Abstract
INTRODUCTION
We examined obstructive sleep apnea (OSA) severity's association with Alzheimer's disease (AD) plasma biomarkers, independent or synergistic with cerebrospinal fluid (CSF) amyloid, and as a proof of concept, whether plasma amyloid beta (Aβ)42/Aβ40 with OSA severity improves detection of amyloidosis and tau pathology.
METHODS
In 120 cognitively normal older adults (70 with CSF data) from New York University sleep and aging studies (2013–2021), OSA severity was measured using apnea/hypopnea index with 4% desaturation; plasma Aβ40, Aβ42, tau, and neurofilament light chain (NfL) via single molecule array; CSF amyloid and tau via enzyme‐linked immunosorbent assay. Associations evaluated adjusted correlations and generalized models; receiver operating characteristic analyses evaluated diagnostic accuracy.
RESULTS
OSA severity correlated with plasma Aβ40 (r = 0.21), Aβ42 (r = 0.26), and Aβ42/Aβ40 (r = 0.20). Plasma tau and NfL associations depended on CSF–Aβ42. OSA severity with Aβ42/Aβ40 improved CSF amyloidosis (area under the curve [AUC] = 0.78) and tau pathology (AUC = 0.71) detection.
DISCUSSION
OSA severity relates to elevated plasma Aβ and, with CSF amyloid, to tau/NfL. Combined plasma and OSA measures aid non‐invasive AD associations’ detection.
Keywords: Alzheimer's disease, amyloid, obstructive sleep apnea, plasma biomarkers
Highlights
Obstructive sleep apnea (OSA) is associated with plasma amyloid beta (Aβ)40, Aβ42, and Aβ42/Aβ40.
OSA in synergism with brain amyloid levels is associated with plasma tau and neurofilament light chain.
Combined with plasma Aβ42/Aβ40, OSA enhances brain amyloidosis and tau pathology detection.
1. BACKGROUND
Alzheimer's disease (AD) is complex and multifactorial, with obstructive sleep apnea (OSA) becoming increasingly recognized as one of the contributors to its pathophysiology. 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 Established molecular markers of AD pathology, such as cerebrospinal fluid (CSF) and positron emission tomography (PET) imaging of amyloid beta (Aβ) and tau, are reliable predictors of amyloid burden and future development of AD. 12 , 13 However, these methods are costly, and require invasive lumbar punctures (LPs), making them less feasible for widespread screening. As a result, blood‐based biomarkers (BBMs) were extensively investigated to validate their use for diagnostic purposes, minimizing costs and invasiveness in the process. 14 , 15
Research has shown that OSA increases the risk of AD and is associated with elevated levels of Aβ and tau proteins, in cognitively normal (NL) older adults and in individuals with mild cognitive impairment (MCI). 5 , 6 , 7 , 8 , 9 , 10 , 11 Severity of OSA has been linked to greater brain Aβ deposition, making it a critical factor in understanding AD pathology. Novel BBMs, such as Aβ42, Aβ40, phosphorylated tau (p‐tau), and total tau (t‐tau), have been shown to mirror the trajectories of CSF biomarkers, with high correspondence to amyloid PET and CSF levels of Aβ42 and p‐tau. 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23
While OSA has been associated with amyloid and tau pathologies, its relationship with plasma amyloid, tau, and markers of neuronal injury, both independently and synergistically with amyloid burden, remains largely unexplored. Understanding these associations is crucial as the use of plasma biomarkers increases in AD research.
This study examined whether OSA influences plasma levels of Aβ40, Aβ42, Aβ42/Aβ40 ratio, t‐tau, t‐tau/Aβ42 ratio, and neurofilament light chain (NfL) independently of and synergistically with amyloid burden. Additionally, as a proof of concept, we determined whether combining OSA severity with plasma Aβ42/Aβ40 ratio improves the predictive power for diagnosing brain amyloidosis, through CSF, and tau pathology. Our central hypothesis is that OSA exhibits independent and synergistic effects on plasma biomarkers, providing critical insights into its role in AD pathology and early diagnosis.
2. METHODS
2.1. Participants
The New York University (NYU) cohort consisted of healthy cognitively unimpaired (CU) older adults recruited from community settings through National Institutes of Health/National Institute on Aging‐supported studies. 10 , 24 , 25 Participants provided written informed consent following NYU Institutional Review Board protocols. Eligibility criteria included ages 55 to 90 years, fluency in English, ≥ 12 years of education, Mini‐Mental State Examination (MMSE) scores >27, 26 a Clinical Dementia Rating (CDR) 27 of 0, and Geriatric Depression Scale scores ≤ 5. 28 A total of 120 participants underwent blood analysis for AD biomarkers, 70 of which also had baseline CSF data. Data collection occurred from 2013 to 2021.
2.2. OSA diagnosis
Participants underwent home monitoring for OSA using ARES Unicorder21 or Embletta MPR Type 3 systems. 29 Both systems contain channels for oxygen saturation (Spo2) and pulse rate and air flow from a nasal cannula and pressure transducer. A certified sleep physician reviewed the recordings. OSA severity was determined using the apnea/hypopnea index with 4% desaturation (AHI4%). Apnea was defined as > 90% airflow reduction for > 10 seconds, and hypopnea as > 30% airflow reduction for 10 seconds with > 4% desaturation. Mild, moderate, and severe OSA corresponded to AHI4% > 5 and < 15, > 15 and < 30, and ≥ 30 events per hour, respectively. 30
2.3. Brain amyloid and tau (CSF–Aβ and CSF–p‐tau)
LPs and subsequent CSF analyses were performed as per previously published protocols. 31 , 32 CSF–Aβ42 and CSF–p‐tau181 concentrations were measured using sandwich enzyme‐linked immunosorbent assay (ELISA), with batch‐wise rescaling reducing variability from 20% to 10%. Amyloid positivity was defined as CSF–Aβ42 ≤ 375.5 pg/mL, and p‐tau positivity as CSF–p‐tau181 ≥ 53.7 pg/mL, 14 using NYU Alzheimer's Disease Research Center data‐driven thresholds.
RESEARCH IN CONTEXT
Systematic review: The authors reviewed the literature using traditional sources like PubMed. Research shows that obstructive sleep apnea (OSA) increases Alzheimer's disease (AD) risk and is linked to elevated amyloid beta (Aβ) and tau proteins in cognitively normal (NL) older adults and individuals with mild cognitive impairment. Novel plasma markers such as Aβ42, Aβ40, phosphorylated tau, and total tau align with cerebrospinal fluid (CSF) biomarkers and amyloid positron emission tomography levels.
Interpretations: Our findings suggest that in NL older adults, OSA severity significantly correlates with higher plasma Aβ40, Aβ42, and Aβ42/Aβ40 ratio. Combined with CSF Aβ42 and plasma Aβ42/Aβ40 ratio, it associates with elevated tau and neurofilament light chain, and enhances brain amyloidosis and tau pathology detection, respectively. These findings are novel and align with recent epidemiological studies.
Future directions: The article proposes a framework for additional research. Examples include exploring (a) OSA–Aβ synergism in identifying high‐risk patients and (b) how OSA severity with plasma Aβ42/Aβ40 associations could enhance participant selection in clinical AD trials.
2.4. Plasma and CSF procedures
In the morning, fasting blood and CSF samples were collected using a standardized protocol. Blood was drawn into six ethylenediaminetetraacetic acid plasma tubes, centrifuged at 2000 × g and 4°C for 10 minutes, and the plasma from all tubes was pooled, mixed, and aliquoted into polypropylene tubes. Plasma samples were stored at −80°C within 30 to 60 minutes of extraction and underwent one to two freeze–thaw cycles prior to analysis (three freeze–thaw cycles have been shown not to affect plasma Aβ or tau values). 33 , 34 CSF was handled per established guidelines. 32 Plasma levels of Aβ40, Aβ42, t‐tau, and NfL were analyzed using ultra‐sensitive single‐molecule array (SIMOA) assays. Data‐driven clinically relevant thresholds were set for plasma Aβ42/Aβ40 at ≤ 0.0513083 pg/mL, indicating Aβ42/Aβ40 positivity. 14
2.5. Covariates/potential confounders
Covariates included age, sex, level of education, body mass index (BMI), apolipoprotein E (APOE) ε4 status (determined by presence of at least one allele), thyroid disease, diabetes, cardiovascular disease (CVD; including conditions like coronary heart disease, heart failure, stroke, and peripheral arterial disease), time between LP and blood draw, and use of antihypertensive medications. Blood was collected before LP, and both were collected on the same day. Plasma was collected after a minimum 6 hour fasting period and processed within 1 hour of collection.
2.6. Statistical analyses
Demographic and clinical characteristics were assessed with t tests, analysis of variance (ANOVA), Kruskal–Wallis tests, or χ 2 tests. Plasma biomarker assays had a coefficient of variation (CV) ≤ 20%. Associations of OSA severity (i.e., AHI4% continuous variable) and plasma AD biomarker levels (n = 120) were assessed using Spearman correlation and/or Pearson correlation analysis after data normalization procedures. Both methods yielded similar findings. The data normalization process included applying the natural log or a square root function to outcome and our main exposure variables, in which data were skewed to reduce right‐skewness, compress large values, stabilize variance, and make distributions more symmetrical for statistical tests like t tests or ANOVA and especially to fulfill the normality assumption associated with these tests. The association between OSA severity and AD plasma biomarkers dependent on CSF–Aβ42 levels (n = 70) was assessed using generalized linear models. We used ANOVA to examine variations in plasma Aβ42/Aβ40 levels by OSA severity status/categories (i.e., normal, mild, moderate, and severe) and within categorized CSF amyloid and p‐tau groups. Data were adjusted for age, sex, race, level of education, BMI, hypertension, and APOE ε4 status. Use of medication, history of thyroid disease, diabetes, and CVD were considered but excluded from the final model due to lack of significance. Receiver operating characteristic (ROC) analyses were performed to evaluate the ability of plasma Aβ42/Aβ40 and OSA severity to diagnose CSF brain amyloidosis and tau pathology in a model with and without age, sex, hypertension, level of education, BMI, and APOE ε4 status. We calculated the Youden index to determine the optimal cutoff for plasma Aβ42/Aβ40. 14
3. RESULTS
3.1. Baseline demographic, clinical, and sleep data
Tables 1 and 2 present a summary of the baseline demographic and sleep characteristics of the study population. The mean (standard deviation) of the participants’ age, BMI, and education was 70.0 (6.8), 27.4 (6.0), and 16.9 (2.2), respectively. More than half of the participants were women (67.5%), and 83.3% were non‐Hispanic Whites. Among the participants, 51 individuals (41%) were classified as healthy control participants without OSA (AHI4% < 5), 32 (26%) had mild OSA (AHI4% 5–14.9), and 40 (33%) had moderate to severe OSA (AHI4% > 15). The prevalence of hypertension (HTN) was higher in individuals with OSA, and its frequency increased with the severity of OSA (72.2% in moderate, 59.1% in severe, and 43.8% in mild OSA, compared to 21.6% in controls, p < 0.001). There were significant differences in APOE ε4 status among the participants, with a higher prevalence observed in individuals with moderate and severe OSA compared to controls (45.5% and 50% vs. 29.4%, p < 0.01). These differences could be potential sources of confounding. Thus, all our analyses were adjusted for age, sex, race, level of education, BMI, hypertension, and APOE ε4 status to address the issue of potential confounding.
TABLE 1.
Baseline demographic, clinical, and sleep characteristics of the participants.
| Characteristic |
All (N = 120) |
No OSA (n = 51) |
Mild OSA (n = 32) |
Moderate OSA (n = 18) |
Severe OSA (n = 22) |
|---|---|---|---|---|---|
| Age, mean (SD), y | 70.0 (6.6) | 70.0 (6.3) | 69.1 (6.3) | 71.7 (6.8) | 70.5 (6.7) |
| Female sex, N (%)*+ | 81 (67.5) | 35 (68.6) | 23 (71.9) | 13 (72.2) | 9 (40.9) |
| Male sex, N (%)*+ | 39 (32.5) | 16 (31.4) | 9 (28.1) | 5 (27.8) | 13 (59.1) |
| Non‐Hispanic White, N (%) | 100 (83.3) | 38 (74.5) | 25 (78.1) | 18 (100) | 19 (86.4) |
| Body mass index, mean (SD) | 27.4 (6.0) | 24.7 (3.9) | 27.7 (6.6) | 30.0 (6.1) | 30.4 (7.2) |
| Education, mean (SD), y | 16.9 (2.2) | 17.0 (2.2) | 16.7 (1.9) | 16.6 (2.2) | 17.2 (2.5) |
| Hypertension, N (%)*** | 51 (42.5) | 11 (21.6) | 14 (43.8) | 13 (72.2) | 13 (59.1) |
| Diabetes, N (%) | 5 (4.1) | 1 (2.0) | 1 (3.1) | 1 (5.5) | 2 (9.0) |
| Cardiovascular disease, N (%) | 6 (5.0) | 1 (2.0) | 4 (12.5) | 1 (5.6) | 0 (0) |
| Thyroid disease, N (%) | 20 (16.7) | 11 (21.6) | 5 (15.6) | 4 (22.2) | 0 (0) |
| APOE ε4 positive+, N (%)** | 44 (36.7) | 15 (29.4) | 10 (31.2) | 9 (50.0) | 10 (45.5) |
| AHI4%, median (IQR) *** | 7.0 (1.8–22.0) | 1.1 (0.6–2.7) | 7.4 (6.0–9.3) | 21.2 (19.8–24.0) | 42.4 (36.2–49.2) |
| RDI_all, median (IQR)*** | 18.5 (10.7–30.7) | 9.0 (6.3–11.8) | 19.0 (16.9–21.2) | 31.0 (27.0–31.0) | 59.0 (43.3–64.0) |
| Average SpO2, median (IQR), % | 94.0 (92.5–95.1) | 94.3 (93.6–95.3) | 93.1 (92.3–95.0) | 93.4 (92.5–93.4) | 93.5 (90.7–94.3) |
| ESS, median (IQR)* | 5.0 (3.0–9.0) | 4.0 (3.0–7.0) | 7.0 (4.0–9.0) | 5.0 (4.0–7.0) | 6.5 (3.5–9.0) |
| Total valid signal time, median (IQR),* | 7.5 (6.5–8.0) | 7.5 (6.5–8.0) | 6.5 (6.0–7.8) | 7.5 (6.8–7.5) | 7.9 (6.5–9.0) |
| MMSE, median (IQR) | 30.0 (29.0–30.0) | 30.0 (29.0–30.0) | 29.0 (29.0–30.0) | 30.0 (28.0–30.0) | 30.0 (29.0–30.0) |
| GDS, median (IQR) | 1.0 (1.0–2.0) | 1.0 (1.0–2.0) | 1.0 (1.0–2.0) | 2.0 (2.0–3.0) | 1.0 (1.0–2.0) |
| CDR, median (IQR) | 0.0 (0.0–0.0) | 0.0 (0.0–0.0) | 0.0 (0.0–0.0) | 0.0 (0.0–0.0) | 0.0 (0.0–0.0) |
Abbreviations: AHI4%, apnea–hypopnea index 4% desaturation; APOE, apolipoprotein E; CDR, Clinical Dementia Rating; ESS, Epworth Sleepiness Scale; GDS, Geriatric Depression Scale; IQR, interquartile range; MMSE, Mini‐Mental State Examination; OSA, obstructive sleep apnea; RDI_all, respiratory disturbance index with 3% desaturation and arousal; SD, standard deviation; y, years.
+ Carrier of at least on ε4 allele.
*+ X2 (chi‐square) p ≤ 0.05.
* p ≤ 0.05.
** p ≤ 0.01.
*** p ≤ 0.001.
TABLE 2.
Baseline descriptive characteristics and biomarker concentration levels of participants by obstructive sleep apnea status.
| Plasma biomarkers (pg/mL), median (IQR) | |||||
|---|---|---|---|---|---|
| Characteristic |
All (N = 120) |
No OSA (n = 51) |
Mild OSA (n = 32) |
Moderate OSA (n = 18) |
Severe OSA (n = 22) |
| Tau | 1.2 (1.0–1.9) | 1.2 (1.0–1.7) | 1.3 (0.9–2.0) | 1.2 (0.9–2.2) | 1.3 (0.78–2.12) |
| Aβ40 | 252.2 (221.7–288.1) | 257.4 (231.6–293.3) | 238.6 (211.4–287.0) | 258.5 (216.6–290.7) | 253.8 (219.3–285.2) |
| Aβ42 | 10.1 (8.1–12.8) | 10.1 (8.1–11.6) | 10.7 (8.1–12.0) | 10.7 (9.0–14.9) | 9.1 (7.8–15.0) |
| Aβ42/Aβ40 ratio | 0.04 (0.03–0.05) | 0.04 (0.03–0.04) | 0.04 (0.03–0.05) | 0.04 (0.03–0.05) | 0.04 (0.03–0.04) |
| Neurofilament light chain | 15.8 (12.5–21.6) | 16.5 (13.7–20.6) | 15.8 (12.1–21.6) | 19.0 (13.5–22.4) | 13.9 (12.5–19.3) |
| CSF Aβ42 and p‐tau181 (pg/mL), median (IQR) | |||||
|---|---|---|---|---|---|
| Characteristic |
All (N = 70) |
No OSA (n = 29) |
Mild OSA (n = 20) |
Moderate OSA (n = 11) |
Severe OSA (n = 10) |
| Aβ42 | 656.2 (443.5–841.9) | 688.7 (482.3–858.3) | 625.6 (427.0–822.6) | 580.5 (380.6–693.8) | 729.2 (570.5–876.7) |
| P‐tau181 | 41.0 (32.0–52.0) | 45.0 (33.0–53.0) | 36.5 (26.0–50.0) | 38.2 (26.0–49.0) | 41.0 (35.0–42.3) |
Abbreviations: Aβ, amyloid beta; AHI4%, apnea–hypopnea index 4% desaturation; CSF, cerebrospinal fluid; IQR, interquartile range; OSA, obstructive sleep apnea; p‐tau, phosphorylated tau.
+ χ 2 (chi‐square) p ≤ 0.05.
* p ≤ 0.05.
p ≤ 0.01.
p ≤ 0.001.
3.2. Association of OSA severity (i.e., AHI4% continuous variable) and AD plasma biomarkers
Figure 1A–F presents the cross‐sectional correlation and regression modeling of associations of OSA severity and plasma Aβ40, Aβ42, Aβ42/Aβ40, and tau for all participants. Increasing OSA severity was associated with higher levels of plasma Aβ40 (r = 0.21), Aβ42 (r = 0.225), and Aβ42/Aβ40 (r = 0.204), with a p value < 0.05 for all. OSA severity was not associated with plasma tau (r = 0.12), plasma tau/Aβ42 (r = −0.02), or plasma NfL (r = −0.056), with p > 0.05 for all.
FIGURE 1.
Association of OSA severity and AD plasma biomarkers. Correlation and regression modeling of OSA severity and plasma amyloid and tau levels. A, Correlation and regression modeling of AHI4% and plasma Aβ40 (r = 0.21, p = 0.024). B, Correlation and regression modeling of AHI4% and plasma Aβ42 (r = 0.23, p = 0.016). C, Correlation and regression modeling of AHI4% and plasma Aβ42/Aβ40 (r = 0.20, p = 0.026). D, Correlation and regression modeling of AHI4% and plasma tau (r = 0.12, p = 0.201). E, correlation and regression modeling of AHI4% and plasma tau/Aβ42 (r = −0.02, p = 0.808). F, Correlation and regression modeling of AHI4% and plasma NfL (r = −0.06, p = 0.561). AHI4% (SQRT) represents the square‐root – transformed AHI4% variable; (Natural Log) indicates that the variable was log transformed. Aβ, amyloid beta; AD, Alzheimer's disease; AHI4%, apnea–hypopnea index 4% desaturation; NfL, neurofilament light chain; OSA, obstructive sleep apnea.
3.3. Association of OSA severity (i.e., AHI4% continuous variable) and plasma levels of tau, tau/Aβ42, or NfL dependent on CSF–Aβ42/p‐tau levels
Table 3 shows the association of OSA severity and plasma levels of t‐tau, t‐tau/Aβ42, and NfL dependent on CSF–Aβ42 levels, revealing significant interactions between CSF–Aβ42 levels and AHI (p < 0.05 for all). Independently, AHI4% showed no significant association with plasma tau (β = 0.009; 95% confidence interval [CI], −0.013 to 0.03; p = 0.432), plasma tau/Aβ42 association (β = 0.002; 95% CI, −0.025 to 0.022; p = 0.867) or plasma NfL (β = 0.006; 95% CI, −0.019 to 0.013; p = 0.752). Independently, CSF Aβ42 levels were significantly associated with plasma tau (β = −0.058; 95% CI, −0.079 to 0.037; p = 0.023), plasma tau/Aβ42 (β = −0.054; 95% CI, −0.075 to 0.033; p < 0.001), and plasma NfL (β = 0.055; 95% CI, −0.086 to 0.024; p = 0.012). The interaction between AHI4% and CSF–Aβ42 was significantly associated with plasma tau, with plasma t‐tau/Aβ42, and with plasma NfL (β = 0.04; 95% CI, 0.018–0.062; p = 0.016; β = 0.015; 95% CI, 0.003–0.039; p = 0.041; and β = 0.033; 95% CI, 0.01–0.056; p = 0.023, respectively), with β estimates suggesting that with combined unit increases in AHI4% and unit decreases in CSF–Aβ42 levels, there were corresponding increases in plasma levels of tau, plasma t‐tau/Aβ42, or plasma NfL.
TABLE 3.
Generalized linear model estimates of the association of OSA severity and plasma levels of tau, tau/Aβ42, or NfL dependent on CSF–Aβ42 levels in NL (cognitively normal) participants.
| Outcome | Model 1 term a |
Standardized estimate (95% CI) |
p‐value |
|---|---|---|---|
| Plasma tau | AHI4% | 0.009 (−0.013 to 0.030) | 0.432 |
| CSF–Aβ42 | −0.058 (−0.079 to −0.037) | 0.023 | |
| AHI4% x CSF–Aβ42 | 0.040 (0.018 to 0.062) | 0.016 | |
| Plasma tau/Aβ42 | AHI4% | 0.002 (−0.025 to 0.022) | 0.867 |
| CSF–Aβ42 | −0.054 (−0.075 to −0.033) | <0.001 | |
| AHI4% x CSF–Aβ42 | 0.015 (0.003 to 0.039) | 0.041 | |
| Plasma NfL | AHI4% | 0.006 (−0.019 to 0.013) | 0.752 |
| CSF–Aβ42 | −0.055 (−0.086 to −0.024) | 0.012 | |
| AHI4% x CSF–Aβ42 | 0.033 (0.010 to 0.056) | 0.023 |
Model adjusted for age, sex, body mass index, education, APOE ε4 status, and hypertension.
Abbreviations: Aβ, amyloid beta; AHI4%, apnea–hypopnea index 4% desaturation; APOE, apolipoprotein E; BMI, body mass index; CI, confidence interval; CSF, cerebrospinal fluid; NfL, neurofilament light chain; OSA, obstructive sleep apnea.
Bolded values are indicative of statistically significant findings.
3.4. ANOVA results examining associations between OSA severity status/categories and plasma Aβ42/Aβ40 levels, within categorized CSF amyloid and p‐tau groups
Overall, baseline plasma Aβ42/Aβ40 did not differ by categorized OSA groups, CSF–amyloid or CSF–p‐tau status (p > 0.05 for all), possibly due to loss of power from categorization. Notably, only a few individuals were positive for CSF–amyloid and CSF–p‐tau; as such, it was not surprising that baseline plasma Aβ42/Aβ40 ratio showed no significant differences by OSA severity status within CSF–amyloid positive and CSF–p‐tau positive groups (p > 0.05 for both). However, compared to the non‐OSA group, plasma Aβ42/Aβ40 ratio differed significantly (p < 0.05) by OSA severity status, with higher levels for mild, moderate, and severe OSA, among patients negative for CSF–Aβ42 and CSF–p‐tau (Figure 2A–G).
FIGURE 2.
Analysis of variance results examining associations between OSA severity status and plasma Aβ42/Aβ40 levels, within categorized CSF amyloid and p‐tau groups. A, Baseline plasma Aβ42/Aβ40 by OSA severity (p = 0.217). B, Baseline plasma Aβ42/Aβ40 by CSF amyloid status (p = 0.208). C, Baseline plasma Aβ42/Aβ40 by CSF–p‐tau status (p = 0.474). D, Baseline plasma Aβ42/Aβ40 by OSA severity (CSF‐amyloid positive; p = 0.371). E, Baseline plasma Aβ42/Aβ40 by OSA severity (CSF‐amyloid negative). Post hoc comparisons significant differences (mild vs. no OSA; moderate OSA vs. no OSA, and severe vs. no OSA). F, Baseline plasma Aβ42/Aβ40 by OSA severity (CSF–p‐tau positive; p = 0.192). G, Baseline plasma Aβ42/Aβ40 by OSA severity (CSF–p‐tau negative). Post hoc comparisons significant differences (Mild vs. no OSA, moderate OSA vs. no OSA, and severe vs. no OSA). Aβ, amyloid beta; CSF, cerebrospinal fluid; OSA, obstructive sleep apnea; p‐tau, phosphorylated tau.
3.5. Correspondence of baseline plasma Aβ42/Aβ40, age, sex, level of education, APOE status, BMI, and OSA severity (i.e., AHI4% continuous variable) to CSFAβ42 and CSFp‐tau (Figure 3A,B)
SIMOA‐assessed plasma Aβ42/Aβ40 alone had a low correspondence with CSF–Aβ42 status (ROC area under the curve [AUC] 0.53 [95% CI = 0.43–0.63]). The ROC AUC was significantly improved by adding AHI4% alone (AUC 0.78 [95% CI: 0.67–0.89]) and outperformed a model that included age, APOE ε4, BMI, sex, and education to an AUC of 0.66 (95% CI = 0.47–0.84; Figure 3A). Combining AHI4% with age, APOE ε4, BMI, sex, and education (AUC 0.75 [95% CI: 0.59–0.91]) did not improve model performance beyond adding AHI4% alone. The baseline plasma Aβ42/Aβ40 ratio model for prediction of CSF–p‐tau status also showed an AUC of 0.53 (95% CI = 0.39–0.66). Once again, addition of AHI4% alone improved model performance (AUC 0.71 [95% CI: 0.61–0.84]), outperforming a model that included age, APOE ε4, BMI, sex, and education AUC to 0.64 (95% CI = 0.51–0.77), with no further model improvement beyond AHI4% alone when AHI4% was combined with these other risk factors (AUC 0.71 [95% CI: 0.52–0.90; Figure 3B).
FIGURE 3.
Correspondence of baseline plasma Aβ42/Aβ40, age, sex, HTN, level of education, APOE status, BMI, and OSA severity to CSF–Aβ42 and CSF–p‐tau using ROC analyses. A, ROC analyses between plasma Aβ42/Aβ40 and CSF–Aβ42. B, ROC analyses between plasma Aβ42/Aβ40 and CSF–p‐tau. Aβ, amyloid beta; APOE, apolipoprotein E; BMI, body mass index; CSF, cerebrospinal fluid; HTN, hypertension; OSA, obstructive sleep apnea; p‐tau, phosphorylated tau; ROC, receiver operating characteristic.
4. DISCUSSION
This study examined whether OSA is associated with AD plasma biomarker levels, independent of or synergistic with amyloid burden, and whether the combination of plasma Aβ42/Aβ40 ratio and OSA severity improves diagnosis of brain amyloidosis and tau pathology, using CSF fluid biomarkers as the standard. Our significant findings were as follows: (1) OSA severity was independently associated with higher levels of plasma Aβ40, Aβ42, Aβ42/Aβ40, but not plasma t‐tau, plasma t‐tau/Aβ42, or plasma NfL; (2) OSA severity and CSF–Aβ42 (amyloid burden) had a synergistic effect on plasma tau, tau/Aβ42, and NfL, revealing significant interactions between CSF–Aβ42 levels and AHI4%; and (3) combining plasma Aβ42/Aβ40 and OSA severity significantly improved correspondence with CSF–Aβ42/p‐tau status, suggesting improved diagnosis of brain amyloidosis and tau pathology.
Our findings showing OSA severity associations with levels of plasma markers of AD pathology are consistent with previous longitudinal and cross‐sectional studies, which demonstrated that OSA increases the risk of AD and is associated with validated AD biomarkers in NL participants. 5 , 9 , 10 , 11 , 12 However, OSA severity alone was not associated with plasma markers of neuronal injury. These discrepancies may also reflect that our cohort was predominantly CSF–amyloid negative. Additionally, plasma and serum Aβ are non‐specific measures representing a small fraction of blood‐soluble Aβ, warranting cautious interpretation when using SIMOA or ELISA assays. Recent investigations have linked OSA and CSF AD biomarkers and neuronal‐derived proteins, but the participants were obtained from a combination of community‐dwelling and sleep clinic patients that tend to be more symptomatic or have greater OSA severity. 6 , 7
Fluid concentrations of Aβ40, Aβ42, and Aβ42/Aβ40 are typically lower in plasma and CSF among individuals with clinical AD, reflecting greater intracerebral amyloid burden. We have previously demonstrated that baseline OSA severity is associated with longitudinal decreases in CSF Aβ42 with the AHI4% as a continuous 10 or categorial 11 , 35 variable. Thus, the observation that AHI4% is positively correlated with plasma amyloid levels at cross‐section may come across as counterintuitive. However, amyloid trajectories across the lifespan remain incompletely characterized. Evidence from the Dominantly Inherited Alzheimer Network cohort suggests that individuals with low amyloid levels later in life may have had higher levels decades earlier, followed by steeper declines. 36 Consequently, a higher concentration of fluid amyloid at cross‐section can reflect either increased or decreased risk for AD depending on the timing with which the fluid is assayed. 37
To further explore whether OSA severity was associated with plasma neuronal injury markers (i.e., plasma tau and NfL), we determined if this association depended on CSF–amyloid levels. We have previously shown synergism between OSA severity and Aβ on the clinical progression of AD. 35 Here, OSA severity and CSF–Aβ42 interacted synergistically with plasma tau, t‐tau/Aβ42, and NfL. Novel plasma Aβ42/Aβ40 markers show strong correspondence with amyloid PET and CSF p‐tau, supporting convergent validity with established biomarkers. 14 , 15 , 19 , 38 Estimates indicated that increasing AHI combined with decreasing CSF–Aβ42 was associated with higher plasma tau, t‐tau/Aβ42, and NfL. These findings support prior work 35 indicating that OSA contributes to AD progression both independently and synergistically with amyloid and tau pathology. This suggests that amyloid‐positive individuals with OSA may represent a high‐risk group for targeted interventions aimed at reducing AD progression, 39 , 40 and can distinguish between older individuals with normal cognition, MCI, or AD. 41 Similar findings are true for plasma NfL, a non‐specific marker of neurodegeneration. Notably, higher levels of plasma NfL inversely correlated with cognitive performance in NL older individuals. 21 , 22 , 23 , 41 , 42 , 43 , 44 , 45 , 46
This study's results highlighted that the plasma Aβ42/Aβ40 ratio was slightly higher based on OSA severity status for mild, moderate, and severe OSA, among participants that are negative for CSF–Aβ42 and CSF–p‐tau. No statistical difference was observed in individuals positive for CSF–Aβ42 and CSF–p‐tau, mainly due to power issues. These findings suggest that plasma Aβ42/Aβ40 may be useful for assessing AD pathology in OSA patients at risk for AD, consistent with prior studies demonstrating strong correspondence with brain amyloidosis. 14 , 15 We found that combining plasma Aβ42/Aβ40 and OSA severity significantly improved correspondence with CSF brain amyloidosis and tau pathology, which is supported by earlier studies regarding the use of plasma Aβ42/Aβ40 as a screening tool for brain amyloidosis detection. 47 Recent studies using lower precision assays have found that plasma Aβ42/Aβ40 ratio (measured by ELISA) is associated with age and amyloid status and that models including age and APOE ε4 status provide better predictions of amyloid status. 48 However, our results indicate that OSA severity combined with plasma Aβ42/Aβ40 performs comparably, even after accounting for sex, education, and BMI, implying that there is no difference between using these factors versus only using AHI indices to determine the relationship between brain amyloidosis and tau pathology.
In a predominantly CSF–amyloid negative cohort, variation in plasma Aβ42/Aβ40 may reflect peripheral or analytical influences rather than central amyloid burden. Nonetheless, our working hypothesis regarding the OSA–AD relationship is multi‐fold. First, OSA may exert direct neurotoxic effects independent of AD pathology, increasing the risk of AD pathology and progression; the observed correlations with plasma Aβ markers in CU individuals support this possibility. Second, OSA may interact with AD pathology, specifically Aβ and tau, to exacerbate disease progression and elevate AD risk in individuals with co‐occurring pathology. Accordingly, our interaction analyses demonstrated synergistic associations between OSA severity (AHI4%) and CSF–Aβ42 on plasma tau, t‐tau/Aβ42, and NfL. Third, OSA–Aβ synergism may occur with or without concurrent p‐tau pathology, suggesting a stepwise process in which OSA increases the likelihood of transitioning from Aβ− to Aβ+, followed by an increased risk of cognitive decline. Finally, AD risk in OSA may also be influenced by comorbid conditions such as hypertension and microvascular disease, which were adjusted for in our analyses, 49 , 50 all of which were adjusted for in our analyses.
Potential mechanisms underlying the combined effects of OSA severity and plasma Aβ42/Aβ40 include sleep fragmentation and intermittent hypoxia, which may promote amyloid accumulation. 51 , 52 , 53 , 54 Disruption of glymphatic clearance, particularly during slow‐wave sleep, is implicated in impaired Aβ removal. 55 In individuals with OSA, slow‐wave sleep is reduced, 56 leading to more accumulation of Aβ and further suppression of the glymphatic system. A combination of OSA and plasma Aβ42/Aβ40 may mirror the synergistic effects of OSA and amyloid pathology that significantly impacts AD stage transition by increasing the risk of AD progression. Chronic intermittent hypoxia, hypercapnia, and hypertension in OSA can also induce neuronal damage, 57 , 58 , 59 suggesting additive neurotoxic mechanisms contributing to AD risk and progression.
ROC analyses did not reveal strong correspondence between plasma and CSF amyloid measures, possibly reflecting assay limitations. Impressive advances in AD plasma assays include the use of ELISA, SIMOA, and liquid chromatography/mass spectrometry (LC/MS). 60 However, Aβ in plasma measured by SIMOA has a relatively low accuracy. 61 Whereas Aβ ELISA immunoassays are highly variable 62 for plasma Aβ42 and/or Aβ42/Aβ40 ratio, LC/MS studies demonstrated lower Aβ42/Aβ40 values from PET Aβ+ individuals 14 , 25 , 61 , 63 and correlated strongly with cortical amyloid PET burden. 19 , 64 , 65 , 66 LC/MS assays for amyloid are also substantially more costly than SIMOA assays. While ELISA remains widely used, chemiluminescent immunoassays (CLEIA or CMIA) are increasingly preferred due to superior sensitivity, wider dynamic range, and automation, offering better low‐level analyte detection and faster turnaround times, increasing reliability for certain diagnostics. ELISA's simplicity keeps it relevant, especially for high‐throughput, less‐sensitive tasks. 67 , 68 Plasma p‐tau181, 217, and 231 can be measured, using available immunoassay platforms, and are more reliable measures of both brain pathology and cognitive status, 17 with 50% to 75% higher mean values for p‐tau with Aβ+ versus Aβ– individuals. 39 , 69 , 70 Future work aims to evaluate broader panels of AD blood‐based biomarkers in relation to OSA.
Strengths of this study include objective cognitive assessment, validated plasma and CSF biomarkers, objective OSA characterization, and robust analytical methods. One major limitation was the lack of an adequate sample size available for CSF–amyloid and tau‐positive individuals, which could directly impact our findings on the synergistic effect of CSF–Aβ42 and OSA severity on plasma Aβ42/Aβ40 levels. Because this was a cross‐sectional study, we could not determine causality as outcome and exposure were measured simultaneously. Our sample mostly consisted of White (83% of the sample) and well‐educated (mean education = 16.9 years) participants, which limits the generalizability of our findings. Another major limitation is that the biomarkers proxy for AD in both plasma and CSF that have been reported to have suboptimal accuracy. CSF Aβ42/Aβ40 outperforms CSF Aβ42 alone, and the Aβ42/Aβ40 ratio in plasma is significantly less accurate than other p‐tau isoforms (181, 217, 231, etc.). These limitations arise due to the extent of information that was available, but efforts were made to account for any effects, thus mitigating their impact on the study results. We acknowledge that the lack of in‐lab polysomnography, the clinical gold standard, is a limitation of our study. However, the validated home sleep apnea testing–derived AHI index aligns with large epidemiologic studies and provides reliable estimates of sleep‐disordered breathing severity.
As the field of sleep medicine moves toward conducting large OSA–AD population‐based trials, identifying at‐risk individuals in the population is critical. Our findings support the use of OSA severity combined with plasma Aβ42/Aβ40 as a feasible, non‐invasive approach for identifying individuals with underlying brain amyloidosis and tau pathology, facilitating efficient screening for future trials.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflicts of interest. Author disclosures are available in the Supporting Information.
CONSENT STATEMENT
All human subjects provided informed consent.
Supporting information
Supporting Information
ACKNOWLEDGMENTS
The authors would like to gratefully acknowledge the dedication, time, and commitment of the research participants and staff of the Aging Research in Sleep Health Equity and Dementia Prevention Program and the Healthy Brain Aging and Sleep Center at NYU Grossman School of Medicine. This work was supported by the National Institutes of Health (NIH/NIA/NHLBI [L30‐AG064670, CIRAD P30AG059303 Pilot, T32HL129953, K23AG068534, R01AG082278, RF1AG083975, R01HL118624, R21AG049348, R21AG055002, R01AG056031, R01AG022374, R01AG066970, R01AG080609, R01AG056531, K07AG05268503, K23HL125939]), Alzheimer's Association grant AARG‐21‐848397, sociation/Michael J. Fox Foundation/CurePSP: SCN‐25‐1474727, and BrightFocus Foundation A2022033S. The funders had no role in the conception or preparation of this manuscript.
Contributor Information
Omonigho Michael Bubu, Email: omonigho.bubu@nyulangone.org.
Ricardo S. Osorio, Email: ricardo.osorio@nyumc.org.
REFERENCES
- 1. Chang WP, Liu ME, Chang WC, et al. Sleep apnea and the risk of dementia: a population‐based 5‐year follow‐up study in Taiwan. PLoS One. 2013;8(10):e78655. doi:10.1371/journal.pone.0078655 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Osorio RS, Gumb T, Pirraglia E, et al. Sleep‐disordered breathing advances cognitive decline in the elderly. Neurology. 2015;84(19):1964‐1971. doi:10.1212/wnl.0000000000001566 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Yaffe K, Laffan AM, Harrison SL, et al. Sleep‐disordered breathing, hypoxia, and risk of mild cognitive impairment and dementia in older women. JAMA. 2011;306(6):613‐619. doi:10.1001/jama.2011.1115 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Shi L, Chen SJ, Ma MY, et al. Sleep disturbances increase the risk of dementia: a systematic review and meta‐analysis. Sleep Med Rev. 2018;40:4‐16. doi:10.1016/j.smrv.2017.06.010 [DOI] [PubMed] [Google Scholar]
- 5. Elias A, Cummins T, Tyrrell R, et al. Risk of Alzheimer's disease in obstructive sleep apnea syndrome: amyloid‐β and tau imaging. J Alzheimers Dis. 2018;66(2):733‐741. doi:10.3233/jad‐180640 [DOI] [PubMed] [Google Scholar]
- 6. Ju YE, Finn MB, Sutphen CL, et al. Obstructive sleep apnea decreases central nervous system‐derived proteins in the cerebrospinal fluid. Ann Neurol. 2016;80(1):154‐159. doi:10.1002/ana.24672 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Liguori C, Mercuri NB, Izzi F, et al. Obstructive sleep apnea is associated with early but possibly modifiable Alzheimer's disease biomarkers changes. Sleep. 2017;40(5):zsx011. doi:10.1093/sleep/zsx011 [DOI] [PubMed] [Google Scholar]
- 8. Osorio RS, Ayappa I, Mantua J, et al. Interaction between sleep‐disordered breathing and apolipoprotein E genotype on cerebrospinal fluid biomarkers for Alzheimer's disease in cognitively normal elderly individuals. Neurobiol Aging. 2014;35(6):1318‐1324. doi:10.1016/j.neurobiolaging.2013.12.030 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Spira AP, Yager C, Brandt J, et al. Objectively measured sleep and β‐amyloid burden in older adults: a pilot study. SAGE Open Med. 2014;2:2050312114546520. doi:10.1177/2050312114546520 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Sharma RA, Varga AW, Bubu OM, et al. Obstructive sleep apnea severity affects amyloid burden in cognitively normal elderly. A longitudinal study. Am J Respir Crit Care Med. 2018;197(7):933‐943. doi:10.1164/rccm.201704‐0704OC [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Bubu OM, Pirraglia E, Andrade AG, et al. Obstructive sleep apnea and longitudinal Alzheimer's disease biomarker changes. Sleep. 2019;42(6):zsz048. doi:10.1093/sleep/zsz048 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Brys M, Pirraglia E, Rich K, et al. Prediction and longitudinal study of CSF biomarkers in mild cognitive impairment. Neurobiol Aging. 2009;30(5):682‐690. doi:10.1016/j.neurobiolaging.2007.08.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Waragai M, Okamura N, Furukawa K, et al. Comparison study of amyloid PET and voxel‐based morphometry analysis in mild cognitive impairment and Alzheimer's disease. J Neurol Sci. 2009;285(1‐2):100‐108. doi:10.1016/j.jns.2009.06.005 [DOI] [PubMed] [Google Scholar]
- 14. Schindler SE, Bollinger JG, Ovod V, et al. High‐precision plasma β‐amyloid 42/40 predicts current and future brain amyloidosis. Neurology. 2019;93(17):e1647‐e1659. doi:10.1212/wnl.0000000000008081 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Palmqvist S, Janelidze S, Stomrud E, et al. Performance of fully automated plasma assays as screening tests for Alzheimer disease‐related β‐amyloid status. JAMA Neurol. 2019;76(9):1060‐1069. doi:10.1001/jamaneurol.2019.1632 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Ashton NJ, Leuzy A, Karikari TK, et al. The validation status of blood biomarkers of amyloid and phospho‐tau assessed with the 5‐phase development framework for AD biomarkers. Eur J Nucl Med Mol Imaging. 2021;48(7):2140‐2156. doi:10.1007/s00259‐021‐05253‐y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Leuzy A, Mattsson‐Carlgren N, Palmqvist S, Janelidze S, Dage JL, Hansson O. Blood‐based biomarkers for Alzheimer's disease. EMBO Mol Med. 2022;14(1):e14408. doi:10.15252/emmm.202114408 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Iaccarino L, Burnham SC, Dell'Agnello G, Dowsett SA, Epelbaum S. Diagnostic biomarkers of amyloid and tau pathology in Alzheimer's disease: an overview of tests for clinical practice in the United States and Europe. J Prev Alzheimer's Dis. 2023;10(3):426‐442. doi:10.14283/jpad.2023.43 [DOI] [PubMed] [Google Scholar]
- 19. Janelidze S, Palmqvist S, Leuzy A, et al. Detecting amyloid positivity in early Alzheimer's disease using combinations of plasma Aβ42/Aβ40 and p‐tau. Alzheimers Dement. 2022;18(2):283‐293. doi:10.1002/alz.12395 [DOI] [PubMed] [Google Scholar]
- 20. Aschenbrenner AJ, Li Y, Henson RL, et al. Comparison of plasma and CSF biomarkers in predicting cognitive decline. Ann Clin Transl Neurol. 2022;9(11):1739‐1751. doi:10.1002/acn3.51670 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Chatterjee P, Goozee K, Sohrabi HR, et al. Association of plasma neurofilament light chain with neocortical amyloid‐β load and cognitive performance in cognitively normal elderly participants. J Alzheimer's Dis. 2018;63(2):479‐487. doi:10.3233/jad‐180025 [DOI] [PubMed] [Google Scholar]
- 22. He L, de Souto Barreto P, Aggarwal G, et al. Plasma Aβ and neurofilament light chain are associated with cognitive and physical function decline in non‐dementia older adults. Alzheimer's Res Ther. 2020;12(1):128. doi:10.1186/s13195‐020‐00697‐0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. He L, Morley JE, Aggarwal G, Nguyen AD, Vellas B, de Souto Barreto P. Plasma neurofilament light chain is associated with cognitive decline in non‐dementia older adults. Sci Rep. 2021;11(1):13394. doi:10.1038/s41598‐021‐91038‐0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Kam K, Jun J, Parekh A, et al. Acute OSA impacts diurnal Alzheimer's biomarkers through nocturnal hypoxemia and state transitions. Am J Respir Crit Care Med. 2022;206(8):1039‐1042. doi:10.1164/rccm.202202‐0262LE [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Bubu OM, Kaur SS, Mbah AK, et al. Obstructive sleep apnea and hypertension with longitudinal amyloid‐β burden and cognitive changes. Am J Respir Crit Care Med. 2022;206(5):632‐636. doi:10.1164/rccm.202201‐0107LE [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Folstein MF, Folstein SE, McHugh PR. “Mini‐mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatry Res. 1975;12(3):189‐198. doi:10.1016/0022‐3956(75)90026‐6 [DOI] [PubMed] [Google Scholar]
- 27. Morris JC. The Clinical Dementia Rating (CDR): current version and scoring rules. Neurology. 1993;43(11):2412‐2414. doi:10.1212/wnl.43.11.2412‐a [DOI] [PubMed] [Google Scholar]
- 28. Yesavage JA, Brink TL, Rose TL, et al. Development and validation of a geriatric depression screening scale: a preliminary report. J Psychiatry Res. 1982;17(1):37‐49. doi:10.1016/0022‐3956(82)90033‐4 [DOI] [PubMed] [Google Scholar]
- 29. Tiihonen P, Hukkanen T, Tuomilehto H, Mervaala E, Töyräs J. Evaluation of a novel ambulatory device for screening of sleep apnea. Telemed J E Health. 2009;15(3):283‐289. doi:10.1089/tmj.2008.0118 [DOI] [PubMed] [Google Scholar]
- 30. Sleep‐related breathing disorders in adults: recommendations for syndrome definition and measurement techniques in clinical research. The Report of an American Academy of Sleep Medicine Task Force. Sleep. 1999;22(5):667‐689. [PubMed] [Google Scholar]
- 31. Spiegel J, Pirraglia E, Osorio RS, et al. Greater specificity for cerebrospinal fluid P‐tau231 over P‐tau181 in the differentiation of healthy controls from Alzheimer's disease. J Alzheimer's Dis. 2016;49(1):93‐100. doi:10.3233/jad‐150167 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Vanderstichele H, Bibl M, Engelborghs S, et al. Standardization of preanalytical aspects of cerebrospinal fluid biomarker testing for Alzheimer's disease diagnosis: a consensus paper from the Alzheimer's Biomarkers Standardization Initiative. Alzheimers Dement. 2012;8(1):65‐73. doi:10.1016/j.jalz.2011.07.004 [DOI] [PubMed] [Google Scholar]
- 33. Liu HC, Chiu MJ, Lin CH, Yang SY. Stability of plasma amyloid‐β 1‐40, amyloid‐β 1‐42, and total tau protein over repeated freeze/thaw cycles. Dement Geriatr Cogn Dis Extra. 2020;10(1):46‐55. doi:10.1159/000506278 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Rózga M, Bittner T, Batrla R, Karl J. Preanalytical sample handling recommendations for Alzheimer's disease plasma biomarkers. Alzheimers Dement (Amst). 2019;11:291‐300. doi:10.1016/j.dadm.2019.02.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Bubu OM, Umasabor‐Bubu OQ, Turner AD, et al. Self‐reported obstructive sleep apnea, amyloid and tau burden, and Alzheimer's disease time‐dependent progression. Alzheimers Dement. 2020;17(2):226‐245. doi:10.1002/alz.12184 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Daniels AJ, McDade E, Llibre‐Guerra JJ, et al. 15 years of longitudinal genetic, clinical, cognitive, imaging, and biochemical measures in DIAN. medRxiv . Published August 9, 2024. doi:10.1101/2024.08.08.24311689 [DOI] [PMC free article] [PubMed]
- 37. Wang HF, Shen XN, Li JQ, et al. Clinical and biomarker trajectories in sporadic Alzheimer's disease: a longitudinal study. Alzheimers Dement (Amst). 2020;12(1):e12095. doi:10.1002/dad2.12095 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Nakamura A, Kaneko N, Villemagne VL, et al. High‐performance plasma amyloid‐β biomarkers for Alzheimer's disease. Nature. 2018;554(7691):249‐254. doi:10.1038/nature25456 [DOI] [PubMed] [Google Scholar]
- 39. Mielke MM, Frank RD, Dage JL, et al. Comparison of plasma phosphorylated tau species with amyloid and tau positron emission tomography, neurodegeneration, vascular pathology, and cognitive outcomes. JAMA Neurol. 2021;78(9):1108‐1117. doi:10.1001/jamaneurol.2021.2293 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Moscoso A, Grothe MJ, Ashton NJ, et al. Longitudinal associations of blood phosphorylated tau181 and neurofilament light chain with neurodegeneration in Alzheimer disease. JAMA Neurol. 2021;78(4):396‐406. doi:10.1001/jamaneurol.2020.4986 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Palmqvist S, Janelidze S, Quiroz YT, et al. Discriminative accuracy of plasma phospho‐tau217 for alzheimer disease vs other neurodegenerative disorders. JAMA. 2020;324(8):772‐781. doi:10.1001/jama.2020.12134 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Nabizadeh F, Balabandian M, Rostami MR, Kankam SB, Ranjbaran F, Pourhamzeh M. Plasma neurofilament light levels correlate with white matter damage prior to Alzheimer's disease: results from ADNI. Aging Clin Exp Res. 2022;34(10):2363‐2372. doi:10.1007/s40520‐022‐02095‐y [DOI] [PubMed] [Google Scholar]
- 43. Rauchmann BS, Schneider‐Axmann T, Perneczky R. Associations of longitudinal plasma p‐tau181 and NfL with tau‐PET, Aβ‐PET and cognition. J Neurol Neurosurgery Psychiatry. 2021;92(12):1289‐1295. doi:10.1136/jnnp‐2020‐325537 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Mattsson N, Cullen NC, Andreasson U, Zetterberg H, Blennow K. Association between longitudinal plasma neurofilament light and neurodegeneration in patients with Alzheimer disease. JAMA Neurol. 2019;76(7):791‐799. doi:10.1001/jamaneurol.2019.0765 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Beydoun MA, Noren Hooten N, Beydoun HA, et al. Plasma neurofilament light as a potential biomarker for cognitive decline in a longitudinal study of middle‐aged urban adults. Transl Psychiatry. 2021;11(1):436. doi:10.1038/s41398‐021‐01563‐9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. West T, Kirmess KM, Meyer MR, et al. A blood‐based diagnostic test incorporating plasma Aβ42/40 ratio, ApoE proteotype, and age accurately identifies brain amyloid status: findings from a multi cohort validity analysis. Mol Neurodegener. 2021;16(1):30. doi:10.1186/s13024‐021‐00451‐6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Verberk IMW, Slot RE, Verfaillie SCJ, et al. Plasma amyloid as prescreened for the earliest Alzheimer pathological changes. Ann Neurol. 2018;84(5):648‐658. doi:10.1002/ana.25334 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Nakamura T, Kawarabayashi T, Seino Y, et al. Aging and APOE‐ε4 are determinative factors of plasma Aβ42 levels. Ann Clin Transl Neurol. 2018;5(10):1184‐1191. doi:10.1002/acn3.635 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Esiri MM, Nagy Z, Smith MZ, Barnetson L, Smith AD. Cerebrovascular disease and threshold for dementia in the early stages of Alzheimer's disease. Lancet. 1999;354(9182):919‐920. [DOI] [PubMed] [Google Scholar]
- 50. Snowdon DA, Greiner LH, Mortimer JA, Riley KP, Greiner PA, Markesbery WR. Brain infarction and the clinical expression of Alzheimer disease. The Nun Study. JAMA. 1997;277(10):813‐817. [PubMed] [Google Scholar]
- 51. Shiota S, Takekawa H, Matsumoto SE, et al. Chronic intermittent hypoxia/reoxygenation facilitate amyloid‐beta generation in mice. Journal of Alzheimer's disease : JAD. 2013;37(2):325‐333. [DOI] [PubMed] [Google Scholar]
- 52. Zhang X, Zhou K, Wang R, et al. Hypoxia‐inducible factor 1alpha (HIF‐1alpha)‐mediated hypoxia increases BACE1 expression and beta‐amyloid generation. J Biol Chem. 2007;282(15):10873‐10880. [DOI] [PubMed] [Google Scholar]
- 53. Lim AS, Kowgier M, Yu L, Buchman AS, Bennett DA. Sleep fragmentation and the risk of incident Alzheimer's disease and cognitive decline in older persons. Sleep. 2013;36(7):1027‐1032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54. Schlosser Covell GE, Dhawan PS, Lee Iannotti JK, et al. Disrupted daytime activity and altered sleep‐wake patterns may predict transition to mild cognitive impairment or dementia: a critically appraised topic. Neurologist. 2012;18(6):426‐429. [DOI] [PubMed] [Google Scholar]
- 55. Cui W, Duan Z, Li Z, Feng J. Assessment of Alzheimer's disease‐related biomarkers in patients with obstructive sleep apnea: a systematic review and meta‐analysis. Front Aging Neurosci. 2022;14:902408. doi:10.3389/fnagi.2022.902408 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56. Ovod V, Ramsey KN, Mawuenyega KG, et al. Amyloid β concentrations and stable isotope labeling kinetics of human plasma specific to central nervous system amyloidosis. Alzheimers Dement. 2017;13(8):841‐849. doi:10.1016/j.jalz.2017.06.2266 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57. Ludemann P, Dziewas R, Soros P, Happe S, Frese A. Axonal polyneuropathy in obstructive sleep apnoea. J Neurol Neurosurg Psychiatry. 2001;70(5):685‐687. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58. Kiernan TE, Capampangan DJ, Hickey MG, Pearce LA, Aguilar MI. Sleep apnea and white matter disease in hypertensive patients: a case series. Neurologist. 2011;17(5):289‐291. [DOI] [PubMed] [Google Scholar]
- 59. Kumar R, Chavez AS, Macey PM, Woo MA, Yan‐Go FL, Harper RM. Altered global and regional brain mean diffusivity in patients with obstructive sleep apnea. J Neurosci Res. 2012;90(10):2043‐2052. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 60. Alawode DOT, Heslegrave AJ, Ashton NJ, et al. Transitioning from cerebrospinal fluid to blood tests to facilitate diagnosis and disease monitoring in Alzheimer's disease. J Intern Med. 2021;290(3):583‐601. doi:10.1111/joim.13332 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61. Janelidze S, Teunissen CE, Zetterberg H, et al. Head‐to‐head comparison of 8 plasma amyloid‐β 42/40 assays in Alzheimer disease. JAMA Neurol. 2021;78(11):1375‐1382. doi:10.1001/jamaneurol.2021.3180 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 62. Toledo JB, Shaw LM, Trojanowski JQ. Plasma amyloid beta measurements—a desired but elusive Alzheimer's disease biomarker. Alzheimers Res Ther. 2013;5(2):8. doi:10.1186/alzrt162 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 63. Verberk IMW, Thijssen E, Koelewijn J, et al. Combination of plasma amyloid beta((1‐42/1‐40)) and glial fibrillary acidic protein strongly associates with cerebral amyloid pathology. Alzheimers Res Ther. 2020;12(1):118. doi:10.1186/s13195‐020‐00682‐7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 64. Doecke JD, Pérez‐Grijalba V, Fandos N, et al. Total Aβ(42)/Aβ(40) ratio in plasma predicts amyloid‐PET status, independent of clinical AD diagnosis. Neurology. 2020;94(15):e1580‐e1591. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 65. Li Y, Schindler SE, Bollinger JG, et al. Validation of plasma amyloid‐β 42/40 for detecting Alzheimer disease amyloid plaques. Neurology. 2022;98(7):e688‐e699. doi:10.1212/wnl.0000000000013211 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 66. Pereira JB, Janelidze S, Stomrud E, et al. Plasma markers predict changes in amyloid, tau, atrophy and cognition in non‐demented subjects. Brain. 2021;144(9):2826‐2836. doi:10.1093/brain/awab163 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 67. Machado KG, Bertelli V, Frassini R, Streck AF, Ely MR. CLIA immunoassay as an alternative and accurate method to detect Sars‐Cov‐2 antigen compared to ELISA. Diagn Microbiol Infect Dis. 2025;112(4):116828. doi:10.1016/j.diagmicrobio.2025.116828 [DOI] [PubMed] [Google Scholar]
- 68. Tanaka R, Takemura M, Sato M, et al. Comparison of chemiluminescence enzyme immunoassay (CLEIA) with ELISA for the determination of anti‐cyclic citrullinated peptide antibodies. Clin Chim Acta. 2010;411(1‐2):22‐25. doi:10.1016/j.cca.2009.09.032 [DOI] [PubMed] [Google Scholar]
- 69. Janelidze S, Berron D, Smith R, et al. Associations of plasma phospho‐tau217 levels with tau positron emission tomography in early Alzheimer disease. JAMA Neurol. 2021;78(2):149‐156. doi:10.1001/jamaneurol.2020.4201 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 70. Mattsson‐Carlgren N, Janelidze S, Bateman RJ, et al. Soluble P‐tau217 reflects amyloid and tau pathology and mediates the association of amyloid with tau. EMBO Mol Med. 2021;13(6):e14022. doi:10.15252/emmm.202114022 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting Information