European Heart Journal (1998) 19, 6373

Chest pain on questionnaire and prediction of major ischaemic heart disease events in men
F. C. Lampe, P. H. Whincup, S. G. Wannamethee, S. Ebrahim, M. Walker, A. G. Shaper
Cardiovascular Research Unit, Department of Primary Care and Population Sciences, Royal Free Hospital School of Medicine, London, U.K.

Objective To examine the prediction of major ischaemic heart disease events by questionnaire-assessed chest pain and other symptoms. Design Population-based prospective study. Subjects 7735 randomly selected men, aged 4059 years at entry. Methods Symptoms and history of diagnosed ischaemic heart disease were ascertained by administered questionnaire at baseline. Follow-up was for an average of 147 years, for rst major ischaemic heart disease event. Results During follow-up, 969 men had a major ischaemic heart disease event. Denite angina (chest pain fullling all WHO criteria) and possible angina (exertional chest pain without all other WHO criteria) were associated with similar ischaemic heart disease outcome, and a single combined angina category was used. In the whole cohort, the relative risks (95% CI) of a major ischaemic heart disease event were 203 (161, 257) for angina only, 213 (172, 263) for possible myocardial infarction only and 450 (357, 566) for angina plus possible myocardial infarction, compared to no chest pain. The relative risk for recall of an ischaemic heart disease diagnosis was 398 (336, 471). Only 33% of men with angina or possible myocardial infarction symptoms recalled a previous ischaemic heart disease diagnosis. In men without recall of an ischaemic heart disease diagnosis (in whom 82% of events during

follow-up occurred), chest pain symptoms remained predictive of major ischaemic heart disease events with relative risks (95% CI) of 169 (127, 224) for angina only, 149 (112, 197) for possible myocardial infarction only and 255 (144, 453) for angina plus possible myocardial infarction. Other chest pain increased risk of a major ischaemic heart disease event by 119 (101, 140) compared to no chest pain. Symptoms of breathlessness or calf pain on walking increased ischaemic heart disease risk in men with other chest pain and in men without chest pain, but had no further eect on ischaemic heart disease risk in men with symptoms of angina or possible myocardial infarction. Conclusions In dening angina by chest pain questionnaire, the exertional component is the crucial criterion. When using questionnaire-assessed symptoms to determine ischaemic heart disease risk, information on previous ischaemic heart disease diagnoses should be taken into account. The majority of men with angina or possible myocardial infarction symptoms do not have a diagnosis of ischaemic heart disease, but they remain at signicantly increased risk of a major ischaemic heart disease event. The value of breathlessness and calf pain on walking in stratifying ischaemic heart disease risk is restricted to men with other chest pain or no chest pain. (Eur Heart J 1998; 19: 6373) Key Words: Chest pain, angina, Rose questionnaire, Ischaemic heart disease, prediction.

Introduction
The WHO (Rose) chest pain questionnaire is used widely in epidemiological studies as a validated and standardized method for dening angina and possible
Revision submitted 26 June 1997, and accepted 10 July 1997. Correspondence: Ms F. Lampe, Cardiovascular Research Unit, Department of Primary Care and Population Sciences, Royal Free Hospital School of Medicine, Rowland Hill Street, London NW3 2PF, U.K. 0195-668X/98/010063+11 $18.00/0 hj970729

myocardial infarction[1,2] and has consistently been shown to be predictive of mortality and major ischaemic heart disease events[313]. Despite extensive use of the questionnaire, several aspects merit further study. First, the need for all of the strict criteria used to dene angina has been questioned, with evidence that chest pain on exertion (without additional criteria) is a satisfactory denition[9,10]. Second, we have already shown that recall of a doctor diagnosis of ischaemic heart disease is among the strongest predictors of subsequent major ischaemic heart disease events[5,6,12], but the extent to
 1998 The European Society of Cardiology

64

F. C. Lampe et al. according to standard criteria as chest pain or discomfort which: (i) was brought on by exertion, (ii) was situated in the central or left anterior chest (site 4, 5 or 8 on diagram), (iii) forced the subject to slow down or stop, (iv) was relieved if the subject did so and (v) was relieved within 10 min. Possible angina was dened as chest pain brought on by exertion, but not fullling all of the four additional criteria for denite angina. Both possible and denite angina could be subdivided according to severity as grade I (chest pain brought on only by walking uphill or hurrying) or grade II (chest pain brought on by walking at an ordinary pace on the level). Chest pain that was not brought on by walking was classied as non-exertional chest pain. Responses regarding exertional chest pain were inconclusive in 14 cases; these were classied as non-exertional chest pain. Information regarding grade of angina was incomplete or contradictory in 16 cases: these were classied as grade I angina. One man gave no information on chest pain. Prolonged severe chest pain Questions were based on the WHO severe chest pain question[1,16]. Men were classied into one of three groups: possible myocardial infarction, other prolonged severe chest pain or no prolonged severe chest pain. Possible myocardial infarction was dened as ever having had an episode of severe pain in the central or left anterior chest (site 4, 5 or 8), lasting for half an hour or more. Other prolonged severe chest pain was dened as severe chest pain in a site not consistent with possible myocardial infarction. One man gave no information on severe chest pain. Chest pain status Chest pain status overall was classied for 7733 men, on the basis of their response to both chest pain and prolonged severe chest pain questions, into ve mutually exclusive groups: (i) No chest pain no chest pain of any duration; (ii) Other chest pain non-exertional chest pain or a history of other prolonged severe chest pain, but no angina or possible myocardial infarction; (iii) Angina only possible or denite angina, without a history of possible myocardial infarction; (iv) possible myocardial infarction only a history of possible myocardial infarction without possible or denite angina; (v) Angina and possible myocardial infarction possible or denite angina with a history of possible myocardial infarction. Calf pain on walking A modied version of the WHO intermittent claudication questionnaire was used[17]. Calf pain on walking was dened as an armative response to either question (see Appendix). The two questions were not administered satisfactorily to the 915 men in the rst three towns and therefore these men were excluded for this variable. Twelve further men did not provide information on calf pain.

which the WHO chest pain questionnaire predicts ischaemic heart disease risk independently of a previous doctor diagnosis of ischaemic heart disease has not been examined. It is of particular interest to assess the usefulness of questionnaire-assessed symptoms in predicting major ischaemic heart disease events in men who are not already known to their doctors to be suering from ischaemic heart disease. Third, the extent to which prediction by chest pain symptoms can be enhanced by other questionnaire-assessed symptoms, such as breathlessness and calf pain on walking, has been little studied. These issues are important both for the use of questionnaires in epidemiological studies and in clinical practice, where it is becoming increasingly important to stratify subjects on the basis of their absolute risk of ischaemic heart disease[14]. We have examined the prediction of major ischaemic heart disease events by questionnaire-assessed chest pain symptoms in middle-aged men in the British Regional Heart Study, using 147 years average followup. Complete and simplied responses to the WHO angina questionnaire have been compared, and prediction by chest pain symptoms assessed in men with and without a pre-existing doctor diagnosis of ischaemic heart disease. We have also examined the contribution made by additional symptoms (breathlessness and calf pain on walking) to the prediction of ischaemic heart disease events in men without diagnosed ischaemic heart disease.

Methods
The British Regional Heart Study examined 7735 men aged 4059 years at entry (197880), randomly selected from the agesex registers of one general practice in each of 24 towns in England, Wales and Scotland. The criteria for selecting the towns, the general practices and the subjects have been reported previously[15]. In brief, the 24 towns were taken from those with populations of 50 000100 000 (1971 census). They were chosen to represent the full range of mortality from cardiovascular disease and included all major geographic regions. The general practice in each town was required to have a social class distribution representative of that town. No attempt was made to exclude men with cardiovascular problems. The average response rate was 78%. Research nurses examined each man, administered symptom questionnaires (see Appendix) and inquired about cardiovascular diagnoses.

Classication of symptoms
Chest pain Chest pain questions were based on the WHO angina questionnaire[1,16]. Men were classied into one of four groups: denite angina, possible angina, non-exertional chest pain or no chest pain. Denite angina was dened
Eur Heart J, Vol. 19, January 1998

Chest pain and prediction of ischaemic heart disease in men

65

Table 1

Major ischaemic heart disease events by chest pain category at baseline: all men
n Prevalence (%) Mean age (years) Event rate /1000/year (number of events) RR age adjusted (95% CI)

Chest pain status No chest pain* Other chest pain Angina only PMI only Angina and PMI Angina or PMI Total

4787 1849 391 490 216 1097 7733

(619) (239) (51) (63) (28) (142)

501 495 524 513 529 520 502

75 86 172 169 390 208 95

(485) (213) (82) (103) (86) (271) (969)

1 119 203 213 450 250

(101, (161, (172, (357, (216,

140) 257) 263) 566) 291)

*Reference category. Relative risks compared to reference category, adjusted for age. PMI=possible myocardial infarction, angina=possible or denite angina.

Breathlessness A modied version of the MRC respiratory questionnaire was used[18]. Breathlessness was dened as an armative response to any of the three questions (see Appendix). Eleven men could not be classied.

Statistical analysis
Event rates per thousand per year were calculated as the number of major ischaemic heart disease events divided by the person-years at risk 1000. The cumulative probability of remaining free of an event during the follow-up period was estimated using the KaplanMeier technique. Cox proportional hazards models were used to assess the eect of symptoms on time free of a major ischaemic heart disease event, adjusted for age. Deaths from causes other than ischaemic heart disease were treated as censored observations. Age was tted as a continuous variable; symptoms and their interactions were tted using dummy variables with appropriate reference categories. Results are presented as relative risks with 95% condence intervals (95% CI). The loglikelihood ratio statistic[21] (LRS) was used to assess statistical signicance.

Recall of ischaemic heart disease diagnosis

Men who reported at initial examination having been told by a doctor that they had either angina or a heart attack (including coronary thrombosis or a myocardial infarction), were classied as having recall of an ischaemic heart disease diagnosis[19]. Nine men could not be classied.

Follow-up
The 7735 men initially examined in 197880 were followed up until the end of December 1993 (mean 147 years, range 133 to 160) for cardiovascular morbidity and all-cause mortality. The established tagging procedure carried out by the National Health Service Central Registers in Southport (for England and Wales) and Edinburgh (for Scotland) was used for notication of deaths, with date and cause from the death certicate. A fatal coronary heart disease event was dened as a death coded to ICD 410414. Information on non-fatal myocardial infarction was obtained from reviews of each subjects general practice records. A non-fatal myocardial infarction was dened as an event associated with at least two features of: severe prolonged chest pain; electrocardiographic evidence of myocardial infarction at the time of the event; cardiac enzyme abnormalities, with survival for at least 28 days[20]. Follow-up was achieved for 99% of the original cohort. The present analysis is concerned with the rst major ischaemic heart disease event (non-fatal or fatal myocardial infarction or sudden cardiac death) occurring during the follow-up period.

Results
Ischaemic heart disease outcome During the period from initial examination to 31 December 1993, 969 (125%) of the 7735 men experienced a major ischaemic heart disease event, with the rst event during follow-up being fatal in 406 men. The overall major ischaemic heart disease event rate was 95/1000/year. Prevalence of chest pain Denite angina was present in 367 (47%) men, and possible angina in 240 (31%). Of the 607 men with questionnaire-assessed angina, 512 (84%) had grade I (milder) angina. Daily or weekly episodes of pain were experienced by the majority (64%) of men with grade II angina, compared to only 30% of those with grade I angina. A single occurrence of pain only was reported by 28 (46%) men with angina symptoms. Table 1 shows the prevalence of chest pain status overall. Symptoms of ischaemic heart disease (either angina or a history of
Eur Heart J, Vol. 19, January 1998

66

F. C. Lampe et al.

1.0 No chest pain 0.9 Non-exertional pain 0.8 Cumulative proportion event free 0.7 0.6 0.5 0.4 0.3 0.2 0.1 Possible angina

Definite angina

8 Years of follow up

10

12

14

16

Figure 1 Proportion free of a major ischaemic heart disease event by WHO angina status. possible myocardial infarction) were found in 1097 (142%) men, but only 216 (20%) of these had both angina and possible myocardial infarction. Of the 1849 men with other chest pain, only 69 had a history of prolonged severe chest pain in an inappropriate site for possible myocardial infarction, the remainder having non-exertional chest pain. Angina symptoms and ischaemic heart disease outcome During follow-up, the major ischaemic heart disease event rates for denite and possible angina were 254/1000/year and 222/1000/year respectively, and the age-adjusted relative risk (95% CI) of denite compared to possible angina was 107 (078, 147) (LRS chisquared (1)=018, P =0.67). The event free survival experience of the two groups over the follow-up period was very similar, with both groups being clearly distinct from non-exertional chest pain (Fig. 1). The groups were also similar at baseline with regard to the proportion with grade II angina (163% and 146% for denite and possible respectively), daily or weekly episodes of pain (345% and 361%, compared to 157% of men with non-exertional pain) and a history of possible myocardial infarction (373% and 329%, compared to 93% of men with non-exertional pain). In the remainder of this paper, angina refers to denite and possible angina combined i.e. to any exertional chest pain. Grade II (severe) angina was associated with a major ischaemic heart disease event rate of 325/1000/ year, compared to a rate of 227/1000/year for grade I
Eur Heart J, Vol. 19, January 1998

angina, but this dierence was not statistically signicant (age adjusted relative risk (95% CI): 133 (090, 195) LRS chi-squared (1)=194, P =016). Neither frequency of chest pain, nor how recently pain had occurred signicantly inuenced risk of a major ischaemic heart disease event in men with angina during the follow-up period. Chest pain status and ischaemic heart disease outcome The inuence of chest pain status at baseline on major ischaemic heart disease events is shown in Table 1. The presence of either angina or possible myocardial infarction at baseline was associated with an age-adjusted relative risk of 25 compared with no chest pain. Angina only and possible myocardial infarction only were each associated with approximately a twofold increase in risk, but the presence of both symptoms together increased risk considerably. Other chest pain marginally but signicantly increased risk of a major ischaemic heart disease event compared to no chest pain. Recall of an ischaemic heart disease diagnosis and chest pain symptoms A doctor diagnosis of ischaemic heart disease (angina or myocardial infarction) was recalled by 424 (55%) men. These men were at extremely high risk of a new major ischaemic heart disease event during follow-up, with an event rate of 391/1000/year, compared to a rate of 82/1000/year for men who did not recall an ischaemic heart disease diagnosis (age-adjusted relative

Chest pain and prediction of ischaemic heart disease in men

67

Table 2 Association between chest pain category and recall of ischaemic heart disease diagnosis: all men
Recall of diagnosis of IHD (angina or MI) n Chest pain status No chest pain Other chest pain Angina only PMI only Angina and PMI Total (%)

4782 1847 390 489 216 7724*

26 37 79 118 164

(05) (20) (203) (241) (759)

diagnosed ischaemic heart disease reported chest pain. In men who did not recall an ischaemic heart disease diagnosis (n=7302), symptoms of angina or possible myocardial infarction were associated with an ageadjusted relative risk of 164 compared to no chest pain. Again, relative risks were similar for angina only and possible myocardial infarction only, and men who had both angina and possible myocardial infarction were at markedly higher risk than those with one symptom alone. Of the 363 men who had questionnaire-assessed angina in this group, 37 (102%) had grade II angina and this increased risk by 136 times that for grade I. Breathlessness, calf pain on walking and chest pain in men without diagnosed ischaemic heart disease Table 4 shows the associations of breathlessness and calf pain on walking with chest pain status in men without recall of an ischaemic heart disease diagnosis. Both breathlessness and calf pain were much more likely to be reported by men who had chest pain. Men with questionnaire-assessed angina were particularly likely to have these additional symptoms, with about half reporting breathlessness and about a third reporting calf pain. Breathlessness and calf pain were less common in men who had possible myocardial infarction only this group having similar rates of these symptoms to men with other chest pain. Breathlessness in addition to chest pain: eect on outcome in men without diagnosed ischaemic heart disease The inuence of breathlessness with chest pain on major ischaemic heart disease events and total mortality was examined in men without recall of an ischaemic heart disease diagnosis (Table 5). Angina or possible myocardial infarction was used as a single group in order to have sucient numbers for analysis. The eect of breathlessness on major ischaemic heart disease

*Two men had missing data for chest pain status and 9 men had missing data for ischaemic heart disease recall. PMI=possible myocardial infarction; angina=possible or denite angina.

risk (95% CI): 398 (336, 471)). There was a very strong association between chest pain symptoms and recall of a doctor diagnosis of ischaemic heart disease (Table 2), although overall only 33% of men with questionnaireassessed angina or possible myocardial infarction symptoms reported having an ischaemic heart disease diagnosis. Chest pain symptoms in men with and without diagnosed ischaemic heart disease: eect on outcome The inuence of questionnaire-assessed chest pain symptoms on major ischaemic heart disease events was examined separately in men with and without recall of an ischaemic heart disease diagnosis (Table 3). In men with an ischaemic heart disease diagnosis (n=424), the presence of angina symptoms or a history of possible myocardial infarction did not appear to greatly inuence risk of a new event, but numbers in the reference group were very small as the vast majority of men with

Table 3 Major ischaemic heart disease events by chest pain category at baseline in men with and without recall of an ischaemic heart disease diagnosis
With IHD diagnosis n Event rate /1000/year (number of events) RR age adjusted (95% CI) n Without IHD diagnosis Event rate /1000/year (number of events) RR age adjusted (95% CI)

Chest pain status No chest pain* Other chest pain Angina only PMI only Angina and PMI Angina or PMI Total

26 37 79 118 164 361 424

409 207 340 389 460 409 391

(11) (8) (28) (49) (74) (151) (170)

1 051 078 095 113 099

(021, (039, (050, (060, (053,

128) 157) 182) 212) 182)

4756 1810 311 371 52 734 7300

74 84 135 112 201 127 82

(474) (205) (53) (54) (12) (119) (798)

1 119 169 149 255 164

(101, (127, (112, (144, (134,

140) 224) 197) 453) 201)

*Reference category. Relative risks compared to reference category, adjusted for age. PMI=possible myocardial infarction, angina=possible or denite angina. Eur Heart J, Vol. 19, January 1998

68

F. C. Lampe et al.

Table 4 Association of chest pain category with breathlessness and calf pain on walking: men without recall of an ischaemic heart disease diagnosis
n Breathlessness n (%) n Calf pain on walking n (%)

Chest pain status No chest pain Other chest pain Angina only PMI only Angina and PMI Total

4751 1808 310 371 52 7292*

502 349 141 83 35 1110

(106) (193) (455) (224) (673) (152)

4193 1568 285 328 47 6421

302 221 70 51 20 664

(72) (141) (246) (155) (426) (103)

*Ten men without an ischaemic heart disease diagnosis had missing data for breathlessness or chest pain status. Eight hundred and eighty-one men without an ischaemic heart disease diagnosis had missing or unusable data for calf pain or had missing data for chest pain status. PMI=possible myocardial infarction, angina=possible or denite angina.

events diered according to category of chest pain. Breathlessness in the absence of chest pain was associated with an age-adjusted relative risk of 136. Similarly, in those with other chest pain, breathlessness increased the risk by a factor of 155. However, in men with angina or possible myocardial infarction, the additional presence of breathlessness did not increase risk of a major ischaemic heart disease event (test for interaction between chest pain status and breathlessness: LRS chisquared (2)=659, P =0037). It can also be seen that, among those without diagnosed ischaemic heart disease, men with other chest pain and breathlessness had an event rate similar to that of men with symptoms of angina or possible myocardial infarction. In all three chest pain groups, breathlessness had a marked eect on total mortality, which was mainly due to an increase in non-cardiovascular deaths. Calf pain on walking in addition to chest pain: eect on outcome in men without diagnosed ischaemic heart disease The eect of calf pain with chest pain on major ischaemic heart disease events in men without an ischaemic heart disease diagnosis showed a similar pattern to that seen for breathlessness (Table 5). Calf pain increased the risk of a major ischaemic heart disease event by 163 and 179 in those with no chest pain and other chest pain respectively, but did not increase the risk in those with angina or possible myocardial infarction (test for interaction LRS chi-squared (2)=714, P =0028). Again, lower ischaemic heart disease event rates were seen only in those who had neither calf pain nor angina or possible myocardial infarction symptoms. The eect of calf pain on total mortality was weaker than that of breathlessness. Calf pain increased total mortality risk in men with no chest pain and other chest pain, but caused only a marginal non-signicant increase in men with angina or possible myocardial infarction.
Eur Heart J, Vol. 19, January 1998

Discussion
This paper has examined prediction of major ischaemic heart disease events by the use of an administered questionnaire for the assessment of chest pain symptoms. We have shown that, in middle-aged British men, denite and possible WHO (Rose) angina were associated with similar ischaemic heart disease outcome over a long follow-up period. A large part of the risk associated with questionnaire-assessed chest pain symptoms in the cohort of men as a whole was attributable to men with a pre-existing ischaemic heart disease diagnosis, who were at very high risk of a new major ischaemic heart disease event. However, a signicant increased risk remained in men with questionnaire-assessed angina and/or possible myocardial infarction who did not have a previous diagnosis of ischaemic heart disease. Breathlessness and calf pain on walking increased ischaemic heart disease risk only in those without symptoms of ischaemic heart disease. Denite and possible angina on questionnaire A previous report from the British Regional Heart Study (BRHS) compared denite and possible questionnaire-assessed angina over 75 years of followup[9]; this present paper has extended the follow-up period and shown that men who had standard WHO (Rose) angina had a similar long-term outcome to those who had chest pain on exertion which did not full all remaining criteria for angina. This nding is supported by a follow-up study in men and women aged 65 and over which found the 3-year coronary heart disease mortality associated with exertional chest pain to be at least as high as that associated with standard WHO angina[10]. In addition, a Norwegian follow-up study reported the site of pain criteria used in WHO angina to be of no value in predicting mortality[8]. These ndings emphasize that the exertional component of chest pain is the crucial feature in the denition of angina by

Table 5 Eect of breathlessness and calf pain on walking with chest pain on major ischaemic heart disease events and total mortality: men without recall of an ischaemic heart disease diagnosis
Major IHD events n Event rate /1000/year (number of events) RR age adjusted (95% CI) Death rate /1000/year (number of deaths) Total mortality RR age adjusted (93% CI)

Chest pain and breathlessness No chest pain (105, 176) (114, 212) (057, 123) (121, 221) (126, 254) (053, 138)

Other chest pain

(141, 208) (165, 280) (117, 220) (134, 218) (128, 243) (078, 166)

Angina or PMI

Chest pain and calf pain No chest pain

no BR* BR no BR* BR no BR* BR

4249 502 1459 349 474 259

70 106 75 124 129 121

(407) (67) (149) (55) (79) (39)

1 136 1 155 1 084

95 190 77 183 122 229

(561) (124) (158) (84) (79) (76)

1 171 1 215 1 161

Other chest pain

Angina or PMI

no CAP* CAP no CAP* CAP no CAP* CAP

3891 302 1347 221 519 141

71 125 76 146 128 122

(370) (47) (137) (41) (84) (21)

1 163 1 179 1 085

101 194 85 167 149 199

(539) (75) (159) (49) (103) (36)

1 171 1 177 1 114

Chest pain and prediction of ischaemic heart disease in men

*Reference category. Relative risk compared to reference category, adjusted for age. CAP=calf pain on walking, BR=breathlessness. PMI=possible myocardial infarction, angina=possible or denite angina.

Eur Heart J, Vol. 19, January 1998

69

70

F. C. Lampe et al. with chest pain (for example because of new, infrequent or less severe symptoms), but also those who have sought medical help but whose symptoms were considered not to be diagnostic of ischaemic heart disease. Several studies have found that men who have WHO (Rose) angina but are not considered to have angina by a simultaneous clinical assessment are at similar risk of coronary heart disease death to those that are clinically judged to have angina[2527]. Breathlessness and calf pain in addition to chest pain, in men without diagnosed ischaemic heart disease Breathlessness has been shown to be a strong predictor of major ischaemic heart disease events[28], but we have previously reported that this association may be restricted to men without any evidence of ischaemic heart disease[18]. This paper shows that breathlessness provided improved prediction of ischaemic heart disease risk only in men with other chest pain or no chest pain but not in men with questionnaire-assessed symptoms of ischaemic heart disease. Breathlessness may precede angina as an important early indicator of unrecognized ischaemic heart disease[29], but this study suggests that once the disease has developed to a symptomatic stage, the additional presence of breathlessness does not increase risk of a major ischaemic heart disease event. Similarly, although calf pain on walking is a predictor of ischaemic heart disease events in men without any evidence of ischaemic heart disease[17,3032], in this study it did not increase risk when questionnaire-assessed ischaemic heart disease symptoms were present. Once indications of ischaemic heart disease are present, it is likely that these, rather than markers of peripheral vascular disease, are the dominating factor in dening risk of a major event. Although breathlessness or calf pain had no further eect on ischaemic heart disease outcome in the presence of angina or possible myocardial infarction symptoms, they did cause an increase in total mortality: for breathlessness this eect was pronounced. The associations of breathlessness with non-cardiovascular disease mortality[33,34], and of calf pain with cardiovascular deaths and stroke[17,3032] have been shown previously.

questionnaire. Some studies have included subjects with possible angina in the other chest pain group, where their considerably worse prognosis is likely to be diluted by large numbers with non-exertional chest pain[4,11]. Chest pain and risk prediction The patterns of risk associated with angina and possible myocardial infarction in this population sample of middle-aged British men are consistent with those seen in other studies[4,7,8,11]. In the Whitehall study of middleaged male civil servants, both angina alone and possible myocardial infarction alone were associated with about a threefold increase in ischaemic heart disease mortality during a 10-year follow-up period. In both the Whitehall Study and the BRHS, the presence of angina and possible myocardial infarction together identied a small group at exceptionally high risk of an adverse ischaemic heart disease outcome, presumably in part due to an increased likelihood of genuine cardiac pain. The observation that angina grade has a discriminating eect on ischaemic heart disease outcome, although weak and non-signicant in the present study, has been reported in epidemiological[4,7,11] and clinical[22] studies. Chest pain which is not consistent with angina or possible myocardial infarction has previously been shown to be associated with a small increase in ischaemic heart disease risk[4,11]. Clinical studies have also suggested the prognosis of atypical chest pain to be relatively benign[23]. Symptom prediction and pre-existing ischaemic heart disease diagnosis Those who recalled a doctor diagnosis of ischaemic heart disease at entry were at very high risk of a new event, irrespective of their response to the chest pain questionnaire. We have suggested previously that recall of an ischaemic heart disease diagnosis, although subjective, is likely to be a powerful indicator of the presence of severe ischaemic heart disease[5]. There was some indication that questionnaire-assessed angina or possible myocardial infarction were not predictive of ischaemic heart disease events in this subgroup, although the power to assess this in our study was very limited; and a test for interaction between chest pain symptoms and a previous ischaemic heart disease diagnosis was not signicant. It is possible that variations in factors such as eective anti-anginal drug treatment and physical activity limit the value of angina symptoms as a marker of prognosis in those with established ischaemic heart disease, and there is some support for this possibility from clinical studies[24]. In men who did not recall a doctor diagnosis of ischaemic heart disease, both angina and a history of possible myocardial infarction on questionnaire were predictive of major ischaemic heart disease events, although the excess risk associated with these symptoms was considerably lower than that seen in the whole population. Men with questionnaireassessed ischaemic heart disease symptoms but without a diagnosis of ischaemic heart disease are likely to consist not only of those who have not presented to their GP
Eur Heart J, Vol. 19, January 1998

Implications for use of the WHO (Rose) questionnaire in epidemiological studies

The similarity of possible and denite angina in terms of ischaemic heart disease outcome suggest that a shortened chest pain questionnaire could be used to dene angina in epidemiological studies. This would consist of questions 1, 5 and 6 in the chest pain questionnaire for angina given in the Appendix. Retaining both questions relating to angina grade allows for stratication on the basis of exercise tolerance. This simplication of the WHO chest pain questionnaire would be of particular benet if the questionnaire is used in a self-administered form, when it is often not satisfactorily completed. A

Chest pain and prediction of ischaemic heart disease in men missing response for one or more questions precludes a classication of angina according to the standard criteria. Angina prevalence may well have been underestimated by earlier studies reporting only denite angina[4,11,35,36]; the inclusion of possible cases considerably increases prevalence[10,37]. In this study, inclusion of possible angina increased the prevalence of questionnaire-assessed angina by 65% (from 47% to 78%) and the overall prevalence estimate of ischaemic heart disease (including ECG abnormalities) by 7% (from 231% to 247%). The chest pain questionnaire should be used in conjunction with information on pre-existing ischaemic heart disease diagnoses, as this substantially aects the level of risk associated with questionnaire-assessed angina or possible myocardial infarction symptoms. The predictive power of the chest pain questionnaire may be limited to those without diagnosed ischaemic heart disease, although this requires further examination in other studies.

71

of ischaemic heart disease. In the absence of these additional symptoms, men with atypical chest pain are at similar overall risk of a major ischaemic heart disease event to men with no chest pain. Identifying men at high risk In this study, the majority of men with symptoms of angina or possible myocardial infarction on questionnaire did not recall a doctor diagnosis of ischaemic heart disease. Should a self-administered chest pain questionnaire be used to identify men with angina or possible myocardial infarction who have not yet been identied by their doctor? This depends on the screening yield, the absolute risk of ischaemic heart disease in men identied and their potential benet from intervention. Our results suggest that screening middle-aged men without a diagnosis of ischaemic heart disease would yield 12% with either grade II angina or angina with possible myocardial infarction and a further 38% with grade I angina. The rst group would have an absolute ischaemic heart disease event rate of 20% per annum, the second a rate of 13% per annum. Both groups would stand to benet from risk factor modication and aspirin treatment. However, much greater yields of men at similar levels of risk could be obtained by using multiple risk factor scoring systems[12,13], which also help to dene a prevention strategy. This suggests that a combination of risk factors and symptoms may be more helpful than symptoms alone in identifying men without a doctor diagnosis who are at high absolute risk of future ischaemic heart disease events.
The British Regional Heart Study is a British Heart Foundation Research Group and receives support from the Department of Health. Fiona Lampe is supported by the Department of Health.

Clinical implications
The results show clearly that recall of a doctor diagnosis of ischaemic heart disease is a more powerful predictor of ischaemic heart disease risk than any combination of chest pain symptoms, breathlessness or calf pain. Overall, men reporting a doctor diagnosis of ischaemic heart disease have an absolute risk of major ischaemic heart disease events of 39% per annum. For men without a history of diagnosed ischaemic heart disease (in whom 82% of all ischaemic heart disease events during followup occurred), the characteristics of chest pain can play an important part in quantifying the absolute level of ischaemic heart disease risk, a process which is also increasingly important in dening the likely benets from treatment[14]. It is clear that in diagnosing angina, the exertional nature of the chest pain is the crucial factor which should be emphasized above the specic site of pain, the patients response to the pain or the timing of disappearance of the pain on resting. More limiting exertional chest pain and, particularly, the presence of both exertional chest pain and possible myocardial infarction, allow the denition of a high risk group with a major ischaemic heart disease event rate of about 20% per year; breathlessness or calf pain do not help further in the stratication of risk in this group. Although the absolute ischaemic heart disease risk in these men remains considerably lower than that in men with a doctor diagnosis of ischaemic heart disease, they are still likely to derive considerable absolute benet from risk factor modication, from aspirin[38] and, in the subgroup with sustained LDL cholesterol levels of 40 mmol or more, from lipid lowering treatment[39]. The longterm prognosis of atypical chest pain (non-exertional chest pain or severe chest pain in an inappropriate site) is relatively good. Enquiry about symptoms of breathlessness and calf pain on walking can help in identifying those men with atypical pain that are at increased risk

References
[1] Rose GA, Blackburn H, Gillum RF, Prineas RJ. Cardiovascular Survey Methods (2nd edn) WHO Geneva 1982. [2] Rose GA. The diagnosis of ischaemic heart pain and intermittent claudication in eld surveys. Bull Wld Hlth Org 1962; 27: 64558. [3] Rose G, McCartney P, Reid DD. Self-administration of a questionnaire on chest pain and intermittent claudication. Br J Prev Soc Med 1977; 31: 428. [4] Rose G, Hamilton PJS, Keen H, Reid DD, McCartney P, Jarrett RJ. Myocardial ischaemia, risk factors and death from coronary heart disease. Lancet 1977; 1: 1059. [5] Shaper AG, Pocock SJ, Walker M, Phillips AN, Whitehead TP, Macfarlane PW. Risk factors for ischaemic heart disease: the prospective phase of the British Regional Heart Study. J Epidemiol Comm Health 1985; 39: 197209. [6] Shaper AG. Risk factors for ischaemic heart disease. Health Trends 1987; 19: 27. [7] Zeiner-Henriksen T. Cardiovascular disease symptoms in Norway a study of prevalence and a mortality follow-up. J Chron Dis 1971; 24: 55367. [8] Zeiner-Henrikson T. Six year mortality related to cardiorespiratory symptoms and environmental risk factors in a sample of the Norwegian population. J Chron Dis 1976; 29: 1533. [9] Cook DG, Shaper AG, Macfarlane PW. Using the WHO (Rose) angina questionnaire in cardiovascular epidemiology. Int J Epid 1989; 18: 60713. Eur Heart J, Vol. 19, January 1998

72

F. C. Lampe et al.

[10] LaCroix AZ, Guralnik JM, Curb D, Wallace RB, Ostfeld AM, Hennekens CH. Chest pain and coronary heart disease mortality among older men and women in three communities. Circulation 1990; 81: 43746. [11] Bulpit CJ, Shipley MJ, Demirovic J, Ebi-Kryston KL, Markowe HLJ, Rose G. Predicting death from coronary heart disease using a questionnaire. Int J Epidemiol 1990; 19: 899904. [12] Shaper AG, Pocock SJ, Phillips AN, Walker M. Identifying men at high risk of heart attacks: strategy for use in general practice. BMJ 1986; 293: 4749. [13] Shaper AG, Pocock SJ, Phillips AN, Walker M. A scoring system to identify men at high risk of a heart attack. Health Trends 1987; 19: 379. [14] Pyo ra la K, DeBacker G, Graham I, Poole Wilson P, Wood D. Prevention of coronary heart disease in clinical practice Recommendations of the Task Force of the European Society of Cardiology, European Atherosclerosis Society and European Society of Hypertension. Eur Heart J 1994; 15: 130031. [15] Shaper AG, Pocock SJ, Walker M, Cohen NM, Wale CJ, Thomson AG. British Regional Heart Study: cardiovascular risk factors in middle-aged men in 24 towns. BMJ 1981; 283: 17986. [16] Shaper AG, Cook DG, Walker M, Macfarlane PW. Prevalence of ischaemic heart disease in middle-aged British men. Br Heart J 1984; 51: 595605. [17] Shaper AG, Wannamethee G, Walker M. Calf pain on walking, risk factors and cardiovascular outcome in middleaged British men. In: Fowkes FGR, ed. Epidemiology of peripheral vascular disease. London: Springer, 1991: 12735. [18] Cook DG, Shaper AG. Breathlessness, lung function and the risk of heart attack. Eur Heart J 1988; 9: 121522. [19] Shaper AG, Cook DG, Walker M, Macfarlane PW. Recall of diagnosis by men with ischaemic heart disease. Br Heart J 1984; 51: 60611. [20] World Health Organisation. Proposal for the multinational monitoring of trends and determinants in cardiovascular disease (MONICA) project and protocol. Geneva: Cardiovascular Diseases Unit, World Health Organisation; 1983. [21] Collett D. Modelling survival date in medical research. London: Chapman and Hall 1995. [22] McNeer JF, Margolis JR, Lee, KL et al. The role of exercise testing in the evaluation of patients with ischaemic heart disease. Circulation 1978; 57: 6470. [23] Wilcox RG, Roland JM, Hampton JR. Prognosis of patients with chest pain? cause BMJ 1981; 282: 431433. [24] Harris PJ, Harrell FE Jr, Lee KJ, Behar VS, Rosati RA. Survival in medically treated coronary artery disease. Circulation 1979; 60: 125969. [25] Blackwelder WC, Kagan A, Gordon T, Rhoads GG. Comparison of methods for diagnosing angina pectoris: The Honolulu Heart Study. Int J Epid 1981; 10: 21115. [26] Friedman LM, Byington RP and The Beta Blocker Heart Attack Trial Research Group. Assessment of angina pectoris after myocardial infarction: comparison of Rose questionnaire with physician judgment in the Beta-Blocker Heart Attack Trial. Am J Epid 1985; 121: 55562. [27] Sigurrdsson E, Sigfusson N, Agnarsson U, Sigvaldason H, Thorgeirsson G. Long-term prognosis of dierent forms of coronary heart disease: The Reykjavik Study. Int J Epid 1995; 24: 5868. [28] Klatsky AL, Friedman GD, Siegelaub AB. Medical history questions predictive of myocardial infarction: results from the Kaiser-Permanente Epidemiologic Study of Myocardial Infarction. J Chron Dis 1976; 29: 68396. [29] Cook DG, Shaper AG. Breathlessness, angina pectoris and coronary artery disease. Am J Cardiol 1989; 63: 92124. [30] Bainton D, Sweetnam P, Baker I, Elwood P. Peripheral vascular disease: consequence for survival and association with risk factors in the Speedwell prospective heart disease study. Br Heart J 1994; 72: 12832. Eur Heart J, Vol. 19, January 1998

[31] Davey Smith G, Shipley M, Rose G. Intermittent claudication, heart disease risk factors and mortality: The Whitehall Study. Circulation 1990; 82: 192531. [32] Criqui MH, Langer RD, Fronek A et al. Mortality over a period of 10 years in patients with peripheral arterial disease. N Engl J Med 1992; 326: 3816. [33] Ebi-Kryston KL. Respiratory symptoms and pulmonary function as predictors of 10-year mortality from respiratory disease, cardiovascular disease and all causes in the Whitehall study. J Clin Epidemiol 1988; 41: 25160. [34] Kaplan GA, Kotler PL. Self-reports predictive of mortality from ischaemic heart disease: a nine-year follow-up of the Human Population Laboratory Cohort. J Chron Dis 1985; 38: 195201. [35] Bainton D, Baker IA, Sweetnam PM, Yarnell JW, Elwood PC. Prevalence of ischaemic heart disease: the Caerphilly and Speedwell surveys. Br Heart J 1988; 59: 2016. [36] Smith WCS, Kenicer MB, Tunstall-Pedoe H, Clark EC, Crombie IK. Prevalence of coronary heart disease in Scotland: Scottish Heart Health Study. Br Heart J 1990; 64: 2958. [37] Reeder BA, Liu L, Horlick L. Sociodemographic variation in the prevalence of cardiovascular disease in Saskatchewan: results from the Saskatchewan Heart Health Survey. Can J Cardiol 1996; 12: 2717. [38] Collaborative overview of randomised trials of antiplatelet therapy-I: Prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. Antiplatelet Trialists Collaboration. BMJ 1994; 308: 81106. [39] Shepherd J, Cobbe SM, Ford I et al. Prevention of coronary heart disease with pravastatin in men with hypercholesterolaemia. West of Scotland Coronary Prevention Study Group. N Engl J Med 1995; 333: 13017.

Appendix:

Symptom questionnaires

SEVERE CHEST PAIN 1 Have you ever had a severe pain in your chest lasting for half an hour or more? (Y/N) If NO, go to next section 2 Where did you get this severe pain? (Show chart) 3 Did you see a doctor because of this pain? CHEST PAIN 1 Do you ever have any pain or discomfort in your chest? If NO, go next section 2 When did you last get the pain Within 1 month 15 months ago 612 months ago Over 1 year ago Occasionally 3 How often do you get it Daily Weekly Monthly Once only Occasionally 4 Where do you get this pain or discomfort? (Show chart) (Y/N)

(Y/N)

1 2 3 4 5 1 2 3 4 5

Chest pain and prediction of ischaemic heart disease in men

73

5 When you walk at an ordinary pace on the level, does this produce the pain? (Y/N) 6 When you walk uphill or hurry, does this (Y/N) produce the pain? 7 When you get any pain or discomfort in your chest on walking, what do you do? Stop 1 Slow down 2 Continue at same 3 pace 8 Does the pain or discomfort in your chest go away if you stand still? (Y/N) 9 How long does it take to go away? 10 minutes or less 1 More than 10 2 minutes BREATHLESSNESS 1 Do you get short of breath walking with people your own age on level ground? (Y/N) 2 On walking up hills or stairs, do you get more breathless than people your own age? (Y/N)

3 Do you ever have to stop walking because of breathlessness? (Y/N)

Right

Left

CALF PAIN 1 Do you ever get pain in your calf muscles on walking at an ordinary pace, on the level? (Y/N) 2 Do you get pain in your calf muscles when you walk uphill or hurry? (Y/N)

Eur Heart J, Vol. 19, January 1998