You are here

Article Text

Article menu

Download PDFPDF

Other topics
Long-term cardiovascular consequences of Rose angina at age 20–54 years: 29-years’ follow-up of the Tromsø Study
  1. Sidsel Graff-Iversen1,2,
  2. Tom Wilsgaard2,
  3. Ellisiv B Mathiesen3,
  4. Inger Njølstad2,
  5. Maja-Lisa Løchen2
  1. 1Norwegian Institute of Public Health, Oslo, Norway
  2. 2Department of Community Medicine, UiT The Arctic University of Norway, Tromsø, Norway
  3. 3Department of Clinical Medicine, UiT The Arctic University of Norway, Tromsø, Norway
  1. Correspondence to Professor Sidsel Graff-Iversen, Norwegian Institute of Public Health, Box 4404, Nydalen, 0403 Oslo, Norway; sgri{at}fhi.no

Abstract

Background The Rose Angina Questionnaire (RAQ) was constructed in the 1960s for assessing the population burden of angina. Studies have found that screening positivity by RAQ conferred an elevated risk of coronary heart disease (CHD). It is, however, not clear to what extent Rose angina represents early CHD in relatively young adults who are free of known CHD. If representing CHD, Rose angina is expected to carry prognostic information in addition to the risk conferred by other risk factors.

Methods The Tromsø Study is a population-based cohort study in Northern Norway. All men aged 20–54 years (n=8238) and women aged 20–49 years (n=8001), free of known cardiovascular disease (CVD), who participated in a survey 1979–1980, were followed throughout 2010 for incident myocardial infarction (MI), and for incident MI or stroke used as proxy for incident CVD. HRs were estimated using a Cox proportional hazard regression model.

Results In age-adjusted analyses, Rose angina predicted MI and CVD in both sexes. The excess risk was substantially accounted for by CVD risk factors, leaving no significantly elevated MI risk above the risk explained by these factors (adjusted HR 1.31; 95% CI 0.95 to 1.80 in men, HR 1.20; 95% CI 0.69 to 2.10 in women). A similar pattern was seen for CVD (adjusted HR 1.16; 95% CI 0.87 to 1.55 in men and 1.30; 95% CI 0.82 to 2.06 in women).

Conclusions Rose angina predicted MI and CVD in a 29-years’ follow-up of a relatively young population. Established CVD risk factors were important mediators.

  • Cardiovascular disease
  • STROKE
  • Cohort studies
  • Epidemiology of cardiovascular disease
  • CORONARY HEART DISEASE

https://doi.org/10.1136/jech-2013-203642

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Two questionnaires were constructed by Geoffrey A Rose in the early 1960s for assessing the population burden of angina pectoris and peripheral artery disease, respectively.1 ,2 Short versions of the Rose Angina Questionnaire (RAQ) have been used in Norwegian health surveys.3–6 The RAQ has been implemented in national surveys worldwide, and it is among the recommended instruments for the assessment of cardiovascular disease (CVD) burden in surveys.7

    Follow-up studies have observed a higher coronary heart disease (CHD) mortality and morbidity in men with Rose angina as compared with men without this symptom,3–6 8–10 even if men with self-reported myocardial infarction (MI) or angina were excluded,4–6 ,9 ,10 if a three-step examination by physicians could not confirm an angina diagnosis,5 or if an electrocardiogram was normal.6 ,8 ,10 Most of these studies found that Rose angina carried an independent risk, above the risk conferred by CVD risk factors.5 ,6 ,8–10 A few long-term follow-up studies found that Rose angina in women, in relative terms, carried about the same CHD risk as in men.8–10 In a Norwegian cohort study, Rose angina conferred elevated CHD mortality, with HR 1.50; 95% CI 1.16 to 1.93 in men and HR 1.98; 95% CI 1.30 to 3.02 in women.9 The finding of an elevated risk in women may surprise, given the universal finding of lower incidence of MI among women than men,11 together with a trend of angina symptoms being more prevalent in women,12 suggesting that the specificity of angina symptoms in women is low. One explanation could be that angina symptoms, particularly in women, may reflect inflammation and endothelial dysfunction as early CHD manifestations,13 by contrast with symptom-free atherosclerosis which has been considered typical for early CHD. If Rose angina represents early pathophysiological processes that to some extent differ by sex, the mediating role of risk factors may show sex differences.

    The long-term cardiovascular prognosis of persons with Rose angina, in the absence of known CHD, is of interest in the perspective of interpretation of chest symptoms, and also for the choice of implementing Rose Questionnaire in surveys. On this background, we aimed to study the long-term cardiovascular consequences of Rose angina in relatively young women and men, including the mediating role of established CVD risk factors.

    Methods

    The Tromsø Study is an ongoing population-based cohort study in the municipality of Tromsø in North Norway.14 The study design includes six repeated health examination surveys, the first in 1974 (men only) and the second in 1979–1980 (both sexes). The baseline for our present study is the second survey (Tromsø 2), to which all men aged 20–54 years (born 1925–1959), and all women aged 20–49 years (born 1930–1959) living in the municipality of Tromsø were invited. All invitees were mailed the main survey questionnaire with questions on prior or present angina, heart attack, other heart disease, stroke, atherosclerotic disease of the legs, diabetes, use of antihypertensive medication and use of nitroglycerine, all with answering alternatives ‘yes’ or ‘no’. Furthermore, they reported on pulmonary symptoms (morning cough and morning phlegm, each yes/no), CHD in parents or siblings (yes/no), current smoking (yes/no) and number of cigarettes per day among smokers, leisure-time physical activity (four levels), some work-related factors, and educational level (years).

    At the examination site, height and weight were measured to the nearest cm and 0.5 kg. Systolic and diastolic blood pressures (SBP and DBP) were measured in sitting position twice after a 4 min rest using a mercury sphygmomanometer.15 Non-fasting serum total cholesterol (TC) and triglycerides were enzymatically determined, and the high-density lipoprotein (HDL)-cholesterol concentration was measured after precipitation of lower-density lipoproteins with heparin and manganese chloride.15

    A four-item version of the WHO RAQ was used: ‘Do you get pain or discomfort in your chest when walking up hills or stairs, or walking fast on level ground?’ (yes or no). ‘Do you get pain or discomfort in your chest when walking at normal pace on level ground?’ (yes or no). ‘If you get pain or discomfort in the chest when walking, do you usually: Stop? Slow down? Carry on with the same pace?’ (mark the alternative best-fitting). ‘If you stop or slow down, does the pain disappear:—within 10 minutes?—after 10 minutes or more?’ (mark the alternative best-fitting). Rose angina was defined by ‘yes’ to any of the first two questions, ‘stop’ or ‘slow down’ on the third and ‘less than 10 minutes’ on the fourth. There was no question on the specific localisation of the pain or discomfort.

    Case identification and validation

    The main endpoints in this study were incident MI, and MI or stroke, as proxy for incident CVD, whereas death from all causes was also assessed. Cases of incident MI and stroke were identified retrospectively by linking the Tromsø Study participant list to the hospital discharge register, including diagnoses from the outpatient clinic, at the University Hospital of North Norway, which is the only local hospital serving the Tromsø population,11 and from the national Causes of Death Registry. To identify all possible first-ever MI and stroke cases, we used a wide search strategy that in 1979 included the International Classification of Diseases (ICD) 8 codes 410–414 and 430–438; from 1980–1998 the ICD 9 codes 410–414 and 430–438, and thereafter ICD 10 codes I20–I25 and I60–I69. Adjudication of hospitalised and out-of-hospital first-ever MI and stroke was performed by an endpoint committee consisting of experienced physicians (including one cardiologist and one neurologist). Medical records were retrieved for case validation. For out-of-hospital deaths, records from prehospital care (ambulance service, general practitioners, nursing homes) and/or death certificates were searched for diagnostic criteria (clinical presentation, diagnostic procedures, laboratory tests, and/or autopsy). Revascularisation procedures were not systematically included as a CVD endpoint in the Tromsø Study endpoint registry.

    Modified MOnica Risks, Genetics, Archiving and Monograph (MORGAM) criteria were used to define MI and included clinical symptoms, findings in ECG, values of cardiac biomarkers and (when available) autopsy reports.16 We did not include silent MIs as defined by ECG as cases, because of difficulties in determining the date of occurrence. Autopsy was performed in around 25% of those who died on the day of MI attack.11 Stroke was defined as a focal or global neurological impairment of sudden onset, and lasting more than 24 h (or leading to death) and of presumed vascular aetiology.17

    Participants

    In Tromsø 2, 11 481 men and 9958 women aged 20–54 years (men), and 20–49 years (women) were invited, and 8477 men (74%) and 8143 women (82%) participated. Of those, 8423 men and 8080 women had complete data for this study. We excluded 185 men and 79 women with prevalent MI, angina (angina alone: 37 men, 10 women), stroke or other CVD at baseline (self-reported or according to the endpoint register), leaving 8001 women and 8238 men for follow-up analyses. The participants were followed for vital status and non-fatal or fatal first-ever MI and stroke throughout 2010. The date of migration from the study area, and date of death from causes other than CVD, were obtained from the Population Registry of Norway and the national Causes of Death Register. In all, 4131 persons migrated from Tromsø during the follow-up period. The mean age of the migrants was lower than of those who stayed, by 4 years in men and 3 years in women. Of men, 24.4% with negative and 15.7% with positive self-reports on CVD at baseline migrated, while there was no such difference in women.

    Statistical analysis

    The Cox proportional hazard regression model was used to assess the association between Rose angina and the endpoints. Participants were censored at time of death, migration, or the end of follow-up at 31 Dec 2010, whichever came first. Results are given as HRs with 95% CIs. Age was included by 1-year groups. Pulmonary symptoms, defined by positive answer to a question on morning cough or a question of morning phlegm, overweight (Body Mass Index (BMI) ≥ 25 kg/m2) and indicators of socioeconomic status (education at basic or higher level, disability pension and poverty during childhood) were considered as potential confounders.

    Daily smoking, family history of MI, use of antihypertensive medication, serum TC, SBP and DBP were included as covariates to evaluate their potential mediating role. HRs for TC, SBP and DBP were reported per SD. Physical inactivity may be a consequence of angina symptoms and was not included. Only factors that influenced the association between Rose angina and the endpoint by 5% or more were included in the final models. The Tromsø Study is approved by the Norwegian Data Inspectorate and the Regional Committee of Medical and Health Research Ethics.

    Results

    The mean length of follow-up until death was 26.8 (SD 4.4, range 0–29) years, while mean follow-up until first MI and stroke was 22.0 (SD 9.4, range 0–29) and 22.3 (SD 9.3, range 0–29) years, respectively. Of the participants included, 168 men (2.0%) and 166 women (2.1%) screened positive for Rose angina. Table 1 shows baseline characteristics, including potential confounders and mediating factors. There was a significant overlap between Rose angina and pulmonary symptoms. Men and women with Rose angina had more often low education and a family history of MI; they were more often overweight and current smokers; and their mean serum TC, triglycerides and DBP were higher, compared with those who screened negative. The use of antihypertensive medication was low-frequent, and self-reported diabetes was uncommon (50 persons in all, not shown in the table). Men with Rose angina were less likely than symptom-free men to migrate during follow-up (18.5% vs 24.5%), while there was no such difference in women (26%).

    View this table:
    Table 1

    Baseline characteristics of men and women with negative self-reports on heart attack, angina or stroke, by Rose angina (the Tromsø Study 1979–80)*

    Mortality and morbidity at follow-up

    By the end of follow-up, 1576 (19.2%) men and 695 (8.7%) women had died. In men, the number of incident MIs was 1095 (13.3% of all men), and in women the corresponding number was 276 (3.4%) (table 2). A total of 506 (6.2%) men and 199 (2.5%) women had a non-fatal or fatal stroke. The incidence of MI events was approximately doubled among those with Rose angina as compared with those free of this symptom (table 2), and in women this pattern was seen also for stroke.

    View this table:
    Table 2

    Deaths and incident myocardial infarction and stroke by lone Rose angina*

    The age-adjusted all-cause mortality was significantly elevated in men with Rose angina at baseline (HR 1.44; 95% CI 1.10 to 1.87). When adjusting for pulmonary symptoms, the HR was lowered (HR 1.25; 95% CI 0.96 to 1.64, model 2). In women, the HR was not significantly elevated when adjusted for age alone (HR 1.37; 95% CI 0.93 to 2.02), and inclusion of pulmonary symptoms and overweight lowered the HR to 1.16; 95% CI 0.79 to 1.72. Inclusion of indicators of socioeconomic status, smoking, or other risk factors, did not change this result.

    In both sexes, the risk of non-fatal or fatal MI was significantly elevated (HR 1.52; 1.11 to 2.08, in men, HR 1.73; 1.01 to 2.98 in women) when adjusting for age alone (table 3, model 1). The association was weaker when adjusted also for pulmonary symptoms and overweight (table 3, model 2). Inclusion of education or other indicators of socioeconomic status did not change this result. In men, the association was non-significant (HR 1.31; 0.95 to 1.80) when including overweight, smoking, family history of MI and serum TC (table 3, model 3). In women, the association was non-significant when adjusted for overweight, smoking and TC, while family history of MI did not attenuate the HR further. Additional inclusion of SBP, DBP, antihypertensive medication or diabetes had no or minimal impact.

    View this table:
    Table 3

    HRs for incident myocardial infarction*

    Three factors had large effects on the association between Rose angina and MI when added to the model that also included age: in men, family history of MI and smoking attenuated the association to HR 1.38; 1.01 to 1.90, and in women the HR was 1.28; 0.74 to 2.23 following the inclusion of smoking and TC. When limiting the analysis to men aged 35–54 years, the multivariable adjusted association between Rose angina and MI was similar to the result in the total age group (HR 1.33; 0.95 to 1.85 adjusted like in table 3, model 3). In women, the corresponding HR was 1.26; 0.70 to 2.25.

    Table 4 shows the association between Rose angina and incident non-fatal or fatal MI or stroke, as a proxy for CVD events. In men, the age-adjusted association was of borderline significance (HR 1.33; 1.00 to 1.78) and weakened when adding overweight and pulmonary symptoms as covariates (model 2). In a model with age, overweight, family history of MI, smoking, TC, DBP and SBP included, the association was attenuated further (HR 1.16; 0.87 to 1.55, model 3). In women, the age-adjusted association between Rose angina and CVD events was significantly elevated (HR 1.73; 1.10 to 2.72) and weakened by additional adjustment for pulmonary symptoms and overweight. When including age, overweight, smoking and TC, the association was further attenuated (HR 1.30; 0.82 to 2.06, model 3). Adjustment for pulmonary symptoms had no impact when the CVD risk factors were added to the model and was for this reason omitted in model 3, tables 3 and 4.

    View this table:
    Table 4

    HRs for incident cardiovascular disease*.

    The exclusion of participants who reported angina as their only CVD manifestation at baseline might potentially bias the results, and we made a supplementary analysis with these 37 men and 10 women included. In men, Rose angina was associated with incident MI by HR 1.30 (0.97 to 1.73) when including education, SBP and DBP in the final model. Other results were similar to those reported in tables 3 and 4.

    Discussion

    This study confirms earlier findings of elevated risk of MI and CVD by Rose angina in both sexes.8–10 The result suggests that established CVD risk factors are important mediating factors. A modest effect of Rose angina per se, in addition to the risk explained by potential confounders or conferred by risk factors, cannot be excluded, given the relatively wide CIs. But in sum, this study does not confirm the findings in other studies8–10 when it comes to a significant ‘independent’ effect with risk factors included in analysis. If Rose angina represented CHD in our cohort, even in early stage, prognostic consequences were to be expected. In the Women's Ischaemia Syndrome Evaluation (WISE) Study, chest symptoms with angiogram defining non-obstructive CHD (1–49% stenosis of coronary arteries) were associated with future CVD events after adjustment for baseline risk factors.18 Similarly, a study of Danish men and women demonstrated more than 50% elevated risk for CVD outcomes by angina symptoms, compared with a reference population without known CVD or angina symptoms.19 In this study, the participants with angina symptoms had normal arteries or diffuse non-obstructive CHD according to angiograms, and adjustment was made for smoking, BMI, diabetes and use of lipid-lowering and antihypertensive medication, which did not attenuate the risk conferred by angina symptoms.19 It should, however, be noted that the mean age was around 55–60 years in these studies,18 ,19 while it was below 40 years in ours. In addition to the younger age, high levels of CVD risk factors may explain why we could not confirm an independent effect of Rose angina. Like several other studies,8–10 ,19 our results do not suggest an important sex difference in the association of Rose angina with the endpoints. The mediating role of CVD risk factors seemed rather similar in women and men.

    Potential non-cardiac causes of Rose angina

    Several studies have examined the RAQ with respect to specificity and sensitivity. Geoffrey Rose noted that the diagnosis of angina presents special problems in populations with a high prevalence of chronic bronchitis,1 and overlap has also been found with asthma.20 In a cross-sectional survey data of men (n=30 658) and women (n=35 733) aged 40–42 years, we found that Rose angina was associated with asthma, anxiousness, depressive complaints, musculoskeletal pain and overweight,21 indicating confounding factors that are hard to fully account for. These health problems may be of particular importance in relatively young populations.

    Methodological strengths and weaknesses

    The study was performed in a general population with a high attendance rate, and the follow-up period covered nearly 30 years. A major strength is the high-quality end-point register with rigorous case validation of non-fatal and fatal events. We observed rather high numbers of events in men. The number of endpoints was by far lower in women than in men, and the CIs were wider, giving less conclusive results. Around 25% of the participants migrated, with a loss of observation years for this reason. Migrants were younger, and among men, there was a ‘healthy migrant’ tendency. However, the rate of migration is a weakness of our study.

    The young age when reporting should be noted, and our results are not necessarily valid for older age groups. The oldest men and women reached 85 and 80 years, respectively, by the end of follow-up, and some non-fatal CVD endpoints might be under-reported if not all cases were admitted to hospital. Silent MIs were not included, even if diagnosed by electrocardiogram, due to difficulties in estimating the exact date for the event. The lack of revascularisations as endpoints is a weakness. But these procedures have been increasingly used in recent years, and could have contributed with events among long-term survivors, mainly.

    The findings are in line with established knowledge on CVD risk factors.22 We find it difficult to draw a sharp line between confounders and mediating factors in the analysis. Overweight may act as a confounder, by abdominal fat mass giving discomfort by exertion, but is also a CVD risk factor.22 Smoking may cause pulmonary symptoms that give rise to chest discomfort, and may, in this way, act as a confounder, in addition to being a mediating factor. Physical inactivity may be a consequence of chest symptoms, but may also be a mediating factor.

    Conclusion

    In this population-based cohort study with nearly 30 years’ follow-up, Rose angina was linked with increased incidence of MI and also with MI or stroke as proxy for CVD. In both sexes, the excess risk was to a considerable degree explained by established CVD risk factors. This study may discourage the use of the RAQ in population surveys in relatively young adult populations.

    What is already known on this subject

    • Follow-up studies have shown that men and women with negative history of coronary heart disease (CHD), but angina symptoms according to the Rose Questionnaire, have elevated CHD morbidity and mortality, compared with people without this symptom. The risk has not been explained by cardiovascular disease risk factors, suggesting that Rose angina indicates CHD in early stage.

    What this study adds

    • The excess risk conferred by Rose angina in this study was substantially accounted for by potential confounding factors and established cardiovascular disease risk factors. The finding may discourage the use of Rose Angina Questionnaire in surveys in the age groups we studied (men 20–54 years, women 20–49 years).

    References

    1. ↵
      1. Rose G
      . The diagnosis of ischemic heart pain and intermittent claudication in field surveys. Bull WHO 1962;27:645–57.
      OpenUrlPubMedWeb of Science
    2. ↵
      1. Rose G,
      2. McCartney P,
      3. Reid DD
      . Self-administration of a questionnaire on chest pain and intermittent claudication. Br J Prev Social Med 1977;31:42–8.
      OpenUrl
    3. ↵
      1. Zeiner-Henriksen T
      . Six year mortality related to cardiorespiratory symptoms and environmental risk factors in a sample of the Norwegian population. J Chron Dis 1976;29:15–33.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Håheim LL,
      2. Holme I,
      3. Hjermann I,
      4. et al
      . Trends in the incidence of acute myocardial infarction and stroke: a 21-year follow-up of the Oslo study. Scand Cardiovasc J 2004;38:216–21.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Wilhelmsen L,
      2. Rosengren A,
      3. Hagman M,
      4. et al
      . “Nonspecific” chest pain associated with long-term mortality: results from the primary prevention study in Göteborg, Sweden. Clin Cardiol 1998:21:477–82.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Bodegård J,
      2. Erikssen G,
      3. Bjørnholt JV,
      4. et al
      . Possible angina detected by the WHO angina questionnaire in apparently healthy men with a normal exercise ECG: coronary heart disease or not? A 26 year follow up study. Heart 2004; 90:627–32.
      OpenUrlAbstract/FREE Full Text
    7. ↵
      1. Primatesta P,
      2. Allender S,
      3. Ciccarelli P,
      4. et al
      . Cardiovascular surveys: manual of operations. Eur J Cardiovasc Prev Rehabil 2007;14(Suppl 3):S43–61.
      OpenUrlCrossRef
    8. ↵
      1. Murphy NF,
      2. Stewart S,
      3. Hart CL,
      4. et al
      . A population study of the long-term consequences of Rose angina: 20-years follow-up of the Renfrew-Paisley study. Heart 2006;92:1739–46.
      OpenUrlAbstract/FREE Full Text
    9. ↵
      1. Graff-Iversen S,
      2. Selmer R,
      3. Løchen ML
      . Rose angina predicts 23-year coronary heart disease mortality in women and men aged 40–49 years. Heart 2008; 94:482–6.
      OpenUrlAbstract/FREE Full Text
    10. ↵
      1. Khalili D,
      2. Hadaegh F,
      3. Fahimfar N,
      4. et al
      . Does an electrocardiogram add predictive value to the rose questionnaire for future coronary disease? 10-year follow-up in a Middle-East population. J Epidemiol Community Health 2012; 66:1104–9.
      OpenUrlAbstract/FREE Full Text
    11. ↵
      1. Mannsverk J,
      2. Wilsgaard T,
      3. Njølstad I,
      4. et al
      . Age and gender differences in incidence and case fatality trends for myocardial infarction: a 30-year follow-up of the Tromsø Study. Eur J Prev Cardiol 2012;19:927–34.
      OpenUrlAbstract/FREE Full Text
    12. ↵
      1. Hemingway H,
      2. Langenberg C,
      3. Damant J,
      4. et al
      . Prevalence of angina in women versus men: a systematic review and meta-analysis of international variations across 31 countries. Circulation 2008:117:1526–36.
      OpenUrlAbstract/FREE Full Text
    13. ↵
      1. Banks K,
      2. Puttagunta D,
      3. Murphy S,
      4. et al
      . Clinical characteristics, vascular function, and inflammation in women with angina in the absence of coronary atherosclerosis. JACC: Cardiovasc Imaging 2011;4:65–73.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Jacobsen BK,
      2. Eggen AE,
      3. Mathiesen EB,
      4. et al
      . Cohort profile: the Tromsø study. Int J Epidemiol 2012;41:961–7.
      OpenUrlFREE Full Text
    15. ↵
      1. Wilsgaard T,
      2. Jacobsen BK,
      3. Shirmer H,
      4. et al
      . Tracking of cardiovascular risk factors. The Tromsø Study, 1979–1995. Am J Epidemiol 2001;154:418–26.
      OpenUrlAbstract/FREE Full Text
    16. ↵
    17. ↵
      WHO MONICA Project Principal Investigators. The World Health organization MONICA Project (monitoring trends and determinants in cardiovascular disease): a major international collaboration. J Clin Epidemiol 1988;41:105–14.
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      1. Gulati M,
      2. Cooper-DeHoff RM,
      3. McClure C,
      4. et al
      . Adverse cardiovasular outcomes in women with non-obstructive coronary artery disease: a report from the Women's Ischemia Syndrome Evaluation Study and the St James Women Take Heart Project. Arch Intern Med 2009;169:843–50.
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Jespersen L,
      2. Hvelplund A,
      3. Abildstrøm SZ,
      4. et al
      . Stable angina pectoris with no obstructive coronary artery disease is associated with risks of major adverse cardiovascular events. Eur Heart J 2012;33:734–44.
      OpenUrlAbstract/FREE Full Text
    20. ↵
      1. Kiss D,
      2. Veegh W,
      3. Schragel D,
      4. et al
      . Bronchial asthma causing symptoms suggestive of angina pectoris. Eur Respir J 2003;21:473–7.
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. Graff-Iversen S,
      2. Selmer R,
      3. Løchen ML
      . Rose Questionnaire in Norwegian Health surveys. Nor J Epidemiol 2003;13:95–102. http://www.ntnu.no/ojs/index.php/norepid/issue/view/65
      OpenUrl
    22. ↵
      1. Perk J,
      2. De Backer G,
      3. Gohlke H,
      4. et al
      . European Guidelines on cardiovascular disease prevention in clinical practice (version 2012). Int J Behav Med 2012; 19:403–88.
      OpenUrlCrossRefPubMed

    Footnotes

    • Contributors The first author drafted the manuscript and made the analysis. The second author gave advice for the analysis. The third, fourth and last author have contributed to the development and quality of the endpoint register. All authors contributed to the project planning and with suggestions and critical review during the research process.

    • Competing interests None.

    • Ethics approval Regional Committee for medical and health research ethics.

    • Provenance and peer review Not commissioned; externally peer reviewed.