062) | .159 |
|
| ||||||
| Trail Making Test, Part B minus A (inverted) | ||||||
| Model 1 | −0.009 (0.009) | .470 | −0.0128 (0.009) | .244 | −0.064 (0.041) | .119 |
| Model 2 | −0.007 (0.010) | .657 | −0.0119 (0.010) | .274 | −0.064 (0.044) | .146 |
|
| ||||||
| Similarities | ||||||
| Model 1 | −0.279 (0.154) | .123 | −0.362 (0.153) | .032 | −1.842 (0.602) | .003 |
| Model 2 | −0.326 (0.162) | .090 | −0.416 (0.159) | .018 | −1.754 (0.635) | .006 |
|
| ||||||
| Hooper Visual Organization Test | ||||||
| Model 1 | −0.002 (0.023) | .935 | −0.001 (0.023) | .953 | −0.082 (0.083) | .323 |
| Model 2 | −0.009 (0.025) | .812 | −0.011 (0.024) | .712 | −0.082 (0.089) | .357 |
Notes: Model 1 is adjusted for age, age squared, sex, time interval between the sleep and neuropsychological assessments, and education. Model 2 is additionally adjusted for body mass index, regular usage of sleeping pills, depressive symptoms, apolipoprotein E ε4 allele carriers, and Framingham Stroke Risk Profile score. Insomnia symptom severity scores were standardized and treated as continuous variables. High frequency of a core insomnia symptom was treated categorically. Participants who answered “almost always” to one of the core insomnia symptom questions (trouble falling asleep, difficulty resuming sleep, and waking up too early) were included in the group with high frequency of a core insomnia symptom (n = 51), and were compared to those that answered either “never” or “rarely” to all of these questions (n = 177). Trail Making Test and Hooper Visual Organization Test were transformed using the natural log. Score units for Logical Memory are number of recalled correct, score units for Similarities and Hooper Visual Organization Test are number correct, and Trail Making Test is time to completion. Scores on the Trail Making Test were multiple by negative 1 such that higher scores indicate better performance.
False discovery rate-corrected.
To provide a comparison with cognitive aging, we regressed delayed memory recall scores on age and compared coefficients with our findings. Each unit increase in the insomnia symptom severity score based on five questions was associated with 1.75 years of cognitive aging as defined by lower delayed recall on Logical Memory, adjusted for Model 2 covariates.
More severe insomnia symptoms were still associated with lower global cognitive functioning and lower immediate and delayed recall when including additional adjustments for OSA, naps, or slow-wave sleep (Table 4).
TABLE 4.
Linear regression models between the severity of insomnia symptoms with worse global cognitive and memory performance including additional adjustments
| Insomnia severity score (five questions), Additionally adjusted for: |
Composite global cognitive score
|
Logical memory, immediate recall
|
Logical memory, delayed recall
|
|||
|---|---|---|---|---|---|---|
| β (standard error) | P | β (standard error) | P | β (standard error) | P | |
| Obstructive sleep apnea | ||||||
|
| ||||||
| Apnea-hypopnea index | −0.096 (0.041) | .020 | −0.399 (0.160) | .013 | −0.483 (0.166) | .004 |
|
| ||||||
| Sleep time with SpO2 < 90% | −0.098 (0.038) | .010 | −0.449 (0.150) | .003 | −0.522 (0.155) | .0008 |
|
| ||||||
| Mild to severe obstructive sleep apnea | −0.086 (0.040) | .032 | −0.357 (0.155) | .022 | −0.430 (0.160) | .007 |
|
| ||||||
| Moderate to severe obstructive sleep apnea | −0.086 (0.040) | .032 | −0.359 (0.155) | .021 | −0.432 (0.160) | .007 |
|
| ||||||
| Naps | ||||||
|
| ||||||
| Nap frequency | −0.101 (0.038) | .009 | −0.441 (0.150) | .004 | −0.511 (0.156) | .001 |
|
| ||||||
| Nappers | −0.100 (0.038) | .009 | −0.434 (0.150) | .004 | −0.507 (0.156) | .001 |
|
| ||||||
| Slow-wave sleep | ||||||
|
| ||||||
| N3 sleep min | −0.096 (0.038) | .012 | −0.442 (0.151) | .004 | −0.518 (0.156) | .001 |
|
| ||||||
| %N3 sleep | −0.098 (0.038) | .010 | −0.444 (0.150) | .003 | −0.524 (0.156) | .0008 |
Notes: Models are adjusted for age, age squared, sex, time interval between the sleep and neuropsychological assessments, education, body mass index, regular usage of sleeping pills, depressive symptoms, apolipoprotein E ε4 allele carriers, and Framingham Stroke Risk Profile score, in addition to the listed variable. Standardization was applied to the insomnia severity score.
Using the alternative insomnia symptom severity score based on three questions rather than five, more severe insomnia symptoms remained associated with lower global cognitive functioning, worse immediate recall and delayed recall, as well as with lower performance on Similarities (Table 3).
3.3 |. Cognitive performance, probable clinical insomnia, and short sleep duration
When using a categorical definition of insomnia (Table 3), participants with high frequency of a core insomnia symptom, that is, who answered “almost always” to one of the core insomnia questions (n = 51), had lower global cognitive functioning, trends for worse immediate and delayed recall, as well as lower performance on Similarities compared to participants who answered “never” or “rarely” to all three questions (n = 177).
Participants who answered “almost always” to one of the core insomnia questions who also reported sleeping < 6 hours per night (n = 20) had lower global cognitive functioning (β[standard error (SE)], −0.909 [0.217], P < .0001), worse immediate recall (β[SE], −2.259 [0.764], P = .004) and delayed recall (β[SE], −3.056 [0.798], P = .0002), and lower performance on Similarities (β[SE], −2.864 [0.949], P = .003) compared to participants who answered “never” or “rarely” to all three questions (n = 177). No association was observed with Trail Making Test or Hooper Visual Organization Test. Of note, these effect sizes were larger compared to when sleep duration was not considered (see Table 3).
Evaluating the moderating effect of habitual sleep duration, we did not observe significant interactions for the immediate and delayed recall. On the other hand, habitual sleep duration moderated the association between insomnia symptom severity (five questions, P = .032; three core questions, P = .021) and global cognition. In participants who reported sleeping < 6 hours per night, more severe insomnia symptoms were associated with lower global cognitive functioning (five questions, (β[SE], −0.297 [0.129], P = .027; three core questions, β[SE], −0.304 [0.114], P = .011). This association was not observed in those who reported sleeping ≥6 hours (β[SE], −0.031 [0.042], P = .450; β[SE], −0.043 [0.042], P = .300; respectively).
3.4 |. Moderating effect of the APOE genotype and CRP levels on the relation between insomnia symptoms and cognitive performance
The severity of insomnia symptoms significantly interacted with APOE genotype in its association with Logical Memory delayed recall (Table 5). While more severe insomnia symptoms (five and three questions) were associated with poorer memory performance in both non-carriers and carriers, this effect was about three times stronger in APOE ε4 carriers. Additionally, a similar interaction was observed for the global composite cognitive score, for which the association between insomnia symptom severity (three core questions) and poorer cognitive performance was stronger in APOE ε4 carriers than in non-carriers.
TABLE 5.
Interactions and stratified linear regression models between the severity of insomnia symptoms and APOE genotype when predicting cognitive performance
| Insomnia symptoms definition | APOE ε4 stratification |
Composite global cognitive score
|
Logical Memory, delayed recall
|
||
|---|---|---|---|---|---|
| β (standard error) | P | β (standard error) | P | ||
| Insomnia symptom severity, five questions | Interaction term | .176 | .053 | ||
| Non-carriers | −0.330 (0.165) | .047 | |||
| APOE ε4 carriers | −1.029 (0.331) | .003 | |||
|
| |||||
| Insomnia symptom severity, three questions | Interaction term | .060 | .035 | ||
| Non-carriers | −0.075 (0.042) | .071 | −0.408 (0.165) | .014 | |
| APOE ε4 carriers | −0.236 (0.068) | .0008 | −1.148 (0.317) | .0005 | |
|
| |||||
| High frequency of a core insomnia symptom | Interaction term | .084 | .049 | ||
| Non-carriers | −0.237 (0.172) | .171 | 0.381 (0.635) | .550 | |
| APOE ε4 carriers | −0.775 (0.233) | .002 | −2.701 (0.945) | .007 | |
Notes: All results are adjusted for age, age squared, sex, time interval between the sleep and neuropsychological assessments, and education. Tests of interactions were considered significant at P < .1, while stratifications were considered significant at P < .05. Insomnia symptom severity scores were standardized and treated as continuous variables. High frequency of a core insomnia symptom was treated categorically. Participants who answered “almost always” to one of the core insomnia symptoms questions (trouble falling asleep, difficulty resuming sleep, and waking up too early) were included in the group with high frequency of a core insomnia symptom (n = 51), and were compared to those that answered either “never” or “rarely” to all of these questions (n = 177).
Abbreviations: APOE, apolipoprotein E; CRP, C-reactive protein.
Moreover, similar interactions were observed when examining the group with high frequency of a core insomnia symptom (Table 5), for which, in APOE ε4 carriers, participants who answered “almost always” to one of the core insomnia symptoms had poorer global cognitive and delayed recall performances compared to participants who answered “rarely” or “never” to all three questions, while it was not the case in non-carriers.
Furthermore, the association between more severe insomnia symptoms and poorer delayed memory recall was only observed in those with average and high CRP levels (Interaction, P = .069; β[SE], −0.760 [0.295], P = .011; β[SE], −0.506 [0.215], P = .019, respectively), while it was not present in the low CRP group (β[SE], −0.089 [0.284], P = .753). However, this interaction was not present when using the insomnia symptom severity score computed on three questions rather than five.
No other interactions were observed between insomnia symptoms and other cognitive tests.
4 |. DISCUSSION
In this community-based cohort study of dementia-free middle-aged to older adults, increased severity of self-reported insomnia symptoms was associated with poorer global cognitive performance and verbal episodic memory measured approximately 3 years later. The association between insomnia symptom severity and verbal episodic memory was especially apparent in APOE ε4 carriers. Given the growing evidence linking sleep disturbances and insomnia to increased risk of incident dementia,8 the present findings suggest that more severe insomnia symptoms might be associated with poorer subsequent cognitive functioning when the brain is especially vulnerable to neurodegenerative processes.
4.1 |. Cognitive performance in middle-aged and older individuals with insomnia symptoms
A recent meta-analysis concluded that insomnia was associated with poorer cognitive functioning in domains such as perceptual processes, alertness, complex attention, executive functions, working memory, and episodic memory.12 The present findings are partially consistent with the meta-analysis, because we observed that more severe insomnia symptoms were associated with poorer global cognitive functioning, verbal memory, and abstract reasoning, although we did not observe effects for attention, processing speed, executive functions, or visuoperceptual organization. However, there are several inconsistencies among previous individual studies evaluating the links between insomnia and cognition. First, many of these reports evaluated participants using varying definitions of insomnia. This led not only to inconsistencies regarding who is considered to have insomnia, but in addition, it does not account for the severity of the condition. In fact, the prevalence of insomnia can vary from 6% to 33% depending on the definition used and the number and severity of symptoms included.32 It was recently hypothesized that assessing insomnia as a continuous dimension of symptoms might be more relevant than categorical diagnoses to the study of insomnia in research contexts.12,33 In the present study performed in a sample of middle-aged and elderly individuals, we show an association suggesting a dose-response relationship between more severe insomnia symptoms and cognitive dysfunction, especially lower verbal episodic memory. Most groups previously evaluated young adults with insomnia, and it is unclear how insomnia affects cognitive functioning in older adults, especially those at higher risk of incident dementia. Insomnia prevalence increases with age1,2 and thus it is especially important to understand how it associates with cognitive functioning in this population. Additionally, insomnia in the elderly might have begun decades earlier, and thus, years of exposure to insomnia could be more strongly associated with poorer cognitive outcomes. Consistently, a few other studies performed in older adults also observed memory impairments with insomnia.34–36
4.2 |. Contribution of other sleep disorders and characteristics
Interestingly, the association between more severe insomnia symptoms and poorer cognitive performance remained significant when adjusting for OSA, and insomnia symptom severity scores did not correlate with the AHI. This suggests that insomnia symptoms were not accounted for by sleep disruption caused by OSA. There is increasing interest in the role of slow-wave sleep loss in cognitive aging and neurodegeneration,37,38 and shorter slow-wave sleep has been reported previously in association with insomnia symptoms.39 However, adjusting for slow-wave sleep in our statistical models did not reduce the strength of our findings with cognitive performance. These data suggest that the association between insomnia symptoms and cognitive performance does not seem to be accounted for by slow-wave sleep, although a single night of polysomnography does not allow us to completely evaluate this question. In addition, our findings remained significant when adjusting for daytime naps, suggesting that the association between insomnia symptoms and cognition is not accounted for by insufficient nighttime sleep leading to elevated daytime sleep propensity.
Insomnia with short sleep duration is a phenotypical presentation that is increasingly hypothesized to be the most detrimental form of insomnia, characterized by a particularly elevated state of activated stress systems and hyperarousal.40 In this study, individuals with high frequency of a core insomnia symptom who also reported short sleep duration presented with the same pattern of cognitive performance as when we did not consider sleep duration, but these deficits were more marked. We also observed that more severe insomnia symptoms were only associated with lower global cognitive functioning in those with short sleep duration. This finding suggests that, in participants who report insufficient short sleep duration, the presence of increasingly severe insomnia symptoms is a predictor of poorer global cognitive functioning. On the other hand, sleep duration was not a moderator when the outcome was memory, suggesting that insomnia symptoms are associated with memory performance whether participants report short sleep duration or not. Overall, our findings support that insomnia itself, independently of many sleep characteristics, is associated with poorer cognitive function, especially when short sleep duration is present.
4.3 |. Moderating effect of biomarkers of dementia risk on the association between insomnia symptoms and memory
An important and novel finding of the present study is that more severe insomnia symptoms were associated differently with memory performance depending on the APOE genotype. The APOE ε4 allele, the strongest genetic risk factor for sporadic AD,41 leads to decreased expression of the apoE protein and impaired cholesterol transport in the central nervous system essential to synapses and dendrites.42 Moreover, the apoE protein is an amyloid scavenger, and APOE ε4 allele carriers show elevated amyloid burden. Insomnia might exacerbate impairments in APOE ε4 allele carriers, leading to poorer cognitive performance. This is consistent with what we observed in the present study: the association between more severe insomnia symptoms with poorer delayed memory recall was markedly stronger in carriers compared to non-carriers. Poorer delayed memory recall is known as a strong predictor of progression to cognitive impairment and AD.43 In fact, insomnia is associated with impaired metabolic function and elevated amyloid production,4,5,7 and thus, more severe insomnia symptoms in APOE ε4 allele carriers might contribute more extensively to the ongoing neurodegenerative pathology and lead to memory impairments. Therefore, in older adults with preexisting risk factors for cognitive decline, insomnia might be an additional insult to the brain and result in lower cognitive performance.
On the other hand, ongoing neurodegenerative processes might affect sleep regulatory structures and lead to sleep disturbances,44 including insomnia. In the age group studied here, ongoing neurodegenerative processes would be expected to be more present in APOE ε4 allele carriers. Thus, this might explain the markedly strong association between insomnia symptoms and cognitive performance in this high-risk group, in which ongoing neurodegeneration processes in APOE ε4 allele carriers could lead to both impaired memory performance and insomnia symptoms. Previous studies have shown that APOE ε4 allele carriers presented with sleep characteristics consistent with insomnia (i.e., insomnia itself, shorter sleep duration, more awakenings, and self-reported daytime and sleep disturbances),45–48 although sleep disturbances of a different nature have been reported in carriers as well (i.e., longer sleep duration).49
To our knowledge, only one previous study evaluated whether the APOE genotype interacts with insomnia to predict brain health outcomes. In cognitively healthy older adults, Grau-Rivera et al. found that the presence of insomnia was associated with worse delayed memory recall in APOE ε4 carriers only, while in non-carriers, insomnia was associated with worse working memory.50 We add to these findings by showing that worse delayed memory recall in APOE ε4 carriers is also observed when we consider insomnia severity rather than a dichotomic diagnosis. Interestingly, other studies suggest that this interaction is not limited to insomnia but is present with other sleep characteristics. OSA was reported to interact with the APOE ε4 allele to predict worse memory performance.51,52 Additionally, better sleep consolidation attenuated the effect of the APOE ε4 allele on AD risk, annual rate of cognitive decline, and post mortem neurofibrillary tangles burden.53
Elevated CRP has been associated with increased risk of dementia or cognitive decline,13,15,54 and has also been reported to be elevated in people with insomnia.7,55,56 In the present study, we observed that more severe insomnia was associated with worse delayed memory recall only in average and high systemic inflammation groups, but not in those exhibiting low inflammatory levels (< 1 mg/L). Insomnia symptoms that associate (or trigger) elevated inflammation may be most strongly predictive of cognitive impairments. On the other hand, elevated inflammation might already be present with ongoing neurodegenerative processes, and lead to damage to sleep regulatory cerebral structures. However, using the insomnia score based on the three core symptoms, we did not observe the same interaction with CRP levels. Because cognitive behavioral therapy for insomnia has been shown to lead to insomnia remission and to lower CRP levels,6 further treatment studies could clarify the causality of the association among insomnia, inflammation, and cognitive aging.
5 |. LIMITATIONS
Although we chose to focus on symptoms of insomnia, we did not have information on the clinical diagnosis of insomnia for the purposes of comparison. In a sensitivity analysis, we observed that our findings were still present in participants most likely to have clinical insomnia (i.e., who answered “almost always” to a core insomnia symptom), suggesting that examining a clinical diagnosis might have led to similar findings. Future studies should aim to replicate our findings in a sample of clinically diagnosed participants with chronic primary insomnia. Nevertheless, diagnostic criteria for insomnia have not been designed with the aim of predicting cognitive impairment or dementia. As stated before, investigating insomnia symptoms in a continuous manner has value in research settings to better understand how insomnia is associated with cognitive aging, especially given the large variability in insomnia severity in diagnosed cases.
The design of our study did not allow us to establish a causal relationship between insomnia symptoms and cognitive functioning. Treatment studies aimed at insomnia remission, especially in APOE ε4 allele carriers, would help to clarify if the presence of insomnia contributes to cognitive decline. Because we did not have circadian data, we were unable to determine whether circadian disorders contributed to the associations between insomnia symptoms and cognition. However, insomnia can lead to circadian misalignment, and thus, future studies should aim to disentangle their specific contributions to cognitive aging. Finally, although effects were larger for delayed than immediate memory recall, we could not determine whether insomnia symptoms were differently associated with encoding and retrieval processes.
6 |. CONCLUSIONS
To our knowledge, this is the first study evaluating whether the APOE genotype modifies the association of insomnia with cognitive performance in older adults. Because insomnia is treatable, its role as a modifiable risk factor for cognitive dysfunction, cognitive decline, and dementia in the older population must be better understood to guide clinical recommendations, especially when it presents with short sleep duration. In this study, we show a linear association between more severe insomnia symptoms and cognitive performance, and that sub-populations at higher risk of AD might be more vulnerable to the effect of insomnia symptoms on cognitive health.
RESEARCH IN CONTEXT.
Systematic review: We reviewed the literature exploring the association between insomnia and cognition in the elderly using PubMed. Evidence is accumulating that insomnia could be associated with Alzheimer’s disease (AD) risk in older individuals, but heterogeneity remains when evaluating the association between cognition and insomnia symptoms.
Interpretation: Our findings show that the severity of insomnia symptoms is linearly associated with poorer cognitive performance, especially memory. Supporting previous work identifying insomnia as a dementia risk factor, we observed that poorer memory recall was associated more strongly with insomnia symptom severity in apolipoprotein E ε4 allele carriers.
Future directions: Additional studies are needed to understand mechanisms underlying the association between insomnia and poor cognitive outcomes to better target individuals at risk. Because insomnia is treatable, future research could evaluate whether an effective insomnia treatment reduces the risk of AD.
HIGHLIGHT.
More severe insomnia symptom severity was associated with poorer cognitive function
Poor memory recall was especially correlated with the severity of insomnia symptoms
This association was stronger in carriers of the apolipoprotein E ε4 allele
Insomnia symptoms might be detrimental to those already at higher Alzheimer’s risk
ACKNOWLEDGMENTS
The authors thank the participants and staff of the Framingham Heart Study and the Sleep Heart Health Study. The Framingham Heart study and Dr. Seshadri are funded by contracts from the National Institutes of Health (N01-HC-25195,HHSN268201500001I, 75N92019D00031) and grants from the National Institute on Aging (AG059421, AG054076, AG049607, AG033090, NS017950). The Sleep Heart Health Study was funded by the National Heart, Lung and Blood Institute (UO1 HL 53941). Dr. Baril is funded by the Fonds de la Recherche du Quebec en Sante (#260192) and the Canadian Institutes of Health Research (#396130). Dr. Redline is partially supported by funds from the National Heart, Lung and Blood Institute (R35 135818). Dr. Pase is funded by a National Heart Foundation of Australia Future Leader Fellowship (ID 102052), the National Health and Medical Research Council (APP1158384), the Alzheimer’s Association (AARG-18-591358), and the Bethlehem Griffiths Research Foundation. Dr. Pase and Dr. Himali are also supported by a grant from the National Institute on Aging (AG062531).
CONFLICTS OF INTEREST
All sources of funding are government granting agencies or foundations, and thus, all authors report no conflicts of interest.
REFERENCES
- 1.Benbir G, Demir AU, Aksu M, et al. Prevalence of insomnia and its clinical correlates in a general population in Turkey. Psychiatry Clin Neurosci 2015;69:543–552. [DOI] [PubMed] [Google Scholar]
- 2.Bhaskar S, Hemavathy D, Prasad S. Prevalence of chronic insomnia in adult patients and its correlation with medical comorbidities. J Family Med Prim Care. 2016;5:780–784. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Brewster GS, Riegel B, Gehrman PR. Insomnia in the older adult. Sleep Med Clin. 2018;13:13–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Chen DW, Wang J, Zhang LL, Wang YJ, Gao CY. Cerebrospinal fluid amyloid-beta levels are increased in patients with insomnia. J Alzheimers Dis. 2018;61:645–651. [DOI] [PubMed] [Google Scholar]
- 5.Chen GH,Xia L,Wang F, Li XW, Jiao CA. Patients with chronic insomnia have selective impairments in memory that are modulated by cortisol. Psychophysiology. 2016;53:1567–1576. [DOI] [PubMed] [Google Scholar]
- 6.Irwin MR, Olmstead R, Carrillo C, et al. Cognitive behavioral therapy vs. Tai Chi for late life insomnia and inflammatory risk: a randomized controlled comparative efficacy trial. Sleep. 2014;37:1543–1552. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Syauqy A, Hsu CY, Rau HH, Kurniawan AL, Chao JC. Association of sleep duration and insomnia symptoms with components of metabolic syndrome and inflammation in middle-aged and older adults with metabolic syndrome in Taiwan. Nutrients. 2019:11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.de Almondes KM, Costa MV, Malloy-Diniz LF, Diniz BS. Insomnia and risk of dementia in older adults: systematic review and meta-analysis. J Psychiatr Res. 2016;77:109–115. [DOI] [PubMed] [Google Scholar]
- 9.Hung CM, Li YC, Chen HJ, et al. Risk of dementia in patients with primary insomnia: a nationwide population-based case-control study. BMC Psychiatry. 2018;18:38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Sindi S, Kareholt I, Johansson L, et al. Sleep disturbances and dementia risk: a multicenter study. Alzheimers Dement 2018;14:1235–1242. [DOI] [PubMed] [Google Scholar]
- 11.Lutsey PL, Misialek JR, Mosley TH, et al. Sleep characteristics and risk of dementia and Alzheimer’s disease: the Atherosclerosis Risk in Communities Study. Alzheimers Dement. 2018;14:157–166. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Wardle-Pinkston S, Slavish DC, Taylor DJ. Insomnia and cognitive performance: a systematic review and meta-analysis. Sleep Med Rev. 2019;48:101205. [DOI] [PubMed] [Google Scholar]
- 13.Kuo HK, Yen CJ, Chang CH, Kuo CK, Chen JH,Sorond F. Relation of C-reactive protein to stroke, cognitive disorders, and depression in the general population: systematic review and meta-analysis. Lancet Neurol. 2005;4:371–380. [DOI] [PubMed] [Google Scholar]
- 14.Poirier J Apolipoprotein E, cholesterol transport and synthesis in sporadic Alzheimer’s disease. Neurobiol Aging. 2005;26:355–361. [DOI] [PubMed] [Google Scholar]
- 15.Tao Q, Ang TFA, DeCarli C, et al. Association of chronic low-grade inflammation with risk of Alzheimer disease in ApoE4 carriers. JAMA Netw Open. 2018;1:e183597. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Feinleib M, Kannel WB, Garrison RJ, McNamara PM, WP C. The Framingham Offspring Study: design and preliminary data. Prev Med. 1975;4:518–525. [DOI] [PubMed] [Google Scholar]
- 17.Salthouse TA. When does age-related cognitive decline begin?. Neurobiol Aging. 2009;30:507–514. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.American Academy of Sleep Medicine. International Classification of Sleep Disorders. 3rd ed. ed. Westchester, IL: American Academy of Sleep Medicine; 2014. [Google Scholar]
- 19.American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 5th ed. Arlington, VA, American Psychiatric Publishing, 2013. [Google Scholar]
- 20.Bastien CH, Vallieres A, Morin CM. Validation of the Insomnia Severity Index as an outcome measure for insomnia research. Sleep Med. 2001;2:297–307. [DOI] [PubMed] [Google Scholar]
- 21.Natale V, Leger D, Martoni M, Bayon V, Erbacci A. The role of actigraphy in the assessment of primary insomnia: a retrospective study. Sleep Med. 2014;15:111–115. [DOI] [PubMed] [Google Scholar]
- 22.Natale V, Plazzi G, Martoni M. Actigraphy in the assessment of insomnia: a quantitative approach. Sleep. 2009;32:767–771. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Quan SF, Howard BV, Iber C, et al. The Sleep Heart Health Study: design, rationale, and methods. Sleep. 1997;20:1077–1085. [PubMed] [Google Scholar]
- 24.Redline S, Sanders MH, Lind BK, et al. Methods for obtaining and analyzing unattended polysomnography data for a multicenter study. Sleep Heart Health Research Group Sleep. 1998;21:759–767. [PubMed] [Google Scholar]
- 25.Whitney CW, Gottlieb DJ, Redline S, et al. Reliability of scoring respiratory disturbance indices and sleep staging. Sleep. 1998;21:749–757. [DOI] [PubMed] [Google Scholar]
- 26.Schutte-Rodin S, Broch L, Buysse D, Dorsey C,Sateia M.Clinical guideline for the evaluation and management of chronic insomnia in adults. J Clin Sleep Med. 2008;4:487–504. [PMC free article] [PubMed] [Google Scholar]
- 27.Au R, Seshadri S, Wolf PA, et al. New norms for a new generation: cognitive performance in the framingham offspring cohort. Exp Aging Res. 2004;30:333–358. [DOI] [PubMed] [Google Scholar]
- 28.Pase MP, Beiser A, Enserro D, et al. Association of ideal cardiovascular health with vascular brain injury and incident dementia. Stroke. 2016;47:1201–1206. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Pearson TA, Mensah GA, Alexander RW, et al. Markers of inflammation and cardiovascular disease: application to clinical and public health practice: a statement for healthcare professionals from the Centers for Disease Control and Prevention and the. American Heart Association Circulation. 2003;107:499–511. [DOI] [PubMed] [Google Scholar]
- 30.Dufouil C, Beiser A, McLure LA, et al. Revised Framingham Stroke Risk Profile to reflect temporal trends. Circulation. 2017;135:1145–1159. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Lewinsohn PM, Seeley JR, Roberts RE, Allen NB. Center for Epidemiologic Studies Depression Scale (CES-D) as a screening instrument for depression among community-residing older adults. Psychol Aging. 1997;12:277–287. [DOI] [PubMed] [Google Scholar]
- 32.Ohayon MM. Epidemiology of insomnia: what we know and what we still need to learn. Sleep Med Rev. 2002;6:97–111. [DOI] [PubMed] [Google Scholar]
- 33.Sexton CE, Sykara K, Karageorgiou E, et al. Connections between insomnia and cognitive aging. Neurosci Bull. 2020;36:77–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Haimov I, Shatil E. Cognitive training improves sleep quality and cognitive function among older adults with insomnia. PLoS One. 2013;8:e61390. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Joo EY, Kim H, Suh S, Hong SB. Hippocampal substructural vulnerability to sleep disturbance and cognitive impairment in patients with chronic primary insomnia: magnetic resonance imaging morphometry. Sleep. 2014;37:1189–1198. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Noh HJ, Joo EY, Kim ST, et al. The relationship between hippocampal volume and cognition in patients with chronic primary insomnia. J Clin Neurol. 2012;8:130–138. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Baril AA, Beiser AS, Mysliwiec V, et al. Slow-wave sleep and MRI markers of brain aging in a community-based sample. Neurology. 2021;96:e1462–e9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Varga AW, Wohlleber ME, Gimenez S, et al. Reduced slow-wave sleep is associated with high cerebrospinal fluid Abeta42 levels in cognitively normal elderly. Sleep. 2016;39:2041–2048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Baglioni C, Regen W, Teghen A, et al. Sleep changes in the disorder of insomnia: a meta-analysis of polysomnographic studies. Sleep Med Rev 2014;18:195–213. [DOI] [PubMed] [Google Scholar]
- 40.Fernandez-Mendoza J The insomnia with short sleep duration phenotype: an update on it’s importance for health and prevention. Curr Opin Psychiatry. 2017;30:56–63. [DOI] [PubMed] [Google Scholar]
- 41.Belloy ME, Napolioni V, Greicius MD. A quarter century of APOE and Alzheimer’s disease: progress to date and the path forward. Neuron. 2019;101:820–838. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Leduc V, Jasmin-Belanger S, Poirier J. APOE and cholesterol homeostasis in Alzheimer’s disease. Trends Mol Med. 2010;16:469–477. [DOI] [PubMed] [Google Scholar]
- 43.Belleville S, Fouquet C, Hudon C, Zomahoun HTV, Croteau J. Consortium for the early identification of Alzheimer’s d-Q. Neuropsychological measures that predict progression from mild cognitive impairment to Alzheimer’s type dementia in older adults: a systematic review and meta-analysis. Neuropsychol Rev. 2017;27:328–353. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Roh JH, Huang Y, Bero AW, et al. Disruption of the sleep-wake cycle and diurnal fluctuation of beta-amyloid in mice with Alzheimer’s disease pathology. Sci Transl Med. 2012;4:150ra22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Kahya M, Vidoni E, Burns JM, Thompson AN, Meyer K, Siengsukon CF. The relationship between apolipoprotein epsilon4 carrier status and sleep characteristics in cognitively normal older adults. J Geriatr Psychiatry Neurol. 2017;30:273–279. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Drogos LL, Gill SJ, Tyndall AV, et al. Evidence of association between sleep quality and APOE epsilon4 in healthy older adults: a pilot study. Neurology. 2016;87:1836–1842. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Burke SL, Maramaldi P, Cadet T, Kukull W. Associations between depression, sleep disturbance, and apolipoprotein E in the development of Alzheimer’s disease: dementia. Int Psychogeriatr. 2016;28:1409–1424. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Wang CC, Lung FW. The role of PGC-1 and Apoepsilon4 in insomnia. Psychiatr Genet. 2012;22:82–87. [DOI] [PubMed] [Google Scholar]
- 49.Basta M,Zaganas I,Simos P, et al. Apolipoprotein E varepsilon4 (APOE varepsilon4) allele is associated with long sleep duration among elderly with cognitive impairment. J Alzheimers Dis. 2021;79:763–771. [DOI] [PubMed] [Google Scholar]
- 50.Grau-Rivera O, Operto G, Falcon C, et al. Association between insomnia and cognitive performance, gray matter volume, and white matter microstructure in cognitively unimpaired adults. Alzheimers Res Ther. 2020;12:4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Cosentino FI, Bosco P, Drago V, et al. The APOE epsilon4 allele increases the risk of impaired spatial working memory in obstructive sleep apnea. Sleep Med. 2008;9:831–839. [DOI] [PubMed] [Google Scholar]
- 52.Nikodemova M, Finn L, Mignot E, Salzieder N, Peppard PE. Association of sleep disordered breathing and cognitive deficit in APOE epsilon4 carriers. Sleep. 2013;36:873–880. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Lim AS, Yu L, Kowgier M, Schneider JA, Buchman AS, Bennett DA. Modification of the relationship of the apolipoprotein E epsilon4 allele to the risk of Alzheimer disease and neurofibrillary tangle density by sleep. JAMA Neurol. 2013;70:1544–1551. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Gabin JM, Saltvedt I,Tambs K, Holmen J. The association of high sensitivity C-reactive protein and incident Alzheimer disease in patients 60 years and older:the HUNT study Norway. Immun Ageing. 2018;15:4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Irwin MR, Olmstead R, Carroll JE. Sleep disturbance, sleep duration, and inflammation: a systematic review and meta-analysis of cohort studies and experimental sleep deprivation. Biol Psychiatry. 2016;80:40–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.Parthasarathy S, Vasquez MM, Halonen M, et al. Persistent insomnia is associated with mortality risk. Am J Med. 2015;128:268–275. [DOI] [PMC free article] [PubMed] [Google Scholar]