The Korean Women’s Chest Pain Registry (KoROSE) is a nation-wide, multicenter, prospective study evaluating patients who visited outpatient clinic with chest pain. The detailed protocol of the registry has been previously published [10,11]. The registration started from January 2012. The registry also enrolled men from February 2014 for comparison between sexes. Patients from 25 tertiary medical centers were registered until May 2018. Exclusion criteria were a history of myocardial infarction, or any coronary revascularization, or malignancy, or inflammatory diseases, or chronic kidney disease on dialysis. This study enrolled consecutive patients in the registry who underwent invasive CAG.

Clinical characteristics including age, gender, anthropometric parameters, and medical history of cardiovascular risk factors were investigated. Hypertension was defined by treatment with anti-hypertensive medications, and a systolic blood pressure (BP) ≥ 140, or diastolic BP ≥ 90 mmHg. Diabetes mellitus (DM) was defined by treatment with insulin or oral hypoglycemic agents or fasting blood glucose ≥ 126 mg/dL. This study protocol was approved by the Institutional Review Boards of all participating centers (IRB No. 2012AN0011). Written informed consent was obtained from all participants.

Detailed chest pain questionnaires were obtained from all the patients by physicians. Six aspects of chest pain were evaluated as follows: (1) pain quality, (2) pain location, (3) pain duration, (4) aggravating factors, (5) location of radiating pain, and (6) associated symptoms. Pain quality was described as squeezing, pressure, and stinging, burning or dull pain. Patients were asked to point to painful locations on the body. The duration of pain was classified into three categories: shorter than 5 minutes, longer than 5 minutes and shorter than 15 minutes, longer than 15 minutes and shorter than 1 hour, and longer than 1 hour. Patients were asked to select the aggravating factors from the list: exercise, meal, emotional stress, low temperature, alcohol or other. The location of radiating pain was investigated among left arm, right arm, neck or back. Patients were also asked to indicate any associated symptoms from the list: palpitation, dizziness, headache, dyspnea, or syncope.

To evaluate OCAD, invasive CAG was performed in each patient for the following indications: (1) typical angina, (2) angina with abnormal function test result, and (3) angina not evaluable with functional testing. Invasive CAG was performed via femoral or radial artery using a standard protocol. The percentage of coronary artery stenosis was determined by the ratio of the diameter of the stenotic lesion and normal site just proximal to the lesion. OCAD was defined as ≥ 70% stenosis in a proximal or middle epicardial coronary artery or major branch.

Categorical variables were presented with frequencies (percentages) and continuous variables were presented with mean ± standard deviation. The differences in baseline and chest pain characteristics were compared between genders and according to the presence of OCAD using the independent t test or chi-square test. Multivariable logistic regression was performed to investigate the impact of each chest pain characteristic for OCAD after adjustment for age, gender, hypertension, DM, dyslipidemia, smoking, and family history of CAD. SPSS version 24 (IBM Co., Armonk, NY, USA) was utilized for statistical analysis. A p < 0.05 was considered statically significant.

The study enrolled 1,549 consecutive patients (mean age 61.1 ± 11.2 years, 514 males and 1,035 females). Table 1 lists the clinical demographics of the total study population. The mean age of men was younger than that of women (59.4 ± 11.1 years vs. 61.9 ± 11.1 years, p < 0.001). The prevalence of current smoking was higher in men compared to women. However, other cardiovascular risk factors including hypertension, DM and dyslipidemia, and cardiovascular medications were not different between the genders.

Six chest pain characteristics including quality, duration, location, radiation, aggravation, and associated symptoms are listed in Table 2. Among all the enrolled patients, 568 patients (30.7%) described squeezing quality, and 817 patients (52.7%) manifested pain in the retrosternal area. Pain duration shorter than 5 minutes was the most common (719 patients, 46.4%), and was aggravated by exercise in 449 patients (29.0%).

When chest pain characteristics were compared between genders, men experienced more squeezing chest pain, while women experienced more dull pain. Men complained more of left-sided chest pain, and women presented with epigastric pain. Pain duration exceeding one hour was more common in women than in men. More women described chest pain radiating to back with accompanying palpitation and headache.

OCAD on CAG was found in 425 patients (27.4%).Chest pain was compared between patients with non-OCAD and OCAD (Table 3). Typical symptoms for angina were observed more in patients with OCAD compared with non-OCAD. In patients with OCAD, squeezing pain was more common than in patients with non-OCAD (171 [40.2%] vs. 397 [35.3%], p = 0.042). Based on location, pain in the retrosternal area was predominant; however, left-sided pain in the chest was less common in patients with OCAD compared with non-OCAD (pain in retrosternal area: 251 [59.1%] vs. 566 [50.4%], p = 0.001; left side of chest: 83 [19.5%] vs. 285 [25.4%], p = 0.009). Duration of chest pain did not differ according to the presence of OCAD. Pain was aggravated by exercise more frequently, and less by emotional stress in patients with OCAD compared with non-OCAD (exercise: 175 [41.2%] vs. 274 [24.4%], p < 0.001; emotional stress: 18 [4.2%] vs. 104 [9.3%], p < 0.001). The location of radiating pain showed no difference according to the presence of OCAD. Among the associated symptoms, patients with OCAD complained of less palpitation and dizziness compared with patients with non-OCAD (palpitation: 26 [6.1%] vs. 142 [12.6%], p < 0.001; dizziness: 26 [6.1%] vs. 138 [12.3%], p < 0.001).

To investigate the independent association between each chest pain characteristic and OCAD, a multivariate logistic regression including age, gender, hypertension, DM, dyslipidemia, smoking, and family history of CAD was performed (Table 4). The analysis revealed that pain in the retrosternal area and aggravation by exercise increased the probability of OCAD, while palpitation decreased the probability of OCAD (pain in retrosternal area: odds ratio [OR], 1.491; 95% confidence interval [CI], 1.178 to 1.887, p = 0.001) (exercise: OR, 2.235; 95% CI, 1.745 to 2.861, p < 0.001) (palpitation: OR, 0.476; 95% CI, 0.304 to 0.746; p = 0.001).

OCAD on CAG was more prevalent in men compared with women (175 [34.0%] vs. 250 [24.2%], p < 0.001). Men and women displayed differences in chest pain characteristics according to the presence of OCAD. In men, typical symptoms for angina were more commonly observed in patients with OCAD compared with non-OCAD, while in women, a component of typical symptom was less commonly observed in patients with OCAD compared with non-OCAD. Quality of chest pain was not related to OCAD in both men and women. Left-sided chest pain was less commonly observed in men with OCAD compared with non-OCAD (42 [24.0%] vs. 118 [34.8%], p = 0.008). However, the location of chest pain was not related to OCAD in women. In terms of duration, only men demonstrated that shorter duration was related to OCAD (97 [55.4%] vs. 152 [44.8%], p = 0.014). Among aggravation factors of chest pain, emotional stress was less frequent in women with OCAD compared with non-OCAD (8 [3.2%] vs. 78 [9.9%], p < 0.001), but the association was not significant in men. Among associated symptoms of chest pain, dyspnea was more common in men with OCAD compared with non-OCAD (49 [28.0%] vs. 66 [19.5%], p = 0.019) and dizziness was less common in women with OCAD compared with non-OCAD (13 [5.2%] vs. 96 [12.2%], p = 0.001).

In men and women, multivariate logistic regression was performed to investigate the association between chest pain and OCAD after adjustment for age, hypertension, DM, dyslipidemia, smoking, and family history of CAD. Shorter duration of chest pain (less than 5 minutes) and dyspnea increased the probability of OCAD only in men (shorter duration: OR, 1.581; 95% CI, 1.086 to 2.303; p = 0.017) (dyspnea: OR, 1.610; 95% CI, 1.040 to 2.490; p = 0.033). Left-sided chest pain decreased the probability of OCAD in men (OR, 0.590; 95% CI, 0.388 to 0.897; p = 0.014), while aggravation by emotional stress and dizziness decreased the probability of OCAD in women (emotional stress: OR, 0.348; 95% CI, 0.162 to 0.746; p = 0.007) (dizziness: OR, 0.457; 95% CI, 0.246 to 0.849; p = 0.013).

Stable angina is characterized by retrosternal squeezing, heaviness, and burning. It may radiate to arms, neck, jaw, and is triggered by exercise or emotional stress [2]. In the total study population, pain in retrosternal area and aggravation by exercise were associated with a high probability of OCAD. The results are compatible with the traditional description of stable angina. However, palpitation was negatively associated with OCAD. Palpitation is a common non-pain symptom [13]. However, the clinical implications of palpitation are unclear in patients with non- acute coronary syndrome. Our findings suggest that palpitation in patients with suspected angina may be an atypical symptom with a low probability of OCAD.

Current practice guidelines in cardiovascular diseases (CVDs) are based on previous studies conducted primarily in men. For example, in the 2007 American Heart Association guidelines for the prevention of CVD in women, women constituted only 30% of all patients in clinical trials [14]. In a decade, numerous remarkable studies focused on clinical characteristics and outcomes of CVD in women [15-18]. In chest pain characteristics, gender differences have been reported variously according to study design (acute coronary syndrome or stable angina setting, patient population and methods to investigate chest pain characteristics). Although there are reports that chest pain characteristics were not significantly different between men and women with OCAD [19], women tend to have less specific and more non-cardiac symptoms compared with men [7,8,13,20]. Compared to men, CAD in women presents later in their lifespan. Because symptoms become less specific with aging, the differences in the distribution of age in men and women with CAD contribute to gender differences in chest pain presentation. The findings of current study that women older than men, and men presents more typical chest pain with squeezing quality and women manifest more dull pain and non-cardiac symptoms such as palpitation and headache is consistent with previous studies.

Although the gender differences in chest pain have been reported in patients with CAD, their clinical implications and association with CAD are scarce. Majority of previous studies were performed in acute coronary syndrome setting, ant these studies demonstrated that diagnostic probability of chest pain characteristics are not significantly different between men and women [21,22]. In the current study among Koreans with stable chest pain in a tertiary care setting, the gender differences in symptoms for including or excluding CAD have been demonstrated.

Typically, a pain duration more than 30 minutes was the strong negative predictor for CAD [23]. In men from this study, shorter pain duration was positively associated with CAD. However, this association was not observed in women from this study. Similarly, the study with 1,212 German patients with chest pain in the primary care setting, also demonstrated that in men, shorter pain duration was positively associated with CAD, while pain duration exceeding 1 hour was associated with CAD in women [24]. Even though our findings were consistent with previous study, the underlying mechanism remains unclear.

In this study, only men showed left-sided chest pain was negatively associated with CAD. Bosner et al. [24] reported localized muscle tension in men was negatively associated with CAD. Because men tend to be more active in physical activity than women, it is possible that localized muscle tension, not CAD, may present as a left-sided chest pain. Simple physical examination such as chest wall tenderness on palpation is required to rule out the probability of CAD in these cases.

In men from this study, dyspnea was positively associated with CAD. The reason of association between dyspnea and CAD in men was not clear. But, smoking rate was much higher in men compared to women. As smoking is a potent common risk factor for both CAD and pulmonary disease, high smoking rate in men may be one cause of the association between dyspnea and CAD.

In women, aggravation by emotional stress was negatively associated with CAD. Although emotional stress is a traditional aggravation factor of angina [1], it is notable that in women from this study, pain worsening with emotional stress suggested a low probability of OCAD. In Western countries, somatization is more prevalent in women compared to men [25]. Higher frequency of depressive disorder in women also support the psychological etiology of chest pain in women [26]. Bosner et al. [24] also reported that in German population, women have more psychogenic disorder as the etiology of chest pain compared to men. Hwabyung is a Korean cultural syndrome and its prevalence in general population is 4.1% [27]. It is prevalent in middle-aged or older women. Chest pain in Korean women without OCAD may be explained as one of the somatization symptoms of Hwabyung. The traditional notion that emotional stress is a predictor of CAD may be a misguided concept that has been derived from previous studies which had been conducted in population mainly composed of men rather than women. Further studies are needed to elucidate the psychological causes of chest pain in women.

Currently, the underlying mechanism of gender difference in chest pain characteristics has not been clearly investigated; however, several mechanisms have been suggested. Firstly, gender differences in pain perception can be one mechanism. In positron emission tomography study, the gender difference of neurophysiological response by painful heat stimulation was observed and women perceive more intense pain than men [28]. In a population study, women tend to have more frequent and severe pain with more referring to greater portion of body region [29]. Second, coronary microvascular dysfunction is reported to be more common in women with suspected angina [30]. This difference of microvascular dysfunction may result in gender difference of chest pain characteristics. Lastly, on the inferior and posterior surface of heart, parasympathetic afferent nerves predominate while sympathetic nerves mainly innervate the anterior surface of heart. In women, right coronary dominance is more prevalent compared to men [31]. The proximal stenosis of a right coronary dominant artery may provoke myocardial ischemia on both anterior and posterior-inferior walls, and activate both sympathetic and parasympathetic nerve. Compared to men with a left coronary dominant system, women with right coronary artery stenosis have more probability to manifest autonomic symptoms such as nausea and emesis [31,32]. These anatomical and physiologic differences may explain the more observed autonomic symptoms such as epigastric pain, palpitation and dizziness in women of present study.

Our study has several limitations. First, despite the growing concern about microvascular angina [33,34], our study did not investigate functional ischemia or microvascular dysfunction. We intend to conduct a coronary flow reserve test for additional patients enrolled. Our future study will investigate chest pain characteristics in patients with microvascular angina. Second, the number of men was relatively smaller than women in our study, suggesting a possible attenuation in the association between chest pain and OCAD in men. Finally, although this study suggested psychological causes as the non-cardiac factors in women’s chest pain, our study did not conduct a psychological analysis.

In conclusion, this is the first study to focus on gender differences in chest pain among Koreans with angina. Symptoms with high probability for OCAD varied between sexes. Our findings suggest that history taking in pretest assessment for OCAD should be individualized according to gender.