Insomnia Symptoms and Biomarkers of Alzheimer’s Disease in the Community

Abstract

Background:

Insomnia is one of the most common sleep disorders yet its relationship to the biology of Alzheimer’s disease remains equivocal.

Objective:

We investigated the cross-sectional relationship between insomnia symptom severity and cerebrospinal fluid (CSF) concentrations of Alzheimer’s disease biomarkers in a cognitively unimpaired middle-aged community sample.

Methods:

A total of 63 participants from the Healthy Brain Project (age = 59±7 years; 67% women) completed a lumbar puncture and two weeks of actigraphy to measure two of insomnia’s core features: difficulty initiating sleep (prolonged sleep onset latency) and difficulty maintaining sleep (wake after sleep onset [WASO] and number of awakenings). Additionally, the Insomnia Severity Index (ISI) was completed by 58 participants. Linear and Tobit regression were used to estimate the associations between each insomnia variable and CSF Aβ₄₂, phosphorylated tau 181 (p-tau181), total-tau, and neurofilament light chain protein (NfL), adjusting for age, sex, and APOE ɛ4 genotype.

Results:

Higher ISI score was associated with greater average levels of CSF Aβ₄₂ (per point: 30.7 pg/mL, 95% CI: 4.17–57.3, p = 0.023), as was higher WASO (per 10 min: 136 pg/mL, 95% CI: 48–223, p = 0.002) and more awakenings (per 5:123 pg/mL, 95% CI = 55–192, p < 0.001). Difficulty initiating sleep was not associated with CSF Aβ₄₂, nor were insomnia features associated with p-tau181, total-tau, or NfL levels.

Conclusion:

Insomnia symptoms were associated with higher CSF Aβ₄₂ levels in this relatively young, cognitively unimpaired sample. These findings may reflect increased amyloid production due to acute sleep disruption.

Keywords

Alzheimer’s disease amyloid dementia insomnia sleep

INTRODUCTION

Sleep plays a fundamental role in maintaining brain health, including synaptic consolidation and memory formation [1]. Of relevance to Alzheimer’s disease (AD) and dementia, slow wave sleep facilitates the removal of neurotoxic waste, including amyloid-β [2]. However, relatively little is known about how sleep disorders relate to AD biomarkers.

Insomnia is one of the most common sleep disorders, experienced in various forms by 1 in 3 people [3]. Insomnia involves a frequent and chronic complaint of dissatisfaction with sleep quantity or quality, associated with difficulty initiating sleep, difficulty maintaining sleep, or early morning awakenings [4]. Several studies have found associations between insomnia and an increased risk of dementia or poorer cognitive function [5 –9]. Insomnia may be associated with AD or neurodegenerative processes, but it is also possible that cognitive impairments in insomnia could stem from decreased alertness and increased fatigue, or shared risk factors, rather than underlying neurodegenerative processes. Better understanding these mechanisms could clarify the role of sleep dysfunction in dementia risk.

Poor sleep and neurodegeneration likely share a bidirectional relationship [10]. Thus, to limit reverse-causation, studies are needed to determine if insomnia symptoms are associated with AD biomarkers in cognitively healthy middle-aged adults who are unlikely to have cognitive impairment or substantial AD related neurodegeneration that could impact sleep. Accordingly, this cross-sectional study examined self-reported insomnia symptom severity and related objective markers of sleep quality (difficulty falling asleep and maintaining sleep on actigraphy) in association with cerebrospinal fluid (CSF) biomarkers that measure different components of AD biology according to the amyloid, tau, and neurodegeneration (ATN) framework [11]. For clinical context, we also examined the relationships between insomnia symptoms and cognitive function. We also explored for interactions by APOE ɛ4 carriage based on evidence that APOE may interact with sleep and cognitive decline in a similarly aged sample [6]. We also explored possible interactions by total sleep time given that short sleep may increase amyloid production (through increased wakefulness) and limit the opportunity for glymphatic clearance. We examined these relationships in a community-based cohort of cognitively healthy middle-aged adults. We hypothesized that greater insomnia symptom severity would be associated with lower CSF Aβ₄₂ and higher phosphorated tau 181 (p-tau181), total-tau (t-tau), and neurofilament light chain (NfL) levels.

METHODS

Participants

Participants were recruited from the Healthy Brain Project (HBP) [12]. The HBP (healthybrainproject.org.au) is a longitudinal community-based online cohort of middle-aged adults (approximately 8,500 enrolments) aged between 40 and 70 years at baseline. The study was designed to investigate the biological, environmental, and psychological factors that affect cognitive aging and dementia risk. Participants with self-reported clinical cognitive impairment and significant neurological disease were excluded.

From this cohort, a subset (n = 82) of participants were recruited for an in-person biomarker sub-study which took place at the Royal Melbourne Hospital in Melbourne, Australia. The subgroup was recruited based on ability to travel to the hospital for the assessment and without conditions that would prevent completing an MRI brain scan and lumbar puncture procedure. The overall study sample was enriched with APOE ɛ4 carriers (38%) to study a group of individuals who would be at higher risk of late-onset AD. In total, 82 participants completed the biomarker sub-study and cognitive assessments; 77 completed a lumbar puncture, and of these, 70 participants completed the sleep assessments. Four participants were excluded for having less than 10 days of actigraphy and three participants did not have APOE ɛ4 genotype data available. Finally, the Insomnia Severity Index (ISI) was available in 58 participants. An analysis sample of 63 was available for actigraphy derived exposures and 58 for the ISI (see Fig. 1).

Fig. 1

Study Flow Diagram. HBP, Healthy Brain Project; ISI, Insomnia Severity Scale. Two participants from the final sample had missing APOE ɛ4 data due to sample processing errors.

The Melbourne Health Human Research Ethics Committee approved the study and all participants provided written informed consent prior to commencing any study assessments (MHREC 2017.302). Data collection occurred between November 2018 and February 2020.

Measures

CSF biomarkers

We measured CSF biomarkers that mapped onto the NIA-AA Amyloid (Aβ₄₂), Tau (p-tau 181), Neurodegeneration (t-tau, NfL) (ATN) Framework [11]. CSF samples were obtained by lumbar puncture in the L3/L4 or L4/L5 interspace, with most samples collected between 13:00 and 14:30 h. CSF samples were transferred for processing on wet ice following well-established guidelines. Samples were spun at 2000× g at +4°C for 10 min. Supernatant was pipetted off to a new polypropylene tube and gently inverted a few times to avoid possible gradient effects. Samples were then aliquoted in 0.5 mL portions into screw-cap polypropylene tubes and stored at –80°C pending biochemical analyses. CSF concentrations of Aβ₄₂, t-tau, and p-tau181 were measured by immunoassay (Roche Elecsys®) and CSF concentrations of NfL were measured using ELISA (UmanDiagnostics, Umeå, Sweden). All analyses were conducted at the National Dementia Diagnostics Laboratory (The Florey Institute, University of Melbourne, Australia). A total of 25 (30%) participants had CSF Aβ₄₂ scores above the maximum limit of detection and were assigned the top range score (1700 pg/mL).

Sleep measures

Insomnia Severity Index

Self-reported insomnia symptom severity over the last two weeks was measured with the 7-item ISI [13]. Each item on the ISI is scored on a 5-point scale (responses range from 0–4). The ISI was scored on a continuous scale with higher scores indicating more severe insomnia symptoms (total score range = 0–28).

Activity monitoring

We measured core features of insomnia from wrist actigraphy (Philips Respironics Actiwatch Spectrum Plus), including difficulty initiating sleep (measured as sleep onset latency) and difficulty maintaining sleep (measured using both wake after sleep onset [WASO] and number of awakenings). While polysomnography is the gold standard for sleep assessment, actigraphy demonstrates high precision in assessing sleep parameters compared with polysomnography and provides a more naturalistic measure of habitual sleeping patterns [14]. Participants were required to wear wrist actigraphy for 17 consecutive days, which measured 24-h activity levels in 30-s epochs. The analysis of actigraphy data was performed according to a standardized protocol by the same trained independent researcher using specialized software (Philips Respironics Actiware 6.0.9). The first two nights of data were excluded due to the possibility of sleep disruptions because of the lumbar puncture, which occurred the day before the first day of actigraphy. Participants also completed a sleep diary, which was used to corroborate bed and rise times generated from actigraphy. Sleep and wake times used in this analysis were derived from the actigraphy. However, the sleep diary time was reported when there was a discrepancy of more than one hour between the time on the diary and the time on the Actiwatch. The Actiware 6.0.9 autonomic wake threshold was used to calculate WASO and awakenings. The wake threshold was calculated by dividing the sum of activity (calculated using an accelerometer) by mobile time (when the number of activity counts recorded in that epoch is greater than or equal to the epoch length in 15-s intervals), multiplied by 0.88888. We averaged all actigraphy measures over at least 10 days to ensure stability in sleeping patterns (i.e., participants with less than 10 days of recording were excluded).

Cognitive measures

Logical Memory I and II

Logical Memory I and II are subtests from the Wechsler Memory Scale IV and were used to access episodic memory. Logical Memory I required participants to listen to a short story read aloud, then repeat as much of the story back as they remembered. After a delay of approximately 20 minutes, participants were again asked to recall the story (Logical Memory II). Higher scores indicated better episodic memory (total score range = 0–25).

Trail Making Test (TMT) Parts A and B

TMT B minus A was used to assess attention and processing speed. TMT A required participants to connect 25 circles (numbered from 1 to 25) in ascending order as fast as possible. TMT B required participants to connect a different set of 25 circles, containing letters (A-L) and numbers (1–13) in an alternating sequence in ascending order as fast as possible. The outcome for both was completion time. The difference between the two scores (TMT B - TMT A) was calculated as a measure of executive function, with lower scores indicating better executive function.

Controlled Oral Word Association Test (COWAT)

The COWAT was used to assess phonemic fluency. Participants were given one minute to produce as many words as possible starting with certain letter of the alphabet (F, A, and S). Higher scores indicated better verbal fluency.

Demographics

Basic demographic information was collected, including age and sex. Participants were also asked if they used any sleep medications or have any medical diagnoses. APOE genotype was determined through TaqMan genotyping assays of saliva samples (Life Technologies).

Statistical methods

Primary analyses

As 30% of the sample had CSF Aβ₄₂ above the limit of detection, Tobit regression models suitable for outcome variables subject to censoring, were used to examine the relationships between sleep metrics (ISI score, WASO, awakenings, and sleep onset latency) and CSF Aβ₄₂. Regression coefficients for these models are interpreted on the latent (uncensored) CSF Aβ₄₂ scale rather than the observed (censored) Aβ₄₂ scale. For the other AD biomarkers, linear regression was used to examine their relationships with the sleep metrics. All exposure and outcome variables were included as continuous variables. Models were adjusted for age, sex, and APOE ɛ4 genotype as these variables are known determinants of sleep and AD biomarkers.

Secondary analysis

To provide comparability to previous studies and for completeness, we performed three secondary analysis. First, we included sleep efficiency as an additional predictor variable in our primary models. Second, we included the CSF tau/Aβ₄₂ ratio as an additional outcome in our primary models. Third, we used linear regressions to examine the relationships between the sleep variables and cognitive scores. All models were adjusted for age, sex, and APOE ɛ4 genotype.

Sensitivity analysis

As a sensitivity analysis, all primary models were further adjusted for actigraphy measured total sleep time (TST). As prior evidence indicates TST has a non-linear relationship with dementia outcomes [15], TST was modelled using a restricted cubic spline with knots at the 5th, 35th, 65th, and 95th percentiles [16]. Finally, an ANCOVA model was performed for the ISI in which the ISI score was split into two groups (normal versus high), where a score greater than 7 indicated high insomnia symptomatology, as utilized previously [13].

Interactions

To examine whether the associations between sleep and AD biomarkers were modified by APOE ɛ4 carriage (ɛ4 carrier versus non-ɛ4 carrier) or TST, interaction terms were added to the primary models.

There were 63 participants (77%) that provided complete data for models including actigraphy exposure variables (WASO, sleep onset latency, awakenings). There were 58 participants (71%) providing complete data for models including the ISI as exposure variable. Conditional on age and sex, missing outcome data was assumed to be missing completely at random. In all models, participants with missing data in predictor or outcome variables were excluded (i.e., we used a complete case analysis).

No p-value adjustments for multiple comparisons were made as our analysis was focused on the estimation of a relatively small set of closely related parameters. As such, it was possible to consider each result within the context of the larger pattern of study results [17]. All analyses were conducted using SPSS Statistical Software (version 26) and R version 4.1.0 [18].

RESULTS

Sample overview

Table 1 displays the demographic characteristics of the sample. On average, participants were 59 (SD, 6) years old and 69% were female. Of the sample, 23 (40%) participants reported insomnia symptom severity above the threshold (ISI score >7). These participants had longer average WASO, more awakenings, and shorter total sleep time. Surprisingly, they also exhibited shorter sleep onset latency and higher sleep efficiency. Additionally, two participants used medications (melatonin supplements) to improve their sleep and two participants reported having sleep apnea. Of the sample, 11 (19%) participants had CSF Aβ₄₂ levels <1000 pg/mL that were indicative of abnormal levels of amyloid [19].

Table 1

Sample demographics stratified by the degree of insomnia severity

		Insomnia Severity Scale score^*
Variable	Overall (N = 58)	Low (N = 35)	High (N = 23)
Age, years	59 (6)	59 (6)	58 (7)
Education, years	17 (4)	17 (4)	16 (4)
Females, N (%)	40 (69)	27 (77)	13 (57)
APOEɛ4 carrier, N (%)	25 (43)	16 (46)	9 (39)
Aβ₄₂, pg/mL	1381 (340)	1310 (355)	1490 (291)
Total-tau, pg/mL	194 (63.3)	197 (69.9)	188 (52.6)
P-tau181, pg/mL	16.2 (5.8)	16.5 (6.6)	15.6 (4.9)
NfL, pg/mL	594 (264)	608 (318)	573 (155)
Sleep onset latency, min¹	18.8 (14.9)	20.7 (17.5)	15.8 (8.88)
WASO, min¹	39 (14)	37 (15)	41 (13)
Awakenings, number¹	33 (9)	30 (8)	35 (10)
Total sleep time, hrs¹	7.07 (0.79)	7.14 (0.71)	6.97 (0.92)
Sleep efficiency, (%)¹	84.5 (5.3)	83.4 (4.7)	85.4 (5.5)
Logical memory I (immediate)	13.0 (3.5)	13.4 (3.4)	12.3 (3.6)
Logical memory II (delayed)	11.3 (4.0)	11.6 (4.0)	10.8 (4.1)
TMT B - TMT A, sec	34.3 (15.5)	35.0 (14.3)	33.2 (17.5)
COWAT total	45.1 (19.5)	45.8 (9.9)	44.0 (11.5)

Only participants with ISI, APOE genotype, and CSF biomarker data available (n = 58) are included in the table. All values are presented as mean (SD) unless stated otherwise. ^*Low ISI = normal score on the ISI, defined as any score of less than 8; High ISI = high score on the ISI, defined as any score of 8 or more. ¹Actigraphy variables additionally available for n = 50 (31 with normal ISI, 19 with high ISI). Cognitive measures were available for all 58 participants. ISI, Insomnia Severity Scale; Aβ₄₂, amyloid-β 42; P-tau181, tau phosphorylated at threonine 181; NfL, neurofilament light chain; WASO, wake after sleep onset; TMT, Trail Making Test; COWAT, Controlled Oral Word Association Test.

Association between insomnia symptoms and AD biomarkers

Higher ISI score was associated with greater average levels of latent (uncensored) CSF Aβ₄₂ (per point: 30.7 pg/mL, 95% CI: 4.17, 57.3, p = 0.023), adjusted for age, sex, and APOE ɛ4 (Fig. 2). Similarly, participants with higher average WASO (per 10 min: 136 pg/mL, 95% CI: 48.4, 224, p = 0.002) and more average awakenings (per 5:123 pg/mL, 95% CI = 54.5, 192, p < 0.001) had greater mean levels of CSF Aβ₄₂.

Fig. 2

A) Association between total ISI score (continuous variable) and Aβ₄₂. B) Association between ISI score (high versus low) and Aβ₄₂. Models are adjusted for age, sex, and APOE ɛ4 genotype. The association between ISI and Aβ₄₂ was estimated using Tobit regression. Consequently, predictions are on the scale of the latent (uncensored) outcome variable and therefore may exceed the limit of detection for the Aβ₄₂ assay (1,700 pg/mL). Low ISI = normal score on the ISI, defined as any score of less than 8; High ISI = high score on the ISI, defined as any score of 8 or more. ISI, Insomnia Severity Scale; Aβ₄₂, amyloid-β 42.

There was no clear evidence that sleep onset latency was associated with CSF Aβ₄₂, nor that any of the sleep characteristics were associated with average levels of CSF total-tau, p-tau181, or NfL (Table 2).

Table 2

Associations between sleep variables and CSF biomarkers

		Aβ₄₂, pg/mL		T-tau, pg/mL		P-tau181, pg/mL		NfL, pg/mL
Variable	N	β (95% CI)	p	β (95% CI)	p	β (95% CI)	p	β (95% CI)	p
ISI, per point	58	30.7 (4.17, 57.3)	0.023	–0.91 (–4.47, 2.65)	0.61	–0.11 (–0.45, 0.23)	0.52	–0.89 (–14.4, 12.6)	0.90
Sleep latency, per 10 min	63	–54.5 (–150, 40.7)	0.26	–7.53 (–19.9, 4.80)	0.23	–0.78 (–1.96, 0.40)	0.19	–7.27 (–44.8, 59.4)	0.78
WASO, per 10 min	63	136 (48.4, 224)	0.002	–2.56 (–14.1, 9.03)	0.66	–0.55 (–1.66, 0.55)	0.32	–1.22 (–47.2, 49.7)	0.96
Awakenings, per 5	63	123 (54.5, 192)	<0.001	–0.92 (–9.93, 8.09)	0.84	–0.33 (–1.19, 0.53)	0.44	–5.96 (–43.5, 31.6)	0.75
Sleep efficiency, per 5%	63	–15.3 (–137, 106)	0.81	8.06 (–7.31, 23.4)	0.30	0.94 (–5.24, 2.41)	0.20	–9.66 (–74.4, 55.0)	0.77

Models adjusted for age, sex, and APOE ɛ4 genotype. Betas are unstandardized. Note: For these analyses, all sleep scores and biomarker variables are included as continuous variables. ISI, Insomnia Severity Scale; Aβ₄₂, amyloid-β 42; T-tau, total tau; P-tau181, tau phosphorylated at threonine 181; NfL, neurofilament light chain; WASO, wake after sleep onset (minutes). Higher scores on the ISI indicate higher insomnia symptom severity. Higher onset latency, WASO, and awakenings indicate poorer sleep quality. Lower sleep efficiency indicates poorer sleep quality.

When investigating differences in Aβ₄₂ between the groups with above and below threshold insomnia symptom severity, we observed that participants with a high ISI score (ISI >7) had greater average levels of CSF Aβ₄₂ (mean difference = 335 pg/mL, 95% CI = 89.0, 580, p = 0.008) when comparing to the remainder of the sample (Fig. 2). Levels of t-tau (mean difference = –3.15 pg/mL, 95% CI = –37.1, 30.8, p = 0.85), p-tau181 (mean difference = –0.47, 95% CI = –3.69, 2.75, p = 0.77), and NfL (mean difference = –11.0, 95% CI = –140, 118, p = 0.87) were not significantly different between insomnia groups.

Secondary analyses

There was no statistically significant associations between sleep efficiency and any of the AD biomarkers (Table 2).

Similar results were obtained we included the CSF tau/Aβ₄₂ ratio as an additional outcome. Participants with higher average WASO (per 10 min: –0.019, 95% CI = –0.033, –0.004, p = 0.012) and more average awakenings (per 5: –0.014, 95% CI-0.026, –0.003, p = 0.012) had lower mean t-tau/Aβ₄₂, adjusted for age, sex, and APOE ɛ4. However, ISI score was no longer significantly associated with latent (uncensored) CSF Aβ₄₂ (per point: –0.003, 95% CI: –0.007, –0,00, p = 0.078). Consistent with our findings from CSF Aβ₄₂ levels, we found no clear evidence that sleep onset latency was associated with tau/Aβ₄₂ (see Supplementary Table 1).

There was no statistically significant associations between Logical Memory I (immediate recall), Logical Memory II (delayed recall), COWAT total score, or TMT B – TMT A, and sleep variables (see Supplementary Table 2).

Sensitivity analyses

Including an additional adjustment for total sleep time did not meaningfully change any of the reported results (see Supplementary Table 3).

Interactions with APOE ɛ4 and TST

The relationships between average WASO and average awakenings and CSF Aβ₄₂ were modified by APOE ɛ4 carriage (p interaction = 0.034 and p interaction = 0.047, respectively). The relationship between average WASO and average awakenings and CSF Aβ₄₂ levels was stronger in ɛ4 allele carriers (Fig. 3).

The relationships between ISI score and average awakenings and CSF Aβ₄₂ were modified by TST (p interaction = 0.036 and p interaction = 0.030, respectively. The relationships between ISI score and average awakenings and CSF Aβ₄₂ levels was stronger in those with short sleep duration (Fig. 4). No other effect modification was found.

Fig. 3

Effect modification by APOE ɛ4 status. A) Estimates of interaction terms (APOE ɛ4 carriage x exposure variable). Interactions represent the estimated difference in the relationship between exposure and biomarker variables between APOE ɛ4 allele carriers and non-carriers. Positive interactions indicate that the relationship between the exposure and biomarker is larger (more positive) among ɛ4 allele carriers. B) Association between average WASO and average awakenings with latent (uncensored) Aβ₄₂, stratified by APOE ɛ4 carriage. Models are adjusted for age and sex. ISI, Insomnia Severity Scale; Aβ₄₂, amyloid-β 42; T-tau, total tau; P-tau181, tau phosphorylated at threonine 181; NfL, neurofilament light chain; WASO, wake after sleep onset (minutes). Higher scores on the ISI indicate higher insomnia symptom severity. Higher onset latency, WASO, and awakenings indicate poorer sleep quality.

Fig. 4

Effect modification by total sleep time (TST). Association between ISI, average awakenings with latent (uncensored) Aβ₄₂, stratified by TST. Models are adjusted for age, sex, and APOE ɛ4 carriage. ISI, Insomnia Severity Scale; Aβ₄₂, amyloid-β 42; WASO, wake after sleep onset (minutes). Higher scores on the ISI indicate higher insomnia symptom severity. Higher onset latency, WASO, and awakenings indicate poorer sleep quality, and lower sleep efficiency indicates poorer sleep quality.

DISCUSSION

In a middle-aged sample, we observed that insomnia symptoms were associated with higher levels of CSF Aβ₄₂. Difficulty maintaining sleep, one of insomnia’s core features, was also associated with higher Aβ₄₂ whereas there was no such evidence for difficulty initiating sleep. Additionally, the strength of the relationship between WASO and number of awakenings and Aβ₄₂ was nominally stronger in APOE ɛ4 carriers compared to non-carriers. Similarly, the relationships between ISI score and average awakenings and CSF Aβ₄₂ levels was stronger in those with short sleep duration. There was no evidence that poor sleep quality was associated with CSF levels of t-tau, p-tau181, or NfL. Overall, these data show that in a group of cognitively unimpaired middle-aged adults enriched for dementia risk, insomnia symptoms are associated with higher levels of CSF Aβ₄₂, though no clear evidence implicated CSF markers of tau or neurodegeneration.

Contrary to our hypothesis, we found a positive association between insomnia symptoms and CSF Aβ₄₂. As low CSF Aβ₄₂ is indicative of greater levels of insoluble cerebral amyloid burden in individuals with AD [20], our results were unexpected. However, neuronal activity increases amyloid production, which is then cleared during sleep [2, 21]. Thus, greater time spent awake and greater sleep disruption may increase Aβ₄₂ production or reduce Aβ₄₂ clearance in persons at risk of dementia. This may also explain why the association between sleep disruption and CSF Aβ₄₂ observed in our study was more pronounced in short sleepers, who may have been unable to compensate for increased awakenings by sleeping longer. Indeed, several studies have shown that acute sleep disruption increases levels of CSF Aβ₄₂ [22, 23]. For our study, however, we examined sleep patterns over two weeks. Therefore, our results might suggest that more sustained sleep disturbance can also increase Aβ₄₂ levels in persons at risk of dementia. In support of this suggestion, one study of middle-aged adults reported that CSF levels of Aβ₄₂ were higher in 23 patients with chronic insomnia as compared to 23 controls [24]. However, this study did not examine other relevant AD biomarkers, or utilize objective sleep measurements. With respect to other sleep disorders, the treatment of obstructive sleep apnea has been shown to result in increased slow-wave sleep which was, in turn, significantly correlated with lower CSF Aβ₄₂ after treatment [25]. Together, these findings suggest that sleep disorders like insomnia and obstructive sleep apnea may result in an increase in CSF Aβ₄₂ levels, particularly when accompanied by short sleep time, perhaps through the effects of sleep disruption and wakefulness on increased amyloid production.

Our results should be interpreted in the context of our middle-aged, cognitively healthy sample. Although we link insomnia symptoms to higher CSF Aβ₄₂ in midlife and others have linked insomnia to a higher risk of dementia [5 , 8], it is unclear if changes in Aβ₄₂ account for the links between poor sleep and dementia. In our study, there was no evidence that sleep was associated with biomarkers of tau or neurodegeneration, which are associated with cognitive decline [26]. Therefore, the significance of the association between insomnia symptoms and Aβ₄₂ for eventual risk of dementia is unclear. If insomnia is a risk factor for AD dementia, we may expect to see an association between insomnia symptoms and tau and NfL, although these associations may only emerge later in life closer to dementia onset. On the other hand, it has been hypothesized that high CSF Aβ₄₂ in cognitively healthy individuals may represent Aβ₄₂ overproduction [27] and that CSF Aβ₄₂ may then begin to fall only once Aβ accumulates in the form of plaques. Thus, one hypothesis arising from these data is that individuals with fragmented sleep and higher amyloid load may be more susceptible to plaque formation, as a mismatch between production and clearance emerges. This can be challenged by future studies that measure longitudinal CSF, imaging, and clinical outcomes to determine the temporal relationship between sleep disturbance and the progression of AD over time. Experimental studies could also examine if acute sleep disruption has the same effects on fluid biomarker levels in persons with and without preclinical AD (e.g., a positive amyloid PET scan).

We did not observe any relationship between insomnia symptoms and cognitive function. However, this finding is consistent with a previous paper from the larger parent cohort of the sample used here [28]. In contrast, Baril et al. showed that in 511 dementia-free Framingham Heart Study (FHS) participants (mean age ±= 62.65±8.7 years; 49% female), insomnia symptom severity was associated with worse subsequent memory performance an average of 3.4 years later. The different findings may be driven by the older FHS cohort, which also had lower logical memory scores and longer TMT scores than the current cohort. Nevertheless, our null cognitive results dovetail with our findings for tau and NfL, given that NfL and tau track more closely with cognitive function than Aβ₄₂ [29].

We observed an interaction with APOE ɛ4 carriage, whereby the positive relationship between sleep fragmentation, as measured by WASO and number of awakenings, and CSF Aβ₄₂ was more pronounced in ɛ4 carriers compared to non-carriers. APOE ɛ4 is the strongest known genetic risk factor for late-onset AD and is involved in the clearance and aggregation of amyloid-β [30]. Thus, sleep disruption and APOE ɛ4 carriage may both impact the glymphatic clearance of Aβ₄₂ [31]. Nevertheless, these interaction results should be interpreted cautiously given the small groups presented in the stratified results and the many interactions assessed (most of which were compatible with no effect modification). External replication would be valuable. Strengths of our study include both the objective and subjective assessment of sleep, and the examination of their relationship with established CSF markers of amyloid, tau, and neurodegeneration. Moreover, since we explored sleep-AD biomarker relationships in a cognitively healthy middle-aged sample, we limited the potential impact of more advanced neurodegeneration on sleep (and thereby limited reverse-causation). Moreover, the use of a midlife sample without cognitive impairment but at high risk of dementia is a strength since understanding sleep and AD relationships in this group could inform future lifestyle-based approaches to prevent dementia. However, our study is not without limitations. Most notably, this was a cross-sectional study, which precluded investigation into whether insomnia symptoms were associated with the progression of biomarker changes preceding dementia. Furthermore, our small and homogenous sample precluded sub-group analysis (e.g., by sex or race). Given the small sample size, our results should be considered as hypothesis generating with further replication of results required.

In conclusion, this study found that in a middle-aged sample, symptoms and features of insomnia were associated with higher CSF Aβ₄₂ levels but no evidence for an association with any CSF marker of tau or neurodegeneration. These findings may reflect increased amyloid production due to sustained sleep disruption and increased wakefulness, an effect that may be magnified in APOE ɛ4 carriers, who have impaired amyloid clearance, and those who have less total sleep. Overall, these data highlight the complex relationships between sleep and fluid AD biomarkers. Prospective studies are needed to ascertain if insomnia is associated with AD progression over time. Such research could inform whether the treatment of insomnia could help protect against the development of dementia.

Footnotes

ACKNOWLEDGMENTS

We thank all HBP participants for their commitment and dedication to helping advance research into the early detection and causation of dementia.

FUNDING

Dr. Pase is supported by a National Heart Foundation of Australia Future Leader Fellowship (GTN102052) with sleep and dementia research funding from the National Health and Medical Research Council of Australia (GTN2009264; GTN1158384), National Institute on Aging (R01 AG062531-01A1), and Alzheimer’s Association (2018-AARG-591358). Dr Cavuoto and Dr Pase are supported by a Dementia Australia Research Foundation award (Lucas’ Papaw Remedies Project Grant). The Healthy Brain Project (healthybrainproject.org.au) is funded by the National Health and Medical Research Council (NHMRC; GNT1158384, GNT1147465, GNT1111603, GNT1105576, GNT1104273, GNT1171816), the Alzheimer’s Association (AARG-17-591424, AARG-18-591358, AARG-19-643133), the Dementia Australia Research Foundation, the Yulgilbar Alzheimer’s Research Program, and the Charleston Conference for Alzheimer’s Disease. Dr Lim is supported by an NHMRC Career Development Fellowship (GNT1162645), and an NHMRC Emerging Leadership Grant (GNT2009550). Dr Buckley is supported by a National Institutes of Health K99-R00 award (K99AG061238) and an Alzheimer’s Association Research Fellowship. Ms. Bransby is supported by a Dementia Australia Research Foundation PhD scholarship.

CONFLICT OF INTEREST

The authors have no conflict of interest to report.

DATA AVAILABILITY

The data supporting the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

The supplementary material is available in the electronic version of this article: .

References

Rasch

, Born

(2013) About sleep’s role in memory. Physiol Rev 93,681–766.

Fultz

, Bonmassar

, Setsompop

, Stickgold

, Rosen

, Polimeni

, Lewis

(2019) Coupled electrophysiological, hemodynamic, and cerebrospinal fluid oscillations in human sleep. Science 366,628–631.

Meadows A, Rehman A, Insomnia, https://www.sleepfoundation.org/insomnia.

(2013) Diagnostic and statistical manual of mental disorders: DSM-5, American Psychiatric Association, Arlington, VA.

de Almondes

, Costa

, Malloy-Diniz

, Diniz

(2016) Insomnia and risk of dementia in older adults: Systematic review and meta-analysis. J Psychiatr Res 77,109–115.

Baril

, Beiser

, Sanchez

, Mysliwiec

, Redline

, Gottlieb

, O’Connor

, Gonzales

, Himali

, Seshadri

, Himali

, Pase

(2021) Insomnia symptom severity and cognitive performance: Moderating role of APOE genotype. Alzheimers Dement 18,408–421.

Hung

, Li

, Chen

, Lu

, Liang

, Liliang

, Tsai

, Wang

(2018) Risk of dementia in patients with primary insomnia: A nationwide population-based case-control study. BMC Psychiatry 18,38.

Lutsey

, Misialek

, Mosley

, Gottesman

, Punjabi

, Shahar

, MacLehose

, Ogilvie

, Knopman

, Alonso

(2018) Sleep characteristics and risk of dementia and Alzheimer’s disease: The atherosclerosis risk in communities study. Alzheimers Dement 14,157–166.

Sindi

, Kåreholt

, Johansson

, Skoog

, Sjöberg

, Wang

, Johansson

, Fratiglioni

, Soininen

, Solomon

, Skoog

, Kivipelto

(2018) Sleep disturbances and dementia risk: A multicenter study. Alzheimers Dement 14,1235–1242.

10.

, Lucey

, Holtzman

(2014) Sleep and Alzheimer disease pathology–a bidirectional relationship. Nat Rev Neurol 10,115–119.

11.

Jack

Jr , . , Bennett

, Blennow

, Carrillo

, Dunn

, Haeberlein

, Holtzman

, Jagust

, Jessen

, Karlawish

, Liu

, Molinuevo

, Montine

, Phelps

, Rankin

, Rowe

, Scheltens

, Siemers

, Snyder

, Sperling

(2018) NIA-AA Research Framework: Toward a biological definition of Alzheimer’s disease. Alzheimers Dement 14,535–562.

12.

Lim

, Yassi

, Bransby

, Properzi

, Buckley

(2019) The Healthy Brain Project: An online platform for the recruitment, assessment, and monitoring of middle-aged adults at risk of developing Alzheimer’s disease. J Alzheimers Dis 68,1211–1228.

13.

Bastien

, Vallières

, Morin

(2001) Validation of the insomnia severity index as an outcome measure for insomnia research. Sleep Med 2,297–307.

14.

Marino

, Li

, Rueschman

, Winkelman

, Ellenbogen

, Solet

, Dulin

, Berkman

, Buxton

(2013) Measuring sleep: Accuracy, sensitivity, and specificity of wrist actigraphy compared to polysomnography. Sleep 36,1747–1755.

15.

, Liang

, Zheng

, Shi

, Zhong

, Xie

(2020) Association between sleep duration and cognitive decline. JAMA Network Open 3,e2013573–e2013573.

16.

Harrell

(2001) Regression modeling strategies: With applications to linear models, logistic regression, and survival analysis , Springer, New York.

17.

Assel

, Sjoberg

, Elders

, Wang

, Huo

, Botchway

, Delfino

, Fan

, Zhao

, Koyama

, Hollenbeck

, Qin

, Zahnd

, Zabor

, Kattan

, Vickers

(2019) Guidelines for reporting of statistics for clinical research in urology. BJU Int 123,401–410.

18.

Team

(2013) R: A language and environment for statistical computing. R Foundation for Statistical Computing 3,201.

19.

Roche Diagnostics Australia Pty Limited (2020) Elecsys® β-Amyloid (1-42) CSF: ElectroChemiLuminescence immunoassay (ECLIA) for the in vitro quantitative determination of ß-Amyloid (1-42) in human cerebrospinal fluid (CSF) [Fact Sheet]. Available from:. https://florey.edu.au/uploads/content/MC-AU-00655_Elecsys_beta_amyloid_(1-42)_CSF_fact_sheet_1.pdf

20.

Fagan

, Mintun

, Mach

, Lee

, Dence

, Shah

, LaRossa

, Spinner

, Klunk

, Mathis

, DeKosky

, Morris

, Holtzman

(2006) Inverse relation between in vivo amyloid imaging load and cerebrospinal fluid Abeta42 in humans. Ann Neurol 59,512–519.

21.

Xie

, Kang

, Xu

, Chen

, Liao

, Thiyagarajan

, O’Donnell

, Christensen

, Nicholson

, Iliff

, Takano

, Deane

, Nedergaard

(2013) Sleep drives metabolite clearance from the adult brain. Science 342,373–377.

22.

Y-ES

, Ooms

, Sutphen

, Macauley

, Zangrilli

, Jerome

, Fagan

, Mignot

, Zempel

, Claassen

JAHR

, Holtzman

(2017) Slow wave sleep disruption increases cerebrospinal fluid amyloid-β levels. Brain 140,2104–2111.

23.

Lucey

, Hicks

, McLeland

, Toedebusch

, Boyd

, Elbert

, Patterson

, Baty

, Morris

, Ovod

, Mawuenyega

, Bateman

(2018) Effect of sleep on overnight cerebrospinal fluid amyloid β kinetics. Ann Neurol 83,197–204.

24.

Chen

, Wang

, Zhang

, Wang

, Gao

(2018) Cerebrospinal fluid amyloid-β levels are increased in patients with insomnia. J Alzheimers Dis 61,645–651.

25.

, Zangrilli

, Finn

, Fagan

, Holtzman

(2019) Obstructive sleep apnea treatment, slow wave activity, and amyloid-β. Ann Neurol 85,291–295.

26.

Dumurgier

, Hanseeuw

, Hatling

, Judge

, Schultz

, Chhatwal

, Blacker

, Sperling

, Johnson

, Hyman

, Gómez-Isla

(2017) Alzheimer’s disease biomarkers and future decline in cognitive normal older adults. J Alzheimers Dis 60,1451–1459.

27.

Holtzman

, Morris

, Goate

(2011) Alzheimer’s disease: The challenge of the second century. Sci Transl Med 3,77sr71.

28.

Nicolazzo

, Xu

, Lavale

, Buckley

, Yassi

, Hamilton

, Maruff

, Baril

, Lim

, Pase

(2021) Sleep symptomatology is associated with greater subjective cognitive concerns: Findings from the community-based Healthy Brain Project. Sleep 44,zsab097.

29.

Jack

Jr , Knopman

, Jagust

, Petersen

, Weiner

, Aisen

, Shaw

, Vemuri

, Wiste

, Weigand

, Lesnick

, Pankratz

, Donohue

, Trojanowski

(2013) Tracking pathophysiological processes in Alzheimer’s disease: An updated hypothetical model of dynamic biomarkers. Lancet Neurol 12,207–216.

30.

Wisniewski

, Drummond

(2020) APOE-amyloid interaction: Therapeutic targets. Neurobiol Dis 138,104784.

31.

Yulug

, Hanoglu

, Kilic

(2017) Does sleep disturbance affect the amyloid clearance mechanisms in Alzheimer’s disease? Psychiatry Clin Neurosci 71,673–677.