Reduction in circulating endothelial progenitor cells caused by obstructive sleep apnea–related hypoxemia and its association with the severity of coronary artery disease

Abstract

Background

Obstructive sleep apnea (OSA) is a risk factor for coronary artery disease (CAD), while endothelial progenitor cells (EPCs) are critical for vascular repair. This study investigated the associations among OSA-related hypoxemia, circulating EPC, and CAD severity.

Methods

This prospective study enrolled patients with unstable angina undergoing coronary angiography. All participants underwent overnight polysomnography to determine the apnea–hypopnea index (AHI) and the sleep apnea–specific hypoxic burden (SASHB). Circulating EPCs were quantified using flow cytometry. CAD severity was assessed via angiography, with a Gensini score greater than or equal to 22 defining severe CAD and the presence of greater than or equal to 2 major vessels with greater than or equal to 50% diameter stenosis defining multivessel CAD.

Results

Among 80 included patients [median age 59 years; 53 (66.3%) male], 42 (52.5%) had OSA (AHI ≥ 15 events/h). Patients with high SASHB exhibited more severe coronary artery lesions than those with low SASHB (Gensini: 33.0 vs. 19.5; P = 0.040). Multivariable linear regression confirmed log10-transformed SASHB as an independent predictor of reduced circulating EPC levels (count: standardized β = –0.37, P = 0.002; percentage: standardized β = –0.40, P < 0.001). Multivariable logistic regression analysis revealed low EPC count [odds ratio (OR) = 3.41, 95% confidence interval (CI): 1.21–9.58, P = 0.020] and low EPC percentage (OR = 2.94, 95% CI: 1.00–8.78, P = 0.049) as independent risk factors for multivessel CAD.

Conclusion

OSA may promote CAD progression by depleting EPCs and hindering vascular repair. Incorporating hypoxemia metrics and EPC levels into risk assessment could help identify patients with OSA-related CAD.

Introduction

Coronary artery disease (CAD), characterized by occlusion of the coronary arteries, remains a global health concern. Obstructive sleep apnea (OSA), marked by intermittent upper airway collapse, is an independent risk factor for CAD [1], with its associated chronic intermittent hypoxemia rather than the frequency of occlusive events, considered the principal predictor of OSA-related cardiovascular injury [2]. Although continuous positive airway pressure (CPAP) reduces respiratory events and alleviates hypoxemia, large clinical trials have not demonstrated that it significantly improves cardiovascular outcomes [3,4]. Moreover, the precise mechanisms by which OSA induces coronary injury remain unclear, hindering the development of targeted therapies for OSA-related CAD.

Endothelial injury initiates atherosclerosis and is pivotal in CAD progression [5,6]. Endothelial progenitor cells (EPCs) are essential for maintaining endothelial homeostasis. Upon endothelial damage, multiple cytokines and growth factors induce EPCs mobilization from the bone marrow into the circulation. These cells migrate to sites of injury and promote repair via paracrine release of vasoactive factors and by differentiating to replace damaged endothelial cells, thereby attenuating atherogenesis [7,8]. While evidence indicates that reduced circulating EPC levels are associated with poor prognosis in patients with CAD [9], to our knowledge, few studies have systematically examined the interrelationships among OSA, EPCs, and coronary impairment.

Therefore, this study recruited patients undergoing coronary angiography to investigate the associations among OSA-related hypoxemia indices, circulating EPC levels, and the invasively assessed severity of coronary artery lesions, to elucidate potential mechanisms of OSA-induced coronary injury and provide new perspectives for CAD prevention and treatment.

Methods

Study population

This study consecutively enrolled patients with typical symptoms of unstable angina admitted to the Department of Cardiology at Beijing Anzhen Hospital and scheduled to undergo coronary angiography between February 2023 and December 2024. The inclusion criteria were (a) clinical stability, defined as stable hemodynamics and daytime oxygen saturation on room air, making them suitable for overnight polysomnography; (b) diagnosis of unstable angina, defined by myocardial ischemic symptoms without biomarker evidence of myocardial necrosis [10]; and (c) availability of research personnel and equipment. The exclusion criteria were (a) diagnosed with central sleep apnea; (b) acute myocardial infarction confirmed by cardiac enzyme levels and electrocardiography findings; (c) prior coronary revascularization, including percutaneous coronary intervention or coronary artery bypass grafting; (d) active infection; (e) severe insomnia, defined as an Insomnia Severity Index score greater than or equal to 22; (f) end-stage chronic diseases, including New York Heart Association class IV heart failure, advanced malignancy, or dialysis-dependent renal failure; (g) cognitive impairment or acute psychiatric episodes; (h) use of CPAP or oral appliance therapy within 6 months before recruitment; and (i) inability to suspend nocturnal oxygen therapy during sleep monitoring. The study protocol was approved by the Ethics Committee of Beijing Anzhen Hospital (approval no. KS2022037) and was registered with the Chinese Clinical Trial Registry (ChiCTR2300067991) on 2 February 2023. Written informed consent was obtained from all patients.

Polysomnography

All patients underwent an 8-h overnight type II polysomnography (Alice PDx; Philips, Murrysville, Pennsylvania, USA) on the night before coronary angiography. The recorded signals included electroencephalography, electromyography, electrooculography, electrocardiography, nasal pressure airflow, oronasal thermistor signals, snoring, thoracoabdominal movements, and pulse oximetry. Two registered sleep technologists, blinded to clinical data, manually scored the recordings according to the American Academy of Sleep Medicine Manual for the Scoring of Sleep and Associated Events, version 3.0 [11]. Apnea was defined as a greater than or equal to 90% reduction in airflow from baseline for greater than or equal to 10 s, with obstructive apneas characterized by persistent respiratory effort, central apneas by the absence of respiratory effort throughout the event, and mixed apneas by an absence of effort during the initial portion of the event followed by its resumption later. Hypopnea was defined as a greater than or equal to 30% reduction in airflow for greater than or equal to 10 s, associated with a greater than or equal to 3% oxygen desaturation or an arousal, and hypopneas were scored as obstructive if snoring, inspiratory flow limitation, or thoracoabdominal paradox was present during the event, and otherwise as central. The apnea–hypopnea index (AHI) was calculated as the total number of apneas and hypopneas per hour of sleep. OSA was defined as an AHI greater than or equal to 15 events/h, with greater than or equal to 50% of events being obstructive. Central sleep apnea was defined as a central AHI greater than or equal to 5 events/h, with more than 50% of events being central in origin [12]. Sleep apnea–specific hypoxic burden (SASHB) was computed with a C++ program based on the algorithm developed by Azarbarzin et al. [13,14]. This metric integrates the duration, depth, and frequency of respiratory event–related oxygen desaturation and was defined as the total area under the desaturation curve divided by the total sleep time. The percentage of sleep time with oxygen saturation less than 90% (T90SpO₂) was extracted directly from the polysomnography report. All patients completed the Epworth Sleepiness Scale and the Insomnia Severity Index questionnaire before polysomnography [15,16].

Endothelial progenitor cells quantification

Fasting peripheral blood was collected into EDTA tubes and analyzed within 4 h. An aliquot of 50 µl anticoagulated blood was then transferred into a Trucount Absolute Counting tube (BD Biosciences, San Jose, California, USA) and incubated with 20 µl of PerCP-cy5.5-conjugated anti-human CD45 (clone 2D1; BD Biosciences), 5 µl of APC-cy7-conjugated anti-human CD34 (clone 581; BioLegend, San Diego, California, USA) and 20 µl of Alexa Fluor 647-conjugated anti-human KDR (clone 89106; BD Biosciences) in the dark for 15 min. BD FACS^TM Lysing Solution (BD Biosciences) was used as the lysing buffer. To minimize cell loss, no wash steps were performed. Flow cytometric acquisition was performed on an LSRFortessa flow cytometer (BD Biosciences). Data were analyzed using BD FACSDiva software (BD Biosciences). First, CD45⁺ was gated and then assayed for expression of CD34 and KDR in the lymphocyte gate. EPCs were defined as CD45^dim, CD34⁺, and KDR⁺ cells. EPC count (cells/ml) was calculated as follows: (number of CD45^dimCD34⁺KDR⁺ events/ number of TruCount bead events) × (number of beads in tube/sample volume in μl) × 1000. The percentage of EPCs among CD45⁺ cells was also calculated and reported as CD45^dimCD34⁺KDR⁺ (% of CD45⁺).

Coronary lesion assessment

Two interventional cardiologists, blinded to polysomnography and EPC data, independently evaluated the coronary angiograms. Multivessel CAD was defined as greater than or equal to 50% diameter stenosis in greater than or equal to 2 major epicardial coronary vessels (left main, left anterior descending, left circumflex, or right coronary artery) [10], with significant branch vessel disease attributed to the parent vessel (e.g. diagonal branch to the left anterior descending). Overall, coronary lesion severity was quantified using the Gensini score, which incorporates both stenosis severity and segment weighting [17]; higher scores indicate more severe disease. A Gensini score greater than or equal to 22 (the cohort median) was used to define severe CAD.

Statistical analysis

Continuous variables that failed the Shapiro–Wilk normality test were presented as median (interquartile range). Categorical variables were summarized as frequencies (percentages). Group differences were assessed using the Mann–Whitney U test for continuous variables and the χ² test for categorical variables. Before correlation and linear regression analyses, nonnormally distributed variables (AHI, T90SpO₂, SASHB) were log10-transformed (log₁₀). For T90SpO₂, which included zero values, the transformation was 1og₁₀ (x + 0.01). The patients were dichotomized based on median values of T90SpO₂ (1%) and SASHB (15 min%/h) to reflect hypoxemia severity. Receiver operating characteristic curves were used to determine optimal cutoff values for EPC levels in predicting multivessel CAD and severe CAD; EPC levels were then dichotomized accordingly. Spearman rank correlations were used to assess the relationships between sleep parameters (AHI, T90SpO₂, SASHB), circulating EPC levels (count and percentage), and Gensini score. Independent pairwise associations among OSA-related indices, EPC measures, and CAD metrics were investigated using multivariable linear (for continuous outcomes) and logistic (for categorical outcomes) regression, adjusted for demographics (age and sex), traditional cardiovascular risk factors (hypertension history, glycated hemoglobin, and low-density lipoprotein) [18], and use of angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers (ACEI/ARB). Statistical analyses were performed using JMP Pro 16.1.0 (SAS Institute, Cary, North Carolina, USA), and a two-sided P value less than 0.05 was considered statistically significant.

Results

Characteristics of the study participants

A total of 80 patients with CAD were enrolled (53 males, 27 females; median age 59 years). Of these, 42 (52.5%) met the diagnostic criteria for OSA. Coronary angiography revealed multivessel disease in 42 (52.5%) patients, while 41 (51.3%) patients were classified as having severe CAD according to a Gensini score greater than or equal to 22. Compared with those without OSA, patients with OSA had higher BMI and glycated hemoglobin levels, a higher prevalence of hypertension, more frequent use of ACEI/ARB, and more severe nocturnal hypoxemia. No other significant differences in clinical characteristics were observed between groups (Table 1 and Supplementary Table S1, Supplemental digital content 1, https://links.lww.com/MCA/A792).

Table 1.

Characteristics of the study population (N = 80)

Basic characteristics	All (N = 80)	OSA (N = 42) (AHI ≥ 15 events/h)	Non-OSA (N = 38) (AHI < 15 events/h)	P value
Demographics
Age (year)	59 (52–67)	63 (53–72)	58 (51–65)	0.071
Male, n (%)	53 (66.3%)	25 (59.5%)	28 (73.7%)	0.181
BMI (kg/m²)	26.4 (24.7–28.5)	27.4 (25.9–29.0)	24.9 (22.9–27.2)	0.002
Neck circumference (cm)	40.0 (37.3–41.5)	40.0 (38.0–42.3)	40.0 (36.0–41.0)	0.358
Medical history
Hypertension, n (%)	52 (65.0%)	35 (83.3%)	17 (44.7%)	<0.001
Diabetes, n (%)	30 (37.5%)	15 (35.7%)	15 (39.5%)	0.729
Stroke, n (%)	4 (5.0%)	2 (4.8%)	2 (5.3%)	0.918
Smoking, n (%)	33 (41.2%)	17 (40.5%)	16 (42.1%)	0.883
Laboratory parameters
Glycated hemoglobin (%)	6.0 (5.7–7.1)	6.2 (5.9–7.6)	5.9 (5.4–6.3)	0.009
Total cholesterol (mmol/l)	3.9 (3.4–4.7)	4.0 (3.5–4.7)	3.8 (3.4–4.62)	0.631
Low-density lipoprotein (mmol/l)	2.1 (1.7–2.8)	2.1 (1.7–2.9)	2.0 (1.6–2.8)	0.996
White blood cell count (10⁹/L)	6.1 (5.1–7.0)	6.2 (5.1–7.5)	6.1 (5.1–6.8)	0.660
High-sensitivity C-reactive protein (mg/l)	1.0 (0.5–1.9)	1.1 (0.7–2.7)	0.8 (0.4–1.2)	0.065
Left ventricular ejection fraction (%)	63 (61–65)	63 (61–65)	64 (61–65)	0.327
Polysomnography
AHI (events/h)	15.7 (6.7–24.4)	23.6 (18.2–33.3)	6.7 (3.5–11.7)	<0.001
T90SpO2 (%)	0.6 (0–3.9)	3.1 (0.5–8.9)	0.1 (0–0.6)	<0.001
SASHB (%min/h)	14.9 (2.4–29.3)	27.5 (17.1–56.7)	2.3 (0.6–10.3)	<0.001
Epworth Sleepiness Scale (score)	6 (3–9)	6 (4–9)	6 (3–10)	0.497
Insomnia Severity Index (score)	4 (1–10)	6 (1–11)	3 (1–8)	0.338

AHI, apnea–hypopnea index; OSA, obstructive sleep apnea; SASHB, sleep apnea–specific hypoxic burden; T90SpO₂, percentage of sleep time with oxygen saturation less than 90%.

Associations between obstructive sleep apnea–related metrics and coronary artery disease

As shown in Fig. 1, log₁₀(AHI) was not significantly correlated with Gensini score (r = 0.197, P = 0.080; Fig. 1a), and Gensini score did not differ significantly between OSA and non-OSA groups (28.0 vs. 20.0, P = 0.367; Fig. 1b). In contrast, log₁₀(T90SpO₂) was significantly positively correlated with Gensini score (r = 0.279, P = 0.012; Fig. 1c). Moreover, multivariable linear regression confirmed log₁₀(T90SpO₂) as an independent predictor of higher Gensini score (β = 7.414, P = 0.027). Patients with high T90SpO₂ (≥1%) had significantly higher Gensini scores than those with low T90SpO₂ (<1%) (36.0 vs. 19.5, P = 0.008; Fig. 1d). Log₁₀(SASHB) showed a positive trend with Gensini score (r = 0.185, P = 0.101; Fig. 1e), and patients with high SASHB (≥15 min%/h) had higher Gensini scores than those with low SASHB (<15 min%/h) (33.0 vs. 19.5, P = 0.040; Fig. 1f). However, the prevalence of multivessel CAD did not differ significantly between patients with OSA, high T90SpO₂, or high SASHB and their respective counterparts.

Fig. 1.

Association of OSA metrics with Gensini score. (a, c, and e) Spearman’s rank-correlation scatterplots of log₁₀(AHI), log₁₀(T90SpO₂), and log₁₀(SASHB) with the Gensini score, respectively. (b, d, and f) Comparisons of Gensini scores between OSA and non-OSA, high and low T90SpO₂, and high and low SASHB groups, respectively. AHI, apnea–hypopnea index; log₁₀: log10-transformed; OSA, obstructive sleep apnea; SASHB, sleep apnea–specific hypoxic burden; T90SpO₂, percentage of sleep time with oxygen saturation less than 90%.

Associations between obstructive sleep apnea–related metrics and circulating endothelial progenitor cell levels

Log₁₀(AHI) and log₁₀(T90SpO₂) showed no significant linear correlations with either EPC count or percentage (Fig. 2a–d). Log₁₀(SASHB) showed a negative trend with EPC count (r = –0.145, P = 0.199; Fig. 2e) and a significantly negative correlation with EPC percentage (r = –0.231, P = 0.039; Fig. 2f). As shown in Table 2, multivariable linear regression analysis further confirmed the independent negative association of log₁₀(SASHB) with EPC levels [count: standardized β (βstd) = –0.37, P = 0.002; percentage: βstd = –0.40, P < 0.001].

Fig. 2.

Correlations between OSA metrics and circulating EPC levels. (a, c, and e) Spearman’s rank-correlation scatterplots between log₁₀(AHI), log₁₀(T90SpO₂), and log₁₀(SASHB) and the EPC count (CD45^dimCD34⁺KDR⁺, cells/ml). (b, d, and f) The corresponding correlations with the EPC percentage (% of CD45⁺). AHI, apnea–hypopnea index; EPC, endothelial progenitor cell; log₁₀, log10-transformed; OSA, obstructive sleep apnea; SASHB, sleep apnea–specific hypoxic burden; T90SpO₂, percentage of sleep time with oxygen saturation less than 90%.

Table 2.

Multivariable linear regression analysis of log₁₀ sleep apnea–specific hypoxic burden and circulating endothelial progenitor cell levels

	Univariate linear regression			Multivariable linear regression model 1a			Multivariable linear regression model 2b
Variables	β	βstd	P value	β	βstd	P value	β	βstd	P value
EPC count
log10SASHB	–258.8	–0.33	0.003	–258.0	–0.33	0.005	–292.4	–0.37	0.002
Age				–1.6	–0.03	0.828	–4.455	–0.07	0.572
Male				–46.5	–0.07	0.570	–45.20	–0.06	0.589
Hypertension							119.1	0.17	0.211
Glycated hemoglobin							50.93	0.11	0.320
Low-density lipoprotein							–164.3	–0.21	0.058
ACEI/ARB							–73.81	–0.10	0.414
EPC percentage
log10SASHB	–0.005	–0.36	<0.001	–0.005	–0.36	0.001	–0.006	–0.40	<0.001
Age				0	–0.03	0.777	0	–0.08	0.520
Male				–0.002	–0.16	0.168	–0.002	–0.17	0.146
Hypertension							0.002	0.15	0.267
Glycated hemoglobin							0	–0.01	0.909
Low-density lipoprotein							–0.002	–0.17	0.126
ACEI/ARB							0	–0.02	0.899

^aModel 1 adjusted for age and sex.

^bModel 2 Adjusted for age, sex, hypertension history, glycated hemoglobin, low-density lipoprotein and use of ACEI/ARB.

ACEI/ARB, angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers; βstd, standardized β; log₁₀, log10-transformed; EPC, endothelial progenitor cell; SASHB, sleep apnea–specific hypoxic burden.

Relationship between circulating endothelial progenitor cell levels and coronary artery disease

Receiver operating characteristic curve analysis was used to identify the optimal EPC cutoffs to predict multivessel CAD (count: 897 cells/ml, area under the curve = 0.595; percentage: 0.019%, area under the curve = 0.578). Multivariable logistic regression based on these cutoffs revealed that patients with low EPC count had a 3.41-fold higher risk of multivessel CAD (odds ratio [OR] = 3.41, 95% confidence interval [CI]: 1.21–9.58, P = 0.020). Similarly, patients with a low EPC percentage had a 2.94-fold higher risk of multivessel CAD (OR = 2.94, 95% CI: 1.00–8.78, P = 0.049; Fig. 3).

Fig. 3.

Multivariable logistic regression analysis of circulating EPC levels and CAD severity. *Univariate logistic regression model. ^†Model 1, adjusted for age and sex. ^‡Model 2, adjusted for age, sex, hypertension history, glycated hemoglobin, low-density lipoprotein, and use of angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers. CAD, coronary artery disease; CI, confidence interval; EPC, endothelial progenitor cell; OR, odds ratio.

Discussion

The results of this study revealed a close association among OSA-related hypoxemia, reduced circulating EPC levels, and CAD severity. The hypoxemia indices SASHB and T90SpO₂ were significantly correlated with the Gensini score, whereas AHI, the conventional OSA measure, was not. Moreover, SASHB was the only OSA metric independently associated with reduced circulating EPC levels, and lower EPC levels were predictive of an increased risk of multivessel CAD. These findings suggest that OSA-related hypoxemia may promote coronary lesion progression by impairing endothelial repair capacity (manifested as EPC depletion), thereby providing new insights into the pathophysiology of OSA-induced coronary injury. Furthermore, our study implies the potential clinical value of integrating hypoxemia severity and EPC levels into CAD risk stratification models.

The lack of association between AHI and CAD severity in this cohort, in contrast to the strong predictive value of hypoxemia metrics (T90SpO₂, SASHB), aligns with prior evidence that hypoxemia metrics better reflect OSA-related cardiovascular damage than AHI [19,20]. Hypoxemia is the principal pathophysiologic basis of coronary damage in OSA [21], whereas AHI merely indexes the frequency of respiratory events and does not reflect hypoxemia severity or pattern. In contrast, T90SpO₂ reflects cumulative exposure to severe hypoxemia, and SASHB integrates the ‘dose’ of hypoxemia (depth × duration) and event frequency, more comprehensively reflecting hypoxemic injury. This supports a phenotype-guided approach, suggesting that patients with high hypoxic burden (rather than merely high AHI) may have a greater risk of cardiovascular diseases and benefit more from interventions such as CPAP. Therefore, future therapeutic strategies should go beyond reducing AHI to include comprehensive mitigation of nocturnal hypoxemia.

EPCs are bone marrow–derived precursor cells characterized by high proliferative potential. Acute vascular injury, such as myocardial infarction, triggers EPC mobilization from the bone marrow via ischemia and cytokine release, leading to a transient elevation in circulating EPC levels that can last for up to 2 months [8,22]. By contrast, in chronic coronary impairment without total occlusion, sustained cardiovascular risk factors such as hypertension, diabetes, and hyperlipidemia contribute to EPC depletion and dysfunction by enhancing oxidative stress, fostering chronic inflammation, reducing nitric oxide bioavailability, and disrupting the bone marrow microenvironment [23,24]. Treating these traditional risk factors by implementing strategies including statin and antihypertensive administration, improved glycemic control, and increased exercise can partially restore EPC levels [25–30]. In the present study, SASHB remained an independent predictor of EPC reduction after adjusting for multiple factors, suggesting that repeated, severe, cumulative hypoxemia in OSA may directly impair EPC homeostasis and endothelial repair. We also noticed that patients with OSA had higher serum concentrations of high-sensitivity C-reactive protein, implying the potential association between OSA-induced inflammation and EPC profile. Studies with small sample sizes have reported no significant improvement in EPC levels after three months of CPAP in patients with moderate-to-severe OSA, but these studies did not adequately report control of traditional risk factors and assessed OSA severity primarily by AHI [31,32]. Future studies should determine whether effective treatment (e.g. CPAP), under optimal traditional risk factor control, can improve EPC quantity and biological function in patients with OSA, particularly in those with severe nocturnal desaturation.

In the present study, we observed that reduced EPC levels were significantly associated with increased risk of multivessel CAD, providing clinical evidence for the ‘OSA-related hypoxemia → impaired endothelial repair → coronary lesions’ pathway (Fig. 4). Unlike patients with acute myocardial infarction, in whom EPC levels may rise reactively during the acute phase [33], the patients in the present study with unstable angina may have experienced chronic endothelial injury, potentially leading to excessive EPC mobilization and eventual depletion. Noticeably, the association between EPC reduction and multivessel CAD, but not Gensini score, suggests that EPC depletion likely contributes more to diffuse, multivessel involvement, whereas focal stenosis may be more influenced by lipid accumulation and plaque burden.

Fig. 4.

Proposed mechanism linking OSA to multivessel CAD. OSA-induced chronic intermittent hypoxia, quantified as SASHB, leads to endothelial injury and EPC reduction both directly and through the mediation of oxidative stress, chronic inflammation, and reduced nitric oxide bioavailability. The resulting impairment in vascular repair capacity contributes to the progression of diffuse, multivessel CAD. CAD, coronary artery disease; EPC, endothelial progenitor cell; OSA, obstructive sleep apnea; SASHB, sleep apnea–specific hypoxic burden.

The value of EPC-based therapies has been demonstrated in cardiovascular diseases, although the efficacy and reliability remain controversial [34]. EPC transplantation and EPC-capturing stents have not consistently yielded clinical benefit in trials [35,36], possibly because of EPC heterogeneity, source variability, and preparation methods [37]. Moreover, persistent cardiovascular risk factors (e.g. hyperglycemia, hypertension, and OSA-related hypoxemia) may create an unfavorable ‘soil’ for transplanted EPCs (‘seed-soil’ mismatch), leading to rapid cell exhaustion or dysfunction and limiting therapeutic efficacy. Given the high prevalence and underdiagnosis of OSA worldwide [38], several incognito CAD risk factors, such as high SASHB, remain unaddressed, limiting the efficacy of EPC-based therapies despite optimal management of traditional risk factors. Future EPC-oriented treatments should incorporate screening for OSA phenotypes to optimize EPC survival and biological function.

Limitations

This study has certain limitations. First, it is a single-center study with a modest sample size, which reduces the statistical power and necessitates caution in extrapolating the findings to broader populations. Second, while flow cytometry was used to quantify EPC number, it did not evaluate functional capacity like migration, adhesion, proliferation, and colony formation, all of which are essential for vascular repair. Third, the cross-sectional design precludes the ability to assess EPC levels as predictors of long-term cardiovascular outcomes. Therefore, larger longitudinal studies are warranted to comprehensively evaluate EPC quantity and function, confirm their prognostic value, and investigate whether an improved EPC profile benefits coronary repair.

Conclusion

OSA-related hypoxemia, typically measured by SASHB, rather than OSA frequency assessed by AHI, is closely associated with lower circulating EPC levels and more severe CAD. EPC depletion independently increases the risk of multivessel CAD, suggesting that EPC reduction may be a mechanistic link between hypoxemia and diffuse coronary injury. These findings provide new evidence for the mechanisms underlying endothelial repair impairment in OSA-induced coronary injury, supporting the inclusion of hypoxemia metrics and EPC levels in risk assessments of the incidence and progression of OSA-related CAD.

Acknowledgements

This research was financially supported by the National Natural Science Foundation of the People’s Republic of China, through grants 82270099, with the funders remaining uninvolved in the research design and conduct.

Conceptualization: J.X. and Y.L. Methodology: J.X., Y.L., and J.Z. Software and visualization: J.Z. Validation and supervision: J.X. and J.Z. Formal analysis and investigation: H.Z. Resources, project administration, and project administration: J.X. Data curation: H.Z., Y.J., S.L., H.W., and X.X. Writing – original draft preparation: H.Z. and Y.L. Writing – review and editing: Y.J., S.L., H.W., X.X., and J.X. All authors have read and agreed to the published version of the manuscript. In addition, they commit to being accountable for all aspects of the work, ensuring that any questions regarding its accuracy or integrity are properly investigated and resolved.

The study was conducted in accordance with the principles of the Declaration of Helsinki and approved by the Ethics Committee of Beijing Anzhen Hospital (approval no. KS2022037). Written informed consent was obtained from all subjects involved in the study.

The raw data supporting the conclusions of this article will be made available by the corresponding author on request.

Conflicts of interest

There are no conflicts of interest.

Supplementary Material

cad-37-385-s001.docx ^{(20KB, docx)}