New-Onset Sleep Disorders before Cardiac Surgery May Indicate an Increased Risk of Postoperative Atrial Fibrillation

Abstract

Purpose: We aimed to determine if sleep disorders before cardiac surgery indicate an increased risk of postoperative atrial fibrillation (POAF).

Methods: In this study, 238 patients undergoing coronary artery bypass grafting in our center were included. Patients were separated into the preoperative sleep disorder group and the control group. The primary endpoint was the incidence of POAF, and the secondary endpoints were the incidence of postoperative stroke, duration of invasive ventilation, length of intensive care unit, and hospitalization stay. Propensity score matching and multivariable logistic regression were used for adjusting potential confounders.

Results: A total of 165 (69.3%) patients had sleep disorders before surgery, and 73 well-matched pairs were generated. A higher incidence of POAF was found in the preoperative sleep disorder group (16.4% versus 5.5%, p = 0.034). In multivariable logistic regression, preoperative sleep disorders were correlated to a higher risk of POAF (odds ratio = 4.627, 95% confidence interval: 1.181–18.123, p = 0.028). In the subgroup of patients without long-term sleep disorders, those who experienced preoperative sleep disorders had a higher incidence of POAF (16.1% versus 4.3%, p = 0.024), meanwhile, no difference was found in the subgroup of long-term sleep disorders.

Conclusion: New-onset sleep disorders before cardiac surgery may indicate a higher incidence of POAF.

Introduction

Postoperative atrial fibrillation (POAF) is prevalent in cardiac surgery, affecting approximately 20% to 40% of patients.¹⁾ The occurrence of POAF has been associated with hemodynamic destabilization, higher risk of postoperative stroke, prolonged hospital and intensive care unit (ICU) stays, and increased medical costs.²⁾

The potential pathogenesis of POAF is complex and incompletely elucidated, including inflammation, myocardial ischemia, and autonomic nervous system activity, and in most cases, it occurs in the presence of existing atrial substrates susceptible to the onset and maintenance of atrial fibrillation such as structural remodelling and electrical remodelling.^3–13)

The association between autonomic nervous system activity and POAF has been shown by several studies,^14–16) but has rarely been measured in clinical practice to assess a patient’s risk of POAF. A correlation between stress and changes in the autonomic nervous system has also been shown.^17–19) Patients undergoing cardiac surgery are often exposed to tremendous mental stress and are more likely to experience persistent sympathetic hyperactivity and acute sleep disturbances before operation,^20,21) resulting in a high rate of perioperative use of benzodiazepines. In the current study, we sought to determine the feasibility of evaluating the occurrence of POAF by the presence of preoperative sleep disorders.

Material and Methods

Study population

Patients undergoing nonemergent, isolated, on-pump coronary artery bypass grafting (CABG) between May 2018 and November 2022 in our center were enrolled in the present study and those with preoperative atrial fibrillation were excluded. Patients were grouped based on the presence of preoperative sleep disorders: those diagnosed with a sleep disorder by the complaint and received benzodiazepines the night before surgery were included in the preoperative sleep disorder group, and patients who did not complain of sleep disturbances or ask for preoperative benzodiazepines served as the control group. In further subgroup analysis for POAF, patients were stratified based on the presence of long-term sleep disorders and β-blocker use.

Baseline characteristics and outcomes

Baseline characteristics, including preoperative demographics, comorbidities, other cardiac medications, echocardiography and laboratory test results, and intraoperative and postoperative variables, were compared between groups. The primary endpoint was the incidence of POAF, and the secondary endpoints included the incidence of postoperative stroke, duration of invasive ventilation, length of ICU, and hospitalization stay. Evaluations of long-term sleep disorders were performed through a nursing assessment at admission. The occurrence of POAF was identified by examining the ICU and medical records retrospectively. Postoperative stroke was confirmed by computed tomography scanning.

Statistical analysis

One-to-one propensity score matching (nearest-neighbor matching, no replacement) was applied to achieve a balanced exposure between groups at baseline. Continuous variables were presented as median (interquartile range) or mean ± standard deviation, and differences were compared by Mann–Whitney U test or Student’s t test, according to the type of distribution of the data. All categorical variables were expressed in terms of number (percentage), and the χ² test and Fisher’s exact test were utilized for the test of between-group differences. Logistic regression was utilized to estimate the correlation between study treatment and POAF; all covariates were introduced into univariable regression, and covariates with p <0.1 were introduced into multivariable regression. We used IBM SPSS Statistics Version 26.0 (IBM Corp, Armonk, NY, USA) for statistical analysis, and differences were considered significant at the level of 0.05 (two tailed).

Results

Baseline characteristics

A total of 238 patients who underwent nonemergent, isolated, on-pump CABG were included in the study. Among them, there were 165 (69.3%) patients in the preoperative sleep disorder group and 73 (30.7%) patients in the control group. A higher rate of long-term sleep disorders (15.2% versus 5.5%, p = 0.035) was observed in the preoperative sleep disorder group, and the residual pump blood transfusion and the cross-clamp time were different between groups before matching. After propensity score matching, 73 pairs of patients with comparable baseline characteristics were generated. The baseline characteristics before and after matching are shown in Tables 1 and 2, respectively.

Table 1. Baseline characteristics of the preoperative sleep disorder and control groups before propensity score matching.

	Sleep disorder (n = 165)	Control (n = 73)	p
Preoperative characteristics
Age (year)	62.0 (53.0–66.0)	60.0 (53.0–67.0)	0.911
Male	146 (88.5%)	64 (87.7%)	0.857
BMI (kg/m2)	24.9 ± 3.1	25.4 ± 3.1	0.210
Smoke	105 (63.6%)	44 (60.3%)	0.621
Alcohol consumption	94 (57.0%)	41 (56.2%)	0.908
Long-term sleep disorders	25 (15.2%)	4 (5.5%)	0.035*
Hypertension	97 (58.8%)	42 (57.5%)	0.856
Diabetes mellitus	55 (33.3%)	24 (32.9%)	0.945
Previous MI	43 (26.1%)	19 (26.0%)	0.996
Previous PCI	15 (9.1%)	9 (12.3%)	0.444
Previous stroke	10 (6.1%)	4 (5.5%)	1.000
Dyslipidemia	98 (59.4%)	42 (57.5%)	0.788
Statins	144 (87.3%)	57 (78.1%)	0.071
β-blocker	106 (64.2%)	38 (52.1%)	0.076
RAAS-blocker	63 (38.2%)	32 (43.8%)	0.411
LVEF (%)	62.0 (52.0–67.0)	63.0 (56.5–68.0)	0.288
LA (mm)	36.0 (32.0–40.0)	36.0 (33.0–38.0)	0.875
LV (mm)	48.0 (46.0–43.0)	49.0 (46.5–51.5)	0.876
Hb (g/L)	137.4 ± 15.6	139.7 ± 15.0	0.288
NT-proBNP (ng/mL)	168.0 (69.5–475.0)	189 (73.5–546.0)	0.624
Cr (μmol/L)	81.0 (72.5–92.5)	83.0 (75.5–94.5)	0.475
Intraoperative characteristics
Bypass vessel	3.0 (3.0–3.0)	3.0 (3.0–3.0)	0.496
Artery bypass vessel	1.0 (1.0–1.0)	1.0 (1.0–1.0)	0.235
RBC transfusion (unit)	0 (0–0)	0 (0–0)	0.375
Plasma transfusion (mL)	0 (0–0)	0 (0–0)	0.944
PLT transfusion (unit)	1.0 (0–1.0)	1.0 (0–1.0)	0.501
Residual pump blood (mL)	400.0 (300.0–400.0)	400.0 (400.0–550.0)	0.036*
Autotransfusion (mL)	300.0 (250.0–500.0)	300.0 (250.0–500.0)	0.843
Rethoracotomy	6 (3.6%)	2 (2.7%)	1.000
CPB time (min)	120.0 (101.5–136.5)	123.0 (108.5–145.5)	0.266
Cross-clamp time (min)	76.0 (64.0–95.0)	83.0 (71.0–100.0)	0.042*
Postoperative characteristics
Vasoactive drug use	2.0 (1.0–3.0)	2.0 (1.0–3.0)	0.967

*p value <0.05. BMI: body mass index; MI: myocardial infarction; PCI: percutaneous coronary intervention; RAAS-blocker: renin–angiotensin–aldosterone system blocker; LVEF: left ventricular ejection fraction; LA: left atrium; LV: left ventricle; Hb: hemoglobin; NT-proBNP: N-terminal pro-brain natriuretic peptide; Cr: serum creatinine; RBC: red blood cell; PLT: platelet; CPB: cardiopulmonary bypass

Table 2. Baseline characteristics of the preoperative sleep disorder and control groups after propensity score matching.

	Sleep disorder (n = 73)	Control (n = 73)	p
Preoperative characteristics
Age (year)	61.0 (53.0–67.0)	60.0 (53.0–67.0)	0.997
Male	64 (87.7%)	64 (87.7%)	1.000
BMI (kg/m2)	25.7 ± 3.1	25.4 ± 3.1	0.541
Smoke	44 (60.3%)	43 (58.9%)	1.000
Alcohol consumption	37 (50.7%)	41 (56.2%)	0.507
Long-term sleep disorders	11 (15.1%)	4 (5.5%)	0.056
Hypertension	41 (56.2%)	42 (57.5%)	0.867
Diabetes mellitus	23 (31.5%)	24 (32.9%)	0.859
Previous MI	16 (21.9%)	19 (26.0%)	0.561
Previous PCI	10 (13.7%)	9 (12.3%)	0.806
Previous stroke	3 (4.1%)	4 (5.5%)	1.000
Dyslipidemia	44 (60.3%)	42 (57.5%)	0.737
Statins	56 (76.7%)	57 (78.1%)	0.843
β-blocker	29 (39.7%)	38 (52.1%)	0.135
RAAS-blocker	35 (47.9%)	32 (43.8%)	0.618
LVEF (%)	64.0 (58.0–69.5)	63.0 (56.5–68.0)	0.439
LA (mm)	36.0 (32.0–39.5)	36.0 (33.0–38.0)	0.831
LV (mm)	48.0 (46.0–51.0)	49.0 (46.5–51.5)	0.277
Hb (g/L)	140.9 ± 17.1	139.7 ± 15.0	0.667
NT-proBNP (ng/mL)	140.0 (66.0–435.5)	189 (73.5–546.0)	0.464
Cr (μmol/L)	86.0 (72.0–99.5)	83.0 (75.5–94.5)	0.637
Intraoperative characteristics
Bypass vessel	3.0 (3.0–3.0)	3.0 (3.0–3.0)	0.901
Artery bypass vessel	1.0 (1.0–1.0)	1.0 (1.0–1.0)	0.997
RBC transfusion (unit)	0 (0–0)	0 (0–0)	0.540
Plasma transfusion (mL)	0 (0–0)	0 (0–0)	0.392
PLT transfusion (unit)	1.0 (0–1.0)	1.0 (0–1.0)	0.665
Residual pump blood (mL)	400.0 (400.0–450.0)	400.0 (400.0–550.0)	0.821
Autotransfusion (mL)	300.0 (250.0–500.0)	300.0 (250.0–500.0)	0.788
Rethoracotomy	4 (5.5%)	2 (2.7%)	0.677
CPB time (min)	120.0 (104.0–142.5)	123.0 (108.5–145.5)	0.726
Cross-clamp time (min)	84.0 (64.5–105.0)	83.0 (71.0–100.0)	0.625
Postoperative characteristics
Vasoactive drug use	2.0 (1.0–3.0)	2.0 (1.0–3.0)	0.924

BMI: body mass index; MI: myocardial infarction; PCI: percutaneous coronary intervention; RAAS-blocker: renin–angiotensin–aldosterone system blocker; LVEF: left ventricular ejection fraction; LA: left atrium; LV: left ventricle; Hb: hemoglobin; NT-proBNP: N-terminal pro-brain natriuretic peptide; Cr: serum creatinine; RBC: red blood cell; PLT: platelet; CPB: cardiopulmonary bypass

Outcomes

Among these patients, one in-hospital death due to gastrointestinal bleeding from the preoperative sleep disorder group was observed. In the comparison of the preoperative sleep disorder group and the control group (Table 3), the preoperative sleep disorder group had an increased incidence of POAF (16.4% versus 5.5%, p = 0.034), and the incidence of postoperative stroke, duration of invasive ventilation, and length of ICU and hospitalization stay were not significantly different. The results of multivariable logistic regression are shown in Fig. 1. In the multivariable logistic regression, preoperative sleep disorder was correlated to a higher risk of POAF (odds ratio = 4.627, 95% confidence interval: 1.181–18.123, p = 0.028). In further subgroup analysis (Table 4), the occurrence of preoperative sleep disorders was associated with a higher incidence of POAF in patients without long-term sleep disorders, but not in patients with long-term sleep disorders. A significantly higher incidence of POAF associated with preoperative sleep disorders was observed in patients who did not receive concomitant β-blockers, but not in patients who received β-blockers.

Table 3. Outcomes of the preoperative sleep disorder and control groups.

	Sleep disorder (n = 73)	Control (n = 73)	p
POAF	12 (16.4%)	4 (5.5%)	0.034*
Postoperative stroke	6 (8.2%)	1 (1.4%)	0.121
Invasive ventilation (min)	1165.0 (897.5–2020.5)	1175.0 (870.0–1514.5)	0.488
Length of ICU stay (min)	3757.0 (2494.0–5522.5)	2767.0 (2557.5–4787.0)	0.399
Length of hospital stay (day)	11.0 (10.0–16.0)	12.0 (9.0–15.0)	0.467

*p value <0.05. POAF: postoperative atrial fibrillation; ICU: intensive care unit

Fig. 1. Multivariable logistic regression of POAF. *p value <0.05. POAF: postoperative atrial fibrillation; LVEF: left ventricular ejection fraction; CPB: cardiopulmonary bypass; OR: odds ratio; CI: confidence interval.

Table 4. Subgroup analysis for POAF.

	Sleep disorder (n = 73)	Control (n = 73)	p
Long-term sleep disorders
Yes	2/11 (18.2%)	1/4 (25.0%)	1.000
No	10/62 (16.1%)	3/69 (4.3%)	0.024*
β-blocker use
Yes	3/29 (10.3%)	3/38 (7.9%)	1.000
No	9/44 (20.5%)	1/35 (2.9%)	0.046*

*p value <0.05. POAF: postoperative atrial fibrillation

Discussion

Activation of the autonomic nervous system has been proven to be one of the underlying mechanisms of occurrence of POAF¹⁾ but is rarely measured in clinical practice to assess a patient’s risk of POAF. Preoperative sleep disorders are common in patients undergoing cardiac surgery and are usually related to stressors, a change in sleep environment or timing, etc.,²²⁾ which may indicate the continuous sympathetic hyperactivity, and has the potential to be an effective indicator to simply assess the status of the autonomic nervous system.

In our study, patients with preoperative sleep disorders seemed to have an increased risk of POAF than those without preoperative sleep disorders. In further analysis, the incidence rates of POAF were significantly different in the subgroup of patients without long-term sleep disorders, but not in the subgroup of patients with long-term sleep disorders. Furthermore, the simultaneous use of β-blockers may eliminate the difference between groups.

Autonomic nervous system activity has been shown to be associated with the occurrence of POAF in previous studies, and is a potential pathogenesis of sleep disorders.²³⁾ In the study of Lai and colleagues,²⁴⁾ an animal model of stress-mediated insomnia was established using case exchange, and it was found that prolonged sleep latency was accompanied by decreased parasympathetic activity and increased sympathetic activity. Novoa et al. found that clinical hypnosis using the Ericksonian technique could lower the risk of POAF in patients undergoing CABG and was hypothesized to be associated with changes in autonomic nervous system activity.²⁵⁾ Existing evidences have suggested that autonomic nervous system activity is an important component in the pathogenesis of sleep disorders.

The association between POAF and preoperative sleep disorders seemed to be only suitable for patients without long-term sleep disorders, perhaps because abnormal autonomic nervous system activity constitutes the main pathogenesis of new-onset preoperative sleep disorders, while long-term sleep disorders have a more complex pathogenesis.^22,26)

β-blockers exert antiarrhythmic effects by blocking β-adrenergic receptors to reduce cardiac sympathetic tone and its mediated triggered activity, conduction velocity, and myocardial oxygen consumption burden and improve ventricular remodelling.²⁷⁾ In the present study, the difference in the incidence of POAF was eliminated in the subgroup of β-blocker use, suggesting that the use of β-blockers may effectively prevent the occurrence of POAF in patients exposed to stress or anxiety.

From the findings of this work, we hypothesize that the occurrence of preoperative new-onset sleep disorders is closely related to autonomic nervous system activity and may serve as a simple and feasible indicator of an increased incidence of POAF.

Our study was subject to some restrictions. First, the sample size was limited and the conclusions should be verified by large-scale prospective studies in the future. Second, only a single disease was studied in the current study, and the results were not representative of other populations. Moreover, we were not able to directly measure the state of the activation of the autonomic nervous system of patients owing to the retrospective nature of the study.

Conclusion

New-onset sleep disorders before cardiac surgery may indicate a higher incidence of POAF; the usage of β-blockers may prevent the increased risk of POAF.

Ethics Approval and Informed Consent

The present study complies with the guidelines for human studies. The research was conducted ethically in accordance with the Declaration of Helsinki. This study was approved by the ethics review board of West China Hospital, Sichuan University (No. 2022-1274, August 24, 2022). Patients were given an opt-out participant information and written informed consent was waived with the approval of the review board.

Data Availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Disclosure Statement

The authors have no relevant financial or non-financial interests to disclose.