Abstract
Background:
We sought to evaluate the associations of chest pain and dyspnea with the long-term risk of cardiovascular disease including coronary disease, heart failure, atrial fibrillation, and stroke.
Methods:
In 13,200 participants without cardiovascular disease in the Atherosclerosis Risk in Communities study (1987–1989), chest pain was categorized into definite angina, possible angina, non-anginal chest pain, and no chest pain using the Rose questionnaire. Dyspnea was categorized into grades 3–4, 2, 1, and 0 by the modified Medical Research Council scale. The associations of chest pain and dyspnea with incident myocardial infarction, heart failure, atrial fibrillation, and stroke over a median follow-up of ~27 years were quantified with multivariable Cox models.
Results:
Definite angina and possible angina were associated with myocardial infarction (adjusted hazard ratios [HR] 1.80 [95%CI 1.45–2.13] and 1.65 [1.27–2.15]). Although lesser magnitude than myocardial infarction, both definite and possible angina were associated with heart failure. For atrial fibrillation, possible angina showed higher HR than definite angina. Dyspnea showed similar HRs for myocardial infarction and heart failure in grades 3–4 (2.00 [1.61–2.49] and 1.94 [1.62–2.32]). Stroke was least associated with chest symptoms. Chest pain and dyspnea significantly improved the discrimination of cardiovascular disease except stroke, beyond traditional risk factors.
Conclusions:
In individuals without cardiovascular disease, chest pain and dyspnea were independently associated with incident cardiovascular disease for about three decades, suggesting the need for evaluating chest pain from a broader perspective of cardiovascular disease beyond coronary disease and the importance of dyspnea for cardiovascular risk assessment including myocardial infarction.
Keywords: Chest pain, dyspnea, myocardial infarction, heart failure, atrial fibrillation
Introduction
Chest pain is one of the most common causes of seeking medical care.1–3 Diagnosis of cardiovascular disease is challenging for many reasons including overlaps of chest pain with numerous, less serious conditions that require different diagnostic and therapeutic approaches. A cornerstone of the current US and European guidelines for chest pain4, 5 is the risk classification of having coronary artery disease (e.g., an assessment for pre-test probability of coronary artery disease) according to demographics, clinical conditions, and types of chest symptoms.
The emphasis on the probability of having coronary artery disease as a cause of chest pain is crucial since coronary artery disease can be lethal and sometimes requires urgent care (e.g., acute coronary syndrome). Simultaneously, it is important to understand the long-term prognostic implications of chest pain since some patients with chest symptoms may have low risk of having coronary disease at that moment but subsequently develop coronary disease or other cardiovascular diseases (e.g., heart failure and atrial fibrillation).
Several community-based studies have explored the prospective association of chest symptoms with the risk of cardiovascular disease,6–11 but there are important caveats. Most studies evaluated coronary artery disease or mortality6–9, 11 but not other relevant cardiovascular outcomes such as heart failure and atrial fibrillation. Those studies also focused on typical angina (e.g., chest pain on exertion localized at anterior chest and relieved by resting),6–11 whereas other chest symptoms (e.g., non-anginal/nonexertional chest pain and dyspnea) may provide unique prognostic information. Furthermore, most studies included homogenous study populations composed of only White individuals or only male participants,6–10 and the follow-up was relatively short (<10 years).6–9, 11
To overcome these caveats, using data from a community-based cohort including Blacks and Whites in the Atherosclerosis Risk in Communities (ARIC) Study, we evaluated different types of chest pain and the severity of dyspnea and their associations with the subsequent risk of cardiovascular outcomes (myocardial infarction, heart failure, atrial fibrillation, and ischemic stroke) during long-term of follow-up.
Methods
Study population
The ARIC Study is a prospective cohort study including 15,792 participants aged 45 to 64 years at visit 1 (1987–1989) from four US communities (Forsyth County, North Carolina; Jackson, Mississippi; suburban Minneapolis, Minnesota; and Washington County, Maryland).12, 13 ARIC was approved by the Institutional Review Board of the participating institutions. All participants provided written informed consent.
Of 15,792 participants at visit 1, a total of 13,200 participants without prior history of myocardial infarction, coronary revascularization, heart failure, atrial fibrillation, or stroke were included in the analysis for chest pain (Supplementary Table 1 and Supplementary Figure 1). Dyspnea grade was not assessed among 573 participants with walking disability due to comorbidities other than heart or lung disease (e.g., arthritis). Thus, 12,627 participants were included in the analysis of dyspnea.
Chest symptoms
Chest pain was categorized into four groups (definite angina, possible angina, non-anginal chest pain, and no chest pain) by the location, severity, frequency of the symptom according to the World Health Organization (WHO) Rose Angina Questionnaire (Supplementary Table 2 and Supplementary Figure 2).14–16 Dyspnea was also categorized into four groups (grade 3–4, 2, 1 and 0) by the severity of the symptom based on the Modified Medical Research Council Dyspnea Scale (Supplementary Table 2 and Supplementary Figure 3).17, 18
Incident cardiovascular disease
We investigated fatal and non-fatal initial events of myocardial infarction, heart failure, atrial fibrillation, and ischemic stroke. Their definitions are summarized in Supplementary Table 3.12, 19–23 Participants were followed up from visit 1 until outcomes of interest, death, loss to follow-up, or the end of follow-up (at the Jackson sites on December 31, 2017 and at the other sites on December 31, 2017 for atrial fibrillation and on December 31, 2019 for the other cardiovascular outcomes).
Covariates
The list of covariates and their definitions are displayed in Supplemental Table 4.
Statistical analysis
We compared baseline characteristics across the groups of chest pain and dyspnea, respectively. Continuous variables were summarized as mean values with standard deviations; categorical variables were described as counts with proportions. We assessed Spearman’s rank correlation coefficient between chest pain and dyspnea.
For the survival analysis, we first estimated the cumulative incidence of cardiovascular outcomes using the cumulative incidence function, accounting non-cardiovascular deaths as a competing outcome with the comparison among groups by Gray’s test.24 Since we were interested in the etiological associations of exposures and outcomes,25 we ran multivariable Cox proportional hazard models. Model 1 was adjusted for age, sex, race and field center and Model 2 was further adjusted for body mass index, educational level, physical activity, estimated glomerular filtration rate, diabetes, smoking history, systolic and diastolic blood pressure, heart rate, total and high-density lipoprotein cholesterol levels, antihypertensive and lipid-lowering medication use. Model 3 included both chest pain and dyspnea to adjust for each other in addition to the covariates in Model 2.
We conducted several sensitivity analyses to confirm the robustness of our primary findings. First, we conducted subgroup analysis stratified by baseline age (< vs. ≥54 years [median]), sex, race, educational levels, and diabetes status. We explored education and diabetes since socioeconomic status may affect healthcare access when chest symptoms occur and since diabetes patients may not manifest chest pain due to autonomic neuropathy.26 To obtain reliable estimates in each subgroup, we modeled chest pain and dyspnea as three groups—instead of the original four groups—chest pain (definite/possible angina, non-anginal chest pain, and no chest pain) and dyspnea (grade 2–4, grade 1, and grade 0). Second, to assess joint associations of chest pain and dyspnea, we assessed their cross-categories and the subsequent risk of cardiovascular outcomes. We explored nine cross-categories by the aforementioned three groups of chest pain and dyspnea, respectively. We tested interaction using likelihood ratio tests. Third, to minimize the potential impact of undiagnosed cardiovascular disease at baseline, we censored incident cardiovascular outcomes within two years from baseline. Finally, although our primary analysis used Cox models, we repeated the analysis using Fine-Gray models accounting for the competing risk of non-cardiovascular death.27
We also explored whether chest symptoms improve the discrimination of future cardiovascular disease beyond traditional risk factors. We evaluated Harrell’s c-statistic using Cox models, with covariates in Model 2 as the base model. C-statistic indicates the probability of cases having higher predicted risk than non-cases, and its value ≥0.7 is often considered acceptable for cardiovascular risk prediction.28 We considered two-sided p<0.05 as statistically significant. All analyses were performed using R 4.2.0 version (The R Foundation for Statistical Computing, Vienna, Austria).
Results
Baseline characteristics
Participants with definite or possible angina were more likely to have cardiovascular risk factors (e.g., older age and diabetes) compared to those with non-anginal chest pain or no chest pain (Table 1), but participants with definite angina and those with possible angina had largely similar characteristics. Similarly, participants with more severe dyspnea generally had more cardiovascular risk factors (Supplementary Table 5). Chest pain and dyspnea were weakly correlated (Spearman’s ρ = 0.20) (Supplementary Table 6).
Table 1.
Baseline characteristics according to chest pain (WHO Rose Angina Questionnaire): The ARIC Study (1987–1989)
| Variables | Overall | No chest pain | Non-anginal chest pain | Possible angina | Definite angina |
|---|---|---|---|---|---|
| Participants, N | 13200 | 8052 | 4327 | 273 | 548 |
| Age, yrs | 53.9 (5.7) | 54.0 (5.8) | 53.7 (5.7) | 54.7 (5.5) | 54.5 (5.6) |
| Male | 5811 (44.0) | 3717 (46.2) | 1858 (42.9) | 84 (30.8) | 152 (27.7) |
| Black | 3322 (25.2) | 1999 (24.8) | 1146 (26.5) | 63 (23.1) | 114 (20.8) |
| Body Mass Index, kg/m2 | 27.5 (5.2) | 27.4 (5.2) | 27.5 (5.1) | 28.0 (5.8) | 28.4 (6.2) |
| Educational level | |||||
| Basic | 2863 (21.7) | 1596 (19.8) | 996 (23.0) | 90 (33.0) | 181 (33.0) |
| Intermediate | 5464 (41.4) | 3331 (41.4) | 1774 (41.0) | 124 (45.4) | 235 (42.9) |
| Advanced | 4873 (36.9) | 3125 (38.8) | 1557 (36.0) | 59 (21.6) | 132 (24.1) |
| Physical activity index | |||||
| Work | 2.2 (0.9) | 2.2 (0.9) | 2.2 (0.9) | 2.2 (1.1) | 2.2 (1.0) |
| Sport | 2.4 (0.8) | 2.5 (0.8) | 2.4 (0.8) | 2.3 (0.7) | 2.2 (0.7) |
| Leisure | 2.4 (0.6) | 2.4 (0.6) | 2.3 (0.6) | 2.3 (0.6) | 2.3 (0.6) |
| Systolic blood pressure, mmHg | 121 (19) | 121 (19) | 120 (18) | 122 (18) | 121 (19) |
| Diastolic blood pressure, mmHg | 74 (11) | 74 (11) | 73 (11) | 73 (11) | 72 (11) |
| Antihypertensive medication | 2816 (21.3) | 1640 (20.4) | 946 (21.9) | 78 (28.6) | 152 (27.7) |
| Heart rate, bpm | 67 (10) | 66 (10) | 67 (10) | 68 (11) | 67 (11) |
| Total cholesterol, mmol/L | 5.5 (1.1) | 5.5 (1.1) | 5.5 (1.1) | 5.6 (1.1) | 5.6 (1.1) |
| HDL cholesterol, mmol/L | 1.4 (0.4) | 1.4 (0.4) | 1.3 (0.4) | 1.4 (0.4) | 1.3 (0.4) |
| Lipid-lowering medication | 322 (2.4) | 175 (2.2) | 113 (2.6) | 13 (4.8) | 21 (3.8) |
| eGFR, ml/min | 102.1 (12.7) | 102.0 (12.7) | 102.3 (12.4) | 102.5 (14.1) | 102.2 (13.5) |
| Diabetes | 1341 (10.2) | 793 (9.8) | 425 (9.8) | 40 (14.7) | 83 (15.1) |
| Current smoker | 3379 (25.6) | 2023 (25.1) | 1130 (26.1) | 72 (26.4) | 154 (28.1) |
| Former smoker | 4184 (31.7) | 2552 (31.7) | 1412 (32.6) | 71 (26.0) | 149 (27.2) |
| Field center | |||||
| Forsyth county, NC | 3479 (26.4) | 2032 (25.2) | 1193 (27.6) | 75 (27.5) | 179 (32.7) |
| Jackson, MS | 2894 (21.9) | 1764 (21.9) | 996 (23.0) | 52 (19.0) | 82 (15.0) |
| Minneapolis, MN | 3559 (27.0) | 2322 (28.8) | 1073 (24.8) | 66 (24.2) | 98 (17.9) |
| Washington county, MD | 3268 (24.8) | 1934 (24.0) | 1065 (24.6) | 80 (29.3) | 189 (34.5) |
Continuous variables with mean (standard deviation) and categorical variables with number (%) were described.
ARIC, Atherosclerosis Risk in Communities; WHO, World Health Organization; HDL, high-density lipoprotein; eGFR, estimated glomerular filtration rate; NC, North Carolina; MS, Mississippi; MN, Minnesota; and MD, Maryland
Chest pain and subsequent cardiovascular events
During a median follow-up of ~27 years (Supplementary Table 7), there were 2,007 cases of myocardial infarction, 3,035 heart failure cases, 2,658 atrial fibrillation cases, and 1,109 ischemic stroke cases. The cumulative incidence of myocardial infarction and heart failure was highest in definite angina, followed by possible angina, non-anginal chest pain, and no chest pain (Figure 1A and 1B). The pattern was similar for atrial fibrillation up to 15 years of follow-up, but then cumulative incidence in possible angina surpassed that in definite angina (Figure 1C). The separation was least evident for ischemic stroke, although cumulative incidences in definite angina and possible angina were higher than that in the remaining two categories (Figure 1D).
Figure 1.
Cumulative incidence of cardiovascular disease according to chest pain Cumulative incidence of cardiovascular disease (A, myocardial infarction; B, heart failure; C, atrial fibrillation; and D, ischemic stroke) accounting for competing risk (death) according to chest pain (lines: orange, definite angina; yellow, possible angina; steel blue, non-anginal chest pain; and light blue, no chest pain).
The associations of chest pain with cardiovascular outcomes remained consistent after adjusting for traditional CVD risk factors (Models 1 and 2 in upper half of Table 2). Both definite angina and possible angina were associated with myocardial infarction, with similar hazard ratios (HRs) in Model 2. Although of a lesser magnitude than myocardial infarction, both definite angina and possible angina were associated with incident heart failure as well. For atrial fibrillation, possible angina demonstrated higher HR than definite angina. For ischemic stroke, only definite angina remained statistically significant in Model 2. Even non-anginal chest pain was significantly associated with myocardial infarction, heart failure, and atrial fibrillation although the HRs were less than 1.2. The further adjustment for dyspnea severity attenuated the associations although the general patterns remained consistent (Model 3 in Table 2).
Table 2.
Chest symptoms and subsequent risk of cardiovascular disease
| Chest pain | No chest pain | Non-anginal chest pain | Possible angina | Definite angina |
|---|---|---|---|---|
|
| ||||
| N event/N total (Models 1 and 2, and 3) and Hazard ratio (95% CI) | ||||
|
|
||||
| Myocardial infarction | 1112/8052 and 1066/7794 | 708/4327 and 656/4085 | 59/273 and 50/252 | 128/548 and 110/496 |
| Model 1† | 1 (reference) | 1.25 (1.13–1.37) | 1.88 (1.44–2.44) | 2.10 (1.75–2.53) |
| Model 2† | 1 (reference) | 1.23 (1.12–1.36) | 1.65 (1.27–2.15) | 1.80 (1.45–2.13) |
| Model 3‡ | 1 (reference) | 1.19 (1.08–1.31) | 1.39 (1.04–1.86) | 1.50 (1.22–1.85) |
| Heart failure | 1737/8052 and 1641/7794 | 1052/4327 and 972/4085 | 75/273 and 63/252 | 171/548 and 151/496 |
| Model 1† | 1 (reference) | 1.19 (1.10–1.28) | 1.48 (1.17–1.86) | 1.73 (1.47–2.02) |
| Model 2† | 1 (reference) | 1.16 (1.07–1.25) | 1.32 (1.04–1.66) | 1.53 (1.31–1.80) |
| Model 3‡ | 1 (reference) | 1.14 (1.05–1.23) | 1.10 (0.85–1.42) | 1.32 (1.11–1.57) |
| Atrial fibrillation | 1539/8052 and 1468/7794 | 916/4327 and 850/4085 | 75/273 and 67/252 | 128/548 and 115/496 |
| Model 1† | 1 (reference) | 1.17 (1.08–1.27) | 1.70 (1.35–2.15) | 1.42 (1.18–1.70) |
| Model 2† | 1 (reference) | 1.16 (1.07–1.26) | 1.60 (1.27–2.02) | 1.32 (1.10–1.59) |
| Model 3‡ | 1 (reference) | 1.15 (1.05–1.25) | 1.50 (1.17–1.92) | 1.21 (0.99–1.47) |
| Ischemic stroke | 653/8052 and 627/7794 | 362/4327 and 329/4085 | 29/273 and 26/252 | 65/548 and 51/496 |
| Model 1† | 1 (reference) | 1.07 (0.94–1.21) | 1.49 (1.03–2.17) | 1.71 (1.33–2.22) |
| Model 2† | 1 (reference) | 1.07 (0.94–1.21) | 1.33 (0.91–1.93) | 1.59 (1.23–2.06) |
| Model 3‡ | 1 (reference) | 1.03 (0.90–1.18) | 1.29 (0.87–1.93) | 1.30 (0.97–1.76) |
|
| ||||
| Dyspnea grade | Grade 0 | Grade 1 | Grade 2 | Grade 3–4 |
|
| ||||
| N event/N total and hazard ratio (95% CI) | ||||
|
|
||||
| Myocardial infarction | 1235/9168 | 462/2609 | 91/480 | 94/370 |
| Model 1‡ | 1 (reference) | 1.53 (1.37–1.71) | 2.02 (1.63–2.51) | 2.87 (2.32–3.54) |
| Model 2‡ | 1 (reference) | 1.24 (1.11–1.39) | 1.33 (1.06–1.65) | 2.00 (1.61–2.49) |
| Model 3‡ | 1 (reference) | 1.19 (1.06–1.33) | 1.21 (0.96–1.51) | 1.81 (1.44–2.26) |
| Heart failure | 1804/9168 | 718/2609 | 169/480 | 136/370 |
| Model 1‡ | 1 (reference) | 1.59 (1.46–1.74) | 2.64 (2.25–3.09) | 2.94 (2.47–3.51) |
| Model 2‡ | 1 (reference) | 1.27 (1.16–1.39) | 1.65 (1.40–1.94) | 1.94 (1.62–2.32) |
| Model 3‡ | 1 (reference) | 1.24 (1.13–1.36) | 1.56 (1.32–1.84) | 1.83 (1.52–2.20) |
| Atrial fibrillation | 1684/9168 | 617/2609 | 115/480 | 84/370 |
| Model 1‡ | 1 (reference) | 1.42 (1.29–1.55) | 1.93 (1.59–2.33) | 2.02 (1.62–2.51) |
| Model 2‡ | 1 (reference) | 1.22 (1.11–1.35) | 1.44 (1.18–1.75) | 1.56 (1.24–1.95) |
| Model 3‡ | 1 (reference) | 1.18 (1.07–1.31) | 1.35 (1.11–1.65) | 1.47 (1.17–1.85) |
| Ischemic stroke | 710/9168 | 236/2609 | 50/480 | 37/370 |
| Model 1‡ | 1 (reference) | 1.31 (1.13–1.52) | 1.84 (1.38–2.46) | 1.75 (1.26–2.45) |
| Model 2‡ | 1 (reference) | 1.11 (0.95–1.29) | 1.36 (1.01–1.83) | 1.33 (0.94–1.86) |
| Model 3‡ | 1 (reference) | 1.08 (0.93–1.26) | 1.28 (0.95–1.74) | 1.24 (0.88–1.76) |
Bold texts indicate statistical significance.
Analyses included participants completed chest pain assessment (N = 13,200).
Analyses included participants completed dyspnea grade assessment (N = 12,627).
Model 1 was adjusted for age, sex, race, and field center; Model 2 was further adjusted for body mass index, educational level, physical activity index (work, sport, and leisure), estimated glomerular filtration rate, diabetes, smoking history, systolic and diastolic blood pressure, heart rate, total and high-density lipoprotein cholesterol levels, antihypertensive and lipid-lowering medication use. Model 3 included both chest pain and dyspnea to adjust for each other on top of the covariates in Model 2.
Dyspnea and subsequent cardiovascular disease
There was a dose-response relationship between dyspnea grade and incident myocardial infarction and heart failure, although the cumulative incidence of heart failure was similar between grade 3–4 and grade 2 dyspnea (Figure 2A and 2B). The pattern was similar for atrial fibrillation up to 15–20 years of follow-up, but then cumulative incidences in grade 3–4, grade 2, and grade 1 became similar (Figure 2C). Again, the separation was least evident for ischemic stroke (Figure 2D).
Figure 2.
Cumulative incidence of cardiovascular disease according to dyspnea grade Cumulative incidence of cardiovascular disease (A, myocardial infarction; B, heart failure; C, atrial fibrillation; and D, ischemic stroke) accounting for competing risk (death) according to dyspnea grade (lines: red, grade 3–4; pink, grade 2; yellow green, grade 1; and green, grade 0).
Even after adjusting for traditional risk factors (Model 2 in lower half of Table 2), there was a dose-response association of dyspnea with incident myocardial infarction and heart failure. With grade 0 as the reference, grade 3–4 demonstrated similar HRs of myocardial infarction and heart failure in Model 2, whereas HR in grade 2 dyspnea tended to be greater for heart failure. Grades 3–4 and grade 2 showed similar HRs for atrial fibrillation and ischemic stroke. Grade 1 dyspnea was also associated with myocardial infarction, heart failure, and atrial fibrillation in Model 2, but HRs were less than 1.25. The further adjustment for chest pain slightly attenuated the results (Model 3 in lower half of Table 2).
Sensitivity analysis
Generally consistent results were seen across different subgroups (Supplementary Figures 4-7). Among those subgroups, non-anginal chest pain was more strongly associated with heart failure in Blacks than in Whites (p-for-interaction 0.04 in Supplementary Figure 5B), whereas the association of grade 2–4 dyspnea with heart failure was weaker in Blacks compared to Whites (p-for-interaction: 0.01 in Supplementary Figure 6B).
Analyzing cross-categories, we confirmed that chest pain and dyspnea were jointly associated with the risk of all cardiovascular outcomes (Figure 3). The HR of myocardial infarction, heart failure, and atrial fibrillation exceeded 2 in the cross-category of definite/possible angina plus grade 2–4 dyspnea compared to the reference of no chest pain plus grade 0 dyspnea. Chest pain and dyspnea showed a significant interaction in their associations with heart failure (p-for-interaction: 0.01) but not with the other cardiovascular outcomes.
Figure 3.
Joint associations of chest pain and dyspnea with cardiovascular disease Each cell of tables (A, myocardial infarction; B, heart failure; C, atrial fibrillation; and D, ischemic stroke) indicates the hazard ratio and 95% CI of cardiovascular disease according to the combination between chest pain and dyspnea compared to both no chest pain and grade 0 dyspnea as reference. The cells are highlighted depending on the hazard ratio (≤1.00, white; 1.01–1.50, yellow; 1.51–2.00, orange; and ≥2.01, red).
The results were largely similar when we censored incident cardiovascular disease within two years after baseline or accounted for competing risk of non-cardiovascular deaths (Supplementary Tables 8 and 9).
Prediction of future cardiovascular events with chest symptoms
For prediction of cardiovascular outcomes, the base model with traditional predictors (i.e., covariates in Model 2) showed a c-statistic ranging from 0.709–0.760 (Supplementary Table 10). Adding either chest pain or dyspnea to the base model slightly but significantly improved the discrimination of cardiovascular outcomes except stroke. Chest pain improved prediction of myocardial infarction most, followed by atrial fibrillation and heart failure; dyspnea improved prediction of heart failure most, followed by myocardial infarction. The improvement of prediction of myocardial infarction was identical when we added chest pain or dyspnea. Simultaneously adding chest pain and dyspnea led to even better improvement in prediction of myocardial infarction, heart failure, and atrial fibrillation.
Discussion
In this community-based study of Blacks and Whites, chest pain was significantly associated with the long-term risk of cardiovascular disease—particularly myocardial infarction. Specifically, definite angina and possible angina showed robust associations with myocardial infarction and heart failure, but for atrial fibrillation, possible angina had a greater HR than definite angina. Dyspnea was also significantly associated with all cardiovascular outcomes, particularly myocardial infarction and heart failure, in a graded manner. Stroke was least associated with both chest symptoms. The associations of chest pain and dyspnea with cardiovascular outcomes were generally independent of each other. The addition of chest symptoms improved cardiovascular risk prediction beyond traditional predictors.
The association of chest pain with the risk of myocardial infarction has been reported previously,6–11 but there are several unique aspects of our study. First, we studied cardiovascular outcomes other than myocardial infarction such as heart failure and atrial fibrillation. Second, we could meticulously evaluate different types of chest pain. Third, we investigated dyspnea, another important chest symptom, in parallel with chest pain. Fourth, we characterized cardiovascular risk according to chest symptoms over three decades. Fifth, we assessed their associations among different demographic and clinical subgroups. Finally, we rigorously evaluated whether chest symptoms can improve cardiovascular risk prediction.
It is not surprising that definite angina was strongly associated with myocardial infarction. However, there are several important findings related to chest pain in our study. First, in the long run, definite angina and possible angina had a similar risk of myocardial infarction. This suggests that clinicians should not downplay too much some deviation from typical chest pain when they follow patients with chest pain. Second, both definite angina and possible angina were associated with heart failure and atrial fibrillation, indicating that clinicians should keep these cardiovascular outcomes in mind when they monitor patients with chest pain. Potential mechanisms behind our observation include coronary artery disease as a risk factor for heart failure 29. Interestingly, for atrial fibrillation, possible angina showed an even stronger association than definite angina, possibly reflecting some individuals feeling arrhythmias as atypical chest pain.30
The association of dyspnea with the risk of heart failure is intuitive and has indeed been previously reported.31 In our study, however, dyspnea was associated with heart failure and myocardial infarction similarly, suggesting the importance of dyspnea for the risk assessment of myocardial infarction. In fact, dyspnea significantly improved the prediction of myocardial infarction beyond traditional risk factors, and the improvement was similar to chest pain. More importantly, the combination of chest pain and dyspnea showed the best prediction improvement for myocardial infarction as well as heart failure and atrial fibrillation. Thus, the simultaneous assessment of chest pain and dyspnea would be optimal for cardiovascular risk management and monitoring.
The evaluation of chest symptoms can be done even in resource-limited settings. However, it may be time-consuming to evaluate these symptoms systematically in every patient. A feasible option would be to ask patients to fill out chest symptom questionnaires while waiting in clinics. Also, smartphone applications that can record chest symptoms may be useful.32 Any additional tests to rule in/out cardiovascular disease confer cost and/or risk of adverse outcomes (e.g., contrast-induced nephropathy or anxiety due to further examination). Thus, optimal approaches to monitor and manage individuals with chest symptoms, especially mild ones, should be based on comprehensive consideration of the benefits and potential harm and should be explored in future investigations.
The present study has several limitations. First, because our study included only Blacks and Whites aged 45 to 64 years, the extrapolation of our findings to other racial or age groups needs careful consideration. Second, since chest symptom assessments were conducted at study visits, our findings will not apply to unstable or persistent chest symptoms in acute settings. Third, heart failure and atrial fibrillation were based on ICD codes and not adjudicated by the physician panel. However, previous ARIC studies have reported positive predictive value of almost 80– 90% compared to physician review.20, 22 Finally, although we did our best to account for potential confounders, we cannot deny the possibility of residual confounding.
In conclusion, chest pain and dyspnea were both associated with the incidence of multiple cardiovascular outcomes, independent of each other and potential confounders. Chest pain was most strongly associated with myocardial infarction, whereas dyspnea was similarly related to myocardial infarction and heart failure. Our findings suggest the need for evaluating chest pain from a broader perspective of cardiovascular disease beyond coronary disease and the importance of recognizing dyspnea in addition to chest pain for the risk management of cardiovascular disease, including myocardial infarction.
Supplementary Material
Acknowledgement
The authors thank the staff and participants of the ARIC study for their important contributions.
Sources of Funding:
The ARIC Study is carried out as a collaborative study supported by National Heart, Lung, and Blood Institute contracts (75N92022D00001, 75N92022D00002, 75N92022D00003, 75N92022D00004, 75N92022D00005, and 75N92022D00005). PLL was partially supported by K24 HL159246. The funders of this study had no role in study design; collection, analysis, and interpretation of data; writing the report; and the decision to submit the report for publication.
Abbreviations:
- ARIC
Atherosclerosis Risk in Communities
- CI
confidence interval
- eGFR
Estimated glomerular filtration rate
- HR
hazard ratio
- WHO
World Health Organization
Footnotes
Conflict of interest disclosure:
Dr Rosamond received grants from National Heart, Lung, and Blood Institute.
Dr Shah received consulting fees from Phillips Ultrasound and Jansen.
Dr Blaha received grants from Amgen, American Heart Association, Bayer, Food and Drug Administration, National Institute of Health, and Novo Nordisk; consulting fees from Agepha, Astra Zeneca, Bayer, Boehringer Ingelheim, Eli Lily, Merck, Novartis, and Novo Nordisk; honoraria for lectures from Novo Nordisk; payment for expert testimony from Novo Nordisk and US FTC.
Dr Mathews received grants from National Heart, Lung, and Blood Institute.
Dr Matsushita received grants from Resolve to Save Lives; consulting fees from AMGA, Kowa Company, and Rhythm X AI; honoraria for lectures fee from Fukuda Denshi.
The other authors have nothing to declare.
Authorship:
We verify that all authors had access to the data and a role in writing the manuscript (see details below).
Drs Matsushita and Ejiri had full access to all the data in the study and take responsibility of the integrity of the data and the accuracy of the data analysis.
Authors roles:
Dr Ejiri: Conceptualization, Data analysis, Interpretation of data, Writing – original draft; Dr Mok: Data acquisition, Interpretation of data, Writing – review & editing; Drs Ding, Chang, Rosamond, Shah, Lutsey, Chen, Blaha and Mathews: Interpretation of data, Writing – review & editing; and Dr Matsushita: Conceptualization, Supervision, Data acquisition, Interpretation of data, Writing – original draft.
Preprint server disclosure:
There is no preprint publication.
Abstract presentation:
The American Heart Association Scientific Session 2022, Chicago, 11/6/2022
Conflict of interest disclosure: None.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Reference
- 1.Rui PK K. National Hospital Ambulatory Medical Care Survey: 2017 emergency department summary tables. Vol 20232021. [Google Scholar]
- 2.Virani SS, Alonso A, Benjamin EJ, et al. Heart Disease and Stroke Statistics-2020 Update: A Report From the American Heart Association. Circulation. 2020;141:e139–e596. doi: 10.1161/cir.0000000000000757. [DOI] [PubMed] [Google Scholar]
- 3.Johnson K, Ghassemzadeh S. Chest Pain. StatPearls [Internet]. Vol 2023. Treasure Island (FL): StatPearls Publishing; 2023. [PubMed] [Google Scholar]
- 4.Gulati M, Levy PD, Mukherjee D, et al. 2021 AHA/ACC/ASE/CHEST/SAEM/SCCT/SCMR Guideline for the Evaluation and Diagnosis of Chest Pain: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation. 2021;144:e368–e454. doi: 10.1161/cir.0000000000001029. [DOI] [PubMed] [Google Scholar]
- 5.Knuuti J, Wijns W, Saraste A, et al. 2019 ESC Guidelines for the diagnosis and management of chronic coronary syndromes. Eur. Heart J. 2020;41:407–477. doi: 10.1093/eurheartj/ehz425. [DOI] [PubMed] [Google Scholar]
- 6.Cook DG, Shaper AG, MacFarlane PW. Using the WHO (Rose) angina questionnaire in cardiovascular epidemiology. Int. J. Epidemiol. 1989;18:607–613. doi: 10.1093/ije/18.3.607. [DOI] [PubMed] [Google Scholar]
- 7.LaCroix AZ, Guralnik JM, Curb JD, Wallace RB, Ostfeld AM, Hennekens CH. Chest pain and coronary heart disease mortality among older men and women in three communities. Circulation. 1990;81:437–446. doi: 10.1161/01.cir.81.2.437. [DOI] [PubMed] [Google Scholar]
- 8.Lampe FC, Whincup PH, Wannamethee SG, Ebrahim S, Walker M, Shaper AG. Chest pain on questionnaire and prediction of major ischaemic heart disease events in men. Eur. Heart J. 1998;19:63–73. doi: 10.1053/euhj.1997.0729. [DOI] [PubMed] [Google Scholar]
- 9.Owen-Smith V, Hannaford PC, Elliott AM. Increased mortality among women with Rose angina who have not presented with ischaemic heart disease. Br. J. Gen. Pract. 2003;53:784–789. [PMC free article] [PubMed] [Google Scholar]
- 10.Murphy NF, Stewart S, Hart CL, MacIntyre K, Hole D, McMurray JJ. A population study of the long-term consequences of Rose angina: 20-year follow-up of the Renfrew-Paisley study. Heart. 2006;92:1739–1746. doi: 10.1136/hrt.2006.090118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Achterberg S, Soedamah-Muthu SS, Cramer MJ, Kappelle LJ, van der Graaf Y, Algra A. Prognostic value of the Rose questionnaire: a validation with future coronary events in the SMART study. Eur J Prev Cardiol. 2012;19:5–14. doi: 10.1177/1741826710391117. [DOI] [PubMed] [Google Scholar]
- 12.The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. The ARIC investigators. Am. J. Epidemiol. 1989;129:687–702. [PubMed] [Google Scholar]
- 13.Wright JD, Folsom AR, Coresh J, 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]
- 14.Rose G, McCartney P, Reid DD. Self-administration of a questionnaire on chest pain and intermittent claudication. Br. J. Prev. Soc. Med. 1977;31:42–48. doi: 10.1136/jech.31.1.42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Rose GA. The diagnosis of ischaemic heart pain and intermittent claudication in field surveys. Bull. World Health Organ. 1962;27:645–658. [PMC free article] [PubMed] [Google Scholar]
- 16.Sorlie PD, Cooper L, Schreiner PJ, Rosamond W, Szklo M. Repeatability and validity of the Rose questionnaire for angina pectoris in the Atherosclerosis Risk in Communities Study. J. Clin. Epidemiol. 1996;49:719–725. doi: 10.1016/0895-4356(96)00022-4. [DOI] [PubMed] [Google Scholar]
- 17.Fletcher CM, Elmes PC, Fairbairn AS, Wood CH. The significance of respiratory symptoms and the diagnosis of chronic bronchitis in a working population. Br. Med. J. 1959;2:257–266. doi: 10.1136/bmj.2.5147.257. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Bestall JC, Paul EA, Garrod R, Garnham R, Jones PW, Wedzicha JA. Usefulness of the Medical Research Council (MRC) dyspnoea scale as a measure of disability in patients with chronic obstructive pulmonary disease. Thorax. 1999;54:581–586. doi: 10.1136/thx.54.7.581. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Astor BC, Coresh J, Heiss G, Pettitt D, Sarnak MJ. Kidney function and anemia as risk factors for coronary heart disease and mortality: the Atherosclerosis Risk in Communities (ARIC) Study. Am. Heart J. 2006;151:492–500. doi: 10.1016/j.ahj.2005.03.055. [DOI] [PubMed] [Google Scholar]
- 20.Shah AM, Claggett B, Loehr LR, et al. Heart Failure Stages Among Older Adults in the Community: The Atherosclerosis Risk in Communities Study. Circulation. 2017;135:224–240. doi: 10.1161/circulationaha.116.023361. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Alonso A, Lopez FL, Matsushita K, et al. Chronic kidney disease is associated with the incidence of atrial fibrillation: the Atherosclerosis Risk in Communities (ARIC) study. Circulation. 2011;123:2946–2953. doi: 10.1161/circulationaha.111.020982. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Chen M, Ding N, Mok Y, et al. Growth Differentiation Factor 15 and the Subsequent Risk of Atrial Fibrillation: The Atherosclerosis Risk in Communities Study. Clin. Chem. 2022;68:1084–1093. doi: 10.1093/clinchem/hvac096. [DOI] [PubMed] [Google Scholar]
- 23.Rathore SS, Hinn AR, Cooper LS, Tyroler HA, Rosamond WD. Characterization of incident stroke signs and symptoms: findings from the atherosclerosis risk in communities study. Stroke. 2002;33:2718–2721. doi: 10.1161/01.str.0000035286.87503.31. [DOI] [PubMed] [Google Scholar]
- 24.Gray RJ. A Class of K-Sample Tests for Comparing the Cumulative Incidence of a Competing Risk. Ann Stat. 1988;16:1141–1154. doi: 10.1214/aos/1176350951. [DOI] [Google Scholar]
- 25.Lau B, Cole SR, Gange SJ. Competing risk regression models for epidemiologic data. Am. J. Epidemiol. 2009;170:244–256. doi: 10.1093/aje/kwp107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Vinik AI, Maser RE, Mitchell BD, Freeman R. Diabetic autonomic neuropathy. Diabetes Care. 2003;26:1553–1579. doi: 10.2337/diacare.26.5.1553. [DOI] [PubMed] [Google Scholar]
- 27.Fine JP, Gray RJ. A Proportional Hazards Model for the Subdistribution of a Competing Risk. J Am Stat Assoc. 1999;94:496–509. doi: 10.2307/2670170. [DOI] [Google Scholar]
- 28.Dziopa K, Asselbergs FW, Gratton J, Chaturvedi N, Schmidt AF. Cardiovascular risk prediction in type 2 diabetes: a comparison of 22 risk scores in primary care settings. Diabetologia. 2022;65:644–656. doi: 10.1007/s00125-021-05640-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Shah SJ, Lam CSP, Svedlund S, et al. Prevalence and correlates of coronary microvascular dysfunction in heart failure with preserved ejection fraction: PROMIS-HFpEF. Eur. Heart J. 2018;39:3439–3450. doi: 10.1093/eurheartj/ehy531. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Freeman JV, Simon DN, Go AS, et al. Association Between Atrial Fibrillation Symptoms, Quality of Life, and Patient Outcomes: Results From the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation (ORBIT-AF). Circ. Cardiovasc. Qual. Outcomes. 2015;8:393–402. doi: 10.1161/circoutcomes.114.001303. [DOI] [PubMed] [Google Scholar]
- 31.Mant J, Doust J, Roalfe A, et al. Systematic review and individual patient data meta-analysis of diagnosis of heart failure, with modelling of implications of different diagnostic strategies in primary care. Health Technol. Assess. 2009;13:1–207, iii. doi: 10.3310/hta13320. [DOI] [PubMed] [Google Scholar]
- 32.Nowbar AN, Howard JP, Shun-Shin MJ, et al. Daily angina documentation versus subsequent recall: development of a symptom smartphone app. Eur Heart J Digit Health. 2022;3:276–283. doi: 10.1093/ehjdh/ztac015. [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.