Endothelial Dysfunction and C-Reactive Protein in Relation with the Severity of Obstructive Sleep Apnea Syndrome

Abstract

Study Objectives:

To investigate flow-mediated dilatation (FMD) and C-reactive protein (CRP) levels in patients with obstructive sleep apnea syndrome (OSAS) in relation with the severity of respiratory disturbances and hypoxemia.

Design:

After subjects had completed nocturnal polysomnography, FMD was measured in the brachial artery, and blood samples were obtained to determine serum CRP levels.

Setting:

Sleep laboratory in Seoul National University Bundang Hospital.

Patients:

Ninety men: 22 normal controls, 28 subjects with mild to moderate OSAS, and 40 with severe OSAS.

Measurements and Results:

FMD was found to be correlated with oxygen desaturation index (ODI), percentage of time below 90% O₂ saturation, average O₂ saturation, lowest O₂ saturation, systolic blood pressure, apnea hypopnea index (AHI), and body mass index. In addition, CRP was correlated with body mass index, waist-to-hip ratio, neck circumference, diastolic pressure, average O₂ saturation and percentage of time below 90% O₂ saturation but not with AHI. Stepwise multiple regression showed that the ODI was a significant determinant of FMD (adjusted R² = 10%, β = −0.33, P < 0.01). In addition, body mass index (β = 0.25, P < 0.05) and waist-to-hip ratio (β = 0.21, P < 0.05) were found to be significantly correlated with CRP (adjusted R² = 12%, P < 0.05), independently of other factors. There was no correlation between FMD and CRP.

Conclusion:

As a marker of nocturnal hypoxemia, ODI rather than AHI might better explain the relationship between OSAS and FMD. Because body mass index and waist-to-hip ratio were identified as risk factors of high serum CRP in OSAS, obesity should be considered when predicting cardiovascular complications in OSAS.

Citation:

Chung S; Yoon IY; Shin YK; Lee CH; Kim JW; Lee T; DJ Choi; HJ Ahn. Endothelial dysfunction and c-reactive protein in relation with the severity of obstructive sleep apnea syndrome. SLEEP 2007;30(8):997-1001.

OBSTRUCTIVE SLEEP APNEA SYNDROME (OSAS) IS A COMMON DISEASE THAT OCCURS IN ABOUT 2% TO 4% OF THE GENERAL POPULATION.^1,2 PATIENTS WITH OSAS suffer from hypoxemia and hypercapnia during sleep caused by repetitive nocturnal respiratory pauses. Moreover, nocturnal hypoxemia is considered to be associated with cardiovascular and cerebrovascular disease,^3,4 although the association has not been fully explained. The relation between OSAS and vascular complications is supported by a high prevalence of OSAS in patients who have experienced a recent stroke or a transient ischemic attack^5,6 and in those with coronary heart disease.⁷

Much attention has been paid to endothelial dysfunction, as opposed to the other various contributors to vascular diseases. Endothelial dysfunction, an early marker of atherosclerosis, develops as a result of arterial endothelial wall injury. Flow-mediated dilatation (FMD) has been used as a noninvasive method to measure endothelial dysfunction. Ip et al showed that FMD is significantly correlated with the apnea-hypopnea index (AHI) in patients with OSAS.⁸ However, conflicting results have been presented by other studies. Kato et al failed to demonstrate a significant difference in FMD between OSAS patients and obese control subjects, although vasodilation was found to be blunted in response to acetylcholine in OSAS patients.⁹ In a large community sample of older adults, a significant linear association was found between the percentage of sleep time with an oxygen saturation below 90% and baseline arterial diameter rather than FMD.¹⁰ Thus, the relationship between FMD and OSAS still requires elucidation.

C-reactive protein (CRP) is an important risk factor in atherosclerosis¹¹ and in other cardiovascular diseases.^12,13 Several studies have reported an increased CRP plasma level in OSAS. Repeated hypoxia in OSAS patients might induce proinflammatory cytokines such as interleukin-6 (IL-6) and tumor necrosis factor-α, which could elevate CRP. However, Guilleminault et al demonstrated no relationship between OSA and CRP in relatively less obese OSA patients¹⁴ and, rather, concluded that only body mass index (BMI) was significantly associated with a high CRP.

Because both FMD and CRP are involved in the development of atherosclerosis, we reasoned that it would be interesting to measure FMD and CRP in OSAS patients to investigate the pathophysiology of cardiovascular complications. In the present study, we aimed (1) to study FMD and CRP in relation with OSAS severity and (2) to identify those variables that most importantly explain FMD and CRP changes in OSAS.

METHODS

Subjects

We enrolled 90 men, aged 19 to 59 years (42.6 ± 9.0; mean ± SD) who were referred to a sleep laboratory in Seoul National University Bundang Hospital to undergo nocturnal polysomnography. Subjects aged older than 60 years, those suffering from inflammatory diseases, and those taking antihypertensives or antihyperlipidemic or diabetes medication were excluded. The study protocol was approved by our Institutional Review Board, and all subjects provided written informed consents. Twenty-two normal control subjects (AHI ≥ 5), 28 subjects with a mild to moderate degree of OSAS (AHI > 5 and AHI ≥ 30), and 40 with a severe degree (AHI > 30) were enrolled in this study.

Polysomnography

We used an Embla™ N 7000 (Embla, Reykjavik, Iceland) and standard electrodes and sensors. Electroencephalography electrodes were applied at C3/A2, O1/A2, and O2/A1, and 2 electrooculography electrodes were applied at the sides of both eyes to record horizontal and vertical eye movements. Submental electromyography electrodes were applied at submentalis muscles, and the electromyograms of both anterior tibialis muscles recorded limb movements during sleep. Strain gauges were used for recording chest and abdominal respiratory movements, and nasal pressure cannulas were used to record airflow. Arterial oxygen saturation was measured using pulse oximeters applied on index fingers. Based on the criteria of Rechtschaffen and Kales,¹⁵ we scored every 30-second epoch of the nocturnal polysomnogram. Apnea was defined as complete cessation of airflow for at least 10 seconds. Hypopnea was defined as a substantial reduction in airflow (> 50%) for at least 10 seconds or a moderate reduction in airflow for at least 10 seconds associated with electroencephalographic arousals or oxygen desaturation (≥ 4%).¹⁶ AHI was defined as the total number of apneas and hypopneas per hour of sleep, and oxygen desaturation index (ODI) was calculated as the number of oxygen desaturations (≥ 4%) per hour of sleep.

FMD Measurement

After nocturnal polysomnography, brachial artery FMD was measured using an Accuvix XQ ultrasound unit (Medison, Seoul, Korea) equipped with a 6–12MHz linear-array transducer and lowerarm occlusion technique. All subjects fasted for at least 8 hours prior to FMD measurement. The baseline diameter of the brachial artery was measured from the anterior intima to the posterior intima. A blood pressure cuff was placed on the lower part of the arm and inflated to 250 mm Hg for 5 minutes. The blood pressure cuff was then released, and brachial artery diameters were measured 3 times at 40, 60, and 80 seconds after release by investigators unaware of subject details. After recording the maximum value of 3 measurements, FMD was calculated as the ratio of change in diameter (maximum - baseline) over baseline value.

Inflammatory markers

Peripheral blood was drawn from patients with OSAS after nocturnal polysomnography. Serum samples were stored at −70°C until required for CRP determinations. High-sensitive CRP was measured by turbidimetric immunoassay using the Vitros 5.1 FS Chemistry System (Ortho-Clinical Diagnostics, Raritan, NJ). The limit of quantification for CRP was 0.01 mg/dL.

Statistical analysis

SPSS Ver. 12.0 for Windows (SPSS Inc, Chicago, IL) was used for the statistical analysis. Results are presented as means ± SD. The Kolmogorov-Smirnov test was used to confirm normality. Differences in clinical variables among the 3 groups were assessed by analysis of variance, with the Tukey posthoc test for parametric variables and the Kruskal-Wallis test for nonparametric variables. CRP levels were transformed into logarithms (logCRP) because of a skewed distribution. Spearman correlation coefficients were calculated to determine relationships between FMD or CRP and clinical variables, and stepwise multiple regression was used to identify factors that contributed to FMD and logCRP. The significance criterion was defined to be P < 0.05 for 2-tailed tests.

RESULTS

Baseline characteristics of the subjects in this study are presented in Table 1. There were no significant differences among 3 groups in age, BMI, neck circumference, waist-to-hip ratio, lipid profiles, or glucose. Significant difference was observed in FMD among 3 groups. The FMD in the severe OSA group was significantly lower than that in the normal control group (P < 0.05), but there was no difference in FMD between the mild to moderate OSAS group and the normal control group (P = 0.70) or between the groups with severe OSAS and mild to moderate OSA (P = 0.11). Serum levels of CRP were not different among the 3 groups. Correlation coefficients of clinical variables are shown in Table 2 and Figure 1. FMD was found to be correlated with BMI (r = −0.21, P < 0.05), systolic blood pressure (r = −0.22, P < 0.05), AHI (r = −0.31, P < 0.01), average O₂ saturation (r = 0.25, P < 0.05), lowest O₂ saturation (r = 0.35, P < 0.01), the percentage of time below 90% O₂ saturation (r = −0.41, P < 0.01), and ODI (r = −0.35, P < 0.01). Also, CRP was correlated with BMI (r = 0.33, P < 0.01), neck circumference (r = 0.29, P < 0.01), waist-to-hip ratio (r = 0.22, P < 0.05), diastolic blood pressure (r = 0.22, P < 0.05), average O₂ saturation (r = −0.25, P < 0.05) and percentage of time below 90% O₂ saturation (r = 0.21, P < 0.05) but not with AHI. FMD was not correlated with CRP (r = −0.06, P = 0.59). Neither lipid profiles nor glucose were significantly correlated with FMD or CRP. Age was not correlated with FMD or CRP.

Table 1.

Baseline Characteristics of Patients with OSAS and Normal Controls

Variables	Normal control n = 22	Mild to moderate OSAS n = 28	Severe OSAS n = 40	P Value
Age, y	42.1 ± 8.7	41.9 ± 9.6	43.3 ± 8.8	0.78
Body mass index, kg/m2	26.2 ± 3.9	26.3 ± 4.1	26.6 ± 2.6	0.88
Neck circumference, cm	38.4 ± 2.6	38.6 ± 2.9	39.4 ± 1.9	0.23
Waist-to-hip ratio	0.92 ± 0.05	0.94 ± 0.04	0.95 ± 0.04	0.14
Blood pressure, mm Hg
Systolica,b	121.8 ± 7.0	127.9 ± 11.6	128.8 ± 7.0	< 0.01
Diastolicb	78.3 ± 6.1	86.9 ± 11.6	84.9 ± 8.2	< 0.01
Average O2 saturationb,c	95.9 ± 1.9	95.6 ± 1.3	93.9 ± 2.2	< 0.01
% of time < 90% O2 saturationa–c	0.02 ± 0.05	1.9 ± 2.2	10.1 ± 12.7	< 0.01
Oxygen desaturation indexa–c	1.1 ± 0.8	11.2 ± 5.3	41.1 ± 18.2	< 0.01
Flow-mediated dilatation, %b	8.1 ± 1.7	7.6 ± 2.7	6.5 ± 2.0	0.02
Cholesterol, mg/dL	205.4 ± 33.0	215.9 ± 52.4	214.3 ± 42.7	0.65
Triglyceride, mg/dL	173.9 ± 89.7	239.6 ± 249.9	189.3 ± 116.1	0.32
HDL cholesterol, mg/dL	45.7 ± 12.8	49.7 ± 12.1	51.5 ±10.5	0.18
Glucose, mg/dL	101.6 ± 22.6	102.9 ± 21.9	107.3 ± 16.7	0.48
C-reactive protein, mg/dL	0.063 ± 0.083	0.115 ± 0.128	0.120 ± 0.146	0.25

Data are presented as mean ± SD.

^aP < 0.05, normal control vs mild to moderate obstructive sleep apnea syndrome (OSAS)

^bP < 0.05, normal control vs severe OSAS

^cP < 0.01, mild to moderate OSAS vs severe OSAS

Table 2.

Correlation Coefficients Between Clinical Variables and FMD or CRP

	FMD	CRP
Age	−0.16	−0.05
Body mass index	−0.21a	0.33b
Neck circumference	−0.13	0.29b
Waist-to-hip ratio	0.07	0.22a
Blood pressure, mm Hg
Systolic	−0.22a	0.21
Diastolic	−0.10	0.22a
Apnea-hypopnea index	−0.31b	0.12
Average O2 saturation	0.25a	−0.25a
Lowest O2 saturation	0.35b	−0.13
Percentage of time below 90% O2 saturation	−0.41b	0.21a
Oxygen desaturation index	−0.35b	0.18
Glucose	−0.01	0.14
Total cholesterol	−0.13	0.05
HDL cholesterol	0.06	−0.18
Triglyceride	0.08	0.21

FMD refers to flow-mediated vascular dilatation; CRP, C-reactive protein; HDL, high-density lipoprotein.

^aP < 0.05

^bP < 0.01

Figure 1.

Correlations between (A) flow-mediated vascular dilatation (FMD) and oxygen desaturation index (ODI) and between (B) C-reactive protein (CRP) and body mass index (BMI).

Stepwise multiple regression was performed to identify factors that contributed to FMD and CRP (Table 3). ODI was a significant determinant of FMD (adjusted R² = 10%, β = −0.33, P < 0.01). BMI (β = 0.25, P < 0.05) and waist-to-hip ratio (β = 0.21, P < 0.05) were significant variables to determine CRP (adjusted R² = 12%, P < 0.05).

Table 3.

Stepwise Multiple Linear Regression Model of FMD and logCRP

Variables	FMDa		logCRPb
	β	P Value	β	P Value
Age	−0.17	0.09	−0.07	0.52
BMI	−0.13	0.21	0.25	0.02
Neck circumference	0.05	0.64	0.14	0.27
Waist-to-hip ratio	0.14	0.18	0.21	0.04
Systolic blood pressure	−0.13	0.21	0.17	0.10
Diastolic blood pressure	−0.09	0.37	0.19	0.07
AHI	0.37	0.21	0.07	0.53
% of time < 90% O2 saturation	−0.06	0.67	−0.04	0.71
ODI	−0.33	< 0.01	0.09	0.37

FMD refers to flow-mediated vascular dilatation; CRP, C-reactive protein; BMI, body mass index; AHI, apnea hypopnea index; ODI, oxygen desaturation index.

^aFMD, adjusted R²= 10%

^blogCRP, adjusted R²= 12%

DISCUSSION

The current study shows that FMD is significantly lower in patients with severe OSA compared with normal controls and that hypoxemia most determines FMD. FMD was found to be correlated with ODI and AHI, but ODI was the only significant variable correlated with FMD by multivariate analysis. In previous studies, endothelial dysfunction, measured by FMD, was found to be associated with the degree of nocturnal hypoxemia¹⁷ or AHI and hypoxemia index in OSAS.¹⁰ Also, endothelial dysfunction has been reported to be improved by treatment with nasal continuous positive airway pressure.^8,18 Endothelium-dependent vasodilation is mediated by nitric oxide, which is the most potent vascular relaxant and an intracellular signaling molecule. Nitric oxide is formed from L-arginine by nitric oxide synthases in endothelium. Because oxygen tension is recognized to regulate the gene expression of nitric oxide synthases,¹⁹ hypoxia may alter nitric oxide biosynthesis, which is an oxygen-dependent process.²⁰ Systemic nitric oxide production has been reported to be impaired in patients with OSAS, and oxygen administration significantly increases serum nitric oxide levels.²¹ Moreover, diminished nitric oxide availability can lead to a prothrombotic and proinflammatory condition.¹⁹ Thus, recurrent hypoxemia may be responsible for triggering endothelial dysfunction in OSAS.

Normally, arterial endothelium contributes to the control of vascular tone and has anticoagulant, antiplatelet, and fibrinolytic properties.²² Arterial endothelial wall injury and resultant endothelial dysfunction might induce atherosclerosis in OSAS. The impairment of endothelium-dependent vasodilatation occurs prior to the atherosclerosis, and atherogenic risk factors, such as hypertension, diabetes mellitus, and dyslipidemia, may impair endothelium-dependent vasodilatation and contribute to cardiovascular disease. In addition, repeated hypoxia and reoxygenation during sleep in OSAS may induce oxidative stress by inducing the production of reactive oxygen species, which play an important role in the development of cardiovascular disease²³ by injuring the vascular endothelium.²⁴ Thus, it may be possible to use antioxidants in patients with OSAS to improve endothelial dysfunction.²⁵

In the present study, a correlation was found between CRP and hypoxemia in OSAS, but multiple regression analysis showed that only BMI and waist-to-hip ratio were associated with a high serum CRP level. CRP is recognized as an important risk factor for the development of atherosclerosis, and CRP elevation is attributed to the increased production of IL-6, which has been found to be correlated with the risk of future cardiovascular disease.²⁶ Moreover, levels of CRP and IL-6 have been reported to be elevated in patients with OSAS^27,28 and may be reduced by treatment with nocturnal continuous positive airway pressure.²⁸ However, these results were not reproduced by the present study. Guilleminault et al reported that RDI is not significantly correlated with CRP, and, in their study, BMI was the only significant variable associated with logCRP by multiple regression analysis.¹⁴ We also found an association between BMI and CRP. Furthermore, the independent relationship between waist-to-hip ratio and CRP could be explained by a report that an elevated CRP is associated with abdominal obesity,^29,30 which can be simply measured by the waist-to-hip ratio.³¹ Adipose tissue is known to play an important role in metabolism and may induce lowgrade systemic inflammation by releasing IL-6, which stimulates CRP production.³⁰ Because subjects in the present study were not morbidly obese but were mildly overweight with a mean BMI of 26.4, we could differentiate the effect of OSAS on serum CRP from the effect of obesity. Asian patients with OSAS have been reported to be less obese than are Caucasian patients.³²

No association was found between FMD and CRP in patients with OSAS in the present study. Verma et al found no relationship between FMD and CRP in healthy subjects³³ but found such an association in patients with metabolic syndrome or cardiovascular disease.^34,35 To our knowledge, no report has been issued on this association in patients with OSAS. Because most of the patients enrolled in the current study did not have overt cardiovascular complications, it is reasonable that we found no correlation between FMD and CRP.

Women with OSAS were excluded from the present study to avoid potential sex-associated endothelial function differences.³⁶ Estrogen might improve endothelial function by inducing nitric oxide bioavailability³⁷ or by enhancing antioxidant activity.³⁸ We also excluded patients with OSAS who were older than 60 years. Aging itself may be an important modulator of endothelial function, as aging has been reported to be associated with a progressive reduction in nitric oxide availability.³⁹ This change in the L-arginine-nitric oxide pathway appears to be related to increases in the activities of nitric oxide synthase inhibitors, such as asymmetric dimethyl-L-arginine.⁴⁰ However, we included OSAS patients with hypertension in the analysis, although endothelial dysfunction has been associated with hypertension.^41,42 Hypertension might occur secondarily to OSAS, and the effect of hypertension on FMD and CRP could be controlled by multivariate analysis.

The subjects enrolled in this study were not evenly distributed in terms of high proportion of patients with severe OSAS, and this constitutes a study limitation, although relatively large numbers of normal controls and OSAS patients were included. In addition, most associations between variables found in the present study were of modest strength. Better matching and more-balanced sampling are required to confirm our findings.

In conclusion, FMD was found to be significantly associated with the severity of OSAS, and the ODI rather than AHI is a more significant variable to explain FMD changes. Moreover, CRP was not found to be associated with OSAS severity, but BMI and waist-to-hip ratio were identified as significant determinants of CRP.