The telephone interview was conducted over a period of two weeks by a professional team of trained telephone interviewers from the Social Sciences Research Centre, the University of Hong Kong in November 2002. The interviewers went through intense training on the delivery of questionnaire to ensure uniformity and the questions to be understood. Random telephone numbers were generated by computer and dialed automatically. Only numbers corresponding to ethnic Chinese households were used in the study. Office numbers, facsimile machines, and non-Chinese households were excluded. Upon identification of target households, the interviewer asked to speak to the household member with the most recent birthday. This aimed to provide randomization among different members of the household. Baseline demographic data, education, occupation and income were assessed, followed by a translated version of the Rose Angina questionnaire[11], a validated GERD questionnaire[12], a validated translated version of the hospital anxiety depression scale[13-15] and assessment of medical care utilization and impact of the disease on social activity as described below. Overall, 3605 ethnic Chinese households were contacted by telephone. The interview was completed in 2209 subjects (response rate = 61.3%, mean age = 40.3 ± 14, 58% female). The demographic characteristics of the study subjects were comparable to the census data of Hong Kong in 2001[16]. This study was approved by the ethics committee of the University of Hong Kong.

A previous study demonstrated that the prevalence of NCCP was approximately 23% in Minnesota, USA[1]. To provide a 95% confidence interval ± 2% and a meaningful comparison between health care seekers and non-health care seekers, we estimated a sample size of around 2000 of successful cases.

The Rose angina questionnaire is a standardized method of measuring angina and myocardial infarction in population surveys and has been validated in different ethnic groups[17-26]. ‘Definite’ angina was defined as chest pain that limits exertion (walking uphill or hurry, or walking at an ordinary pace on the level), is situated over the sternum or in the left chest and left arm, and is relieved within 10 min by rest[11]. ‘Possible’ angina was defined as chest pain that limits exertion and other criteria for definite angina not fulfilled. NCCP was defined as non-exertional chest pain according to the Rose angina questionnaire and had not been diagnosed as ischaemic heart diseases by a physician[27]. Musculoskeletal-like chest pain was defined as chest pain that worsens on breathing, movement or the presence of chest wall tenderness. The duration and characteristics of the chest pain were recorded. In the development of the Chinese version of the questionnaire, the original instrument was translated, back translated and tested for reproducibility in a sample of thirty patients attending the gastrointestinal clinic. The intraclass correlation coefficient of the translated questionnaire was 0.91. Furthermore, we tested the Rose angina questionnaire in a pilot of 50 patients with known coronary heart diseases proven by coronary angiography and 100 healthy controls. The sensitivity and specificity determined by the pilot study was 68% and 95% for the diagnosis of ischemic heart disease respectively.

Gastrointestinal symptoms were assessed by a translated Chinese version of a validated GERD questionnaire[12]. The GERD questionnaire examined the symptoms of heartburn, acid regurgitation, dyspepsia, dysphagia, globus, odynophagia, hoarseness of voice, chronic cough, asthma and pneumonia in details. In addition, past medical history, medication use, past history of esophageal, gastric, cardiac or pulmonary disease; smoking, alcohol intake, and the intake of tea and coffee were assessed. Part of the results on GERD had been presented elsewhere[28]. The severity and frequency of chest pain and other gastrointestinal symptoms were graded on a five-point Likert scale as follows: 1 (none- no symptoms / none in the past year), 2 (mild- symptoms can be easily ignored / less than once per mo), 3 (moderate- awareness of symptoms but easily tolerated / ≥ once per month), 4 (severe- symptoms sufficient to cause an interference with normal activities / ≥ once weekly) and 5 (incapacitating- incapacitating symptoms with an inability to perform daily activities or require day-off / ≥ once daily)[29]. GERD was defined as heartburn and/or acid regurgitation over the past year, which has been shown to be specific for the diagnosis of GERD[30,31]. Patients who had used non-steroidal anti-inflammatory drugs (NSAIDs) / aspirin for at least 3 d at any dosage within 3 mo prior to the survey were considered to be NSAIDs / aspirin users[32]. Anxiety and depression were assessed by the hospital anxiety and depression scale[13]. The Chinese version of this questionnaire has previously been validated[14,15] which consists of 7 questions on anxiety and 7 questions on depression.

Medical care utilization of subjects with NCCP was classified into categories including the use of over-the-counter medication, community based medicine, accident and emergency department and admission to hospital. Impact of disease was measured by the proportion of subjects requiring days-off work due to gastrointestinal complaints, and whether subjects reported an adverse effect of the chest pain on their normal social life (social life being affected), i.e. symptoms sufficient to cause an interference with normal daily and social activity.

Univariate analysis was performed by Student’s t test for continuous variables and by chi-square test for categorical variables to assess the risk factors associated with NCCP. Multiple logistic regression analysis with sex and age adjustment was then performed to determine the risk factors associated with NCCP. Furthermore, a multiple logistic regression model was designed to determine the factors (severity and frequency of chest pain, presence of heartburn and/or acid regurgitation over the past year, gender, age, educational level (primary school or below, secondary or matriculation and tertiary), occupation (3 levels), anxiety and depression scores and social life being affected by chest pain) associated with health seeking behaviour in NCCP. To find the best model, a backward elimination stepwise procedure was carried out in a way that the factor would be eliminated from the analysis if the corresponding P value was greater than 0.2, in order not to miss out too much information in view of the small sample size. A P value of 0.05 or less was considered statistically significant and all reported P values were 2 sided.

Chest pain over the past year was present in 454 subjects (20.6%, 95%CI 19-22), of which 147 (6.7%) had ‘possible’ angina by the Rose angina questionnaire and/or ischaemic heart diseases diagnosed by a physician[11,27].

NCCP over the past year was present in 307 subjects (13.9%, 95%CI 13-15). The demographic characteristics of subjects with NCCP, cardiac chest pain or no chest pain are given in Table 1. The prevalence of NCCP was higher in men than in women (16.6% vs 11.9%, P = 0.002). Subjects with NCCP were significantly younger than subjects with no chest pain. However, for subjects with cardiac chest pain, the prevalence was similar between men and women (7.5% vs 8%, P = 0.68) and the mean age was similar to subjects with no chest pain. Aspirin usage was significantly higher in subjects with cardiac chest pain than in subjects with no chest pain (20.6% vs 7.7%, P < 0.001). NSAIDs usage was significantly higher in subjects with either NCCP or cardiac chest pain than in subjects with no chest pain. Cigarette and coffee consumption were the highest in subjects with NCCP, while alcohol consumption was similar between subjects with NCCP and subjects with cardiac chest pain. The education level and socioeconomic status were similar in subjects with NCCP, cardiac chest pain or no chest pain (Table 1). The median duration of NCCP was 24 mo (range 0.1 to 360 mo). Most (96%) subjects with NCCP had mild to moderate chest pain over the central chest area (50%) (Table 2). The frequency of chest pain was less than once per mo in three quarters of the subjects. Of the 307 subjects with NCCP, 155 (50.5%) could be classified as having GERD, while 79 (25.7%) had musculoskeletal-like chest pain. But GERD and musculoskeletal-like chest pain were overlapped in 15.6% (48/307) of subjects. The gender difference in NCCP between men and women persisted after exclusion of subjects with musculoskeletal-like chest pain (13.1% vs 8.5%, P = 0.001). Figure 1 shows the age and sex- prevalence rates of NCCP and musculoskeletal-like chest pain over the past year. For men with NCCP over the past year, the prevalence was the highest in 18-24 age group, less common in 25-54 age group and increased again after the age of 55 (Figure 1). In contrast, the prevalence of NCCP in women was almost constant from age 18 to 44, and then dropped gradually with age. The overall prevalence of musculoskeletal-like chest pain was similar between men and women (4.2% vs 4.0%, P = NS), but men aged 18-24 had a higher prevalence when compared to women of similar age (7.9% vs 2.6%, P = 0.034).

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Age and sex-specific prevalence rates (per 100) of noncardiac chest pain (NCCP) and musculoskeletal-like chest pain (MSL-CP) over the past year.

By univariate analysis, NCCP was associated with heartburn and/or acid regurgitation, globus, dyspepsia, feeling of acidity in stomach, alcohol and NSAID usage but not with dysphagia, odynophagia, chronic cough, hoarseness of voice, asthma, pneumonia, history of smoking, aspirin usage and coffee intake (Tables 1 and 3). By multiple logistic regression analysis, heartburn and/or acid regurgitation (OR 2.3, 95%CI 1.7-3.1), globus (OR 1.9, 95%CI 1.3-2.8), and NSAIDs use (OR 1.9, 95%CI 1.2-2.9) were independent risk factors associated with NCCP.

Thirty-nine percent of subjects with NCCP over the past year used one or more forms of treatment for their problems. Among the 307 subjects with NCCP over the past year, 14 (5%) purchased over the counter medication, 85 (28%) visited an outpatient clinic; 10 (3%) visited the accident and emergency department, and 11 (4%) were admitted to regional hospitals for further management. NCCP over the past year was significantly correlated with increased health care utilization (P < 0.001, OR 3.2, 95%CI 2.4-4.3). A significantly higher proportion of health seekers with NCCP had moderate to severe chest pain when compared to non-health seekers (39% vs 27%, P = 0.025) (Table 4). However, the proportion of subjects with at least monthly chest pain (28% vs 22% P = 0.265) was similar between health seekers and non-health seekers. GERD (heartburn and/or acid regurgitation) over the past year was more prevalent in NCCP health seekers than in NCCP non-health seekers (67.5% vs 40.1%, P < 0.001). A significantly higher proportion of subjects with NCCP reported their social life was affected by the chest pain (17% vs 9%, P < 0.001, OR 1.9, 95%CI 1.4-2.5) and required days-off (17% vs 12%, P = 0.021, OR 1.4, 95%CI 1.1-1.9) when compared to subjects with no chest pain.

Subjects with NCCP symptoms over the past year had higher average anxiety (5.5 vs 4.1, P < 0.001) and depression (4.4 vs 3.8, P = 0.007) scores when compared to subjects with no chest pain. However, for subjects with NCCP symptoms, the mean anxiety score and the mean depression score were similar between NCCP subjects who had sought any medical consultation and those who had not (Table 4).

By multiple logistic regression analysis, female gender (OR 1.9, 95%CI 1.1-3.2, P = 0.023), the presence of GERD (heartburn and/or acid regurgitation) symptoms over the past year (OR 2.8, 95%CI 1.6-4.8, P < 0.001) and social life being affected by chest pain (OR 6.9, 95%CI 3.3-15.9, P < 0.001) were independent factors associated with health seeking behaviour of subjects with NCCP.