Abstract
Background
Coronary artery calcium testing using noncontrast cardiac computed tomography is a guideline‐indicated test to help refine eligibility for aspirin in primary prevention. However, access to cardiac computed tomography remains limited, with carotid ultrasound used much more often internationally. We sought to update the role of aspirin allocation in primary prevention as a function of subclinical carotid atherosclerosis.
Methods and Results
The study included 11 379 participants from the MESA (Multi‐Ethnic Study of Atherosclerosis) and ARIC (Atherosclerosis Risk in Communities) studies. A harmonized carotid plaque score (range, 0–6) was derived using the number of anatomic sites with plaque from the left and right common, bifurcation, and internal carotid artery on ultrasound. The 5‐year number needed to treat and number needed to harm as a function of the carotid plaque score were calculated by applying a 12% relative risk reduction in atherosclerotic cardiovascular disease (ASCVD) events and 42% relative increase in major bleeding events related to aspirin use, respectively. The mean age was 57 years, 57% were women, 23% were Black, and the median 10‐year ASCVD risk was 12.8%. The 5‐year incidence rates (per 1000 person‐years) were 5.5 (4.9–6.2) for ASCVD and 1.8 (1.5–2.2) for major bleeding events. The overall 5‐year number needed to treat with aspirin was 306 but was 2‐fold lower for individuals with carotid plaque versus those without carotid plaque (212 versus 448). The 5‐year number needed to treat was less than the 5‐year number needed to harm when the carotid plaque score was ≥2 for individuals with ASCVD risk 5% to 20%, whereas the presence of any carotid plaque demarcated a favorable risk–benefit for individuals with ASCVD risk >20%.
Conclusions
Quantification of subclinical carotid atherosclerosis can help improve the allocation of aspirin therapy.
Keywords: aspirin, calcium, cardiovascular disease, carotid ultrasound, CPS, hemorrhage, risk, safety
Subject Categories: Cardiovascular Disease, Epidemiology
Nonstandard Abbreviations and Acronyms
- ARIC
Atherosclerosis Risk in Communities
- CPS
carotid plaque score
- MESA
Multi‐Ethnic Study of Atherosclerosis
Clinical Perspective.
What Is New?
An ultrasound‐based carotid plaque score strongly predicted atherosclerotic cardiovascular disease events and could be used to model the expected benefit and lack of benefit from aspirin therapy in primary prevention.
Individuals with a carotid plaque score ≥2 (at least 2 carotid segments with plaque) without high bleeding risk might receive a net benefit from low‐dose aspirin.
What Are the Clinical Implications?
Patients with advanced subclinical atherosclerosis may benefit from aspirin therapy in primary prevention.
The role of aspirin in the primary prevention of atherosclerotic cardiovascular disease (ASCVD) events is currently controversial. 1 , 2 Since 2018, 3 landmark randomized controlled trials 3 , 4 , 5 and 2 large meta‐analyses 6 , 7 have suggested a limited benefit of low‐dose aspirin for the primary prevention of ASCVD events, which on average is offset by an increased risk of bleeding, particularly in older individuals and in those at higher baseline hemorrhagic risk. Still, in the largest primary prevention meta‐analysis available to date, a significant 12% relative risk reduction in ASCVD events was observed for aspirin. 7 Given these updated data, the 2019 American College of Cardiology/American Heart Association 8 and 2021 European Society of Cardiology guidelines 9 for primary prevention convey a tentative IIb recommendation for aspirin therapy among: (1) “select adults 40 to 70 years of age who are at higher ASCVD risk but not at increased bleeding risk,” and (2) “patients with type 2 diabetes at high or very high ASCVD risk in the absence of clear contraindications.” However, it remains unclear how clinicians should best identify patients who are likely to derive a net benefit from aspirin therapy in routine primary prevention. For example, the United States Preventive Services Task Force provides only a grade C recommendation to initiate low‐dose aspirin for adults 40 to 59 years of age with a 10‐year ASCVD risk of >10% using the pooled cohort equations risk calculator, stating that the net benefit of aspirin in this group may be small. However, a key limitation with the pooled cohort equations and similar risk calculators (eg, Framingham Risk Score) are that they are heavily reliant on age and therefore may overestimate risk in middle‐aged to older adults, 10 , 11 which has implications for defining the aspirin‐eligible patient population.
Integrating subclinical atherosclerosis imaging with risk calculators has shown promise for improving ASCVD risk assessment and guiding the allocation of primary prevention therapies. In the past decade, key observational evidence from MESA (Multi‐Ethnic Study of Atherosclerosis) has demonstrated a role of coronary artery calcium (CAC) scoring to help identify patients who are most likely to benefit from long‐term aspirin therapy. For example, the 5‐year number needed to treat (NNT5) to prevent 1 ASCVD event is estimated to be much lower when CAC ≥100 versus CAC = 0 for both individuals with a Framingham Risk Score < 10% (NNT5: 173 versus 2036) and ≥10% (NNT5: 92 versus 808). 12 Furthermore, the NNT5 is considerably lower than the 5‐year number needed to harm (NNH5) when CAC ≥100, regardless of the pooled cohort equations calculated 10‐year risk (NNT5 = 140 versus NNH5 = 518), whereas an opposite trend is observed when CAC = 0 (NNT5 = 1190 versus NNH5 = 567). 13
Although the previous analyses involving CAC and aspirin eligibility have been highly influential, with the conclusions, for example, incorporated into a new Scientific Statement from the National Lipid Association, 14 not all clinicians around the world have access to CAC. In many countries, including many from South America, Africa, Asia, and Europe, use of carotid ultrasound is much more prevalent than CAC. For example, the most recent European Society of Cardiology/European Atherosclerosis Society dyslipidemia guidelines provide a IIa recommendation to use carotid ultrasound for primary prevention ASCVD risk stratification, whereas CAC received a IIb recommendation. 15
Given the interest in using imaging biomarkers to inform precise use of pharmacotherapy in primary prevention, we aimed to estimate the net benefit of aspirin in primary prevention as a function of the degree of subclinical carotid artery atherosclerosis. To enhance statistical power, we combined data from 2 landmark National Heart, Lung, and Blood Institute prospective cohorts, MESA and the ARIC (Atherosclerosis Risk in Communities) study.
Methods
Data Availability Statement
Per National Heart, Lung, and Blood Institute policy, all MESA and ARIC data is available online via the Biologic Specimen and Data Repository Information Coordinating Center (https://biolincc.nhlbi.nih.gov/home/).
MESA and ARIC Studies
The MESA and ARIC studies are both US community‐based prospective cohort studies, and their details have been previously described. Briefly, MESA enrolled 6814 adults 45 to 84 years of age who were free of clinical ASCVD, including White, Black, Hispanic, and Chinese participants. 16 All MESA participants underwent carotid ultrasound at the baseline visit (Visit 1, 2000–2002). The ARIC study enrolled 15 792 men and women between 45 and 64 years of age in 1987 to 1989. 17 A more detailed description of the design for these community‐based prospective cohort studies is available elsewhere. 16 , 17 Because of the detailed carotid ultrasound data available at Visit 2 (1990–1992), we used this visit as the ARIC study baseline (n=14 348). 18 Institutional review board approval was obtained at each study site for MESA and ARIC; all participants provided informed consent.
Study Population and High‐Risk Bleeding Exclusions
For the current study, we included MESA and ARIC participants (for MESA Visit 1 and ARIC Visit 2) who underwent baseline carotid ultrasound, were free of ASCVD at the time of evaluation, and had information on baseline aspirin use. Participants with missing information necessary to calculate a carotid plaque score (CPS) (n=7275), with existing ASCVD (n=557), or reporting baseline daily aspirin use (n=1210) were excluded.
The key exclusion criterion was presence of a condition placing a participant at high risk of bleeding. High‐risk bleeding conditions were defined at baseline as follows: systolic blood pressure >180 mm Hg, estimated glomerular filtration rate <45 mL/min per 1.73 m2, use of corticosteroids, use of oral anticoagulation, or use of proton pump inhibitors, a marker of a high‐risk gastrointestinal condition. After excluding 726 additional participants with a high risk of bleeding, the final study population for analysis was 11 379.
Of note, MESA Visit 1 and ARIC Visit 2 largely preceded the use of proton pump inhibitors. Therefore, in sensitivity analysis, we further excluded participants taking histamine‐2 receptor (H2) antagonists (H2 blockers, n=569) as well as those reporting daily use of NSAIDs (n=889). The inclusion, exclusions, and determination of the final study population is presented in Figure S1.
Carotid Ultrasound and CPS Derivation
The carotid ultrasound exams in MESA and ARIC have been previously described. 19 , 20 In both studies, B‐mode ultrasound was performed to obtain images of the right and left common, bifurcation, and internal carotid artery segments. In MESA, carotid intima‐media thickness and plaque were evaluated via ultrasound using an M12L transducer (General Electric Medical Systems, Waukesha, WI; common carotid artery frequency, 13 MHz). 19 In ARIC, a Biosound 2000IISA system was used, recorded on a video home system tape, and images were reviewed at the ARIC Ultrasound Reading Center.
The presence of a carotid artery atherosclerotic plaque was defined by a distinct, focal wall thickening ≥1.5 mm or thickening ≥50% than the surrounding common carotid intima‐media thickness. Both cohorts defined plaque similarly, including specifications for abnormal wall thickness as above, as well as abnormal shape or abnormal wall texture from the surrounding vascular tissue. The intrareader variability for carotid plaque was robust across both studies (MESA: 0.83, ARIC: 0.76).
We sought to harmonize carotid ultrasound‐defined plaque burden from both ARIC and MESA. Among participants with adequate imaging, a total CPS (range, 0–6) was used to semiquantify carotid plaque burden. For the 3 segments analyzed, considering from both left and right sides of the body for each participant, 1 point was assigned for each segment with plaque. We considered all imaging data available for each carotid segment, including from the near and far walls, and for anterior and posterior imaging angles as available from each study imaging protocol.
Outcome Definitions and Event Ascertainment
The outcomes for the study were hard coronary heart disease and ASCVD events. Hard coronary heart disease events were defined as a nonfatal myocardial infarction, death from coronary heart disease, or resuscitated cardiac arrest. ASCVD events were defined as hard coronary heart disease events plus fatal/nonfatal stroke, other atherosclerotic death, or other cardiovascular disease‐related death. 12
Major bleeding events were defined by using hospitalization International Classification of Diseases, Ninth Revision and Tenth Revision (ICD‐9 and ICD‐10) codes at discharge summary (Table S1). To increase the robustness of our findings, we used a primary definition and 2 alternative definitions of major bleeding. The primary definition was the modified Food and Drug Administration definition derived from the Mini‐Sentinel initiative protocol. 21 The Food and Drug Administration definition, which is a balance of sensitivity and specificity, considers nontrauma intracranial hemorrhage in any position on the discharge summary, and nonintracranial hemorrhage ICD‐9 and ICD‐10 codes if they appear in a primary position. Here we considered the first 3 positions on the discharge summary as a primary position defining a major bleeding event. 22 The Cunningham definition, which is the most specific, considers both intracranial hemorrhage and nonintracranial hemorrhage ICD‐9 and ICD‐10 codes in the first position only on the discharge summary as major bleeding events. We also considered an unrestricted definition considering the above major bleeding codes in any position on the discharge summary.
General Clinical Examination and Measurement of ASCVD Risk Factors
Standardized methods were used to collect demographic and clinical information, including sex, race and ethnicity, education status (post‐high school education versus high school education or less). Smoking status was categorized as current, former, or never for cigarette smoking. Body mass index was defined as kilograms per square meter using height and weight measurements from physical exams.
Blood pressure was measured in triplicate from the brachial artery while participants were in a seated resting position, and the average of the second and third readings was used. Hypertension was defined as a systolic blood pressure ≥130 mm Hg or diastolic blood pressure ≥80 mm Hg. 23 Fasting blood glucose was measured using a hexokinase/glucose‐6‐phosphate dehydrogenase method. 24 Type 2 diabetes was defined as self‐reported history of diagnosis of diabetes by a physician, the use of glucose‐lowering medications, and/or baseline fasting blood glucose levels ≥126 mg/dL (all at baseline). Total cholesterol and high‐density lipoprotein cholesterol were measured enzymatically, 24 and low‐density lipoprotein cholesterol values were calculated using the Friedewald equation. 25
Baseline 10‐year ASCVD risk was calculated using the pooled cohort equations calculator. 26 As per the guidelines, 27 three 10‐year ASCVD risk strata were defined: <5%, 5% to 20%, and >20%.
Statistical Analysis
Study sample characteristics were presented as mean±SD for continuous variables, and categorical variables were presented as percentages. Continuous variables that were not normally distributed were presented as median (Q1–Q3). Descriptive statistics were presented for the overall sample and stratified by those without carotid plaque versus those with carotid plaque. Differences between absence of carotid plaque versus presence of carotid plaque for normally and nonnormally distributed variables were assessed through the Student t test and Wilcoxon signed rank test, respectively. Differences between categorical variables for those with absence of carotid plaque versus presence of carotid plaque were evaluated through the χ2 test. In sensitivity analyses, we presented descriptive statistics according to the 6‐point CPS.
To examine the distribution of CPSs, histograms were conducted across the 6‐point CPS for both the overall study population and by individual cohort.
Crude incidence rates of 5‐year ASCVD and major bleeding events over 5 years were calculated as the number of events per 1000 person‐years for individuals with absence of carotid plaque versus presence of carotid plaque, and across the 6‐point CPS, with additional stratification by 10‐year predicted risk <5%, 5% to 20%, and ≥20%. Multivariable Cox proportional hazards regression assessed the association of the 6‐point CPS with incident ASCVD and major bleeding events over the entire follow‐up of MESA and ARIC. The proportional hazards assumption was verified using log–log plots and scaled Schoenfeld residuals for key variables. These models were adjusted for age, sex, race and ethnicity, cohort, current smoking, type 2 diabetes, statin medication use, total cholesterol, high‐density lipoprotein cholesterol, systolic blood pressure, blood pressure‐lowering medication use, H2 receptor blocker use, and daily NSAID use.
Next, to assess the estimated benefit of aspirin, we applied pooled estimates from an aspirin primary prevention trial meta‐analysis that demonstrated a 12% relative risk reduction in ASCVD events with daily low‐dose aspirin therapy to the observed 5‐year ASCVD event rates, identical to the approach of Cainzos‐Achirica et al. 13 Using the reciprocal of this calculated absolute risk reduction, the NNT5 was then calculated. These calculations were performed both in the overall sample and the group stratified by CPS. Correspondingly, for estimated harm of aspirin, we applied the meta‐analysis based on 42% relative risk increase in bleeding events for aspirin to the observed 5 years of major bleeding events to calculate absolute risk increase and the NNH5 for major bleeding events. 7 The NNT5 and NNH5 were then compared graphically to assess the benefit and harm of aspirin therapy in the overall sample and according to CPS groups. Because the major bleeding rates differed somewhat between those with and without carotid plaque, we calculated the NNH5 separately according to the presence versus absence of carotid plaque. Given the small number of participants with a CPS of 6, NNT5 in this group was modeled as a linear function of risk across the remaining range of CPS (0–5).
To maximize the robustness of our results, we conducted several key sensitivity analyses. To increase statistical power, we repeated the above calculations using a 10‐year follow‐up for incident ASCVD and major bleeding events, and then used the Anderson‐Altman method to scale back 10‐year observed incidence to 5‐year NNT5 and NNH5 calculations. We also repeated all analyses after removing those taking H2 blockers or daily NSAIDs. Finally, we repeated analyses using the Cunningham bleeding definition instead of the modified Food and Drug Administration definition.
Results
The mean age was 57.3±6.7 years, 57.2% were women, 22.5% were Black (Table 1), and the median 10‐year ASCVD risk was 12.8% (6.2%–24.4%). A total of 42.9% of individuals had carotid plaque (Figure 1). In total, 20.2% had a CPS of 1, 12.1% had a CPS of 2, 6.1% had a CPS of 3, 3.1% had a CPS of 4, 1.0% had a CPS of 5, and 0.4% had a CPS of 6. Carotid plaque was associated with a higher burden of all ASCVD risk factors, except for diastolic blood pressure, body mass index, and triglycerides. Individuals with carotid plaque also reported higher use of lipid‐lowering medication, daily NSAIDs, and H2 blockers. Among those with prevalent carotid plaque, an overall similar pattern was observed across higher CPS 1 through 6 (Table S2). There were no differences in carotid plaque burden according to the MESA versus ARIC study cohorts (Figure 1).
Table 1.
Characteristics of 11 379 MESA and ARIC Participants Who Underwent CPS Assessment, Stratified by Absence Versus Presence of Carotid Plaque
| Characteristic | All (n=11 379) | No carotid plaque (n=6669) | Carotid plaque (n=4710) |
|---|---|---|---|
| Demographics | |||
| Age, y, mean±SD | 57.3±6.7 | 56.1±6.3 | 59.0±6.8 |
| Women, n (%) | 6503 (57.2) | 4082 (61.2) | 2421 (51.4) |
| Race and ethnicity*, n (%) | |||
| White | 7900 (68.7) | 4580 (68.7) | 3320 (70.5) |
| Black | 2518 (22.5) | 1502 (22.5) | 1016 (21.6) |
| Chinese | 364 (3.2) | 241 (3.6) | 123 (2.6) |
| Hispanic | 597 (5.3) | 346 (5.2) | 251 (5.3) |
| Post‐high school education, n (%) | 9266 (81.4) | 5598 (83.9) | 3668 (77.9) |
| ASCVD risk factors | |||
| PCE 10‐y ASCVD risk, % | 7.9±7.8 | 6.4±6.5 | 10.1±8.9 |
| <7.5% | 5201 (46.1) | 3642 (55.1) | 1559 (33.4) |
| 7.5%–19.9% | 5230 (46.4) | 2671 (40.4) | 2559 (54.8) |
| ≥20% | 846 (7.5) | 296 (4.5) | 550 (11.8) |
| Systolic blood pressure, mm Hg, mean±SD | 120.1±17.4 | 118.3±16.7 | 122.8±18.1 |
| Diastolic blood pressure, mm Hg, mean±SD | 71.4±9.6 | 71.4±9.5 | 71.4±9.8 |
| Hypertension, n (%) | 3982 (35.0) | 2075 (31.1) | 1907 (40.5) |
| Antihypertensive medication, n (%) | 3087 (27.1) | 1642 (24.6) | 1445 (30.7) |
| Diabetes, n (%) | 980 (8.6) | 462 (7.0) | 518 (11.0) |
| Current cigarette smoking, n (%) | 2303 (20.3) | 1143 (17.2) | 1160 (24.6) |
| Body mass index, kg/m2, mean±SD | 27.4±4.9 | 27.4±5.0 | 27.2±4.7 |
| Total cholesterol, mg/dL, mean±SD | 206.2±38.5 | 203.7±37.2 | 209.8±40.0 |
| LDL cholesterol, mg/dL, mean±SD | 129.7±36.0 | 126.9±34.9 | 133.6±37.2 |
| HDL cholesterol, mg/dL, mean±SD | 50.8±16.5 | 51.8±16.7 | 49.5±16.2 |
| Triglycerides, mg/dL, median (Q1–Q3) | 110.9 (79.9–157.9) | 107.0 (77.9–153.9) | 115.9 (83.9–163.9) |
| Lipid‐lowering medication, n (%) | 3456 (30.5) | 1863 (28.1) | 750 (32.8) |
| NSAID use, n (%) | 889 (7.8) | 494 (7.4) | 395 (8.4) |
| H2 blocker/proton pump inhibitor | 566 (5.0) | 314 (4.7) | 252 (5.4) |
ARIC indicates Atherosclerosis Risk in Communities; ASCVD, atherosclerotic cardiovascular disease; CPS, carotid plaque score; H2, histamine‐2 receptor; HDL, high‐density lipoprotein; LDL, low‐density lipoprotein; MESA, Multi‐Ethnic Study of Atherosclerosis; and PCE, pooled cohort equations.
Race and ethnicity categories are those collected from MESA and do not reflect US Census Bureau definitions.
Figure 1. CPS prevalence overall (A) and by cohort (B).
Values are presented as percent of individuals. ARIC indicates Atherosclerosis Risk in Communities; CPS, carotid plaque score; and MESA, Multi‐Ethnic Study of Atherosclerosis.
The overall crude 5‐year incidence rates per 1000 person‐years for ASCVD and major bleeding events in the first 5 years were 5.5 (4.9–6.2) and 1.8 (1.5–2.2), respectively (Table 2). Individuals with carotid plaque had both higher ASCVD and major bleeding events compared with those without carotid plaque. The highest ASCVD event rates in the entire follow‐up time were observed for those with a CPS of 5 (31.9 [95% CI, 24.0–42.3]) and 6 (26.5 [95% CI, 15.7–44.8]; Table 3). Regardless of CPS group, higher 5‐year and 10‐year incident ASCVD were observed across higher ASCVD risk groups (Table S3 and Table S4).
Table 2.
Crude 5‐Year Incidence Rates (95% CI) per 1000 Person‐Years for ASCVD and Major Bleeding Events in the First 5 Years
| Variable | All | Carotid plaque absent | Carotid plaque present |
|---|---|---|---|
| All | |||
| ASCVD | 5.5 (4.9–6.2) | 3.8 (3.2–4.5) | 8.0 (6.9–9.3) |
| Major bleeding events* | 1.8 (1.5–2.2) | 1.5 (1.2–2.0) | 2.1 (1.6–2.8) |
| Major bleeding events† | 2.5 (2.1–2.9) | 2.0 (1.6–2.5) | 3.1 (2.5–3.9) |
| ASCVD risk <5% | |||
| ASCVD | 0.8 (0.4–1.6) | 0.4 (0.1–1.2) | 2.0 (0.9–4.5) |
| Major bleeding events* | 0.6 (0.3–1.3) | 0.5 (0.2–1.3) | 1.0 (0.3–3.1) |
| Major bleeding events† | 0.7 (0.4–1.5) | 0.5 (0.2–1.3) | 1.3 (0.5–3.6) |
| ASCVD risk 5%–20% | |||
| ASCVD | 3.6 (2.9–4.4) | 3.2 (2.4–4.2) | 4.2 (3.1–5.7) |
| Major bleeding events* | 1.3 (0.9–1.8) | 1.4 (0.9–2.2) | 1.2 (0.7–2.1) |
| Major bleeding events† | 1.8 (1.4–2.4) | 1.9 (1.3–2.7) | 1.7 (1.1–2.7) |
| ASCVD risk >20% | |||
| ASCVD | 11.4 (9.9–13.2) | 7.9 (6.1–10.1) | 14.7 (12.4–17.6) |
| Major bleeding events* | 3.2 (2.4–4.2) | 2.5 (1.6–3.9) | 3.8 (2.7–5.4) |
| Major bleeding events† | 4.4 (3.5–5.5) | 3.2 (2.2–4.7) | 5.4 (4.1–7.2) |
ASCVD risk was calculated with the pooled cohort equations calculator.
Total person‐time: 55 515.3 person‐years.
ASCVD indicates atherosclerotic cardiovascular disease.
Food and Drug Administration definition.
Unrestricted definition.
Table 3.
Hazard Ratios (95% CIs) and Incidence Rate (95% CIs) per 1000 Person‐Years for ASCVD and Major Bleeding Events, Stratified by CPS
| CPS | ASCVD* | Major bleeding events* (FDA definition) | ||||
|---|---|---|---|---|---|---|
| HR (95% CI) | P trend | Incidence rate (95% CI) per 1000 person‐years | HR (95% CI) | Incidence rate (95% CI) per 1000 person‐years | P trend | |
| 0 | Ref | <0.001 | 8.5 (8.0–8.9) | Ref | 3.9 (3.6–4.3) | <0.001 |
| 1 | 1.24 (1.12–1.38) | 12.1 (11.1–13.2) | 1.13 (0.97–1.32) | 5.0 (4.4–5.7) | ||
| 2 | 1.57 (1.39–1.76) | 16.6 (15.1–18.4) | 1.19 (0.98–1.44) | 5.5 (4.6–6.4) | ||
| 3 | 1.69 (1.46–1.95) | 20.8 (18.3–23.7) | 1.26 (0.98–1.61) | 6.3 (5.0–7.9) | ||
| 4 | 2.05 (1.70–2.48) | 25.3 (21.4–30.1) | 1.71 (1.28–2.30) | 9.6 (7.3–12.5) | ||
| 5 | 2.34 (1.75–3.15) | 31.9 (24.0–42.3) | 1.25 (0.70–2.23) | 7.0 (4.0–12.3) | ||
| 6 | 1.70 (0.98–2.95) | 26.5 (15.7–44.8) | 1.82 (0.81–4.09) | 14.3 (7.2–28.6) | ||
Incidence rates and HRs are derived here using the entire follow‐up period of MESA and ARIC.
ARIC indicates Atherosclerosis Risk in Communities; ASCVD, atherosclerotic cardiovascular disease; CPS, carotid plaque score; FDA, Food and Drug Administration; HR, hazard ratio; and MESA, Multi‐Ethnic Study of Atherosclerosis.
Adjusted for age, sex, ethnicity, cohort, statin medication, total cholesterol, high‐density lipoprotein cholesterol, systolic blood pressure, type 2 diabetes, blood pressure‐lowering medication, cigarette smoking, histamine‐2 receptor blockers/proton pump inhibitors, nonsteroidal anti‐inflammatory drugs.
In multivariable Cox regression analyses, there was a strong stepwise higher hazard of ASCVD across higher CPS groups that was statistically significant at all levels of CPS except the small CPS of 6 group (Table 3). Compared with individuals without carotid plaque, individuals with a CPS of 3 (hazard ratio [HR], 1.69 [95% CI, 1.46–1.95]), 4 (HR, 2.05 [95% CI, 1.70–2.48]), and 5 (HR, 2.34 [95% CI, 1.75–3.15]) had between a 1.7‐ and 2.3‐fold higher hazard of ASCVD events. For all CPS groups, the relative association with major bleeding events was lower compared with the relative association with ASCVD events. For example, the only CPS category that was observed to have a significantly higher risk of major bleeding events was CPS 4 (HR, 1.71 [95% CI, 1.28–2.30]). Unadjusted point estimates for ASCVD and major bleeding events were similar and higher in magnitude compared with adjusted point estimates (Table S5). Kaplan‐Meier plots according to carotid plaque burden are presented in Figure S2A and S2B.
The overall estimated NNT5 with aspirin was 306 (Table 4), which was 2‐fold lower for individuals with carotid plaque versus those without carotid plaque (212 versus 448). A graded lower NNT5 was observed for individuals with higher CPS, because those with a CPS of 5 had a NNT5 of 118. There was no clear pattern of association between non‐0 CPS and the NNH5 for major bleeding events (Table 5).
Table 4.
Number Needed to Treat With Aspirin During 5 Years With Aspirin to Prevent 1 ASCVD Event
| Variable | Observed events at 5 years | Aspirin (assuming 12% relative risk reduction) | |||
|---|---|---|---|---|---|
| N | Incidence (%) | Modeled incidence* (%) | Absolute risk reduction (%) | No. needed to treat (n) | |
| All | 306 | 2.72 | 2.39 | 0.29 | 306 (273–345) |
| Carotid plaque | |||||
| Absent | 123 | 1.86 | 1.64 | 0.20 | 448 (377–531) |
| Present | 183 | 3.94 | 3.47 | 0.42 | 212 (183–246) |
| CPS | |||||
| 1 | 67 | 2.95 | 2.60 | 0.31 | 282 (259–308) |
| 2 | 57 | 4.43 | 3.90 | 0.47 | 188 (171–208) |
| 3 | 30 | 4.71 | 4.15 | 0.50 | 177 (156–202) |
| 4 | 21 | 6.54 | 5.76 | 0.69 | 127 (107–151) |
| 5 | 7 | 7.09 | 6.24 | 0.75 | 118 (89–156) |
| 6 | 1 | 2.86 | 2.52 | 0.30 | … |
ASCVD indicates atherosclerotic cardiovascular disease; and CPS, carotid plaque score.
Modeled data apply at 12% risk reduction with use of aspirin to calculate expected event incidence, absolute risk reduction, and therefore number needed to treat. Due to the small sample size, number needed to treat calculations for CPS 6 were unable to be performed.
Table 5.
Number Needed to Harm With Aspirin During 5 Years With Aspirin to Cause 1 Major Bleeding Event (Food and Drug Administration Definition)
| Variable | Observed events at 5 years | Aspirin (assuming 42% relative risk increase) | |||
|---|---|---|---|---|---|
| N | Incidence (%) | Modeled incidence* (%) | Absolute risk increase (%) | No. needed to harm (n) | |
| All | 99 | 0.88 | 1.22 | 0.51 | 291 (243–356) |
| Carotid plaque | |||||
| Absent | 50 | 0.76 | 1.06 | 0.45 | 337 (270–449) |
| Present | 49 | 1.06 | 1.47 | 0.62 | 242 (184–323) |
| CPS | |||||
| 1 | 20 | 0.88 | 1.22 | 0.51 | 291 (256–332) |
| 2 | 19 | 1.48 | 2.06 | 0.87 | 173 (145–201) |
| 3 | 5 | 0.79 | 1.09 | 0.46 | 325 (258–408) |
| 4 | 4 | 1.25 | 1.74 | 0.73 | 205 (156–267) |
| 5 | 1 | 1.04 | 1.45 | 0.61 | 247 (141–434) |
| 6 | 0 | 0 | 0 | 0 | … |
CPS indicates carotid plaque score.
Modeled data apply at 42% relative risk increase with use of aspirin to calculate expected event incidence, absolute risk increase, and therefore number needed to harm. Due to the small sample size, number needed to harm calculations for CPS 6 were unable to be performed.
In general, the NNT5 was consistently lower than the NNH5, suggesting a net benefit of aspirin, when the CPS was ≥2 (Figure 2), which did not change after the exclusion of individuals on NSAIDs or H2 blockers (Figure S3). The margin between NNT5 and NNH5 was smaller for CPS ≥2 to 4 when major bleeding events were defined according to the unrestricted bleeding definition (Figure S4), and wider considering the Cunningham definition (Figure S5). The NNT5 was lower than NNH5 when the CPS was ≥2 for individuals with an ASCVD risk of 5% to 20%, whereas the presence of any carotid plaque demarcated a favorable risk–benefit for individuals with ASCVD risk >20% (Figure 3A through 3C). In contrast, the NNH5 was lower than the NNT5, suggesting net harm from aspirin, when there was no carotid plaque or very mild carotid plaque (CPS of 1).
Figure 2. Different NNT with low‐dose aspirin during 5 years to prevent 1 CVD event and number needed to cause a major bleeding event by baseline CPS in the total study population.
Values are presented as number of individuals. Horizontal lines represent NNH thresholds. CPS indicates carotid plaque score; CVD, cardiovascular disease; NNH, number needed to harm; and NNT, number needed to treat.
Figure 3. Fixed number needed to treat with aspirin during 5 years to prevent 1 CVD event and number needed to cause a major bleeding event, stratified by ASCVD risk and CPS.
A, Fixed number needed to treat with aspirin during 5 years to prevent 1 CVD event and number needed to cause a major bleeding event, by estimated ASCVD risk and CPS. B, Fixed number needed to treat with aspirin during 5 years to prevent 1 CVD event and number needed to cause a major bleeding event, by estimated ASCVD risk 5% to 20% and CPS. C, Fixed number needed to treat with aspirin during 5 years to prevent 1 CVD event and number needed to cause a major bleeding event, by estimated ASCVD risk >20% and CPS. Values are presented as number of individuals. Ten‐year ASCVD risk was estimated by using the pooled cohort equations (A) ASCVD risk <5% vs ASCVD risk 5% to 20% vs ASCVD risk >20%, (B) ASCVD risk 5% to 20%, and (C) ASCVD risk >20%). The red horizontal lines represent the NNH threshold for each ASCVD risk stratum. ASCVD indicates atherosclerotic cardiovascular disease; CPS, carotid plaque score; CVD, cardiovascular disease; and NNH, number needed to harm.
When considering the 10‐year follow‐up, the conclusions were largely similar. For example, NNT5 was less than the NNH5 when the CPS ≥2 (Figure 4). Differences in the NNT5 for those without carotid plaque versus those with carotid plaque (364 versus 173) were similar when applying 10‐year incidence data for ASCVD events (Table S6). Similar to the findings on 5‐year incidence, no strong pattern of association between CPS and major bleeding events (and thus NNH5) was observed when considering the 10‐year follow‐up (Table S7).
Figure 4. Different NNT with low‐dose aspirin during 10 years to prevent 1 CVD event and number needed to cause a major bleeding event (10‐year ASCVD and major bleeding events rates) by baseline CPS in the total study population.
Values are presented as number of individuals. Horizontal lines represent NNH thresholds. ASCVD indicates atherosclerotic cardiovascular disease; CPS, carotid plaque score; CVD, cardiovascular disease; NNH, number needed to harm; and NNT, number needed to treat.
Discussion
In a sample of >11 000 US adults in primary prevention who were not taking aspirin and were free of high‐risk bleeding features, we observed a general stepwise higher risk of ASCVD across increasing carotid plaque burden such that the estimated NNT5 to prevent 1 ASCVD event with aspirin was approximately 50% to 70% lower among individuals with prevalent or moderate carotid plaque (CPS ≥2) versus absent carotid plaque (NNT5 = ≈150–212 versus 448). Although the estimated risk of major bleeding with aspirin exceeded the benefit on ASCVD in the total population or when carotid plaque was absent, the net benefit of aspirin appeared evident when at least moderate carotid plaque was present, defined as a CPS ≥2. Overall, consistent with prior results examining CAC to guide aspirin allocation, these results suggest a usefulness of carotid plaque burden assessment using ultrasound to improve the short‐term allocation of aspirin therapy in primary prevention.
The findings of this study are clinically useful for several reasons. As reflected in recent guidelines, the approach to using aspirin as primary prevention therapy is continuously evolving and requires personalization; therefore, there is an urgent need to identify the patients who are most and least likely to derive net benefit from daily aspirin therapy. Importantly, there are limitations with current approaches in ASCVD risk assessment for guiding aspirin therapy, primarily overreliance on age, 10 , 11 which may lead to a substantially higher pool of eligible participants who may not uniformly benefit from primary prevention aspirin therapy. Subclinical atherosclerosis imaging may be able to help overcome this major limitation because it is able to measure cumulative lifetime risk and the contribution of genetics, lifestyle, and risk factor burden exposure. Given the interest in imaging to guide clinical risk assessment and the widespread international use of ultrasound, there is an immediate path for translating our findings linking carotid plaque burden assessment to estimated net benefit of aspirin. Here, our potentially immediately actionable tipping point analysis suggests that individuals with an ASCVD risk ≤20%, a CPS ≥2 correlated with net benefit; however, any carotid plaque in a high‐risk patient also was linked with net benefit. In general, the number needed to treat to prevent 1 event for aspirin would be between 150 and 160 for a CPS ≥2 for those at low‐intermediate risk and any carotid plaque for individuals at high risk.
Our results on carotid plaque quantification to guide the eligibility for daily aspirin therapy should be compared with other imaging modalities, particularly noncontrast cardiac computed tomography to measure CAC. Two high‐profile previous analyses using similar methods have demonstrated that ASCVD risk reduction outweighs risk for major bleeding events for individuals with CAC ≥100. 13 , 28 We found similar NNT5 for individuals with a CPS ≥2 when compared with CAC ≥100 (118 versus 140), which may suggest a complementary and/or potentially synergistic relationship between the 2 imaging modalities. Similarly, daily aspirin use appears highly unfavorable when CAC = 0 or when carotid plaque is absent. Our results extend previous studies that have shown that the measurement of CAC and/or carotid plaque can both help to improve the precision of statin therapy for primary prevention. 12 , 13 , 29 Overall, the usefulness of carotid ultrasound to guide aspirin and statin therapy may be especially helpful in resource‐limited health care systems that do not yet have access to computed tomography technology. Although CAC scoring is endorsed by the American College of Cardiology/American Heart Association 27 due to superior risk stratification and being easier to quantify than carotid plaque, carotid atherosclerosis quantification on ultrasound provides the benefit of having no radiation, wider international availability, and the unique possibility of bringing risk assessment to the bedside.
Our findings indicate a potential role for carotid atherosclerosis imaging to help guide aspirin use, but also that future studies that concurrently assess CAC and carotid atherosclerosis for aspirin use are necessary. For example, previous evidence suggests that the presence of carotid plaque, even in the absence of CAC = 0, still confers a 66% higher relative risk for incident coronary heart disease. 30 The net clinical benefit of primary prevention aspirin therapy across carotid plaque and CAC burden is unknown and requires further study.
Methodological approaches are an important consideration for the interpretation of our results. We chose to apply a 12% relative risk reduction in ASCVD and a 42% relative risk increase in major bleeding events based on the highest‐quality meta‐analysis of randomized controlled trials. 7 However, other estimates have been published. For example, the United States Preventive Services Task Force used a slightly different approach to arrive at a 10% reduced odds of ASCVD and 44% increased odds of major bleeding. This small difference would not be sufficient to change our overall results. Another methodological consideration is the definition of major bleeding, which remains somewhat controversial. We used a modified Food and Drug Administration definition, but our sensitivity analysis using the unrestricted and Cunningham definitions should be considered bookends of possible scenarios based on different approaches for major bleeding case finding.
Our results build upon previous studies that have attempted to model the net benefit versus risk of aspirin therapy. In 2016, Mora and Manson created the aspirin guide based on 11 randomized clinical trials, and found that the net benefit of primary prevention aspirin therapy generally exceeded the risk of major bleeding events at a predicted 10‐year risk ≥10%. 31 In a separate meta‐analysis, 21% of women and 41% of men were found to derive net benefit from aspirin over 5 years if 1 ASCVD was modeled and assumed to be equivalent to 2 major bleeding events. 32
Our study should be interpreted in the setting of certain limitations. First, we used ICD codes to define major bleeding events, which were nonadjudicated and did not capture minor bleeding events. The best approach to quantifying major bleeding events from ICD codes remains uncertain; therefore, we tested 3 definitions that provide a range of major bleeding estimates, with aspirin ranging from most to least sensitive, across which our conclusions remained robust. Second, we did not have access to a history of peptic ulcer disease and/or Helicobacter pylori status on participants included in the analysis, both of which would have been important covariables to consider as a part of the high‐risk bleeding restriction. Notably, we removed individuals on proton pump inhibitors as a surrogate of a high‐risk bleeding condition; however, clinical data have suggested a potential 75% reduction in aspirin‐related gastrointestinal bleeding with proton pump inhibitor treatment, leaving this clinical scenario uncertain. Third, our study did not have information on platelet‐specific factors, including hyperreactivity, which could have differentially influenced CPS groups. Future translational studies in this research space should strive to assess the role of platelet‐specific factors to help inform the use of primary prevention aspirin therapy across subclinical atherosclerosis burden. Furthermore, our data were limited by the small number of participants with a high CPS; therefore, our results, for example with CPS 6, required modeling, and therefore estimates for NNT5 should be considered to have wide confidence bounds. Lastly, our CPS variable was rederived from the raw MESA and ARIC ultrasound data, because these cohorts were not designed with carotid ultrasound protocols focused on plaque burden. Nevertheless, this strategy is both timely and appropriate, with recent suggestions that plaque burden, not stenosis per se, is the predominant driver of ASCVD risk. 33 , 34
In summary, our results demonstrate that semiquantification by carotid plaque scoring may help to improve short‐term precision and personalization in primary prevention aspirin therapy allocation. Similar to CAC, carotid ultrasound imaging can help identify patients at higher risk of ASCVD, without a strong increase in the risk of bleeding. Therefore, it appears that carotid ultrasound can be used as an alternative to CAC to personalize aspirin allocation in primary prevention. Our results have implications for everyday practice around the world, with particular interest in the context of coming innovations in 3‐dimensional ultrasound 29 and point‐of‐care ultrasound. 35
Sources of Funding
This research was supported by an investigator‐initiated grant from Bayer. The ARIC study has been funded in whole or in part with federal funds from the National Heart, Lung, and Blood Institute, National Institutes of Health, Department of Health and Human Services, under contract numbers 75N92022D00001, 75N92022D00002, 75N92022D00003, 75N92022D00004, 75N92022D00005. Computed tomography scans to evaluate CAC in ARIC were supported by R01HL136592 (co‐principle investigators: Drs Matsushita and Blaha). Dr Dzaye received support from National Institutes of Health grant T32 HL007227. The MESA study was supported by contracts HHSN268201500003I, N01‐HC‐95159, N01‐HC‐95160, N01‐HC‐95161, N01‐HC‐95162, N01‐HC‐95163, N01‐HC‐95164, N01‐HC‐95165, N01‐HC‐95166, N01‐HC‐95167, N01‐HC‐95168, and N01‐HC‐95169 from the National Heart, Lung, and Blood Institute, and by grants UL1‐TR‐000040, UL1‐TR‐001079, and UL1‐TR‐001420 from the National Center for Advancing Translational Sciences. This publication was developed under the Science to Achieve Results research assistance agreements, number RD‐831697 (MESA Air) and RD‐83830001 (MESA Air Next Stage), awarded by the US Environmental Protection Agency. It has not been formally reviewed by the US Environmental Protection Agency. The views expressed in this document are solely those of the authors, and the US Environmental Protection Agency does not endorse any products or commercial services mentioned in this publication.
Disclosures
Dr Blaha reports grants from the National Institutes of Health, US Food and Drug Administration, American Heart Association, and Aetna Foundation; grants and advisory board activity with Amgen, Novo Nordisk, and Bayer; and advisory board activity from Novartis, Merck, Roche, Boehringer Ingelheim, Vectura, and Agepha. The remaining authors have no disclosures to report.
Supporting information
Data S1
Acknowledgments
The authors thank the other investigators, staff, and participants of the MESA study and ARIC study for their valuable contributions. A full list of participating MESA investigators and institutions can be found at http://www.mesa‐nhlbi.org.
This article was sent to Erik B. Schelbert, MD, MS, Associate Editor, for review by expert referees, editorial decision, and final disposition.
Supplemental Material is available at https://www.ahajournals.org/doi/suppl/10.1161/JAHA.123.034718
For Sources of Funding and Disclosures, see page 12.
References
- 1. Fernandes A, McEvoy JW, Halvorsen S. Doctor, should I keep taking an aspirin a day? N Engl J Med. 2019;380:1967–1970. doi: 10.1056/NEJMclde1903004 [DOI] [PubMed] [Google Scholar]
- 2. Raber I, McCarthy CP, Vaduganathan M, Bhatt DL, Wood DA, Cleland JGF, Blumenthal RS, McEvoy JW. The rise and fall of aspirin in the primary prevention of cardiovascular disease. Lancet. 2019;393:2155–2167. doi: 10.1016/S0140-6736(19)30541-0 [DOI] [PubMed] [Google Scholar]
- 3. Gaziano JM, Brotons C, Coppolecchia R, Cricelli C, Darius H, Gorelick PB, Howard G, Pearson TA, Rothwell PM, Ruilope LM, et al. Use of aspirin to reduce risk of initial vascular events in patients at moderate risk of cardiovascular disease (ARRIVE): a randomised, double‐blind, placebo‐controlled trial. Lancet. 2018;392:1036–1046. doi: 10.1016/S0140-6736(18)31924-X [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Group TASC . Effects of aspirin for primary prevention in persons with diabetes mellitus. N Engl J Med. 2018;379:1529–1539. doi: 10.1056/NEJMoa1804988 [DOI] [PubMed] [Google Scholar]
- 5. McNeil JJ, Woods RL, Nelson MR, Reid CM, Kirpach B, Wolfe R, Storey E, Shah RC, Lockery JE, Tonkin AM, et al. Effect of aspirin on disability‐free survival in the healthy elderly. N Engl J Med. 2018;379:1499–1508. doi: 10.1056/NEJMoa1800722 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Mahmoud AN, Gad MM, Elgendy AY, Elgendy IY, Bavry AA. Efficacy and safety of aspirin for primary prevention of cardiovascular events: a meta‐analysis and trial sequential analysis of randomized controlled trials. Eur Heart J. 2019;40:607–617. doi: 10.1093/eurheartj/ehy813 [DOI] [PubMed] [Google Scholar]
- 7. Zheng SL, Roddick AJ. Association of Aspirin use for primary prevention with cardiovascular events and bleeding events: a systematic review and meta‐analysis. JAMA. 2019;321:2245. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Grundy S, Stone N, Bailey A, Beam C, Birtcher K, Blumenthal R, Braun L, de Ferranti S, Faiella‐Tommasino J, Forman D, et al. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA guideline on the Management of Blood Cholesterol: a report of the American College of Cardiology/American Heart Association task force on clinical practice guidelines. Circulation. 2019;139:e1082–e1143. doi: 10.1161/CIR.0000000000000625 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Visseren FLJ, Mach F, Smulders YM, Carballo D, Koskinas KC, Bäck M, Benetos A, Biffi A, Boavida J‐M, Capodanno D, et al. ESC guidelines on cardiovascular disease prevention in clinical practice: developed by the task force for cardiovascular disease prevention in clinical practice with representatives of the European Society of Cardiology and 12 medical societies with the s. Eur Heart J. 2022;42:3227–3337. doi: 10.1093/eurheartj/ehab484 [DOI] [Google Scholar]
- 10. Rana JS, Tabada GH, Solomon MD, Lo JC, Jaffe MG, Sung SH, Ballantyne CM, Go AS. Accuracy of the atherosclerotic cardiovascular risk equation in a large contemporary, multiethnic population. J Am Coll Cardiol. 2016;10:2118–2130. doi: 10.1016/j.jacc.2016.02.055 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Blaha MJ, Blankstein R, Nasir K. Coronary artery calcium scores of zero and establishing the concept of negative risk factors. J Am Coll Cardiol. 2019;74:12–14. doi: 10.1016/j.jacc.2019.05.032 [DOI] [PubMed] [Google Scholar]
- 12. Miedema MD, Duprez DA, Misialek JR, Blaha MJ, Nasir K, Silverman MG, Blankstein R, Budoff MJ, Greenland P, Folsom AR. Use of coronary artery calcium testing to guide aspirin utilization for primary prevention: estimates from the multi‐ethnic study of atherosclerosis. Circ Cardiovasc Qual Outcomes. 2014;7:453–460. doi: 10.1161/CIRCOUTCOMES.113.000690 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Cainzos‐Achirica M, Miedema MD, McEvoy JW, Al Rifai M, Greenland P, Dardari Z, Budoff M, Blumenthal RS, Yeboah J, Duprez DA, et al. Coronary artery calcium for personalized allocation of aspirin in primary prevention of cardiovascular disease in 2019: the MESA study (multi‐ethnic study of atherosclerosis). Circulation. 2020;141:1541–1553. doi: 10.1161/CIRCULATIONAHA.119.045010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Orringer CE, Blaha MJ, Blankstein R, Budoff MJ, Goldberg RB, Gill EA, Maki KC, Mehta L, Jacobson TA. The National Lipid Association scientific statement on coronary artery calcium scoring to guide preventive strategies for ASCVD risk reduction. J Clin Lipidol. 2020;15:33–60. doi: 10.1016/j.jacl.2020.12.005 [DOI] [PubMed] [Google Scholar]
- 15. Mach F, Baigent C, Catapano AL, Koskinas KC, Casula M, Badimon L, Chapman MJ, De Backer GG, Delgado V, Ference BA, et al. 2019 ESC/EAS guidelines for the management of dyslipidaemias: lipid modification to reduce cardiovascular risk. Eur Heart J. 2020;41:111–188. doi: 10.1093/eurheartj/ehz455 [DOI] [PubMed] [Google Scholar]
- 16. Bild DE, Bluemke DA, Burke GL, Detrano R, Diez Roux AV, Folsom AR, Greenland P, Jacob DR, Kronmal R, Liu K, et al. Multi‐ethnic study of atherosclerosis: objectives and design. Am J Epidemiol. 2002;156:871–881. doi: 10.1093/aje/kwf113 [DOI] [PubMed] [Google Scholar]
- 17. Hill C, Gerardo D, James F, Tyroler HA, Chambless LE, Romm J, Disanto AR, Barr K, Bergsten J, Conrad J, et al. The atherosclerosis risk in communities (ARIC) study: design and objectives. Am J Epidemiol. 1989;129:687–702. doi: 10.1093/oxfordjournals.aje.a115184 [DOI] [PubMed] [Google Scholar]
- 18. Wright JD, Folsom AR, Coresh J, Sharrett AR, Couper D, Wagenknecht LE, Mosley TH, Ballantyne CM, Boerwinkle EA, Rosamond WD, et al. The ARIC (atherosclerosis risk in communities) study: JACC focus seminar 3/8. J Am Coll Cardiol. 2021;77:2939–2959. doi: 10.1016/j.jacc.2021.04.035 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Tattersall MC, Gassett A, Korcarz CE, Gepner AD, Kaufman JD, Liu KJ, Astor BC, Sheppard L, Kronmal RA, Stein JH. Predictors of carotid thickness and plaque progression during a decade: the multi‐ethnic study of atherosclerosis. Stroke. 2014;45:3257–3262. doi: 10.1161/STROKEAHA.114.005669 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Riley WA, Barnes RW, Bond MG, Evans G, Chambless LE, Heiss G. High‐resolution B‐mode ultrasound Reading methods in the atherosclerosis risk in communities (ARIC) cohort. J Neuroimaging. 1991;1:168–172. doi: 10.1111/jon199114168 [DOI] [PubMed] [Google Scholar]
- 21. Platt R, Carnahan RM, Brown JS, Chrischilles E, Curtis LH, Hennessy S, Nelson JC, Racoosin JA, Robb M, Schneeweiss S, et al. The U.S. Food and Drug Administration's mini‐sentinel program: status and direction. Pharmacoepidemiol Drug Saf. 2012;21:1–8. doi: 10.1002/pds.2343 [DOI] [PubMed] [Google Scholar]
- 22. Mehran R, Rao SV, Bhatt DL, Gibson CM, Caixeta A, Eikelboom J, Kaul S, Wiviott SD, Menon V, Nikolsky E, et al. Standardized bleeding definitions for cardiovascular clinical trials: a consensus report from the bleeding academic research consortium. Circulation. 2011;123:2736–2747. doi: 10.1161/CIRCULATIONAHA.110.009449 [DOI] [PubMed] [Google Scholar]
- 23. Whelton PK, Carey RM, Aronow WS, Casey DE, Collins KJ, Dennison Himmelfarb C, DePalma SM, Gidding S, Jamerson KA, Jones DW, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and Management of High Blood Pressure in adults: a report of the American College of Cardiology/American Heart Association task force on clinical Pr. Hypertension. 2018;71:e13–e115. doi: 10.1161/HYP.0000000000000065 [DOI] [PubMed] [Google Scholar]
- 24. Levin G, Kestenbaum B, Der Ida Chen Y, Jacobs DR, Psaty BM, Rotter JI, Siscovick DS, De Boer IH. Glucose, insulin, and incident hypertension in the multi‐ethnic study of atherosclerosis. Am J Epidemiol. 2010;172:1144–1154. doi: 10.1093/aje/kwq266 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low‐density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18:499–502. doi: 10.1093/clinchem/18.6.499 [DOI] [PubMed] [Google Scholar]
- 26. Goff DC, Lloyd‐Jones DM, Bennett G, Coady S, D'Agostino RB, Gibbons R, Greenland P, Lackland DT, Levy D, O'Donnell CJ, et al. 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American college of cardiology/American heart association task force on practice guidelines. Circulation. 2014;129:S49–S73. doi: 10.1161/01.cir.0000437741.48606.98 [DOI] [PubMed] [Google Scholar]
- 27. Arnett DK, Blumenthal RS, Albert MA, Buroker AB, Goldberger ZD, Hahn EJ, Himmelfarb CD, Khera A, Lloyd‐Jones D, McEvoy JW, et al. 2019 ACC/AHA guideline on the primary prevention of cardiovascular disease. Circulation. 2019;140:e596–e646. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Ajufo E, Ayers CR, Vigen R, Joshi PH, Rohatgi A, De Lemos JA, Khera A. Value of coronary artery calcium scanning in association with the net benefit of aspirin in primary prevention of atherosclerotic cardiovascular disease. JAMA Cardiol. 2020;6:246. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Mortensen MB, Fuster V, Muntendam P, Mehran R, Baber U, Sartori S, Falk E. A simple disease‐guided approach to personalize ACC/AHA‐recommended statin allocation in elderly people: the BioImage study. J Am Coll Cardiol. 2016;68:881–889. doi: 10.1016/j.jacc.2016.05.084 [DOI] [PubMed] [Google Scholar]
- 30. Mehta A, Rigdon J, Tattersall MC, German CA, Barringer TA, Joshi PH, Sperling LS, Budoff MJ, Bertoni A, Michos ED, et al. Association of carotid artery plaque with cardiovascular events and incident coronary artery calcium in individuals with absent coronary calcification: the MESA. Circ Cardiovasc Imaging. 2021;14:e011701. doi: 10.1161/CIRCIMAGING.120.011701 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Mora S, Manson JAE. Aspirin for primary prevention of atherosclerotic cardiovascular disease: advances in diagnosis and treatment. JAMA Intern Med. 2016;176:1195–1204. doi: 10.1001/jamainternmed.2016.2648 [DOI] [PubMed] [Google Scholar]
- 32. Selak V, Jackson R, Poppe K, Wu B, Harwood M, Grey C, Pylypchuk R, Mehta S, Choi YH, Kerr A, et al. Personalized prediction of cardiovascular benefits and bleeding harms from aspirin for primary prevention a benefit–harm analysis. Ann Intern Med. 2019;171:529–539. doi: 10.7326/M19-1132 [DOI] [PubMed] [Google Scholar]
- 33. Mortensen MB, Dzaye O, Steffensen FH, Bøtker HE, Jensen JM, Rønnow Sand NP, Kragholm KH, Sørensen HT, Leipsic J, Mæng M, et al. Impact of plaque burden versus stenosis on ischemic events in patients with coronary atherosclerosis. J Am Coll Cardiol. 2020;76:2803–2813. doi: 10.1016/j.jacc.2020.10.021 [DOI] [PubMed] [Google Scholar]
- 34. Dzaye O, Razavi AC, Blaha MJ, Mortensen MB. Evaluation of coronary stenosis versus plaque burden for atherosclerotic cardiovascular disease risk assessment and management. Curr Opin Cardiol. 2021;36:769–775. doi: 10.1097/HCO.0000000000000911 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Johri AM, Glass C, Hill B, Jensen T, Puentes W, Olusanya O, Capizzano JN, Dancel R, Reierson K, Reisinger N, et al. The evolution of cardiovascular ultrasound: a review of cardiac point‐of‐care ultrasound (POCUS) across specialties. Am J Med. 2023;136:621–628. doi: 10.1016/j.amjmed.2023.02.020 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data S1
Data Availability Statement
Per National Heart, Lung, and Blood Institute policy, all MESA and ARIC data is available online via the Biologic Specimen and Data Repository Information Coordinating Center (https://biolincc.nhlbi.nih.gov/home/).