Carotid Atherosclerosis and Stroke in Atrial Fibrillation
The Atherosclerosis Risk in Communities Study
Background and Purpose—Whether consideration of carotid intima-media thickness (cIMT) and carotid plaque would improve risk prediction of ischemic stroke in persons with atrial fibrillation (AF) is unknown. The purpose of this study was to assess the improvement in risk prediction of stroke by adding cIMT and carotid plaque to the CHA2DS2-VASc (variables age, heart failure, hypertension, diabetes mellitus, myocardial infarction, and peripheral arterial disease) score.
Methods—We included participants from the Atherosclerosis Risk in Communities (ARIC) study (mean age, 63 years) who developed AF within 5 years after carotid measurement, were not on warfarin, and had no prior stroke at AF diagnosis. AF was ascertained from study ECGs and diagnosis codes, and stroke was physician adjudicated. Multivariable Cox models were used to assess association between carotid indices and ischemic stroke. Improvement in 10-year risk prediction of stroke was assessed by the C-statistic, net reclassification improvement, and relative integrated discrimination improvement.
Results—There were 81 (11.2%) stroke events that occurred among 724 participants with AF during a mean follow-up of 8.5 years. Increased cIMT and presence of carotid plaque were significantly associated with increased stroke risk. The addition of cIMT+plaque to the CHA2DS2-VASc score marginally increased the C-statistic (95% confidence interval) from 0.685 (0.623−0.747) to 0.698 (0.638−0.759). The net reclassification improvement and integrated discrimination improvement for cIMT+plaque were 0.091 (95% confidence interval, 0.012−0.170) and 0.101 (95% confidence interval, 0.002−0.226), respectively.
Conclusions—Increased cIMT and presence of carotid plaque are associated with increased risk of ischemic stroke in individuals with AF. Furthermore, they may improve risk prediction of stroke, over and above the CHA2DS2-VASc score.
The association of carotid atherosclerosis (carotid intima-media thickness [cIMT] and the presence of carotid plaque) with the risk of stroke in a large cohort of community-dwelling adults with atrial fibrillation (AF) has not been examined. It is unknown whether addition of cIMT and carotid plaque would improve risk prediction of stroke over the CHA2DS2-VASc (variables age, heart failure, hypertension, diabetes mellitus, myocardial infarction, and peripheral arterial disease) risk score1 among individuals with AF.
We evaluated the association of carotid atherosclerosis with ischemic stroke among participants with AF in the Atherosclerosis Risk in Communities (ARIC) study.
The ARIC study is a population-based prospective study of cardiovascular disease in a biracial cohort of 15 792 participants aged 45 to 64 years at enrollment (1987–1989) sampled from 4 US communities.2
AF was ascertained using standardized methods.3 cIMT and carotid plaque were measured by ultrasound.4 Carotid indices and covariates were taken from the most recent visit before AF ascertainment. Covariates included age at time of AF ascertainment, sex, race, ARIC field center, heart failure, hypertension, diabetes mellitus, coronary heart disease, peripheral arterial disease, and warfarin use.
cIMT was evaluated in quintiles and as a continuous variable. To estimate the association of cIMT and carotid plaque with time to incident ischemic stroke, we calculated hazard ratios and 95% confidence intervals (CIs) using multivariable Cox models. Person-years at risk were calculated from AF ascertainment until date of development of stroke, death, loss to follow-up, or end of follow-up (December 2011), whichever occurred first. The first multivariate model was adjusted for age, sex, race, and field center. The second model additionally adjusted for CHA2DS2-VASc variables (heart failure, hypertension, diabetes mellitus, coronary heart disease, and peripheral arterial disease). Race was included in the models because AF risk factors, incidence, and outcomes differ by race.5
We utilized the CHA2DS2-VASc score just before AF diagnosis as the benchmark to assess the role of arterial indices in enhancing risk prediction for stroke in AF. We considered 4 models: the benchmark alone, addition of cIMT or carotid plaque alone, and addition of both cIMT and carotid plaque. To assess model discrimination, we computed the C-statistic. We also calculated net reclassification improvement, Grønnesby–Borgan statistic, and the integrated discrimination improvement. Statistical analyses were performed with SAS.version 9.3 (SAS Inc, Cary, NC).
Participants who had strokes during follow-up were more likely to be women, black, had traditional atherosclerotic risk factors, higher cIMT, and presence of carotid plaque (Table 1).
Association of cIMT and Carotid Plaque With Ischemic Stroke Risk in AF
There were 81 (11.2%) stroke events occurring among 724 participants with AF during a mean follow-up of 8.5 years (Figure I in the online-only Data Supplement). Compared with participants in the lowest quintile, the hazard ratio (95% CI) of stroke for those in the highest quintile of cIMT was 2.30 (1.12–4.74), P value for trend=0.005, after adjustment for age, sex, race, and field center (Figure II in the online-only Data Supplement). This association was attenuated after additional adjustment for CHA2DS2-VASc variables. Each 1 SD increase in cIMT was associated with a 23% increased risk of stroke. The relationship between cIMT and stroke did not differ by sex (interaction P=0.25) or race (P=0.53; Table 2).
Compared with participants without plaque, the presence of plaque was associated with a 56% increased risk of stroke after adjustment for age, sex, race, and field center. These results remained unchanged after additional adjustment for CHA2DS2-VASc variables, hazard ratio (95% CI), 1.56 (1.00–2.45). There was no interaction of the relationship between carotid plaque and stroke by sex (P=0.31) or race (P=0.43).
Model Discrimination, Calibration, and Reclassification
C-statistic increased from 0.685 (95% CI, 0.623–0.747) for the model including race and CHA2DS2-VASc variables to 0.698 (95% CI, 0.638–0.759) after addition of both cIMT and plaque information. Addition of both plaque and cIMT to this model resulted in a significant net reclassification improvement, 0.091 (95% CI, 0.012–0.170). Overall, the risk levels for participants reclassified because of cIMT and carotid plaque data were more accurate (Table I in the online-only Data Supplement; Table 3).
The integrated discrimination improvement showed improved risk classification after the addition of cIMT and plaque information. The Grønnesby–Borgan statistic also showed good model fit, after the addition of cIMT and plaque information. Overall, the best model for prediction of stroke in AF was one comprising race and CHA2DS2-VASc variables, with addition of cIMT and plaque information.
In this population-based prospective study of participants with incident AF, cIMT and carotid plaque were found to be independent predictors of 10-year ischemic stroke risk in middle-aged adults. The addition of cIMT and carotid plaque information provided incremental predictive value for risk of stroke in adults with AF, over the CHA2DS2-VASc risk score. Direct evaluation of carotid plaque and cIMT using noninvasive and readily available imaging modalities like carotid ultrasonography may contribute to improvement in risk prediction of stroke in AF over the CHA2DS2-VASc risk score that uses clinical variables alone.
For patients with intermediate risk of stroke (CHA2DS2-VASc score of 1), the 2012 ESC guidelines6 recommend oral anticoagulation; whereas the 2014 American Heart Association/ACC/HRS guidelines recommend no antithrombotic therapy, oral anticoagulant, or aspirin.7 Hence, there is equipoise in management of these patients. Our findings provide preliminary evidence that carotid ultrasonography may be useful in reclassifying these patients. These results also reinforce the concept that atherosclerosis plays an important role in the cause of stroke in AF.
Strengths of this study include the extensive measurement of covariates, large number of AF cases, >40% women, and inclusion of nonwhite participants. There are however, a few limitations. First, it is possible that nonhospitalized stroke events that were not validated in the study could influence the results. However, the magnitude of any potential underestimation of the rate of stroke is likely to be small (<5%).8 Second, there may be misclassification of AF. However, prior analysis within the ARIC cohort to determine the validity of hospital discharge diagnoses for AF reported 84% sensitivity and 98% specificity in AF ascertainment.3 Third, our study population has a mean age of 63 years followed by an 8.5-year mean follow-up. This may limit generalizability to the elderly.
Additional studies are needed to validate these findings in other populations and develop a scoring system that would enhance stroke risk prediction by incorporating cIMT and carotid plaque into the CHA2DS2-VASc score.
We thank the staff and participants of the Atherosclerosis Risk in Communities (ARIC) study.
Sources of Funding
This work was supported by a grant from the National Institute of Aging (R21AG042660-01A1) awarded to Dr Chen. In addition, the Atherosclerosis Risk in Communities Study was performed as a collaborative study supported by National Heart, Lung, and Blood Institute contracts (HHSN268201100005C, HHSN268201100006C, HHSN268201100007C, HHSN268201100008C, HHSN268201100009C, HHSN268201100010C, HHSN268201100011C, and HHSN268201100012C).
Dr Lip is a consultant for Bayer/Janssen, Astellas, Merck, Sanofi, BMS/Pfizer, Biotronik, Medtronic, Portola, Boehringer Ingelheim, Microlife and Daiichi-Sankyo. Speaker for Bayer, BMS/Pfizer, Medtronic, Boehringer Ingelheim, Microlife, Roche and Daiichi-Sankyo. The other authors report no conflicts.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.116.013133/-/DC1.
- Received February 12, 2016.
- Revision received March 8, 2016.
- Accepted April 5, 2016.
- © 2016 American Heart Association, Inc.
- Lip GY,
- Nieuwlaat R,
- Pisters R,
- Lane DA,
- Crijns HJ
- 2.↵The Atherosclerosis Risk in Communities (ARIC) study: design and objectives. The ARIC Investigators. Am J Epidemiol. 1989;129:687–702.
- Soliman EZ,
- Prineas RJ,
- Case LD,
- Zhang ZM,
- Goff DC Jr
- Bekwelem W,
- Misialek JR,
- Konety S,
- Solomon SD,
- Soliman EZ,
- Loehr LR,
- et al
- Camm AJ,
- Lip GY,
- De Caterina R,
- Savelieva I,
- Atar D,
- Hohnloser SH,
- et al
- January CT,
- Wann LS,
- Alpert JS,
- Calkins H,
- Cigarroa JE,
- Cleveland JC Jr,
- et al
- Sacco RL,
- Boden-Albala B,
- Gan R,
- Chen X,
- Kargman DE,
- Shea S,
- et al