J Clin Epidemic-d Printed in Great

Vol. 43, No. 7, pp. 637-643, Britain. All rights reserved

0895-4356/90$3.00+ 0.00 Copyright 0 1990Pergamon Press plc

1990

CARDIOVASCULAR DISEASE IN EDINBURGH AND NORTH GLASGOW-A TALE OF TWO CITIES W. C. S. SMITH,*M. C. SHEWRY,HUGH TUNSTALL-PEDOE, I. K. CROMBIE and R. TAVENDALE Cardiovascular

Epidemiology Unit, Ninewells Hospital & Medical School, Dundee DDl 9SY. Scotland (Received in revised form 4 January 1990)

Abstract-Scotland has one of the highest mortality rates from coronary heart disease for both men and women. However within Scotland there are major geographical differences in the mortality rates, exemplified by the differences between the two largest Scottish cities. Glasgow on the west has a much higher mortality rate for cardiovascular diseases than Edinburgh in the east. During 1986 coronary risk factor population surveys were conducted simultaneously in Edinburgh and North Glasgow as part of the WHO MONICA study. These surveys employed standardized methods and the central quality control of the WHO project. Measures of coronary heart disease morbidity were higher in North Glasgow, except for electrocardiographic evidence of ischaemia, consistent with the mortality rates. The major coronary risk factors were uniformly higher in North Glasgow than in Edinburgh, except for serum lipids which were not significantly different. The risk factors, except the lipids, showed a gradient by socioeconomic status, so that when the risk factors levels were standardized for housing tenure the significant differences between the cities largely disappeared, and the serum cholesterol levels in women become significantly higher in Edinburgh because of their slight negative relationship with social status. This study shows that the socioeconomic differences between Edinburgh and North Glasgow largely explain the coronary risk factor differences between the cities. The socioeconomic differences in coronary disease and its major risk factors require further investigation and may be more fundamentally important than the geographical differences in the patterns of coronary heart disease.

INTRODUCTION

have major socioeconomic differences [3] as well as historical differences. Population surveys of CHD prevalence and risk factors were conducted simultaneously in 1986 in both Edinburgh and North Glasgow as part of the World Health Organization’s MONICA project [4] using the same methods. The coronary event registration aspect of the project was not continued in Edinburgh but the population survey was completed according to the study protocol. The data from these surveys in Edinburgh and North Glasgow have been used to address three important questions about CHD in these two cities. Firstly, does the CHD prevalence in the two cities correspond to the

The coronary heart disease (CHD) mortality rates in both men and women in Scotland are amongst the highest in the world [l]. There is considerable variation in CHD mortality rates

in Scotland although none of the rates can be regarded as low by international standards. This geographical variation is seen in the male standardized mortality ratios in the two Scottish cities of Edinburgh (81) and North Glasgow (117) based on the Scottish national age specific mortality rates [2]. However these two cities *Author for correspondence. 637

638

w. c. s.

differences in CHD mortality? Secondly, are the differences in CHD risk factors consistent with the difference in CHD mortality, and finally to what extent do the socioeconomic differences between the cities explain the risk factor and CHD mortality differences? METHODS

The two project areas are defined as the local government districts of Edinburgh City and the part of Glasgow City lying north of the River Clyde. Assessment of the coronary heart disease mortality experience of the two project areas was based on standardized mortality ratios calculated for men and women aged 35-64 years for the years 1979-83 using the age specific Scottish rates for ICD 410-414 as the standard. Census data on the socioeconomic characteristics of the two areas are derived from the 1981 national census [3]. A two stage sampling process was used to select a representative sample from each city. In both project areas a random sample of 30 general practitioners was drawn from the list of practitioners in the areas. A random sample of patients listed with each of these practitioners was selected proportional to the size of each IO-year age-sex group of each list to achieve an examination target of 200 men and women in each IO-year age group between the ages of 25 and 64 years. The list of selected individuals was checked by the practitioner before subjects were approached. Subjects were then sent a letter inviting them to participate in the study by completing a questionnaire and attending a local clinic for examination. Results of the examination were sent, with the individual’s consent, to the subject’s own practitioner. The questionnaire was made up of questions on sociodemography, medical history, physical activity, smoking, diet, alcohol intake, Bortner questionnaire and health knowledge [5]. Each participant underwent a standardized examination by a team of specially trained nurses which included measurement of height, weight, blood pressure, expiratory carbon monoxide, the recording of a 12-lead electrocardiogram and collection of a sample of venous blood. Blood samples were analysed in Edinburgh from the Edinburgh area and in Dundee from the Glasgow City. The age standardization used is based on world population standards [6]. Social class groupings have been derived from occupation

hITI

el a/.

[7] and housing tenure was classified into owner occupiers, council tenants including housing associations, and private rented tenants. Standardization by housing tenure was preferred to social class because of anomalies in social class in women; this was carried out by combining the housing tenure distributions of the two cities.

RESULTS

(i) Study population

The number of participants and the response rates in the study in both cities are shown in Table 1. A larger number of invitations ‘in Glasgow than in Edinburgh were returned by the Post Office because the subjects were no longer resident at a given address. The response rates were then calculated excluding those not resident in the project areas at the time of the survey. The response rates are shown to vary by age and sex in a similar pattern in both cities with the lowest response rates in the youngest age group and response higher in Edinburgh than Glasgow. The employment characteristics of participants in the two cities differed with 5.6% of men and 2.2% of women in Edinbugh being unemployed compared to 20.6% and 3.5% of men and women in North Glasgow. These unemployment rates were slightly lower than the corresponding rates for the two populations found at the 1981 census. 42.5% of men in Edinburgh were in manual social classes compared to 72.4% of men in North Glasgow. The participants in the two cities also varied by housing tenure where 74% in Edinburgh were owner occupiers compared to 30% in North Glasgow. Table 1. Responses to the invitation in Edinburgh North Glasgow Edinburgh Age group

Participated

and

North Glasgow

Response rate (%) Participated

Response rate (%)

25-34 3544 45-54 5564 Total

158 188 197 213 756

Men 57.0 68.2 70.0 71.3 67.0

145 154 175 176 650

61.5 63.6 66.9 66.5 64.9

25-34 34-44 45-54 55-64 Total

162 195 219, 211 787

Women 59.3 70.5 72.6 66.5 67.2

128 151 162 171 612

57.3 63.6 64.5 61.4 61.9

Cardiovascular Disease in Scotland

(ii) Cardiovascular disease rates

639

project areas is also reflected in reported passive smoking: 34% of men and 35% of women in North Glasgow reporting a lot of exposure to cigarette smoke from others compared to 21 and 19% respectively in Edinburgh. The smoking differences between the two city areas is confirmed by significantly higher proportions of subjects in North Glasgow having expiratory CO levels above 7 ppm and serum thiocyanate levels above 80 mmol/l compared with Edinburgh.

The all causes mortality rates for men and women in North Glasgow were considerably higher than those in Edinburgh, as were the rates for all cardiovascular diseases, coronary heart disease and cerebrovascular disease (Table 2). The age-standardized prevalence rates for various measures of coronary heart disease morbidity obtained from this survey are also presented. The rates for all measures are higher in North Glasgow than Edinburgh except the rates for ECG ischaemia using the Whitehall criteria [8]. The rates of angina in women in both cities are similar or slightly higher than those in men whereas the rates of myocardial infarction are higher in men.

(iv) Other risk factors

(iii) Smoking

The cigarette smoking patterns by age and sex in the two city project areas are shown in Fig. 1. This reveals very different patterns of cigarette smoking behaviour between men and women and between both cities. There is a large difference between the age standarization percentage of regular cigarette smokers in both cities in men (49 and 35%) and in women (51 and 31%) in North Glasgow and Edinburgh, respectively. However, within age groups, the pattern varies between men and women with a higher proportion of ever smoked (current and ex-smokers) in the older men than the older women and a higher porportion in the younger women compared to the younger men. The highest proportions of smokers were in the 35-44 year age group. In Edinburgh 3.9% of men smoke cigars regularly compared to 2.5% in North Glasgow, and 3.2% compared with 2.4% of men in the two cities regularly smoke a pipe. No women in the study smoked cigars or a pipe. The differences in cigarette smoking between the two

The age-standardized levels of other coronary heart disease risk factors in men and women in the two areas are shown in Table 3. The significance level of p < 0.01 has been used because of the multiple comparisons undertaken. There are significant differences between the cities in the levels of leisure activity, vigorous activity at work, percentage of alcohol consumers, type of alcohol consumed, Bortner type A personality score, height and blood pressure in both men and women. Weight and body mass index (BMI) are significantly higher in Glasgow women and the median alcohol consumption in units per week is significantly higher in Glasgow men compared with their Edinburgh counterparts. There were no significant differences in the levels of total cholesterol and HDL-cholesterol between the cities in either men or women. (v) Risk factors and social status The differences in risk factors by socioeconomic factors was examined to investigate whether the influence of these factors might explain the risk factor differences between the two cities. Gradients in risk factor levels by housing tenure and social class were observed for most risk factors with higher levels in the

Table 2. Cardiovascular mortality and morbidity rates in Edinburgh and North Glasgow Men Age standardized mortality rates 35-64 years for 1984 per 100,000 All causes Cardiovascular disease Coronary heart disease Cerebrovascular disease

Edinburgh

Women North Glasgow

Edinburgh

North Glasgow

851 362 287 43

1298 565 434 67

439 122 72 33

714 237 134 61

2.1 3.7 8.5

4.2 8.5 8.5

2.2 5.1 9.5

3.0 10.3 7.9

2.0

2.7

0.7

1.2

5.2 0.7

9.8 2.3

3.2 0.3

5.7 0.9

Prevalence rates (%)

History of angina Rose angina ECG ischaemia History of myocardial infarction

Rose possible infarction ECG Q/QS pattern



640

W. C. S.

SMITHet al.

Edinburgh

Never smoked

Ex-smoker

< 20 / day > 20 / day 25-34

35-44

45-54

55-64

25-34

35-44

45-54

55-64

Women

Men

North Glasgow 100 -

Never smoked

80 -

60 Ex-smoker 40 < 20 / day 20 > 20 / day o25-34

35-44

45-54

55-64

Men

25_3435-44

45-54

55-64

Women

Fig. 1. Cigarette smoking behaviour in men and women in Edinburgh and North Glasgow.

lower socioeconomic groups. There was little effect of social status on BMI, and general levels of physical activity showed a slight inverse relationship as did total cholesterol levels. The effect of housing status on the percentage of current regular smokers in men and women in both cities is presented in Table 4. It can be seen that much of the difference in smoking status between the cities can be explained by such a socioeconomic factor. However, there is a residual difference which is greater for women than for men, and this difference is largely due to smoking behaviour differences in owner occupiers and those living in private rented accom-

modation. Similarly, with other risk factors where residual differences between the cities were observed, these small differences tended to be in the higher social classes and owner occupier groups. More detailed analysis of the composition of the higher social classes in the two cities showed that they were made up of different occupational groups with higher proportions of lawyers and civil servants in Edinburgh but lower proportions of engineers and architects than in North Glasgow. Risk factor rates standardized by housing have been calculated for all risk factors by applying the city specific rates by housing tenure to the combined popu-

641

Cardiovascular Disease in Scotland Table 3. Age standardized risk factor levels in men and women (25-64yr) in Edinburgh and North Glasgow Men Risk factors

Edinburgh

Women North Glasgow

North Glasgow

Edinburgh

1. Biochemistry 6.16 1.38

Total cholesterol HDL cholesterol

6.05 1.39

6.22 1.64

6.04 1.59

2. Anthropometry Weight Height BMI

16.3 174.5** 25.1

75.1 112.1 25.4

62.8** 161.3** 24.1**

64.4 159.4 25.4

3. Blood Pressure Systolic Diastolic

130.0** 81.0**

134.0 84.4

123.3** 17.0+*

128.4 80.1

88.1

84.7

17.6**

16.1 11.3’2 11.2 61.7

17.7 5.5 14.8 60.2

3;.;** 24:9** 7.1

66.4 3.3 18.3 35.0 1.4

38.1 11.8** 26.4** 18.0

42.1 21.3 19.9 18.2

45.3 8.9** 23.4** 12.1

41.1 19.1 14.8 13.0

175.3**

165.0

172.8**

167.2

4. Alcohol % Consumers Median units/week % mostly wine % mostly spirits % mostly beer 5. Physical Activity % active at work % vigorous at work % active at leisure % vigorous at leisure 6. Personality Mean Bortner score

Significant difference between the cities, **p -C0.01.

lation samples from Edinburgh and North Glasgow in each of the housing categories. The resultant risk factor levels standardized for housing tenure are presented in Table 5. The inverse relationship between social factors and cholesterol increased the differences between the cities when standardized for housing tenure. However the significant differences between the cities for most of the other risk factors disappear except those for wine drinking in women and diastolic blood pressure in men.

of the WHO MONICA project [4]. These data allow comparisons to be made not only between Edinburgh and North Glasgow which have different rates of coronary disease, but also with the other centres around the world participating in the WHO project. The study methods have been based on an internationally standardized protocol and have met the WHO study standard for quality control. The response rates in North Glasgow were slightly lower than those in Edinburgh but these small differences are unlikely to be a major source of bias between the two population samples. The sampling frame used in both cities was general practitioner patient lists and these were less accurate in North Glasgow where 23% of invitations were returned by the Post Office

DISCUSSION

This study was based on population surveys which were carried out simultaneously in Edinburgh and North Glasgow in 1986 as part

Table 4. Cigarette smoking by housing tenure in Edinburgh and North Glasgow Men Edinburgh n % Owner occupier 563 27.1 Council tenant 135 58.9 Private rent 56 33.3 Total 755 33.8 Percent smokers standardized for housine tenure 39.8

Women North Glasgow n %

198 404 42 641

32.4 56.8 49.1 48.5 36.8

Edinburgh n % 580 161 44 787

23.2 55.2 36.6 30.8 43.0

North Glasgow n % 183 401 25 612

31.3 60.7 48.7 51.2 44.0

w. c. s. SMITH et al.

642

Table 5. Risk factor levels in Edinburgh and North Glasgow standardized for age and housing Men Risk factors

Edinburgh

Women North Glasgow

Edinburgh

North Glasgow

1. Biochemistry

Total cholesterol (mmol/l) HDL cholesterol (mmol/l) 2. Anthropometry Weight (kg) Height (cm) BMf (kg/m) 3. Blood Pressure Systolic (mmHg) Diastolic (mmHg) 4. Alcohol % Consumers Median units/week % mostly wine % mostly spirits % mostly beer 5. Physical Activity % active at work % vigorous at work % active at leisure % vigorous at leisure 6. Personality Mean Bortner score

6.24 1.37

6.05 1.40

6.44** 1.63

6.15 1.65

76.3 173.3 25.4

76.1 172.8 25.5

63.6 160.5 24.7

64.4 159.6 25.3

132.8 81.8**

134.7 84.6

127.6 78.6

129.5 80.5

86.8 16.3

84.2 16.7

72.1 4.0

68.2 3.6

9.9

7.5

32.7**

22.0

12.7 59.6

17.4 55.5

26.9 6.9

32.5 6.5

41.4 15.3 24.0 15.3

41.3 18.7 19.5 15.1

45.5 11.0 21.7 10.3

40.8 15.3 15.8 10.7

169.0

167.0

168.2

169.5

Significant difference **p < 0.01.

as incorrect addresses compared to only 13% in Edinburgh. Response rates were calculated excluding those not resident at the stated addresses, so that the study sample in North Glasgow was smaller than that for Edinburgh, although the response rates were not too dissimilar. Measures of coronary heart disease morbidity were uniformly higher in North Glasgow, except for electrocardiographic evidence of the coronary ischaemia, which confirmed patterns seen with mortality rates. The coronary differences between the two cities are real and cannot be explained by differences in certification of cause of death. The differences in coronary risk factors are consistent with the differences in coronary heart disease except for total cholesterol and HDLcholesterol in both men and women. The differences in cigarette smoking are particularly marked with about one third of the Edinburgh sample smoking compared with over half of the participants in North Glasgow. The percentage of female cigarette smokers in North Glasgow was very high particularly at the younger age groups of 25-34 years and 35-4 years where the levels exceed those for men in the same age group almost reaching 60%. A marked social gradient in cigarette smoking was seen in both

cities, and similar gradients were seen with all risk factors except blood lipids. These observations raised the possibility that the socioeconomic differences between the cities may explain their risk factor differences and also the differences in coronary heart disease. Most of the significant risk factor differences between Edinburgh and North Glasgow disappeared when the risk factor levels were standardized for housing tenure although differences remained which did not reach statistical significance. The differences which remained were in diastolic blood pressure in men and wine drinking in women. The cholesterol levels in Edinburgh women were significantly higher than those for North Glasgow after standardizing for housing tenure because of the slight inverse relationship between socioeconomic status and cholesterol levels. These risk factor data are valuable as a baseline measure for evaluation of the Good Hearted Glasgow cardiovascular disease intervention project. The similarity of risk factor levels between Edinburgh and North Glasgow for specific socioeconomic groups may make it possible to use Edinburgh as a control population for this intervention. These analyses of coronary risk factor data collected simultaneously in Edinburgh and North Glasgow using the same method show

Cardiovascular

Disease in Scotland

that the major differences in risk factor levels between the two cities can largely be explained on the basis of socioeconomic factors. Large differences in risk factor levels in different social groupings exist within each city which have importance for targeting of intervention strategies. The importance of socioeconomic factors in explaining the geographical distribution in coronary heart disease in the U.K. has been noted in other studies [9, lo]. Other explanations for the differences such as ethnic group are unlikely to be important since they represent only a very small proportion of either city. The hardness of the water in both cities is similar and comes largely from a single source [I I]. Our findings confirm this in terms of risk factor patterns and suggest that socioeconomic differences are more important than the geographical differences and deserve further investigation particularly as to why different social groupings have different risk factor levels and whether these can be explained by differences in knowledge and attitudes to health related behaviours. Acknowledgements-This study was funded by the Chief Scientist Office of the Scottish Home and Health Department; however, the conclusions of this communication are those of the authors, not of the Scottish Home and Health Department. Laboratory analyses in Edinburgh were done by Dr Rudolph Riemersma and fibrinogen was assayed in Glasgow by Dr Gosdon Lowe.

643 REFERENCES

1. Tunstall-Pedoe HD, Crombie IK, Smith WCS. Levels and trends of coronary heart disease mortality in Scotland compared with other countries. Health Bull 1986; 44: 153-161. 2. Crombie IK, Smith WCS, Kenicer MB, TunstallPedoe HD. Geographical variation in coronary heart disease mortality in Scotland. Health Bull 1986; 44: 193-202. 3. Registrar General Scotland. Census 1981 Scotland. Edinburgh: HMSO; 1982. 4. WHO MONICA Project Principal Investigators (Prepared by H Tunstall-Pedoe). The WHO MONICA Project (monitoring trends and determinants in cardiovascular disease)--a major international collaboration. J Clin Epidemiol 1988; 41: 105-114. 5. Smith WCS, Crombie IK, Tavendale R, Irving JM, Kenicer MB, Tunstall-Pedoe H. The Scottish Heart Health Study: objectives and development of methods. Health Bull 1987; 45: 211-217. 6. World Health Organisation. In: Waterhouse J, Muir C, Shanmugaratnam K, Powell J, Eds. Cancer Incidence in-Five Continents. IARC Scientific Publications No. 42. Vol. IV. Lvon: IARC: 1982. I. Office of Population Censuses and Surveys. Classification of Occupations 1980. London: HMSO; 1980. 8. Rose G, Baxter PJ, Reid DD, McCartney P. Prevalence and prognosis of electrocardiographic findings in middle-aged men. Br Heart J 1978; 40: 636-643. 9. Crombie IK, Kenicer MB, Smith WCS, TunstallPedoe HD. Unemployment, socio-economic factors and the distribution of coronary heart disease in Scotland. Br Heart J 1989; 61: 172-177. 10. Pocock SJ, Cook DG, Shaper AG el al. British Regional Heart Study: geographical variations in cardiovascular mortality, aid iherole of water quality. Br Med J 1980: 280: 1243-1249. 11. Smith WCS, drombie IK. Coronary heart disease and water hardness in Scotland-is there a relationship? J Epidemiol Commun Health 1987: 41: 227-228.

Cardiovascular disease in Edinburgh and north Glasgow--a tale of two cities.

Scotland has one of the highest mortality rates from coronary heart disease for both men and women. However within Scotland there are major geographic...
636KB Sizes 0 Downloads 0 Views