Previous studies have reported differences in presenting symptoms and angiographic characteristics between women and men undergoing evaluation for suspected coronary artery disease (CAD). We examined the relation between symptoms and extent of CAD in patients with type 2 diabetes mellitus and known CAD enrolled in the Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI 2D) trial. Of 1,775 patients (533 women, 30%, and 1,242 men, 70%), women were more likely than men to have angina (65% vs 56%, p <0.001) or an atypical angina/anginal equivalent (71% vs 58%, p <0.001). More women reported unstable angina (17% vs 13%, p = 0.047) or were in a higher Canadian Cardiology Society class compared to men (Canadian Cardiology Society classes II to IV 78% vs 68%, p = 0.002). Fewer women than men had no symptoms (14% vs 22%, p <0.001). Women had a lower mean myocardial jeopardy index (42.5 ± 24.3 vs 47.9 ± 24.3, p <0.001), smaller number of total significant lesions (2.3 ± 1.7 vs 2.7 ± 1.8, p <0.001), and fewer jeopardized left ventricular regions (p <0.001 for distribution) or long-term occlusions (29% vs 42%, p <0.001). After adjustment for relevant covariates, the odds of having CAD symptoms were still higher in women than men (odds ratio for angina 1.31, 95% confidence interval 1.02 to 1.69; odds ratio for atypical angina 1.52, 95% confidence interval 1.17 to 1.96). In conclusion, in a high-risk group of patients with known CAD and diabetes mellitus, women were more symptomatic than men but had less obstructive CAD. These data suggest that factors other than epicardial CAD severity influence symptom presentation in women in this population.
Most studies evaluating patients with stable and unstable angina pectoris have demonstrated less extensive or less severe disease in women than in men. One might expect that the presence of less severe disease on coronary angiogram would correlate with fewer symptoms in women. However, previous reports have shown the opposite findings. In the first Bypass Angioplasty Revascularization Investigation (BARI) study, which examined patients with symptomatic multivessel coronary artery disease (CAD) amenable to revascularization with percutaneous coronary intervention (PCI) or coronary artery bypass grafting, women had more severe symptoms despite a similar burden of disease detected by coronary angiography. The Coronary Artery Surgery Study (CASS) registry demonstrated that for any given extent of disease (1 vessel or 2 or 3 vessels), women were more symptomatic compared to men. Although this information strongly suggests sex differences in symptoms related to CAD, little is known regarding the relation of symptoms to extent of CAD after adjusting for other confounding factors. The BARI 2 Diabetes 2D (BARI 2D) trial baseline data provide a unique opportunity to examine sex differences in the relation between presenting long-term symptoms and severity and extent of CAD in a population of patients with type 2 diabetes mellitus (DM) and established CAD.
Methods
The BARI 2D trial is a multicenter randomized clinical trial designed to determine the optimal treatment for patients with type 2 DM and documented CAD that is suitable for elective revascularization. A detailed description of the study design and patient population has been published previously. Briefly, 2,368 participants were enrolled from 49 clinical sites from January 1, 2001 through March 31, 2005. Patients were randomized in a 2 × 2 factorial design to a strategy of immediate revascularization plus aggressive medical therapy versus aggressive medical therapy alone for treatment of CAD and an insulin-providing strategy versus an insulin-sensitizing strategy for management of type 2 DM.
Patients were eligible for enrollment in the study if they had type 2 DM and angiographically documented CAD involving ≥1 coronary vessel that was amenable to treatment with aggressive medical therapy or elective revascularization with PCI or coronary artery bypass grafting. Exclusion criteria included unstable symptoms requiring revascularization, severe left main coronary artery disease (≥50% stenosis), PCI or coronary artery bypass grafting in the previous 12 months, or history of chronic kidney disease with serum creatinine level >2.0 mg/dl (176.8 μmol/L).
Trained BARI 2D trial coordinators collected all baseline clinical data for enrolled patients before randomization. A detailed instruction manual in addition to training sessions provided definitions for ischemic chest pain, unstable and stable symptoms, and atypical angina/anginal equivalents. Before randomization, patients were queried about whether they had chest pain or other associated symptoms at any time or within 6 weeks of randomization. Clinical co-ordinators used this information to document the presence of ischemic-like chest pain, atypical angina, and/or anginal equivalents in the case-report form. Atypical angina was defined as (1) chest discomfort that was not characteristic for angina but with precipitating factors that typically triggered myocardial ischemia (i.e., exercise, emotional stress, cold exposure, etc.) or (2) chest discomfort consistent with angina but with unusual precipitating triggers. Anginal equivalents included symptoms of myocardial ischemia other than angina such as shortness of breath (defined as the need to exert an increased amount of effort to breath at rest or with minimal activity), exertional dyspnea (defined as labored or difficult breathing only evident with effort greater than normal ambulatory activity), exertional fatigue (defined as easy fatigue with minimal activity or with effort greater than normal activity), nausea (defined as nausea at rest or with activity not associated with an identifiable cause other than typical angina), or unexplained diaphoresis (defined as diaphoresis at rest or with minimal activity not attributed to an identifiable cause). A patient could have ischemic chest pain and/or atypical angina/anginal equivalents. In addition, a patient may have reported ≥1 angina equivalent. In patients with ischemic chest pain, symptoms were further categorized into 1 of 2 mutually exclusive groups: a history of unstable angina (defined as angina at rest or precipitated by minimal activity) or stable angina. Stable angina was then classified by Canadian Cardiology Society functional class. The present analysis focused on symptoms within 6 weeks of study entry.
Baseline angiograms were assessed locally at clinical sites and at a centralized core laboratory (Stanford University, Stanford, California). Data recorded from the centralized core laboratory were used to analyze this paper. Stenosis ≥50% diameter was defined as a significant lesion. Burden of CAD was measured using the myocardial jeopardy index (MJI) score and total number of significant lesions. Total MJI is percent distal myocardium that is jeopardized by lesions ≥50% downstream in any of the 3 main coronary arteries or their branches. Number of diseased left ventricular (LV) regions (defined as LV regions subtended by arteries with stenoses ≥50%) was recorded. Diseased LV regions ranged from 0 to 3 (anterior, lateral, and inferoposterior). To estimate extent of diffuse disease, we evaluated the total number of lesions with >20% arterial diameter narrowing.
Of the 2,368 randomized patients, 2,321 patients had >80% of baseline information available. Because the purpose of the study was to examine the relation between MJI and symptoms and because MJI does not accurately account for total burden of CAD in patients with previous revascularization, we excluded 546 patients with previous revascularization (PCI or coronary artery bypass grafting) from our analysis. The remaining 1,775 patients formed the subset for the present study.
Chi-square tests were performed to test general associations between sex and categorical characteristics. Two-sample t tests were conducted for comparisons of means of continuous variables between women and men. A multivariate logistic regression adjusting for age, race, ethnicity, education level, geographic region, physical activity, current cigarette smoking, number of antianginal agents, history of hypertension, intermittent claudication, hemoglobin A 1c ≥7%, low-density lipoprotein ≥100 mg/dl, blood pressure >140/90 mm Hg, ankle–brachial index categories, body mass index, duration of DM, metabolic syndrome, Michigan Neuropathy Screening Instrument score >2, taking diuretics, and taking insulin at baseline was performed to assess for differences by sex in severity and extent of angiographic disease.
We compared symptoms between women and men after stratifying by severity of CAD. We chose 3 angiographic characteristics representing severity or extent of CAD: number of diseased LV regions, MJI score, and number of significant lesions. Each angiographic characteristic was stratified into 3 strata representing degree of severity or extent of CAD. Symptom differences were further analyzed using multivariate logistic regression adjusting for the same variables used for analysis of angiographic disease plus number of jeopardized LV regions, number of significant lesions, presence of total occlusion, and LV ejection fraction <50%. A p value <0.05 was considered statistically significant. Analyses were performed by SAS 9.1.3 (SAS Institute, Cary, North Carolina). Figures were plotted by R software (R Development Core Team, 2006, available at: http://www.R-project.org ).
Results
Of the 1,775 patients examined, 533 were women (30%) and 1,242 were men (70%; Table 1 ). There were notable differences in baseline demographic and clinical variables between women and men as presented Table 1 .
| Baseline Characteristics | Women | Men | p Value |
|---|---|---|---|
| (n = 533) | (n = 1,242) | ||
| Age at entry (years) | 62.8 ± 9.3 | 62.0 ± 8.8 | 0.111 |
| Race/ethnicity | |||
| White non-Hispanic | 299 (56%) | 849 (68%) | <0.001 |
| Black non-Hispanic | 145 (27%) | 156 (13%) | |
| Hispanic | 68 (13%) | 167 (13%) | |
| Asian non-Hispanic | 21 (4%) | 70 (6%) | |
| Education | |||
| Not beyond high school | 383 (72%) | 690 (56%) | <0.001 |
| Beyond high school | 149 (28%) | 549 (44%) | |
| Region | |||
| United States | 340 (64%) | 698 (56%) | <0.001 |
| Canada | 45 (8%) | 256 (21%) | |
| Mexico | 19 (4%) | 60 (5%) | |
| Brazil | 109 (20%) | 184 (15%) | |
| Czech Republic/Austria | 20 (4%) | 44 (4%) | |
| Previous myocardial infarction | 116 (22%) | 311 (25%) | 0.130 |
| Congestive heart failure | 33 (6%) | 60 (5%) | 0.236 |
| Hypertension | 460 (87%) | 971 (79%) | <0.001 |
| Hypercholesterolemia | 422 (80%) | 962 (78%) | 0.367 |
| Intermittent claudication | 103 (19%) | 187 (15%) | 0.026 |
| Duration of diabetes (years) | 12.0 ± 9.6 | 9.5 ± 8.0 | <0.001 |
| Body mass index (kg/m 2 ) | 32.6 ± 7.0 | 31.2 ± 5.5 | <0.001 |
| Michigan Neuropathy Screening Instrument >2 | 243 (46%) | 647 (52%) | 0.013 |
| Cigarette smoking | |||
| Former | 189 (35%) | 737 (59%) | <0.001 |
| Current | 55 (10%) | 167 (13%) | |
| Physical activity | |||
| Sedentary | 150 (28%) | 245 (20%) | <0.001 |
| Mild | 230 (43%) | 499 (40%) | |
| Moderate | 140 (26%) | 454 (37%) | |
| Strenuous | 12 (2%) | 42 (3%) | |
| Hemoglobin A 1c ≥7 (%) | 360 (68%) | 718 (58%) | <0.001 |
| Low-density lipoprotein ≥100 (mg/dl) | 265 (50%) | 465 (38%) | <0.001 |
| Sitting blood pressure >140/90 (mm Hg) | 178 (34%) | 343 (28%) | 0.013 |
| Albumin/creatinine ratio >30 | 166 (34%) | 376 (32%) | 0.626 |
| Ankle–brachial index | |||
| ≤0.90 | 125 (26%) | 186 (17%) | <0.001 |
| 0.90–1.3 | 330 (69%) | 797 (72%) | |
| >1.3 | 23 (5%) | 120 (11%) | |
| Metabolic syndrome | 474 (89%) | 961 (77%) | <0.001 |
| Number of antianginal agents | 1.6 ± 1.0 | 1.5 ± 1.0 | 0.332 |
| Any antianginal agent | 457 (86%) | 1,044 (84%) | 0.368 |
| β Blocker | 377 (71%) | 873 (70%) | 0.908 |
| Nonsublingual nitrate | 181 (34%) | 334 (27%) | 0.003 |
| Sublingual nitrates or nitro spray | 108 (20%) | 308 (25%) | 0.037 |
| Calcium channel blocker | 168 (32%) | 363 (29%) | 0.344 |
| Diuretic | 253 (47%) | 410 (33%) | <0.001 |
| Insulin | 188 (35%) | 284 (23%) | <0.001 |
Table 2 presents angiographic data stratified by sex for the 1,775 patients. Women had less severe and significant CAD. Even after adjustment for other confounding variables, women were less likely to have severe disease on coronary angiogram as defined by multiple diseased LV regions (odds ratio of having multiple diseased region for women vs men 0.59, p <0.0001) or MJI (beta coefficient −5.2, p = 0.0001).
| Angiographic Characteristics | Women | Men | p Value |
|---|---|---|---|
| (n = 533) | (n = 1,242) | ||
| Number of significant lesions | 2.3 ± 1.7 | 2.7 ± 1.8 | <0.001 |
| Number of lesions >20% | 4.3 ± 2.2 | 4.8 ± 2.3 | <0.001 |
| Number of jeopardized left ventricular regions | |||
| 0–1 | 207 (39%) | 361 (29%) | <0.001 |
| 2 | 193 (36%) | 450 (36%) | |
| 3 | 132 (25%) | 431 (35%) | |
| Myocardial jeopardy index | 42.5 ± 24.3 | 47.9 ± 24.3 | <0.001 |
| Total occlusion | 155 (29%) | 526 (42%) | <0.001 |
| Significant lesions in proximal left anterior descending coronary artery | 58 (11%) | 160 (13%) | 0.242 |
| Left ventricular ejection fraction | 59.3 ± 10.5 | 56.8 ± 10.9 | <0.001 |
| Left ventricular ejection fraction <50% | 57 (11%) | 225 (19%) | <0.001 |
Table 3 presents the presence of CAD symptoms for women and men. Women were significantly more likely than men to have ischemic chest pain or atypical angina/anginal equivalent. Of patients with ischemic chest pain, more women reported a history of unstable angina. Of patients with stable angina, more women were in a higher Canadian Cardiology Society functional class (II to IV) compared to men. Women were also less likely than men to have no symptoms at presentation.
| Symptom | Women | Men | p Value |
|---|---|---|---|
| (n = 533) | (n = 1,242) | ||
| Ischemic chest pain | 347 (65%) | 692 (56%) | <0.001 |
| Unstable angina | 60 (17%) | 88 (13%) | 0.047 |
| Stable angina | 287 (83%) | 604 (87%) | |
| Class I | 63 (22%) | 195 (32%) | 0.004 |
| Class II | 167 (58%) | 330 (55%) | |
| Class III | 47 (16%) | 66 (11%) | |
| Class IV | 9 (3%) | 12 (2%) | |
| Anginal equivalents or atypical angina | 373 (71%) | 712 (58%) | <0.001 |
| Shortness of breath | 265 (71%) | 438 (62%) | 0.002 |
| Dyspnea on exertion | 271 (73%) | 507 (71%) | 0.615 |
| Exertional fatigue | 244 (66%) | 399 (56%) | 0.003 |
| Nausea | 88 (24%) | 75 (11%) | <0.001 |
| Unexplained diaphoresis | 108 (29%) | 156 (22%) | 0.010 |
| No angina/equivalents/atypical angina | 75 (14%) | 276 (22%) | <0.001 |
When stratified for severity/extent of CAD, a pattern of sex-specific difference in symptoms was observed for the 3 degrees of severity/extent ( Table 4 ). Within the least severe/extensive strata (the first 2 strata of each angiographic factor), women were more likely to report ischemic chest pain and/or atypical angina/anginal equivalents than men. This difference by sex in symptoms was no longer seen for the group of patients with the most severe/extensive CAD findings (i.e., stratum of 3 diseased LV regions or MJI ≥65% or >3 significant lesions).
