Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Jun 29.
Published in final edited form as: Int J Cardiovasc Imaging. 2022 Oct 31;39(3):621–630. doi: 10.1007/s10554-022-02743-4

Sleep apnea and carotid atherosclerosis in the Multi-Ethnic Study of Atherosclerosis (MESA): leveraging state-of-the-art vascular imaging

Neomi Shah 1, Michelle Reid 2, Venkatesh Mani 3, Vaishnavi Kundel 1, Robert C Kaplan 4,5, Jorge R Kizer 6, Zahi A Fayad 3, Steven Shea 7, Susan Redline 2
PMCID: PMC10309069  NIHMSID: NIHMS1909419  PMID: 36316593

Abstract

Purpose

To further characterize the relationship between obstructive sleep apnea (OSA) and carotid atherosclerosis, we examined the structural and metabolic features of carotid plaque using hybrid 18-F-fluorodeoxyglucose (FDG) Positron Emission Tomography/Magnetic Resonance Imaging (PET/MRI) in the Multi-Ethnic Study of Atherosclerosis (MESA).

Methods

We studied 46 individuals from the MESA-PET and MESA-Sleep ancillary studies. OSA was defined as an apnea hypopnea index [AHI] ≥ 15 events per hour (4% desaturation). PET/MRI was used to measure carotid plaque inflammation (using target-to-background-ratios [TBR]) and carotid wall thickness (CWT). Linear regression was used to assess the associations between OSA, CWT and TBR.

Results

The mean age was 67.9 years (SD 8.53) and the mean BMI was 28.9 kg/m2 (SD 4.47). There was a trend toward a higher mean CWT in the OSA (n = 11) vs. non-OSA group (n = 35), 1.51 vs. 1.41 (p = 0.098). TBR did not differ by OSA groups, and there was no significant association between OSA and carotid plaque inflammation (TBR) in adjusted analyses. Although there was a significant interaction between OSA and obesity, there were no statistically significant associations between OSA and vascular inflammation in stratified analysis by obesity.

Conclusion

Despite a trend toward a higher carotid wall thickness in OSA vs. non-OSA participants, we did not find an independent association between OSA and carotid plaque inflammation using PET/MRI in MESA. Our findings suggest that simultaneous assessments of structural and metabolic features of atherosclerosis may fill current knowledge gaps pertaining to the influence of OSA on atherosclerosis prevalence and progression.

Keywords: Obstructive sleep apnea, Vascular inflammation, Atherosclerosis, MESA, OSA, Carotid wall thickness, Obesity, PET/MRI

Introduction

Obstructive sleep apnea (OSA) is a common sleep disorder that affects at least 25 million Americans [1, 2]. The prevalence of OSA is even higher among individuals with underlying cardiovascular disease (CVD) [3–9]. Atherosclerosis is a key intermediate mechanism linking OSA to CVD [10–12]. However, this mechanistic link is poorly understood.

Although OSA has been associated with increased carotid intima media thickness (CIMT) [13], reduction of CIMT with drug therapy does not reduce future CVD event risk [14] suggesting that measures of atherosclerosis other than CIMT are likely relevant for CVD event risk. Plaque inflammatory activity is one such relevant measure which has been identified as the strongest predictor of future vascular events [15]. Plaque vulnerability is associated with metabolic activity [16, 17]. Therefore, simultaneous assessment of carotid wall thickness (structure) and carotid plaque inflammation (metabolic activity) in OSA patients may help shed light on the mechanistic underpinnings of the link between OSA and atherosclerosis.

In this study, we therefore leverage state-of-the-art vascular imaging in the MESA-PET and MESA-Sleep ancillary studies to examine the cross-sectional association between OSA and carotid wall thickness (CWT) and carotid plaque inflammation. We utilize hybrid Positron Emission Tomography/Magnetic Resonance Imaging (PET/MRI) [18]. We use fluorodeoxyglucose (FDG) radiotracer that localizes in activated macrophages within carotid plaques and is a validated measure of vascular inflammation [19]. Based on our pilot study [20] and prior evidence [13]. We hypothesized that OSA is associated with both increased CWT and carotid plaque inflammation. We also test for effect modification by age, sex, statin use, and obesity.

Participants and methods

Study sample

MESA is a longitudinal study that recruited participants from six U.S. communities from 2000 to 2002 to assess risk factors for the incidence and progression of CVD. The cohort consisted of 6,814 men and women aged 45–84 years who were either Non-Hispanic white, Chinese-American, Hispanic, or black and were free of clinically-recognized CVD at baseline. Participants were studied every two years with in-clinic examinations, starting in July 2000.

During MESA Exam 5 (2010–2012), a total of 2261 subjects participated in the sleep ancillary study, including in-home overnight polysomnography (PSG), 7-day wrist actigraphy, and sleep questionnaires. PSG was performed in 2166, of which 2060 met minimum quality criteria for determining the apnea–hypopnea index (AHI). During MESA Exam 6 (2016–2018), 82 subjects participated in MESA-PET an ongoing ancillary study of carotid PET/MRI at one MESA Field Center (Columbia University). Of the 82 participants in MESA-PET, 46 have polysomnography data making up the analytical sample for this manuscript. Institutional Review Board approval was obtained by the Columbia University institutional review board (IRB AAAQ1318). Informed consent was obtained from all participants.

Sleep data

Participants who reported not regularly using treatment of sleep disordered breathing (SDB) were invited to participate in the MESA Sleep study [21]. The Compumedics Somte devices (Compumedics Ltd, Abbotsville, Australia) were used to conduct the polysomnograms using procedures previously described [22–24]. An apnea was defined as 90% reduction in airflow lasting for 10 s or longer, and further distinguished as central or obstructive based on respiratory effort detected using inductance plethysmography. Hypopneas were defined as a 30% or more reduction in airflow for 10 s or longer in association with at least a 4% desaturation. The AHI was defined as the sum of all apneas plus hypopneas divided by total sleep time. SDB was categorized according to the following clinical cutoffs: no-SDB when AHI < 5, mild SDB when 5 ≤ AHI < 15 and moderate-severe SDB when AHI ≥ 15.

Carotid imaging and image analysis

Image acquisition and analysis was conducted at the Biomedical Engineering Imaging Institute (BMEII) at the Icahn School of Medicine at Mount Sinai. Participants attending Exam 6 (2016–2018) underwent carotid imaging using a single hybrid PET/MRI machine (Siemens BiographTM 3T mMR scanner) which acquired the PET and MRI images simultaneously. Prior to the imaging procedure, a blood test was performed to ensure that the subjects’ glucose levels met the inclusion criteria. A urine pregnancy test was performed for all female participants. After overnight fasting, 18F-FDG tracer (10 mCi [milliecurie]) was injected 90 min prior to PET scanning. The circulation time was based on previous studies as giving an optimal signal to background FDG uptake ratio [25]. MR sequences to localize and for PET attenuation correction were obtained. PET data was then obtained in 3-D mode for both carotid arteries. Acquired data was reconstructed to 5mm3 voxel size using the Fourier rebinning-iterative algorithm [26].

Carotid vascular inflammation was measured using target-to-background-ratio (TBR), a well validated method of measuring global vascular inflammation [27, 28]. For the right and left carotid arteries, image analysis started 2 cm below the bifurcation and moved 3 mm superiorly up until 2 cm into the internal carotid artery. Along the aforementioned length of the carotid vessels, mean and maximal standardized uptake value (SUV) metrics for FDG (SUVmean, SUVmax) were measured by drawing circular regions of interest (ROI) around the arterial walls on 5 mm-thick axial slices. Then, the mean and maximum TBR (TBRmean, TBRmax) were calculated from the respective ratios of the S UVmean and S UVmax of the target ROIs divided by the SUVmean of the background ROI, the latter derived from the jugular vein (Fig. 1). Previous analysis by BMEII has demonstrated high intraobserver (0.93 and 0.98) and interobserver (0.90 to 0.97) agreement for this approach [27]. The above TBRmean and TBRmax were calculated for all imaging visits. T2 weighted MR images were used to manually trace the outer and inner carotid vessel wall boundaries. Carotid wall area (mm2), wall thickness (mm), and total vessel area (mm2) were calculated (semi) automatically as previously described [29].

Fig. 1.

Fig. 1

Open in a new tab

A This figure shows an image of the left carotid artery on PET/MRI. Circular regions of interest (ROIs) are drawn around the target vessel wall on anatomical MR (for each axial slice) and copied onto simultaneously acquired and co-registered PET images. B Carotid artery diagram: Standardized uptake value (SUV) metrics are measured for each segment/slice of the target vessel. Mean (average of all measured SUVs, 1.43) and maximum SUV (1.8) are recorded for the target vessel. C Mean and maximum TBR values are calculated after dividing SUVmean or SUVmax values of the target vessel by the background (jugular vein) SUVmean uptake values

Covariates

Height and weight were measured using standardized approaches, and sociodemographic and comorbidity data were obtained from questionnaires [30]. Age was assessed as a continuous variable. Race/ethnicity was categorized into four groups: Non-Hispanic white (ref), non-Hispanic black, Hispanic, and Chinese-American. Smoking use (self-reported) was categorized into three groups: Never smoker (ref), Former smoker, and Current smoker. Body mass index (BMI) was derived from measured height and weight (kg/m2). Presence of hypertension was a yes/no variable based on the sixth report of the Joint National Committee [31] criteria. Dyslipidemia was categorized as a yes/no variable based on a low-density lipoprotein level ≥ 160 mg/dl, high-density lipoprotein level < 40 mg/dl in men and < 50 mg/dl in women, triglycerides ≥ 150 mg/dl or use of statins. Coronary heart disease (CHD) was a yes/no variable based on the presence of CHD (Myocardial Infarction), resuscitated cardiac arrest, definite angina, or probable angina [if followed by revascularization] [30]. Diabetes was categorized into a yes/no variable defined as untreated or treated diabetes based on the 2003 American Diabetes Association (ADA) fasting criteria [32].

Statistical analysis

OSA was examined as a dichotomous variable (AHI ≥ 15 events per hour; hypopnea with 4% desaturation). We chose this cutoff for OSA to focus on clinically significant disease. Carotid vascular inflammation was quantified as TBR and measured as a continuous variable using TBR (TBRmean, TBRmax). The baseline characteristics and the primary and secondary outcomes such as TBR and CWT were summarized by OSA category and compared using unpaired t-tests. Continuous skewed variables were log transformed. Linear regression analysis was performed to assess the independent association between OSA (as a continuous and categorical variable) and CWT and TBR adjusted for age, sex, race/ethnicity, BMI, smoking status, hypertension, dyslipidemia. Linear regression analysis was also performed to determine the independent association between other measures of OSA such as arousal index and hypoxemia and TBR and CWT. Interaction with OSA (categorical) and statin use, obesity [yes/no; BMI ≥ 30], and sex were also tested. Analysis was conducted with alpha set to 0.05.

Results

Baseline characteristics (Tables 1, 2)

Table 1.

Baseline characteristics of the cohort

Total sample (n = 46) AHI < 15 (n = 35) AHI ≥ 15 (n = 11) p-value
Mean age, years (SD) 67.93 (8.53) 68.09 (1.49) 67.45 (7.87) 0.83
Female, n (%) 31 (67.39%) 26 (74.29%) 5 (45.45%) 0.14
Race, n (%) 0.31
 White 14 (30.43%) 9 (25.71%) 5 (45.45%)
 Black 12 (26.09%) 11 (31.43%) 1 (9.09%)
 Hispanic 20 (43.48%) 15 (42.86) 5 (45.45%)
Mean BMI, kg/m2 (SD) 28.90 (4.47) 28.22 (4.12) 31.08 (5.03) 0.06
Waist-hip circumference, cm (SD) 0.92 (0.07) 0.91 (0.06) 0.97 (0.06) 0.01
HTN, n (%) 27 (58.70%) 20 (57.14%) 7 (63.64%) 0.70
Diabetes, n (%) 11 (23.91%) 6 (17.14%) 5 (45.45%) 0.08
Pre-diabetes, n (%) 10 (27.74%) 7 (20.0%) 3 (27.27%) 0.08
Dyslipidemia, n (%) 22 (47.83%) 15 (42.86%) 7 (63.64%) 0.31
Smoking, n (%) 0.88
 Never 17 (36.96%) 13 (37.14%) 4 (36.36%)
 Former 26 (56.52%) 19 (54.29%) 7 (63.64%)
 Current 3 (6.52%) 3 (8.57%) 0 (0%)
Open in a new tab

AHI apnea–hypopnea index, BMI body mass index, HTN hypertension, SD standard deviation

Table 2.

Baseline sleep characteristics of the cohort

Total sample (n = 46) AHI < 15 (n = 35) AHI ≥ 15 (n = 11) p-value
AHI 13.12 (17.67) 5.77 (3.92) 36.50 (23.71) < 0.01
REM AHI 21.13 (18.56) 17.79 (14.98) 32.82 (25.33) 0.10
Sleep maintenance efficiency, % 76.79 (14.49) 76.76 (15.43) 76.91 (11.67) 0.98
Arousal index (event/hour) 19.49 (13.83) 15.06 (6.93) 33.58 (20.25) 0.01
Time in REM sleep (minutes) 67.87(27.52) 71.91 (26.60) 55.0 (27.6) 0.08
Total Sleep Time (minutes) 365.33 (78.14) 360.0 (82.23) 382.2 (63.8) 0.42
SaO2 average—REM, % (SD) 94.24 (2.08) 94.83 (1.36) 92.20 (2.86) 0.02
SaO2 average—NREM, % (SD) 94.70 (1.38) 94.97 (1.42) 93.82 (0.75) < 0.01
SaO2 minimum—REM, % (SD) 83.44 (8.93) 86.26 (4.94) 73.60 (12.62) 0.01
SaO2 minimum—NREM, % (SD) 84.89 (7.97) 88.37 (3.34) 73.82 (8.40) < 0.01
% Time spent with SaO2 less than 90% (SD) 2.68 (5.21) 0.78 (1.13) 8.71 (8.07) < 0.01
Open in a new tab

AHI apnea–hypopnea index, REM rapid eye movement, NREM non-rapid eye movement, SaO2 oxygen saturation, SD standard deviation

Our analytical sample (n = 46) had a mean age of 67.9 years (SD 8.53) and a mean BMI of 28.9 (SD 4.47). Our sample is composed of all participants in the MESA-PET ancillary study who have objective data on OSA, CWT and carotid plaque inflammation or TBR. The sample was predominantly female (67.4%) and 43.5% of the sample was Hispanic whereas 30.4% was White. The overall prevalence of moderate to severe OSA was 23.9% (n = 11/46). As expected, obesity and cardiometabolic conditions (hypertension, diabetes, prediabetes, and dyslipidemia) were more common in the moderate to severe OSA group vs. none to mild OSA group (Table 1). Table 2 provides the baseline sleep characteristics of the study sample by OSA category. The mean AHI was 5.77 (SD 3.92) in none-mild OSA and 36.50 (SD 23.71) in moderate to severe OSA. The mean arousal index was 15.06 (SD 6.93) in the none-mild OSA group and 33.58 (SD 20.25) in the moderate to severe OSA group. Individuals with moderate to severe OSA spent less time in REM sleep (55 vs. 71.9 min, p = 0.08) and spent more time with oxygen levels less than 90% compared to those with none to mild OSA (8.71 vs. 0.78 min, p = 0.009).

Carotid vessel imaging characteristics by OSA status (Table 3)

Table 3.

Carotid vessel imaging characteristics by OSA category

Total sample (n = 46) AHI < 15 (n = 35) AHI ≥ 15 (n = 11) p-value
Carotid arteries
 Avg. TBRmean (SD) 1.44 (0.19) 1.44 (0.20) 1.43 (0.19) 0.86
 Avg. TBRmax (SD) 1.70 (0.24) 1.69 (0.24) 1.71 (0.24) 0.77
 Avg. Wall area (SD) mm2 33.70 (6.03) 33.05 (6.09) 35.96 (5.51) 0.18
 Avg. Total vessel area (SD) mm2 62.36 (11.18) 61.54 (11.56) 65.25 (9.72) 0.36
 Avg. Carotid wall thickness (SD) mm 1.43 (0.16) 1.41 (0.15) 1.51 (0.16) 0.10
Open in a new tab

OSA obstructive sleep apnea, TBR target-to-background ratio, SD standard deviation

Table 3 describes the vessel imaging characteristics by OSA group. Average CWT trended higher in mod-severe OSA vs. none-mild OSA group [1.51 mm vs. 1.41 mm (p = 0.098)], though this did not reach statistical significance. There were no statistically significant differences in average total wall area and vessel area in moderate-severe OSA vs. none-mild OSA group [35.96 mm2 vs. 33.05 mm2 (p = 0.18), and 65.25 mm2 vs. 61.54 mm2 (p = 0.36)]. The TBRs were similar in the two groups i.e. none-mild vs. mod-severe (TBRmean 1.44 vs. 1.43; TBRmax 1.69 vs. 1.71).

In regression analysis, we did not find any statistically significant associations between OSA (continuous and categorical) and the carotid atherosclerosis parameters (both TBR and CWT). In adjusted models (Table 4) (age, sex, race/ethnicity, BMI, smoking, hypertension, dyslipidemia) OSA (categorical) was not associated with average CWT [Beta (SD) = 0.08 (0.06), p = 0.205], or with carotid plaque inflammation [TBRmean Beta (SD) = − 0.02 (0.07), p = 0.730; TBRmax Beta (SD) = − 0.002 (0.08), p = 0.984].

Table 4.

Association between OSA (continuous and categorical) and carotid atherosclerosis

Beta (SD) p-value (AHI < 15) (n = 35) (AHI ≥ 15) (n = 11) p-value
Vascular inflammation (Avg. TBRmean)
 Model 1 0.0093 (0.0232) 0.69 REF −0.0124 (0.0663) 0.85
 Model 2 −0.0117 (0.0235) 0.62 REF −0.0539 (0.0654) 0.41
 Model 3 −0.0005 (0.0251) 0.98 REF −0.0242 (0.0700) 0.73
Vascular inflammation (Avg. TBRmax)
 Model 1 0.0158 (0.0284) 0.58 REF 0.0240 (0.0816) 0.77
 Model 2 −0.0080 (0.0287) 0.78 REF −0.0392 (0.0798) 0.62
 Model 3 0.0057 (0.0305) 0.85 REF −0.0018 (0.0853) 0.98
Avg. Carotid wall thickness
 Model 1 0.0206 (0.0188) 0.27 REF 0.0923 (0.0534) 0.08
 Model 2 0.0099 (0.0185) 0.59 REF 0.0574 (0.0534) 0.28
 Model 3 0.0136 (0.0203) 0.50 REF 0.0767 (0.0605) 0.21
Open in a new tab

Model 1: Unadjusted

Model 2: Adjusted for age, sex, race/ethnicity

Model 3: Adjusted for Model 2 + BMI, smoking, hypertension, dyslipidemia

OSA obstructive sleep apnea, AHI apnea–hypopnea index, TBR target-to-background ratio

*

Continuous AHI was log transformed due to skewness

In fully adjusted models (Table 5), we did not find a significant association between arousal index and plaque inflammation. Further, duration of hypoxemia (% time spent with SaO2 less than 90%) was weakly associated with CWT [Beta (SD) = 0.027 (0.0152), p = 0.08] after adjustment for covariates. However, there was no significant association between duration of hypoxemia and carotid plaque inflammation (TBR) in both unadjusted and adjusted analyses (Table 5).

Table 5.

Association between arousal index, nocturnal hypoxemia and carotid atherosclerosis

Arousal Index* Beta (SD) P-value % time spent in SaO2* < 90% Beta (SD) p-value
Vascular inflammation (Avg. TBRmean)
 Model 1 0.0336 (0.0428) 0.43 −0.0009 (0.0184) 0.96
 Model 2 0.0121 (0.0432) 0.78 −0.0065 (0.0178) 0.72
 Model 3 0.0414 (0.0432) 0.34 −0.0012 (0.0214) 0.95
Vascular inflammation (Avg. TBRmax)
 Model 1 0.0200 (0.0530) 0.71 0.0040 (0.0230) 0.86
 Model 2 0.0049 (0.0525) 0.93 −0.0062 (0.0219) 0.78
 Model 3 0.0377 (0.0523) 0.47 −0.0018 (0.0258) 0.94
Avg. Carotid wall thickness
 Model 1 0.0090 (0.0353) 0.80 0.0239 (0.0135) 0.08
 Model 2 0.0117 (0.0336) 0.73 0.0163 (0.0125) 0.19
 Model 3 0.0280 (0.0343) 0.41 0.0266 (0.0152) 0.08
Open in a new tab

Model 1: Unadjusted

Model 2: Adjusted for age, sex, race/ethnicity

Model 3: Adjusted for Model 2 + BMI, smoking, hypertension, dyslipidemia

TBR target-to-background ratio, SaO2 oxygen saturation, SD standard deviation

*

Continuous Arousal index and percent time spent in SaO2 < 90% was log transformed due to skewness

In addition, obesity was a significant effect modifier for the association between OSA and carotid plaque inflammation: TBRmean (pinteraction = 0.028) and TBRmax (pinteraction = 0.032). However, stratified analyses by obesity (Table S1) did not reveal any statistically significant associations between OSA and carotid plaque inflammation. Similarly, obesity did not modify the association between OSA and CWT (pinteraction = 0.403). The interaction of statin-use and sex with OSA (categorical) were not significant for TBRmean, TBRmax, or CWT.

Discussion

Our analysis from the MESA-PET and MESA-Sleep ancillary studies is the first to simultaneously evaluate the association between OSA and structural as well as metabolic features of carotid atherosclerosis. In our study, although moderate to severe OSA was not associated with carotid plaque inflammation, there was a trend toward a higher mean carotid wall thickness in the OSA vs. non-OSA group.

Our study—although limited by small sample size—provides important insights into the complex association between OSA and structural (CWT), as well as metabolic (TBR) indicators of carotid atherosclerosis. First, we demonstrate that structural assessments of carotid atherosclerosis may not correlate with metabolic activity of carotid plaque in individuals with OSA. While there was a trend toward a higher CWT in individuals with moderate to severe OSA, there was no difference in carotid vascular inflammation between the two groups. Carotid plaque inflammation and wall thickness are distinct manifestations of atherosclerosis and may respond differently to OSA and other cardiometabolic risk factors. Second, our study suggests that OSA-associated measures such as hypoxemia and arousals may also have unique and contrasting effects on CWT and carotid plaque inflammation. Third, obesity may potentially interact with the characteristic physiologic perturbations of OSA to produce unique and unexpected vascular effects [33] especially as it pertains to plaque vulnerability, a key factor in the risk of acute coronary events. Although we did not find significant associations between OSA and vascular inflammation in stratified analysis by obesity (likely due to limited sample size), this should be further explored in larger, longitudinal studies.

Our results are supported by another study [34] that leveraged the larger MESA sample (n = 1615) to investigate the association between OSA and subclinical carotid atherosclerosis using high-resolution b-mode ultrasound. This study measured CIMT and carotid plaque. Like us, they did not find a significant association between OSA and indicators of subclinical atherosclerosis in the overall cohort. However, findings from their exploratory analysis suggested that the mechanisms responsible for the association between OSA and carotid atherosclerosis may be different for carotid plaque vs. CIMT. In their study, hypoxemia was associated with increased CIMT, whereas higher arousal index was associated with carotid plaque in young individuals. This study [34] however did not measure metabolic activity of carotid plaque. Similarly, Drager et al. [13] assessed early signs of carotid atherosclerosis in a group of 42 patients (n = 12 controls; n = 30 OSA) and demonstrated that OSA was independently associated with both carotid-femoral pulse wave velocity (PWV) and CIMT. The same group then conducted a randomized controlled trial [35] of 24 patients with OSA (and no comorbidities) to receive CPAP vs. no treatment for 16 weeks and measured PWV and CIMT before and after 16 weeks of CPAP. They found a significant reduction in PWV and CIMT in the group randomized to CPAP for 16 weeks. This important study was limited by sample size and inclusion of only severe OSA patients free of comorbidities, therefore limiting generalizability. In fact, when replicated in a larger sample of over 200 patients with moderate to severe OSA, CIMT did not respond to 4-months of CPAP [33]. More importantly, these studies did not measure carotid plaque and inflammation.

Although CIMT is a marker of atherosclerosis and is associated with an increased risk for CVD events [36], evidence suggests that carotid plaque predicts CVD risk more accurately than CIMT [37]. Furthermore, combining carotid plaque assessments with CIMT enhanced the risk-prediction for CVD events when compared to CIMT alone [38]. As such, although Drager et al. demonstrated improvement in CIMT with short-term CPAP, his findings may not imply long-term CVD risk reduction. This was shown by Costanzo et al. [14] where reduction or regression of CIMT with drug therapy did not reduce future CVD event risk. Contrary to CIMT, vascular inflammatory activity as measured in our study has been identified as one of the strongest predictors of future vascular events [15]. Utilizing imaging modalities aimed at detecting plaque inflammation/activity has shown that morphologically similar appearing plaques can be metabolically quite different, i.e. either active or inactive. Plaques that are metabolically active better predict vulnerability and rupture vs. those plaques that are metabolically inactive [16, 17]. Therefore, it is crucial to assess vascular inflammation/plaque activity in OSA patients to better understand the link between OSA and vascular events.

Moreover, recent trials have called into question the overall cardiovascular benefit of CPAP in reducing composite CVD events such as myocardial infarction and stroke [39–41], highlighting the need to identify sleep apnea phenotypes at increased risk of atherosclerosis and CVD. We believe that identifying subgroups of OSA patients with increased vascular inflammation, CWT, or presence of plaque may allow us to identify those at increased risk for future CVD events, and risk-stratify these patients for future prospective trials as potential “CPAP-responders.”

Our study had several strengths. It utilized state-of-the-art vascular imaging that combined MRI with PET to obtain structural and metabolic measurements of carotid arteries in individuals with and without OSA. While the cost of imaging, participant burden, and radiation exposure are important considerations when assessing vascular inflammation with PET imaging, our study was the first and largest study to-date to assess the independent association of OSA and carotid plaque inflammation. Furthermore, due to the robust sleep measurements obtained from the MESA-Sleep ancillary study we were able to examine the associations between hypoxemia, arousals during sleep, and CWT/plaque inflammation. We showed that OSA and features such as arousal index and duration of hypoxemia may be distinctly associated with various measures of carotid atherosclerosis obtained simultaneously from the same patient. Finally, despite limited sample size, we were able to explore whether obesity, statin use, and sex are effect modifiers of the association between OSA and atherosclerosis, which can guide future subgroup analyses in larger studies.

Our study also had several limitations. First, our study was cross-sectional. However, based on pathophysiological considerations our underlying causal model was that OSA contributes to plaque progression and atherosclerosis, not the reverse. Second, due to limited sample size, we had limited power to determine independent and significant associations between OSA and carotid atherosclerosis measures (especially vascular inflammation). While the carotid vascular inflammation parameters were not significantly different between the none to mild OSA vs. moderate to severe groups, those with moderate to severe OSA on average had a trend towards a higher CWT by 0.1mm2 [1.51 (SD 0.16) vs. 1.41 (SD 0.15)] and increased carotid wall area by approximately 3 mm2 [35.96 1.41 (SD 5.51) vs. 33.05 (SD 6.09)] based on effect sizes, although this did not reach statistical significance. As a comparison, this average difference in CWT and carotid wall areas observed between the two OSA groups was not significantly different from the magnitude of reduction in carotid wall thickness/wall area (on MRI) observed following use of angiotensin II type-1 receptor blockade (ARB) [42]. Therefore, the differences in the CWT and carotid wall areas between the two OSA groups may potentially represent a clinically significant difference, although these results must be confirmed in larger prospective studies. Additionally, it’s important to point out that the proportion of patients with CVD risk factors was higher in the OSA group, and therefore these differences may be due to confounding factors rather than OSA itself.

Yet, we believe our results are hypothesis-generating. Ours is the first and largest study to date to evaluate the link between OSA, CWT and carotid vascular inflammation in the same individuals using advanced vascular imaging. Despite the limitation in sample size, our work is an important step forward in emphasizing the need for future studies aimed at (1) understanding the association between OSA and atherosclerosis, and (2) conducting a comprehensive assessment of both structural and metabolic features of atherosclerosis to understand the mechanistic underpinnings of the complex relationship between OSA and atherosclerosis. Such studies will become necessary to help understand the neutral results from randomized clinical trials assessing the impact of OSA treatment on cardiovascular risk reduction [39–41].

Conclusion

Despite a trend toward a higher carotid wall thickness in OSA vs. non-OSA participants, we did not find an independent association between OSA and carotid plaque inflammation using PET/MRI in MESA. However, we found suggestive evidence that OSA may have unique associations with structural and metabolic features of atherosclerosis. Our findings suggest that simultaneous assessments of structural and metabolic features of atherosclerosis may fill current knowledge gaps pertaining to the influence of OSA on carotid atherosclerosis prevalence and progression. Larger, longitudinal studies are needed to understand how OSA and its treatment influences structural and metabolic features of atherosclerosis to better predict cardiovascular event risk in OSA.

Supplementary Material

Supplement Table S1

Acknowledgements

The authors thank the investigators, staff, and participants of the MESA study for their valuable contributions. A full list of participating MESA investigators and institutions can be found at http://www.mesa-nhlbi.org. Imaging studies were performed at the Biomedical Engineering and Imaging Institute (BMEII) at Mount Sinai, NY. Participants for the MESA-PET Ancillary Study were recruited at Columbia University, NY by Dr. Steven Shea (R01HL127637—MESA Exam 6). Manuscript preparation was conducted at the Icahn School of Medicine at Mount Sinai, NY (Dr. Neomi Shah—R01HL143221).

Funding

This research was supported by contracts with University of Washington Coordinating Center (HHSN268201500003I, N01-HC-95159), UCLA Field Center (N01-HC-95160), Columbia University Field Center ( N01-HC-95161), Johns Hopkins University Field Center (N01-HC-95162), University of Minnesota Field Center (N01-HC-95163), Northwestern University Field Center (N01-HC-95164), Wake Forest University Field Center (N01-HC-95165),Central Laboratory (N01-HC-95166), Ultrasound Reading Center (N01-HC-95167), MRI Reading Center (N01-HC-95168) and CT Reading Center (N01-HC-95169) from the National Heart, Lung, and Blood Institute, and by grants UL1-TR-000040 (Columbia CTSA), UL1-TR-001079 (Johns Hopkins ICTR), and UL1-TR-001420 (Wake Forest University CTSA) from NCATS. The MESA Sleep study was support by NHLBI grant R01HL56984. The MESA-PET study was supported by NHLBI grant R01HL127637. Dr. Susan Redline was partially supported by R35 HL135818. Dr. Neomi A Shah has funding from the National Institute of Health/National Heart, Lung, and Blood Institute (1R03HL140273-01, 1R01HL143221-01). Dr. Kundel has funding from the National Institute of Health/National Heart, Lung, and Blood Institute (1K23HL161324) and the American Academy of Sleep Medicine Foundation (274-BS-22).

Abbreviations

AASM

American Academy of Sleep Medicine

ADA

American Diabetes Association

AHI

Apnea–hypopnea index

BMI

Body mass index

CAD

Coronary artery disease

CHD

Coronary heart disease

CIMT

Carotid intima media thickness

CMR

Cardiac magnetic resonance

CWT

Carotid wall thickness

CPAP

Continuous positive airway pressure

CVD

Cardiovascular disease

ECG

Electrocardiogram

EMG

Electromyography

HDL

High density lipoprotein

LDL

Low density lipoprotein

LGE

Late-gadolinium enhancement

LV

Left ventricular

MESA

Multi-Ethnic Study of Atherosclerosis

MI

Myocardial infarction

OSA

Obstructive sleep apnea

PET/MRI

Positron emission tomography/magnetic resonance imaging

PSG

Polysomnography

PWV

Pulse wave velocity

REM

Rapid eye movement

SA

Sleep apnea

SCD

Sudden cardiac death

SDB

Sleep disordered breathing

SUV

Standardized uptake value

TBR

Target-to-background ratio

TST

Total sleep time

TWPAS

Typical week physical activity survey

Footnotes

Conflict of interest Ms. Michelle Reid declares that she has no conflict of interest. Dr. Venkatesh Mani declares that he has no conflict of interest. Dr. Vaishnavi Kundel declares that she has no conflict of interest. Dr. Robert Kaplan declares that he has no conflict of interest. Dr. Zahi Fayad declares that he has no conflict of interest. Dr. Steven Shea declares that he has no conflict of interest. Dr. Neomi Shah reports receiving funding and consulting funds from Itamar and Respicardia, unrelated to this project and stock ownership in Abbott Laboratories. Dr. Susan Redline reports receiving grants and consulting funds from Jazz Pharmaceuticals and consulting fees from Eisai Pharmaceuticals, unrelated to this project. Dr. Jorge Kizer reports stock ownership in Abbott, Bristol-Myers Squibb, Johnson & Johnson, Medtronic, Merck and Pfizer.

Research involving human participants and/or animals All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. The MESA protocol was approved by the Institutional Review Board from each participating institution [Wake Forest University (IRB00008492), Columbia University (IRB00002973), Johns Hopkins University (IRB00001656), University of Minnesota (IRB00000438), Northwestern University (IRB00005003), University of California Los Angeles (IRB00000172), and University of Washington (IRB00005647)]. Additionally, Institutional Review Board approval was obtained by the Columbia University institutional review board (IRB AAAQ1318) for the MESA PET ancillary study.

Informed consent Informed consent was obtained from all individual participants included in the study.

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s10554-022-02743-4.

Data availability

The data analyzed in this study is subject to the following licenses/restrictions: MESA data can be accessed after submitting a manuscript proposal to MESA Publications & Presentations Committee and obtaining its approval to conduct the research work. Requests to access these datasets should be directed to Karen Hansen, hansenk3@u.washington.edu.

References

  • 1.Morgenthaler TI, Croft JB, Dort LC et al. (2015) Development of the National Healthy Sleep Awareness Project Sleep Health Surveillance Questions. J Clin Sleep Med 11:1057–1062 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Jung Y, Junna MR, Mandrekar JN, Morgenthaler TI (2017) The National Healthy Sleep Awareness Project Sleep Health Surveillance Questionnaire as an obstructive sleep apnea surveillance tool. J Clin Sleep Med 13:1067. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Mooe T, Wiklund U, Franklin KA, Eriksson PRT (1996) Sleep-disordered breathing in men with coronary artery disease. Chest 109:659–663 [DOI] [PubMed] [Google Scholar]
  • 4.Mooe T, Rabben T, Wiklund U et al. (1996) Sleep-disordered breathing in men with coronary artery disease. Chest 109:659–663 [DOI] [PubMed] [Google Scholar]
  • 5.Carlson JT, Hedner JA, Ejnell H, Peterson LE (1994) High prevalence of hypertension in sleep apnea patients independent of obesity. Am J Respir Crit Care Med 150:72–77 [DOI] [PubMed] [Google Scholar]
  • 6.Ludka O, Stepanova R, Vyskocilova M et al. (2014) Sleep apnea prevalence in acute myocardial infarction–the Sleep Apnea in Post-acute Myocardial Infarction Patients (SAPAMI) Study. Int J Cardiol 176:13–19. 10.1016/j.ijcard.2014.06.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Somers VK, Amin R, Abraham WT, Costa F, Culebras A, Daniels S, Floras JS, Hunt CE, Olson LJ, Pickering TG, Russell R, Woo M, Young TWDP (2008) Sleep apnea and cardiovascular disease: an American Heart Association/American College of Cardiology Foundation Scientific Statement from the American Heart Association Council for High Blood Pressure Research Professional Education Committee, Council. J Am Coll Cardiol 52:686–717 [DOI] [PubMed] [Google Scholar]
  • 8.Gonzaga C, Bertolami A, Bertolami M et al. (2015) Obstructive sleep apnea, hypertension and cardiovascular diseases. J Hum Hypertens 29:705–712. 10.1038/JHH.2015.15 [DOI] [PubMed] [Google Scholar]
  • 9.Floras JS (2014) Sleep apnea and cardiovascular risk. J Cardiol 63:3–8. 10.1016/J.JJCC.2013.08.009 [DOI] [PubMed] [Google Scholar]
  • 10.Drager LF, Yao Q, Hernandez KL et al. (2013) Chronic intermittent hypoxia induces atherosclerosis via activation of adipose angiopoietin-like 4. Am J Respir Crit Care Med 188:240–248. 10.1164/rccm.201209-1688OC [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Jun J, Reinke C, Bedja D et al. (2010) Effect of intermittent hypoxia on atherosclerosis in apolipoprotein E-deficient mice. Atherosclerosis 209:381–386. 10.1016/j.atherosclerosis.2009.10.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Ryan S, Taylor CT, McNicholas WT (2005) Selective activation of inflammatory pathways by intermittent hypoxia in obstructive sleep apnea syndrome. Circulation 112:2660–2667 [DOI] [PubMed] [Google Scholar]
  • 13.Drager LF, Bortolotto LA, Lorenzi MC et al. (2005) Early signs of atherosclerosis in obstructive sleep apnea. Am J Respir Crit Care Med 172:613–618. 10.1164/rccm.200503-340OC [DOI] [PubMed] [Google Scholar]
  • 14.Costanzo P, Perrone-Filardi P, Vassallo E et al. (2010) Does carotid intima-media thickness regression predict reduction of cardiovascular events? A meta-analysis of 41 randomized trials. J Am Coll Cardiol 56:2006–2020. 10.1016/j.jacc.2010.05.059 [DOI] [PubMed] [Google Scholar]
  • 15.Rominger A, Saam T, Wolpers S et al. (2009) 18F-FDG PET/CT identifies patients at risk for future vascular events in an otherwise asymptomatic cohort with neoplastic disease. J Nucl Med 50:1611–1620. 10.2967/jnumed.109.065151 [DOI] [PubMed] [Google Scholar]
  • 16.Cocker MS, Mc Ardle B, Spence JD et al. (2012) Imaging atherosclerosis with hybrid [18F]fluorodeoxyglucose positron emission tomography/computed tomography imaging: what Leonardo da Vinci could not see. J Nucl Cardiol 19:1211–1225. 10.1007/s12350-012-9631-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Tarkin JM, Joshi FR, Rudd JHF (2014) PET imaging of inflammation in atherosclerosis. Nat Rev Cardiol 11:443–457 [DOI] [PubMed] [Google Scholar]
  • 18.Fayad ZA, Mani V, Woodward M et al. (2011) Safety and efficacy of dalcetrapib on atherosclerotic disease using novel non-invasive multimodality imaging (dal-PLAQUE): a randomised clinical trial. Lancet (Lond, Engl) 378:1547–1559. 10.1016/S0140-6736(11)61383-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Rudd JHF, Warburton EA, Fryer TD et al. (2002) Imaging atherosclerotic plaque inflammation with [18F]-fluorodeoxyglucose positron emission tomography. Circulation 105:2708–2711 [DOI] [PubMed] [Google Scholar]
  • 20.Kundel V, Trivieri MG, Karakatsanis NA et al. (2018) Assessment of atherosclerotic plaque activity in patients with sleep apnea using hybrid positron emission tomography/magnetic resonance imaging (PET/MRI): a feasibility study. Sleep Breath. 10.1007/s11325-018-1646-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Lutsey PL, Mcclelland RL, Duprez D et al. (2015) Objectively measured sleep characteristics and prevalence of coronary artery calcification: the Multi-Ethnic Study of Atherosclerosis Sleep study. Thorax 70:880–887. 10.1136/thoraxjnl-2015-206871 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Redline S, Sanders MH, Lind BK et al. (1998) Methods for obtaining and analyzing unattended polysomnography data for a multi-center study. Sleep Heart Health Res Group Sleep 21:759–767 [PubMed] [Google Scholar]
  • 23.ASDA, Anonymous, Bonnet M, et al. (1992) EEG arousals: scoring rules and examples: a preliminary report from the Sleep Disorders Atlas Task Force of the American Sleep Disorders Association. Sleep 15:173–184. 10.1093/sleep/15.2.174 [DOI] [PubMed] [Google Scholar]
  • 24.Redline S, Budhiraja R, Kapur V et al. (2007) The scoring of respiratory events in sleep: reliability and validity. J Clin Sleep Med 3:169–200 [PubMed] [Google Scholar]
  • 25.Rudd JH, Myers KS, Bansilal S et al. (2007) (18)Fluorodeoxyglucose positron emission tomography imaging of atherosclerotic plaque inflammation is highly reproducible: implications for atherosclerosis therapy trials. J Am Coll Cardiol 50:892–896 [DOI] [PubMed] [Google Scholar]
  • 26.Ben Bouallegue F, Crouzet J-F, Comtat C et al. (2007) Exact and approximate fourier rebinning algorithms for the solution of the data truncation problem in 3-D PET. IEEE Trans Med Imaging 26:1001–1009. 10.1109/TMI.2007.897362 [DOI] [PubMed] [Google Scholar]
  • 27.Rudd JHF, Myers KS, Bansilal S et al. (2008) Atherosclerosis inflammation imaging with 18F-FDG PET: carotid, iliac, and femoral uptake reproducibility, quantification methods, and recommendations. J Nucl Med 49:871–878. 10.2967/jnumed.107.050294 [DOI] [PubMed] [Google Scholar]
  • 28.Bucerius J, Hyafil F, Verberne HJ et al. (2016) Position paper of the Cardiovascular Committee of the European Association of Nuclear Medicine (EANM) on PET imaging of atherosclerosis. Eur J Nucl Med Mol Imaging 43:780–792. 10.1007/s00259-015-3259-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.El AH, Mani V, Weinshelbaum KB et al. (2012) MRI Plaque burden of the carotid arteries and aorta: Reproducibility, age, sex and systemic distribution. Nat Clin Pr Cardiovasc Med 6:219. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Bild DE, Bluemke DA, Burke GL et al. (2002) Multi-ethnic study of atherosclerosis: objectives and design. Am J Epidemiol. 10.1093/aje/kwf113 [DOI] [PubMed] [Google Scholar]
  • 31.Kaplan NM (1998) The 6th Joint National Committee report (JNC-6): new guidelines for hypertension therapy from the USA. Keio J Med. 10.2302/kjm.47.99 [DOI] [PubMed] [Google Scholar]
  • 32.Association AD (2003) Standards of medical care for patients with diabetes mellitus. Diabetes Care 26:s33–s50. 10.2337/DIACARE.26.2007.S33 [DOI] [PubMed] [Google Scholar]
  • 33.Kim J, Mohler ER, Keenan BT et al. (2017) Carotid artery wall thickness in obese and non-obese adults with obstructive sleep apnea before and following positive airway pressure treatment. Sleep. 10.1093/sleep/zsx126 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Zhao YY, Javaheri S, Wang R et al. (2019) Associations between sleep apnea and subclinical carotid atherosclerosis the multi-ethnic study of atherosclerosis. Stroke. 10.1161/STROKEAHA.118.022184 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Drager LF, Bortolotto LA, Figueiredo AC et al. (2007) Effects of continuous positive airway pressure on early signs of atherosclerosis in obstructive sleep apnea. Am J Respir Crit Care Med 176:706–712. 10.1164/rccm.200703-500OC [DOI] [PubMed] [Google Scholar]
  • 36.Polak JF, Pencina MJ, Pencina KM et al. (2011) Carotid-wall intima-media thickness and cardiovascular events. N Engl J Med 365:213–221. 10.1056/NEJMoa1012592 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Inaba Y, Chen JA, Bergmann SR (2012) Carotid plaque, compared with carotid intima-media thickness, more accurately predicts coronary artery disease events: a meta-analysis. Atherosclerosis 220:128–133. 10.1016/j.atherosclerosis.2011.06.044 [DOI] [PubMed] [Google Scholar]
  • 38.Nambi V, Chambless L, Folsom AR et al. (2010) Carotid intima-media thickness and presence or absence of plaque improves prediction of coronary heart disease risk: the ARIC (Atherosclerosis Risk in Communities) study. J Am Coll Cardiol 55:1600–1607. 10.1016/j.jacc.2009.11.075 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.McEvoy RD, Antic NA, Heeley E et al. (2016) CPAP for prevention of cardiovascular events in obstructive sleep apnea. N Engl J Med 375:919–931. 10.1056/NEJMoa1606599 [DOI] [PubMed] [Google Scholar]
  • 40.Peker Y, Glantz H, Eulenburg C et al. (2016) Effect of positive airway pressure on cardiovascular outcomes in coronary artery disease patients with nonsleepy obstructive sleep apnea. The RICCADSA Randomized Controlled Trial. Am J Respir Crit Care Med 194:613–620. 10.1164/rccm.201601-0088OC [DOI] [PubMed] [Google Scholar]
  • 41.Yu J, Zhou Z, Mcevoy RD et al. (2017) Association of positive airway pressure with cardiovascular events and death in adults with sleep apnea a systematic review and meta-analysis. JAMA 318:156–166. 10.1001/jama.2017.7967 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Ramadan R, Dhawan SS, Binongo JNG et al. (2016) Effect of angiotensin II Type I receptor blockade with valsartan on carotid artery atherosclerosis: a double blind randomized clinical trial comparing valsartan and placebo (EFFERVESCENT). Am Heart J 174:68–79. 10.1016/J.AHJ.2015.12.021 [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

Supplement Table S1
NIHMS1909419-supplement-Supplement_Table_S1.docx (16.1KB, docx)

Data Availability Statement

The data analyzed in this study is subject to the following licenses/restrictions: MESA data can be accessed after submitting a manuscript proposal to MESA Publications & Presentations Committee and obtaining its approval to conduct the research work. Requests to access these datasets should be directed to Karen Hansen, hansenk3@u.washington.edu.

RESOURCES