Original Article | Published:

Is the blood pressure of people from African origin adults in the UK higher or lower than that in European origin white people? A review of cross-sectional data

Subjects

Abstract

The aim of the study was to review published evidence on whether blood pressure (BP) levels and the prevalence of hypertension are higher in adult populations of African descent living in the UK as compared to the white population. A systematic literature review was carried out using MEDLINE 1966–2002 and EMBASE 1980–2002 and citations from references. In all, 14 studies were identified. Nearly all studies were carried out in the London area. The data showed important differences between studies in terms of age and sex of samples, definition of African/black and methods of evaluating BP. A total of 10 studies reported higher mean systolic BPs, while 11 studies reported higher mean diastolic BPs in men from African descent compared to white men. In women, 10 of 12 studies reported higher systolic, and 10 of 12 studies reported higher diastolic BPs. For prevalence of hypertension, eight of 10 studies reported higher rates in men from African descent; eight of nine studies showed higher rates of hypertension in women from African descent. Overall, the most representative sample and up-to-date data came from the Health Survey for England '99. Ethnic group differences in BP were not present in the younger age groups. Women of African descent had higher BP and higher body mass index (BMI). In men of African descent high BP did not coincide with higher BMI. In conclusion, the reported higher rates of hypertension in people from African descent in the UK are confirmatory of the USA African-American and white comparisons. Variations in study methods, size and body composition, and in the mix of Afro-Caribbean and West African groups explain much of the inconsistent results in the UK studies.

Introduction

Hypertensive diseases and stroke are among the dominant causes of death in people from African descent in the UK with rates even higher than in the European origin white population (henceforth white). As a major cardiovascular risk factor, hypertension needs to be carefully managed in UK's African origin populations. There is widespread acceptance that in people from African descent blood pressure (BP) levels are comparatively high. Studies in the UK on differences between people from African descent and white populations in BP and prevalence of hypertension have, however, not always shown consistent results with, for example, Cruickshank et al1 finding no difference in Birmingham. Such inconsistency prompted us to undertake this research study. We apply the concept of ethnicity as discussed by Senior and Bhopal,2 and given in the glossary.

Methods

Search strategy

EMBASE and MEDLINE were used to identify papers published from 1980–2002 and 1966–2002, respectively. Medical Subject Heading ‘African Caribbean’, or ‘West Africans’, or ‘blacks’, or ‘ethnic minority population’ were combined with BP and hypertension. The search was limited to the United Kingdom and papers published in the English language only. We included population-based studies that reported BP and/or prevalence data on people from African descent in comparison with a white or general population. All studies based on clinically selected patients and children were excluded. The reference list of all known primary studies and review articles were scrutinised and additional relevant citations were identified. In all, 14 studies were found that were relevant.3,4,5,6,7,8,9,10,11,12,13,14,15,16 Two authors supplied unpublished data.11,15

Study selection and data extraction

Of 14 papers included in the final analysis, 11 were cross-sectional surveys, most in the community, but some in occupational settings. Lane et al's16 study includes some of the research participants in the study of Cruickshank et al.5 Meade et al's3 was a cohort study and Whitty et al12 reported cross-sectional data from a cohort design. Confidence intervals of 95%, P-values and odd ratios are reported as given in the paper. Mean systolic and diastolic BPs of the total study sample in two studies were calculated as the results were based on age-specific groups.5,15

Results

Methods of the reviewed studies

Table 1 shows that most reports relate to the 1970s to 1990s and all studies were carried out in England. With five exceptions,5,13,14,15,16 two of which were in Birmingham, one in Manchester, and two national surveys,13,14 all studies were carried out in and around London. Six studies6,7,8,10,11,15 were based on general practices' lists, three studies3,4,5 were factory based, two studies were from the population-based health surveys for England,13,14 and one study was based on civil servants.12 Two studies had samples from different sources.9,16 Three studies5,10,16 were designed specifically to compare BP levels; the rest had broad aims relating to a range of cardiovascular risk factors. Name analysis,9,11 self-reported origin,9,13,14 country of birth,3,6,7,9 parental origin,8,10,11,15 and observer classification5,12,16 were used as indicators of ethnicity. One study did not state which indicator of ethnicity was used.4 The age range varied widely. In one study, the Afro-Caribbean group was significantly older than their white counterparts (P<0.0001).16 Most studies assessed BP levels and prevalence of hypertension in men and women separately, but two only studied men.7,9 Response rates varied widely, ranging from 58%10 to 81%.7 Sample size varied widely. The largest, most representative and most up-to-date study was the Health Survey for England '99.14

Table 1 Contextual details—publication, location, timing, design, sampling frame, aims, sample identification and size and response rate

BP levels

Table 2 shows that BPs were measured in three studies in a workplace,3,4,5 in three studies at the participants' own homes,7,13,14 in four studies in a health centre,6,8,10,15 in one study in a hospital clinic11 and in one study in a hospital clinic and in a workplace.9 In six studies, BP was measured with a random zero sphygmomanometer.5,6,7,8,9,16 Sever et al4 used a Bosomat automatic BP recorder, Cappuccio et al11 used an automated ultrasound sphygmomanometer, whereas Primastesta et al13 and Karlsen et al14 used a Dinamap 8100 monitor. Most studies measured BPs in a sitting position. Two studies measured participants' BP in a supine position.7,11 The readings reported varied. For instance, some studies measured BP twice and used the mean readings for analysis.6,8,10,12,16 In another three studies, BPs were measured thrice and the averages of the last two readings were taken for analysis,11,13,14,15 whereas in some studies either the first5 or the lowest of the three4 readings were taken for analysis. Two studies did not indicate which measurements were taken for analysis.3,9 Two studies reported actual mean systolic and diastolic BPs,3,4 two age-adjusted median systolic and diastolic BPs9,10 and the rest of the studies published age-adjusted mean systolic and diastolic BPs.5,8,11,12,13,14,15,16

Table 2 Setting, measurement technique and mean systolic and diastolic blood pressures by sex and ethnic group

In 10 of 14 studies, mean systolic BP was higher in men from African descent than in white men.3,4,7,8,9,10,11,12,16 In Primatesta et al's national study, mean systolic BP was lower in the black group aged 16–39 years; however, in those aged 40 years and above, mean systolic BP was higher than in the white group.13 In Karlsen et al's national study,14 the observed mean systolic BP was lower in Afro-Caribbean; however, once the effect of age had been controlled for and results were presented as standardised ratios of means, there was no difference between the two groups. In Cruickshank et al's Birmingham Factory study and Haines et al's study, the mean systolic BP was lower in black than in white populations.5,6 For diastolic BP, all the studies reported higher mean values in black than in white groups, except Primatesta et al13 in the 16–39 year age group and Karlsen et al.14

A total of 12 studies included women.3,4,5,6,8,10,11,12,13,14,15,16 Of these, 10 studies3,4,5,8,10,11,12,14,15,16 reported a higher and one study6 reported a lower systolic BP in black than in white women. In Primatesta et al,13 the younger black group had a lower and the older black group a higher mean systolic BP than the white group. For diastolic BP, with two exceptions,6,14 one in which the levels were not given,14 but stated as showing no difference, all the studies reported higher levels in black than in white groups.

The only study that compared systolic BP levels between Afro-Caribbeans and West Africans reported similar levels but the number of West Africans (men n=26, women n=25) was relatively small compared to Afro-Caribbeans (men n=211, women n=303).10

Prevalence of hypertension

Table 3 shows 12 studies4,5,6,7,8,10,11,12,13,14,15,16 that reported on the prevalence of hypertension using various cutoff points. Most studies reported on a cutoff point of systolic BP of 160 mmHg or more and/or diastolic BP of 95 mmHg or more, and defining as hypertensive those receiving hypertension treatment. Three studies did not indicate the number of people who were receiving treatment for hypertension.4,7,16 Two studies combined men and women.4,12 One study did not standardise the prevalence rates for age.4 Karlsen et al14 reported on a cutoff point of 140/90 mmHg.

Table 3 Criteria for diagnosis, age adjustment, receipt of treatment and prevalence of hypertension by sex and ethnic group

With one exception in which the treatment rate was slightly lower in Afro-Caribbean women than in white women,8 all the studies reported higher treatment rates in men and women of African descent compared to white men and women.

A total of 10 studies reported on men.5,6,7,8,10,11,13,14,15,16 Of these, eight studies reported higher prevalence rates in people from African descent than in white men.5,6,8,10,11,14,15,16 In Primatesta et al,13 the prevalence rate was lower in the 16–39 years age group, but higher in the 40 years and above age group in black compared to white men. Miller et al7 reported lower prevalence rates in black men than in white men, but the number of black men were relatively small (N=24) compared to white men (N=68). Of the nine studies that included women,5,6,8,10,11,13,14,15,16 seven studies5,8,10,11,13,15,16 reported higher and one study6 reported lower prevalence rates in people from African origin compared to white women. In Karlsen et al,14 Black Caribbean women had lower observed prevalence rate than white women; however, once the effect of age had been standardised and presented in risk ratios, the Black Caribbean women had higher rate than white women. Two studies' results were based on both men and women combined and reported higher prevalence rates in black than in white populations.4,12

Anthropometry data and mean difference in BP

Table 4 shows, in studies ranked by African origin sample size, body mass index (BMI), waist-to-hip ratio (WHR), and mean difference in systolic and diastolic BP. In all, 11 studies reported on overall mean BMI on men.3,6,7,8,9,10,11,13,14,15,16 Of these, six studies3,9,11,13,14,16 reported higher BMI and five studies6,7,8,10,15 reported lower BMI in African descent compared to white men. Overall, differences in BMI were small. Five studies reported on WHR on men.8,9,11,14,15 Of these, three studies11,14,15 reported lower, one study9 reported the same and one study8 reported higher ratios in African descent men compared to white men. Again, the differences were small. Nine studies reported on BMI3,6,8,10,11,13,14,15,16 and four studies reported on WHR8,11,13,15 on women and all reported higher BMI and WHR in Africans than in white population. The mean differences in systolic and diastolic BPs varied widely. The mean differences were particularly marked in women from African descent reaching 10.6–17 mmHg higher in some studies.4,10,11,16 The patterns of BP were partly associated with BMI, the fit being better in women of African descent. For example, in four studies,7,8,10,15 the mean BMI was lower in black men but mean BP levels were higher than those in white men.

Table 4 Mean BMI, WHR and mean difference in blood pressures by sex and ethnic group ordered by ranking on size of the African descent sample

Discussion

Key findings

Nearly all the studies were carried out in the London area where BP levels and prevalence of hypertension were higher in men and women of African descent as compared to the white population. The BP pattern was different in the younger age group than in older age groups in African compared to the white groups.13 Treatment rates were higher in people from African descent compared to white people. Women of African descent consistently had comparatively high BMI and WHR but this did not apply to men.

In three studies of clinically selected adults17,18,19 (hence excluded from the tables), two reported higher hypertension prevalence rates17,18 and one reported similar BP levels19 in people from African descent compared to white adults.

Limitation of the review and included studies

In some studies, Afro-Caribbean and West Africans were combined as one ethnic group.3,10,11,13 This method of combination is subjective, imprecise, and unreliable. Afro-Caribbean people have cultural values that are different from people of West Africa. In fact, the term West Africans is inappropriate because there are major ethnic differences between countries and also within countries in West Africa in terms of culture such as language, diet, religion, geography, and socioeconomic variations. All these factors are important determinants of health and have been shown to influence BP.20 UK mortality statistics show that black Africans had higher mortality rates from chronic rheumatic heart disease and hypertensive disease than Afro-Caribbeans.21 The studies span about 20 years, which potentially affects the comparability and interpretability of the results. Middle-age people of African descent at the end of the 20th century were likely to be different from their counterparts in the early 1980s, in terms of immigration, habits, and socioeconomic status. There are also many black people in the UK who have some white ancestry (mixed race), which further complicates interpretation. With one exception,5 none of the authors in these reviewed papers stated whether mixed race people were included or excluded.

Huge variations in methods of measurement of BP limit the capacity to synthesise the data. For example, in seven studies participants' BPs were measured with a Hawksley random zero sphygmomanometer.5,6,7,8,9,10,16 In other studies,13,14 a Dinamap 8100 monitor was used that tends to provide higher systolic and lower diastolic BP levels than mercury sphygmomanometer readings.22,23 Participants' BPs were measured in different locations. The readings analysed also varied. The estimated white coat hypertension ranges from 12 to 53% depending on the population studied and the definition used.24,25 It is difficult to account for the effect of white coat hypertension in each study. It is also not clear whether the effect of white coat hypertension differs in different ethnic groups. Some studies did not take the effect of age on BP into account in their analysis,3,4 which might have given misleading results. Karlsen et al's14 study highlights the importance of age adjustment when comparing BP levels between different ethnic groups. In this study, Black Caribbean men had lower observed BPs. However, once age was controlled for, the levels between Black Caribbean and the general population were no longer different.14 Many people from African descent in the UK fall into the lower end of the socio-economic structure and this is a major potential confounding factor in most studies on ethnic differences in health. The few studies that reported on socioeconomic characteristics showed that people from African descent were clustered to the lower end of the social structure11 or grade structure even in the same industry.3,5,12,13,14 Studies did not formally adjust for socioeconomic confounders. Differences in BMIs and sex composition of the samples further complicate interpretation of these studies. These differences in study methods means that between study differences are not reliable and focus attention on within study variations.

Some authors used name analysis to identify subjects with origins in West Africa.9,11 West Africans and Afro-Caribbean people who share christian names with the white population are likely to be missed by this technique. In Cappuccio et al,11 for example, using name analysis identified 80% of white people, 94% of West Africans but only 51% of the Afro-Caribbeans. The knowledge of practice staff was used to improve identification of this group. This could have resulted in selection bias, if people who were well known to the practices were more likely to be selected. Some studies3,5,7,9 used country of birth as an indicator of ethnicity. This technique may exclude Afro-Caribbeans and West Africans born in the UK.

Huge variations in study methods might make it problematic for researchers to conduct a meta-analysis, which would require collecting the original data. Whether such a task is feasible and worthwhile needs to be assessed in further research.

Discussion of the key results

The findings of higher BP levels and a higher prevalence of hypertension in the people of African descent in the UK are confirmatory of the USA's African-American and white comparisons.

Higher BP levels reported in most studies did not reflect the patterns shown by the Health Survey for England '99.14 In Health Survey for England '99, standardised mean systolic and diastolic BP levels were similar in men in both ethnic groups.14 The study also reported no significant difference in standardised mean diastolic BP in women. Little difference in BP between African Caribbeans and Europeans in this study probably reflects better treatment in African Caribbeans across the UK and this is indicated by the higher proportions on therapy (Table 3). The fact that the African Caribbeans included in the study came from all social strata may also be relevant.

A higher prevalence of hypertension was not always associated with higher BP in that population. The explanations for this are either that the percentage on antihypertensive treatment differs or the distribution of BP is skewed.26 The former clearly applied. The higher rates of antihypertensive treatment reported in people from African descent is encouraging, given that there is concern about reduced access to health-care services among ethnic minority groups in the UK.21 Higher rates of use of antihypertensives among people from African descent probably reflect awareness among doctors about the risk of hypertension and stroke in this ethnic group.

The findings of higher BP levels and higher prevalence rates of hypertension in African descent are clearcut in the older age groups but not the younger age groups. In Primatesta et al's study, the younger black group had lower but the older black group had higher BP levels compared to whites. Also, in Cruickshanks et al's study, the younger black group aged 15–24 years had a lower BP than the white age group.5 Again, in Lane et al's study, mean systolic and diastolic BPs were marginally lower in Afro-Caribbeans than in the white group in the <30-year-old age group.15 In another two studies on school leavers (hence excluded from the tables), mean systolic and diastolic BPs were lower in the black groups than in the white groups.27,28

Epidemiological studies have consistently shown an important and independent association between increased BP and cardiovascular disease, especially stroke and coronary heart disease.29 Whether the physiological optimal level of BP is the same in each ethnic group is unknown.30 Black hypertensives may have higher death rates from all causes than their white counterparts.31 However, evidence to support this has not been consistent. Findings from the St James Study in Trinidad showed ethnic similarities, for example, the attributable mortality after 8 years from a systolic BP between 155 and 179 mmHg was 7.9 deaths/1000 person year for black and 8.2 for European men, and above 180 mmHg, 14.7 and 15.1, respectively.32 Whether the arbitrary cutoff point for defining hypertension, for example the WHO criteria, will overestimate33 or underestimate individual risk in different ethnic groups remains an open question.

In conclusion, BP levels and prevalence of hypertension are clearly higher in UK residents of African descent. Some of the inconsistent results in the UK studies could be explained by the variations in the methods of measurement of BP and the classification of black groups in some studies. Differences in BMI, age, and sex also further confound findings. Different BP patterns between younger and older age groups in people from African descent in the UK means that there is a case for epidemiological research, including a meta-analysis, examining BP levels and prevalence of hypertension in younger populations (including a review of data in children), and how this pattern may be changing. More research studies are needed in cities outside London. Future research must recognise the various subgroups of West Africans. They should be designed in such a way that data can be combined easily for future systematic reviews, for example by standardising the way in which BP is measured.

Additional information

Glossary of ethnicity termsIn extracting information from the original papers, whenever appropriate, we have taken the terms used by the authors.

Ethnicity—refers to the group individuals belong to as a result of their culture, which includes language, religion, diet, and ancestry.2

Ethnic minority group—refers to minority non-European non-white populations.

White people—refers to people with European ancestral origin.

Black people—refers to people with African ancestral origin.

Afro-Caribbean—refers to people, and their offspring, with African ancestral origin but migrating to Britain via the Caribbean islands.

References

  1. 1

    Cruickshank JK et al. Similarity of blood pressure in blacks, whites and Asians in England: The Birmingham Factory Study. J Hypertens 1985; 3: 365–371.

  2. 2

    Senior PA, Bhopal R . Ethnicity as variable in epidemiological research. BMJ 1994; 309: 327–330.

  3. 3

    Meade TW et al. Ethnic groups comparisons of variables associated with ischaemic heart disease. Br Heart J 1978; 40: 789–795.

  4. 4

    Sever PS et al. Ethnic differences in blood pressure with observation on nor-adrenaline and renin. Clin Exp Hypertens 1979; 1: 733–744.

  5. 5

    Cruickshank JK et al. Blood pressure in black, white and Asian factory workers in Birmingham. Postgrad Med J 1983; 59: 622–626.

  6. 6

    Haines AP, Boorof A, Goldenberg E . Blood pressure, smoking, obesity and alcohol consumption in blacks and whites in general practice. J Hum Hypertens 1987; 1: 39–46.

  7. 7

    Miller GJ et al. Dietary and other characteristics relevant for coronary heart disease in men of Indian, West Indian and European descent in London. Atherosclerosis 1988; 70: 63–72.

  8. 8

    Cruickshank JK et al. Ethnic differences in fasting plasma C-peptide and insulin in relation to glucose tolerance and blood pressure. Lancet 1991; 338: 842–847.

  9. 9

    McKeigue PM, Shah B, Marmot MG . Relation of central obesity and insulin resistance with high diabetes prevalence and cardiovascular risk in South Asians. Lancet 1991; 337: 382–386.

  10. 10

    Chaturvedi N, McKeigue PM, Marmot MG . Resting and ambulatory Blood pressure differences in Afro-Caribbeans and Europeans. Hypertension 1993; 22: 90–96.

  11. 11

    Cappuccio FP, Cooj DG, Atkinson RW, Wicks PD . The Wandsworth Heart and Stroke Study. A population-based survey of cardiovascular risk factors in different ethnic groups. Methods and baseline findings. Nutr Metab Cardiovasc Dis 1998; 8: 371–385.

  12. 12

    Whitty CJM et al. Differences in biological risk factors for cardiovascular disease between three ethnic groups in the Whitehall II study. Atherosclerosis 1999; 142: 279–286.

  13. 13

    Primatesta P, Bost L, Poulter NR . Blood pressure levels and hypertension status among ethnic groups in England. J Hum Hypertens 2000; 14: 143–148.

  14. 14

    Karlsen S, Primatesta P, McMum A . Blood Pressure. Chapter 7 In: Erens B, Primatesta P, Prior G. Health Survey for England—The Health of Minority Ethnic Groups '99. London, The Stationary Office, 2001, pp 175–197.

  15. 15

    Cruickshank JK et al. Hypertension in four African-origin populations: current ‘Rule of Halves’, quality of blood pressure control and attributable risk of cardiovascular disease. J Hypertens 2001; 19: 41–46.

  16. 16

    Lane D, Beevers DG, Lip GYH . Ethnic differences in blood pressure and prevalence of hypertension in England. J Hum Hypertens 2002; 16: 267–273.

  17. 17

    Pacy PJ et al. Prevalence of hypertension in white, black and Asian diabetic in a district hospital diabetic clinic. Diabetic Med 1984; 2: 125–130.

  18. 18

    Khattar RS, Swales JD, Senior R, Lahiri A . Racial variation in cardiovascular morbidity and mortality in essential hypertension. Heart 2000; 83: 267–271.

  19. 19

    Rowlands DB . Intra-arterial blood pressure and cardiovascular responses in black and white hypertensives. Postgrad Med J 1981; 57: 772–773.

  20. 20

    World Health Organisation. Hypertension Control: Report of a WHO Expert Committee. Geneva: WHO, 1996.

  21. 21

    Gill PS, Kai J, Bhopal RS, Wild S . Health care needs assessment: black and minority ethnic groups. In: Raftery J (ed). Health Care Needs Assessment. The Epidemiologically Based Needs Assessment Reviews. Third Series. Abingdon: Radcliffe Medical Press Ltd. (in press), available at http://hcna.radcliffe-online.com/bemgframe.htm

  22. 22

    Weaver MG, Park MK, Lee DH . Difference in blood pressure levels obtained by auscultatory and oscillometric methods. Am J Dis Child 1990; 144: 911–914.

  23. 23

    Bolling K . Dinamap 8100 Calibration Study. OPCS: London, HMSO, 1994.

  24. 24

    Pickering TG et al. How common is white coat hypertension? J Am Med Assoc 1988; 259: 225–228.

  25. 25

    Verdecchia P . Variability between current definition of ‘normal’ ambulatory blood pressure: implications in the assessment of white coat hypertension. Hypertension 1992; 20: 555–562.

  26. 26

    Agyemang C, Bhopal R S . Is blood pressure of South Asian adults in the UK higher or lower than that in European white adults? A review of cross-sectional data. J Hum Hypertens 2002; 16: 739–751.

  27. 27

    De Giovanni JV, Pentecost BL, Beevers DG . The Birmingham blood pressure school study. Postgrad Med J 1983; 59: 627–629.

  28. 28

    Khaw KT, Marmot MG . Blood pressure in 15 to 16-year-old adolescents of different ethnic groups in two London Schools. Postgrad Med J 1983; 59: 630–631.

  29. 29

    MacMahon S, Peto R, Cutler J . Blood pressure, stroke, and coronary heart disease, Part 1, prolonged differences in blood pressure: prospective observational studies corrected for the regression dilution biases. Lancet 1990; 335: 765–774.

  30. 30

    McKeigue PM, Sevak L . Coronary Heart Disease in South Asian Communities. London: Health Education Authority, 1994.

  31. 31

    Hypertension Detection and Follow-up Program Co-operative Group Group. Five-year findings of the Hypertension Detection and Follow-up Program, II. Mortality by race, sex and age. J Am Med Assoc 1979; 242: 2562–2577.

  32. 32

    Miller GJ et al. Adult male all-cause, cardiovascular and cerebrovascular mortality in relation to ethnic group, systolic blood pressure and blood glucose concentration in Trinidad, West Indies. Int J Epidemiol 1988; 17: 62.

  33. 33

    Cruickshank JK, Beevers DG . Epidemiology of hypertension: blood pressure in blacks and whites. Clin Sci 1982; 62: 1–6.

Download references

Acknowledgements

We thank A de Jonge for her valuable advice throughout this work, S Sengupta-Wiebe for her advice during the initial literature search and Hazel King for secretarial support. We also thank Dr Kennedy Cruickshank and Dr Nishi Chaturvedi for their useful comments and Dr Lisa Riste for supplying unpublished data. We thank the anonymous referee, who provided comments that helped improve an earlier version of this paper.

Author information

Correspondence to Raj Bhopal.

Rights and permissions

To obtain permission to re-use content from this article visit RightsLink.

About this article

Further reading