Background
Alzheimer’s disease (AD) is the most prevalent neurodegenerative disorder in the world, currently affecting more than 55 million people [
1]. Female sex and the presence of the apolipoprotein E (APOE) ε4 allele are two non-modifiable features associated with an increased risk of developing the disease [
2]. Given that aging is also a relevant non-modifiable risk factor, estimations are that the prevalence of AD will increase in the next decades, especially considering the growing number of older adults in the nowadays society [
1,
3]. Aiming to prevent this and/or reduce the progression of the disease, considerable attention has been dedicated to the modifiable factors associated with increased risk of AD such as hypertension [
4], diabetes [
5], obesity [
6], and obstructive sleep apnea (OSA) [
7], among others.
The presence of OSA among cognitively healthy subjects is associated with atypical levels of AD pathological markers [
8‐
10], increased cognitive decline [
11,
12], and a higher risk of mild cognitive impairment/dementia [
13‐
16]. Nevertheless, previous studies suggest a different scenario once patients are diagnosed with AD. Accordingly, brain amyloid burden or cerebrospinal fluid (CSF) levels of phosphorylated-tau (P-tau) and total-tau (T-tau) presented similar alterations among AD patients with and without OSA after 2.52 ± 0.51 years of follow-up [
17]. In addition, the presence of OSA was not associated with an increase in the magnitude of cognitive decline among patients with mild-moderate AD after three years of follow-up [
18]. Yet, OSA recognizably disrupts the sleep structure, and previous research reported relevant associations between sleep lightening, atypical levels of AD pathological markers, and increased cognitive decline [
19,
20].
OSA is currently diagnosed based on the apnea–hypopnea index (AHI), a metric composed of distinct events which are not properly evaluated for their individual contribution in terms of adverse outcomes. To address this and improve the understanding concerning the impact of OSA among patients diagnosed with AD, we first performed a comprehensive characterization of the breathing cessation events that compose the AHI among our cohort of consecutive patients with mild-moderate AD. In the sequence, we evaluated the impact of each of these events on sleep structure. Finally, we investigated whether the events associated with increased disruption of the sleep structure would be the ones related to atypical levels of AD pathological markers and higher cognitive decline.
Methods
Study population
This is an ancillary study of a prospective trial designed to evaluate the influence of OSA on the cognitive decline of AD patients after one year of follow-up (NCT02814045). The recruitment started in 2015 and finished in 2019. We included acetylcholinesterase inhibitor-naïve individuals aged over 60 years who were diagnosed with AD according to the National Institute on Aging and Alzheimer’s Association (NIA-AA) clinical criteria [
21] and presented a Mini-Mental State Examination (MMSE) score ≥ 20. We excluded individuals presenting at least one of the following characteristics: (i) presence of visual and/or communication difficulties that could influence the compliance with the procedures of the study; (ii) presence of a previously diagnosed sleep disturbance; (iii) comorbidities such as cancer, severe renal or hepatic insufficiency, severe cardiac or respiratory failure; (iv) excessive alcohol intake (> 280 g/week); (v) MRI evidence of hydrocephalus, stroke, space-occupying lesion, or any clinically relevant disease of the central nervous system other than AD; (vi) presence of mental disorders according to DSM-V-TR™ criteria; (vii) use of medications under investigation or use of beta-blockers, antidepressants, neuroleptics, or hypnotics during the 15 days previous to polysomnography (PSG); (viii) presence of untreated (or treated for less than 3 months prior to the screening visit) vitamin B12 or folate deficiency; and (ix) presence of untreated thyroid disease.
The study was approved by the ethics committee of the Hospital Universitari Arnau de Vilanova-Santa Maria (Lleida, Spain) (CE-1218) and conducted according to the Declaration of Helsinki. The patient, the responsible caregiver, and the legal representative (when different from the responsible caregiver) signed an informed consent form.
Study design
Consecutive patients arrived at the Cognitive Disorders Unit of the Hospital Universitari de Santa Maria (Lleida, Spain) and were assessed for their eligibility. A clinical evaluation was performed to investigate associated comorbidities and to collect sociodemographic and anthropometric data. Blood and CSF were obtained to respectively determine ApoE genotypes and the levels of AD pathological markers. Sleep-related evaluations included the application of the Epworth Sleepiness Scale (ESS) and an overnight PSG. Neuropsychological assessment was performed at baseline and after 12 months of follow-up.
Clinical variables
The following variables were collected: sex, age, years of education, comorbidities (hypertension, diabetes mellitus, cardiopathy, periodic limb movements), and personal and family psychiatric and neurological history. Body mass index (BMI) was calculated as body weight (in kg)/height (in m2).
Apolipoprotein E (ApoE) genotype
DNA was extracted from buffy coat cells using a Maxwell® RCS blood DNA kit (Promega, USA). Twenty microliters of DNA were used for ApoE genotyping by polymerase chain reaction (PCR). ApoE genotype was dichotomized as ApoE-ε4 homozygous or heterozygous carrier (ApoEε4 +) or not (ApoEε4 −).
CSF biomarkers
The CSF samples were collected at baseline between 8:00 and 10:00 a.m. They were placed in polypropylene tubes, centrifuged at 2000 × g for 10 min at 4 °C, immediately frozen, and stored within 4 h in a − 80 °C freezer. The measurement of amyloid-beta protein (Aß42), T-tau, and P-tau was performed using commercial kits (Innotest® β-Amyloid-42; Innotest® hTAU Ag; and Innotest® Phospho-TAU181P, Fujirebio-Europe, Gent, Belgium). Neurofilament light (NF-L) was measured by commercial ELISA kit (Quidel, San Diego, CA). All determinations were performed in duplicate and in one round of experiments using one batch of reagents by board-certified laboratory technicians who were blinded to the clinical data. The intra-assay coefficients of variation were lower than 10% for internal quality control samples (two per plate). Based on previous data obtained by the research group, Aβ42 levels < 600 pg/ml were considered pathological [
22].
Epworth Sleepiness Scale (ESS)
Excessive daytime somnolence was assessed by the ESS. This questionnaire is composed of eight questions to assess the chance of falling asleep during different daily situations. Each question is rated on a 3-point scale, in which 0 represents no chance of occurrence, and 3 indicates a high chance of occurrence. The overall score ranges from 0 to 24 points. Higher scores represent increased daytime somnolence [
23‐
25].
Polysomnography (PSG)
The overnight PSG (Philips Respironics Alice 6 LDx, Philips, Murrysville, USA) was performed to assess the following variables: time spent in bed (in hours), total sleep time (in hours), sleep efficiency (in %, defined as the ratio between total sleep time and the time spent in bed), latency to N1 (in minutes, defined as the time spent awake until the first sleep episode), latency to REM sleep (in minutes, defined as the time spent asleep until the first REM sleep episode), the time spent in N1 stage (in %, defined as the percentage of time spent in N1 while asleep), the time spent in N2 stage (in %, defined as the percentage of time spent in N2 while asleep), the time spent in slow-wave sleep (SWS, also known as N3) (in %, defined as the percentage of time spent in SWS while asleep), the time spent in REM sleep (in %, defined as the percentage of time spent in REM sleep while asleep), the arousals index (defined as the mean number of awakening events per hour after the sleep onset), and the AHI (defined as the mean number of apnea and hypopnea events per hour during the time spent asleep).
The sleep staging and classification of the breathing cessation events that compose the AHI were performed according to the American Academy of Sleep Medicine (AASM) manual [
26] by an experienced sleep technician who was blinded to this study. In this study, hypopneas were defined as the reduction of airflow that lasted more than 10 s leading to arousal or oxygen desaturation (represented by a decrease in the oxygen saturation greater than 3%).
The indexes of each one of the breathing cessation events that compose the AHI such as obstructive apneas, central apneas, mixed apneas, and hypopneas were calculated in three different ways: (i) considering the total sleep time ([number of the event of interest × 60]/total sleep time), (ii) considering only NREM sleep ([number of the event of interest during NREM sleep × 60]/time spent in NREM sleep), and (iii) considering only REM sleep ([number of the event of interest during REM sleep × 60]/time spent in REM sleep).
Neuropsychological assessment
Patients underwent a neuropsychological evaluation through the MMSE at the beginning of the study and after 12 months of follow-up. The MMSE includes questions to evaluate different domains, such as attention, time and place orientation, and word recall. The scores of this test range from 0 to 30, and a higher score indicates better cognitive function [
27,
28].
Statistical analysis
Descriptive statistics were performed to report sociodemographic, clinical, and AD- and sleep-related data. Absolute and relative frequencies were used for qualitative data. The means (standard deviation (SD)) and medians (25th percentile; 75th percentile [p25;p75]) were estimated for quantitative variables with normal and nonnormal distributions, respectively. The normality of the distribution was assessed by the Shapiro–Wilk test.
The associations between the breathing cessation events that compose the AHI, sleep structure, and AD pathological markers were evaluated through linear regression models adjusted by age, sex, and BMI. Given the high collinearity among the breathing cessation events, we performed additional methods of data integration using partial least squares (PLS) regression to identify the pattern, in terms of AHI events, that better explained the relationship among the variables related to each outcome (sleep structure and AD pathological markers).
The associations between the breathing cessation events that compose the AHI and the cognitive decline at 12-month follow-up were evaluated through linear regression models adjusted by age, sex, BMI, years of education, and MMSE at baseline. Additionally, we performed multivariate analyses using PLS regression to identify the pattern, in terms of AHI events, most associated with the cognitive decline. The association between the first component of the PLS regression and cognitive decline was evaluated with Generalized Additive Model (GAM) adjusted for age, sex, BMI, years of education, and MMSE at baseline.
The p-value threshold defining statistical significance was set at < 0.05. Data management and statistical analyses were performed using R (version 4.0.1).
Discussion
In the current study, we first characterized the breathing cessation events composing the AHI among our cohort of mild-moderate AD patients. Such analysis revealed a high frequency of hypopneas followed by the presence of obstructive apneas. The distribution of the events along the sleep stages presented substantial variability among the patients. Furthermore, each event presented a distinct pattern of associations with sleep. Apneas were associated with arousals and sleep lightening whereas hypopneas, despite being related to an increased number of arousals, did not affect the sleep structure. Nevertheless, hypopneas were associated with the levels of the classical markers of AD. More importantly, the hypopnea index was the most relevant factor to predict an increased cognitive decline at the 12-month follow-up.
The inclusion of apneas and hypopneas in the same metric to diagnose and establish the severity of the so-called OSA generates an implicit assumption that these events are similar in relation to their clinical impact. While some studies corroborate this line [
29], others challenge the absence of distinctiveness among them [
30,
31]. Kulkas and collaborators (2017) demonstrated that obstructive apneas are associated with more severe SpO2 desaturation compared to hypopneas, suggesting a higher relevance of the first when estimating the severity of OSA and associated long-term cardiovascular outcomes [
30]. Several studies also reveal distinct outcomes according to the used criteria for the classification of hypopneas, highlighting the relevance of a detailed analysis of the events composing the AHI [
31‐
36]. To our knowledge, this is the first study to address this matter, demonstrating the distinctiveness between hypopneas and obstructive apneas in terms of associated clinical outcomes among AD patients.
We observed that obstructive apneas were related to an increase in the time spent in N1 in detriment of the time spent in SWS, besides an association with an increased number of arousals. Also, these events were related to increased levels of NF-L, a marker for the axonal damage and cognitive decline of patients with AD [
37‐
39]. Accordingly, we have previously demonstrated that patients with a propensity to spend most of the time in the lighter sleep stage (light sleepers) present an increased probability of having high NF-L levels compared with those individuals with a propensity to deepen their sleep (deep sleepers) [
40]. Differently, despite the association with an increased number of arousals, the hypopneas were not related to alterations in the sleep structure. Nevertheless, the hypopnea index was associated with the levels of classical markers of AD and was the most relevant breathing cessation event to predict an increased cognitive decline at the 12-month follow-up. This suggests a possible contribution of arousals within this context regardless of whether alterations in the sleep structure are present or detected.
Based on the guidelines of the American Academy of Sleep Medicine (AASM), hypopneas and obstructive apneas are differentiated by the degree of airflow decrease (30% to 89% versus ≥ 90%, respectively) [
26]. In consequence, obstructive apneas are associated with higher oxygen desaturation, which suggests a greater clinical impact in comparison to hypopneas [
30]. Similarly, our findings revealed a stronger association between obstructive apneas and arousals than that involving hypopneas, which reinforces the greater risk of clinical consequences related to the apneic events. Still, our data indicate a distinguished role of hypopneas on the cognitive decline after 12 months of follow-up. A possible explanation for such outcome could be related to the 3-times higher frequency of hypopneas among our population compared to that of obstructive apneas. Future studies will be necessary to evaluate such interesting finding, confirming whether there are relationships of causality in this regard. Also, larger cohorts will provide the required variability to investigate differences between patients who have a pattern of breathing cessation events mostly composed of hypopneas and those in whom obstructive apneas are the most prevalent.
Limitations
The current findings should be interpreted in light of some aspects: (i) due to the sample size and the exploratory nature of this study, our data were not adjusted for multiple comparisons; (ii) given previous demonstrations of a differential effect of OSA on cognitive function depending on the cognitive status of the subjects and/or presence of AD pathology [
8,
12,
17,
18], the associations herein observed cannot be extrapolated to cognitively healthy subjects or cognitively unimpaired patients with AD pathology; (iii) the sleep evaluation revealed a low percentage of time spent in REM sleep, which might have prevented the observation of associations in this regard; (iv) the discreet prevalence of central and mixed apneas in addition to the high collinearity among the components of the AHI may have led to associations influenced by the most prevalent events such as hypopneas and obstructive apneas; (v) although the objective of this study was limited to investigate whether the AHI components associated with increased sleep disruption would be the ones linked to worse outcomes, the influence of the respiratory burden related to each event should not be ruled out; (vi) given the observational design of this study, it is not possible to confirm causality and directionality between the associations of interest.
Conclusions
In summary, we demonstrated that the hypopneas and the obstructive apneas are the components that define the AHI among consecutive patients with mild-moderate AD. These events were distinctively associated with the sleep structure, the levels of pathological markers of AD, and cognitive decline, suggesting that the OSA does have an influence once patients are diagnosed with symptomatic disease up to a certain extent. This altogether highlights the importance of a patient-centered approach, with a comprehensive and detailed analysis of the AHI to effectively predict the different outcomes and tailor the appropriate therapeutic strategies.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.