The Epidemiology of Elevated Blood Pressure As an Estimate For

Home , Epidemiology

Journal of Human Hypertension (1999) 13, 399–404  1999 Stockton Press. All rights reserved 0950-9240/99 $12.00 http://www.stockton-press.co.uk/jhh ORIGINAL ARTICLE The epidemiology of elevated blood pressure as an estimate for hypertension in Aydın, Turkey · HM So¨nmez1, O Bas¸ak2, C Camcı1, R Baltacı3,HS¸ Karazeybek4, F Yazgan4,IErtin5 and S¸Ç Çelik6 Departments of 1Internal Medicine and 2Family Practice, School of Medicine, 3Students Health Centre, Adnan Menderes University, Aydın; 4Social Insurance Hospital, Aydın; 5Kuyucak Health Centre, Aydın; 6Gu¨ llu¨ bahçe Health Centre, Aydın, Turkey

Background: Hypertension is an important public health (for BP у140/90 mm Hg or on treatment). Hypertension problem, with some variability of its epidemiological prevalence increased progressively with age, from 9% properties in different populations. in 18- to 29-year-olds to 70.6% in those 70–79 years of Objectives: The purpose of this study was to estimate age. Women had a signiﬁcantly higher prevalence than the prevalence of hypertension and to determine the men (34.1% vs 26.0% respectively). Overall, 57.9% of hypertension awareness, treatment and control rates in hypertensive individuals were aware that they had high Aydın, a Turkish province. BP, and 82.1% of aware hypertensives were being Methods: Of 1600 coincidentally selected people aged treated with antihypertensive medications, but only over 18 years in Aydın, 1480 (92.5%) had their blood 19.8% of treated hypertensives were under control pressure (BP) measured and answered a standard ques- (systolic pressure Ͻ140 mm Hg and diastolic pressure tionnaire in 1995. Ͻ90 mm Hg). In addition, housewives, unemployed, and Results: Estimates of the prevalence of hypertension the less educated individuals had greater mean systolic and its control were computed using two different cri- and diastolic BP. teria to deﬁne hypertension: BP у140/90 mm Hg or on Conclusions: Our results indicate that hypertension is treatment and BP у160/95 mm Hg or on treatment. Over- highly prevalent in Aydın, Turkey, and the detection and all, the estimated prevalence of hypertension was 29.6% control of hypertension is unsatisfactory.

Keywords: hypertension; prevalence; detection; awareness; control; Turkey

Introduction The Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure Hypertension (HT) is an important public health (JNC V) have suggested a new classification of adult problem. High blood pressure (BP) is one of the BP and have divided high BP into four stages that major risk factors for cardiovascular disease and the convey more efficiently the major impact of HT on most important risk factor for cerebrovascular dis- 1,2 the risk of cardiovascular disease. It has been eases. The data from various epidemiological defined that even high-normal BP, as well as high studies about HT have shown that 15–25% of the BP stage 1 (previously termed mild) results in target- population have high BP, and that detection, treat- organ damage. People with high-normal BP are at ment and control of HT decrease the mortality and increased risk of experiencing cardiovascular 3–7 morbidity from stroke and coronary heart disease. events, compared to otherwise similar persons with According to the estimates derived from the lower blood pressures.1,10 However, most hyperten- National Health and Nutrition Examination Surveys sives are unaware of their condition, aware but III, there are 50 million people with elevated BP or untreated, or treated inadequately. taking antihypertensive medication in the USA The epidemiological properties of HT show some resulting in a loss of 2 million years of human life a 3,6,11–13 1,8 variability in different populations. In year or 500 000 productive years per annum. The addition, each country has different socio-economic, mortality from stroke has decreased approximately cultural, and health care capabilities. However, few 1% annually possibly as a result of improvement in national or regional data have been published on the taking antihypertensive treatment and controlling 14–18 9 epidemiology of HT in Turkey. HT during the last two decades. With the purpose of identifying individuals with elevated BP previously detected or not and those Correspondence: Yrd. Doç. Dr Okay Bas¸ak, Adnan Menderes inadequately treated· or uncontrolled,· we carried out Universitesi, Tıp Faku¨ ltesi Aile Hekimlig˘i Anabilim Dalı 09100, the AYDIN HI·PERTANSIYON ARAS¸TIRMASI sur- Aydın, Turkey vey (AYDINHIP) in our university region, Aydın, Received 13 January 1999; accepted 3 February 1999 Turkey, in 1995. The epidemiology of hypertension in Aydın, Turkey HM So¨nmez et al 400 Subjects and methods Blood pressure measurement After the subjects had answered the questions in the Study design questionnaire and rested for 5 min, two sitting BP measurements were taken in the right arm with a This was a cross-sectional survey. Data collection pretested aneroid sphygmomanometer approxi- was obtained by structured interviews using preco- mately 2 min apart. The accuracy of eight aneroid ded questionnaires administered by trained inter- sphygmomanometers used in the screening were viewers in the subjects’ homes and at worksites. checked at different pressure levels by connecting Interviewers were trained according to a standard- them with a metal T-connector to the tubing of a ised protocol for the measurement of BP. Standard- standardised mercury column manometer.21,22 The ised questionnaires were administered prior to the average of two readings was used to determine the measurement of BP. The questionnaire included BP level. The cuff size was 23 × 12.5 cm. The sys- questions on demographic information, and on the tolic (SBP) and diastolic BP (DBP) readings were subjects’ knowledge of BP and treatment status as defined as corresponding to the first of two consecu- well as questions on their health status. tive Korotkof sounds and the disappearance of A kind of systematic sampling procedure was sound, respectively. BP measurements were taken used in selecting the study population. Screening on one occasion only and recorded to the nearest 2 was conducted in the city centre, and in two towns mm Hg. BP measurement procedures recommended and two villages from both the west and east sides by WHO/ISH were followed.4 of the city centre, covering eight geographic areas. The sampling was based on household and worksite selection. The households and worksites on either Statistics the right side or the left side only of a street were All recorded data was transferred to a PC media and visited. Each investigator visited 200 persons who analysed using the Statistical Package for Social met the criteria of inclusion and invited them to par- Sciences (SPSS) software.23 Frequency, cross-table, ticipate in the study. Screening was performed Levene’s test, Student’s t-test, Pearson’s bivariate between March and May (in the Spring season) and correlation test, analysis of variance (ANOVA), between 8 and 11 am every investigation day. We analysis of covariance (ANCOVA) and Tukey-HSD did not include in the study those who were not test were used in statistical assessment with signifi- resident of the region, who had a mental disorder, cance level of 0.05. severe obesity or a chronic metabolic illness such as chronic renal failure, hypothyroidism, and hyper- thyroidism and who were under 18 years old. Results Decision to exclude was based on the history taking General findings from the individuals visited and on the observations of the interviewers. Subjects were assessed to be sev- Of the 1466 subjects included in the study, 44.5% erely obese if they were unable to participate in were women and 55.5% men. Thirty-four percent of daily activities due to obesity. the people screened had an educational level higher than primary school and 58% were non-smokers. Thirty-one percent of the study subjects were housewives, 12% self-employed, 20% farmers, 18% offi- Population cers (not army officers), 11% workers and 8% unemployed. This survey was carried out in Aydın (a south-west- Two hundred and eighty-four participants ern province in Turkey) with about 540 000 inhabi- reported having HT previously. But, 32 of these tants over 18 years of age. We visited 1600 persons individuals had BP within the normal range (50 having study criteria of whom 120 persons refused individuals for 160/95 mm Hg cut-off point), an invitation to participate in the study. Non- although they did not use antihypertensive medi- response rate was 7.5%. Fourteen participants were cation. Therefore, these people have not been excluded from the study at the stage of analysing included in the estimate for HT. Among the subjects data obtained because of the reasons mentioned reporting a diagnosis of HT, 51% were practising above. Therefore, we screened 1480 subjects and salt restriction. While 77 previous hypertensives recruited 1466 of them (813 males, 653 females). A (27%) were not on antihypertensive medication and total of 970 participants (66%) were from urban 84 previous hypertensives (30%) were taking drug centres (with inhabitants above 10 000) and 496 treatment inconsistently, only 123 previous hyper- (34%) from the rural areas. The average age was tensives (43%) reported taking regular antihyperten- 47.6 ± 15 years and median 44 years (46.2 ± 15.2 and sive medication. 45 for men, and 44.9 ± 17 and 43 for women). Sixty- three percent of participants were aged 18–50 years. Blood pressure findings The histogram of age was slightly depressed from right to left (Kurtosis: −0.78 ± 0.1 and Skewness: The prevalence of HT was estimated using the level 0.37 ± 0.06), reflecting the predominance of young of 140/90 mm Hg, as recommended by the JNC V. In people in Turkey. About half of the population of addition, it was assessed according to criteria of the Turkey are between 15 and 44 years old, and the WHO/ISH (BP у160/95 mm Hg). Subjects who ratio of elderly is rather low.19,20 reported a diagnosis of HT and taking antihyperten- The epidemiology of hypertension in Aydın, Turkey HM So¨nmez et al 401 sive medication but with normal blood pressures at ones (119.6 ± 20.6, 76.3 ± 12.8) (t = 10.09 and the time of survey were also included in an estimate P = 0.000, t = 3.25 and P = 0.001 respectively). of the overall prevalence of HT. Analysis with one-way ANOVA indicated a statisti- A systolic and/or diastolic BP у140/90 mm Hg cally significant difference of mean SBP and DBP was found in 394 individuals, ie 26.8% of parti- levels among some socioeconomic categories = = = cipants (31.1% of women, 23.7% of men). Forty-one (F6,1456 19.57, P 0.000 and F6,1456 10.02, individuals (2.8% of study population) reporting a P = 0.000 respectively). When analysis of covariance diagnosis of HT and receiving antihypertensive was made using age as a covariate, it was felt that treatment had normal BP at the time of survey. Thus, systolic and diastolic BP increased with age the overall prevalence of elevated BP was 29.6% (P = 0.000) and while controlling for the effects of (34.1% in women, 26.0% in men). For the 160/95 age on BP levels the difference among socio-econ- mm Hg cut-off point, the numbers were 251 individ- omic categories continued to be significant uals and 17.1%, 79 individuals and 5.4%, and (P = 0.000). Further analysis with Tukey’s pairwise 22.5% (28.3% in women, 17.8% in men) respect- comparison procedure to control for multiple testing ively. HT prevalence rates by sex and ages in the revealed that the mean systolic and diastolic BP lev- study population (over 18 years of age) have been els of the categories of housewife and unemployed shown in Table 1. As seen, in young adulthood (18– were significantly higher than those of workers, 29 years of age) the prevalence of elevated BP is officers, farmers and self-employed with an overall greater for men than for women; thereafter, the alpha level of 0.05 (Table 2). In addition, as shown reverse is true. in Table 3, for the 35–64 age group age-specific For the purpose of comparison directly with other mean systolic and diastolic BP levels of the house- European studies, age-specific prevalence rates for wives were significantly higher than those of other the 40–59 age group as considered in the ERICA Pro- female participants (P = 0.000). But there were no ject were also determined.24 Assuming a cut-off for significant differences among the blood pressures of HT of 160/95 mm Hg, we found a prevalence of the other working females (P Ͼ 0.05). 26.0% (33.0% in women and 21.1% in men). Considering settlement status, SBP levels were The prevalence of elevated BP was significantly significantly higher in rural populations higher in women than in men (␹2 = 11.126, P = 0.000 (132.6 ± 29.3 mm Hg) than in urban populations for the cut-off point of 140/90 mm Hg and (128.7 ± 26.1 mm Hg) (t = 2.58, P = 0.010). In con- ␹2 = 22.858, P = 0.000 for the cut-off point of 160/95 trast, DBP levels were higher in urban populations mm Hg). Systolic and diastolic BPs were also sig- (78.9 ± 14.7 mm Hg) than in rural populations nificantly higher in women than in men. The mean (77.1 ± 16.0 mm Hg) (t = 2.11, P = 0.035). SBP in women was 132.0 ± 31.5 mm Hg compared with 128.4 ± 23.2 mm Hg in men (t = 2.45, P = 0.015) Hypertension awareness, treatment and control and the mean DBP in women was 79.4 ± 16.3 mm Hg rates compared with 77.4 ± 14.1 mm Hg in men (t = 2.39, P = 0.017). Age-specific mean systolic and diastolic Participants were asked about hypertension aware- BP for the 35–64 age group according to WHO stan- ness and antihypertensive medication. Overall, dardisation were 131.0 ± 25.1/79.8 ± 14.8 mm Hg 57.9% of hypertensive individuals (252/435) were (134.2 ± 28.3/81.4 ± 15.3 mm Hg in women and aware that they had high BP. The percentage of 128.8 ± 22.3/78.6 ± 14.4 mm Hg in men).25 aware hypertensives who were under antihyperten- As expected, there was a positive correlation sive drug treatment (regularly or inconsistently) was between the age distribution and BP levels of the 82.1% (207/252), and the percentage of treated population studied (r = 0.52, P = 0.000 for SBP and hypertensives achieving an adequate BP control r = 0.29, P = 0.000 for DBP). The SBP and DBP levels (systolic pressure Ͻ140 mm Hg and diastolic press- were greater in less educated participants ure Ͻ90 mm Hg) was 19.8% (41/207). Hypertension (133.5 ± 28.3, 79.1 ± 15.8) than in more educated awareness, treatment and control rates for BP

Table 1 Hypertension prevalence rates by sex and ages in the study population (over 18 years of age), 1995

Age (year) HT prevalence estimates for 140/90 mm Hg HT prevalence estimates for 160/95 mm Hg cut-off point* cut-off point**

Total % Women % Men % Total % Women % Men %

18 and above 29.6 34.1 26.0 22.5 28.3 17.8 18–29 9.0 5.9 12.1 3.6 3.3 4.0 30–39 13.1 19.4 8.1 8.2 11.1 5.9 40–49 27.1 30.5 24.7 20.0 22.9 17.9 50–59 44.3 53.3 38.1 33.9 46.6 25.1 60–69 54.5 65.6 46.4 45.0 59.8 34.2 70–79 70.6 80.7 62.5 60.3 80.7 43.7 80 and above 54.5 73.3 14.2 50.0 66.6 14.2

*HT defined as a SBP у140 mm Hg and/or DBP у90 mm Hg and/or reported treatment with one or more antihypertensive medications. **HT defined as a SBP у160 mm Hg and/or a DBP у95 mm Hg and/or reported treatment with one or more antihypertensive medications. The epidemiology of hypertension in Aydın, Turkey HM So¨nmez et al 402 Table 2 Mean systolic and diastolic BP levels of some socio-econ- cant difference between the mean BP levels of the omic categories previously hypertensive subjects (regardless of their actual BP measurement) on regular medication and n Mean Mean Ͼ SBP ± s.d. DBP ± s.d. of those taking no medication (P 0.05). (mm Hg) (mm Hg) Discussion Workers 132 118.8 ± 20.4 72.8 ± 14.1 Officers 252 119.9 ± 22.5 76.1 ± 13.3 With this study, we aimed to determine some epide- Self-employed 179 126.5 ± 23.1 77.1 ± 13.4 miological aspects of HT in our university region. Managers 29 128.3 ± 19.3 81.5 ± 8.9 The data reported here are derived from a cross- ± ± Farmers 292 130.6 25.4 76.5 15.3 sectional survey and are likely to have several limi- Housewives 455 137.7 ± 31.7 81.6 ± 16.7 Unemployed 123 139.1 ± 25.1 82.1 ± 13.8 tations for estimating the prevalence of HT in the Overall study 1462 130.1 ± 27.3 78.3 ± 15.1 region. population First of all, the estimates are based on one- occasion measurements and follow-up information Analysis of covariance. Age as a covariate (F = 451.83, P = 0.000 has not been obtained. Using a single visit estimate and F = 114.04, P = 0.000). = to determine control of clinically relevant HT might Main effects of some socio-economic categories (F 6.60, 26 P = 0.000 and F = 5.96, P = 0.000). have led to overestimation of the prevalence of HT. The use of aneroid sphygmomanometers for BP measurements can be another limitation of the у140/90 mm Hg or on treatment, and for BP study. However, because the accuracy of these у160/95 mm Hg or on treatment have been shown sphygmomanometers has been checked by using a in Table 4. As seen, women had a better awareness closed system designed according to the recommen- (␹2 = 25.52, P = 0.000), and treatment (P Ͼ 0.05) but dations of the American Heart Association (AHA) worse control (P Ͼ 0.05) status for HT than men. In and British Hypertension Society (BHS), they are the 40–59 age group the numbers were 64.2% for likely to be reliable, their measurements reproduc- awareness (63.9% in women, 64.6% in men), 86.4% ible within machine and comparable between for treatment (89.1% in women, 83.3% in men) and machines. On the other hand, only one cuff size has 51.3% for control (53.7% in women, 48.6% in men) been used for all individuals in the study and the for BP у160/95 mm Hg. length of 23 cm does not meet present recommen- Participants who were diagnosed as hypertensive dations of the AHA and BHS that a normal adult- previously had significantly higher systolic and sized bladder should have at least dimensions of diastolic BP levels (157.0 ± 31.4 mm Hg and 13 × 24 cm and 12 × 26 cm respectively.21,22 89.6 ± 16.8 mm Hg) than the rest of the study popu- Although severe obese individuals were excluded, lation (123.6 ± 21.8 mm Hg and 75.7 ± 13.4 mm Hg) the availability of one cuff size might have affected (t = 16.81 and P = 0.000, t = 12.88 and P = 0.000, measurements at the extremes of the arm circumfer- respectively). But there was no statistically signifi- ence size. Considering that a bladder measuring

Table 3 Differences of mean systolic and diastolic BP levels between housewives and other female participants (35–64 years of age)

n Mean SBP ± s.d. Signiﬁcance Mean DBP ± s.d. Signiﬁcance Aged 35–64 (mm Hg) value* (mm Hg) value*

Housewives 267 137.6 ± 28.6 82.6 ± 15.6 t = 8.57, P = 0.000 t = 4.86, P = 0.000 Other women 76 122.5 ± 23.9 77.3 ± 14.2

*Student’s t-test.

Table 4 Awareness, treatment and control of high BP in the study population (over 18 years of age), Aydın, Turkey, 1995

140/90 mm Hg cut-off point 160/95 mm Hg cut-off point

Total % (n) Women % (n) Men % (n) Total % (n) Women % (n) Men % (n)

Awareness of high 57.9 69.5 45.7 70.9 78.9 60.7 BP* (252/435) (155/223) (97/212) (234/330) (146/185) (88/145)

Treated** 82.1 85.1 77.3 88.4 90.4 85.2 (207/252) (132/155) (75/97) (207/234) (132/146) (75/88)

Controlled 19.8 15.9 26.6 38.1 34.1 45.3 Adequately*** (41/207) (21/132) (20/75) (79/207) (45/132) (34/75)

*Data are percentages of hypertensive individuals who are aware that they have high BP. **Data are percentages of aware hypertensives who are under antihypertensive drug treatment. ***Data are percentage of treated hypertensives achieving an adequate BP control: systolic BP below 140 mm Hg (or 160 mm Hg) and diastolic BP below 90 mm Hg (or 95 mm Hg). The epidemiology of hypertension in Aydın, Turkey HM So¨nmez et al 403 12 × 26 cm would correctly cuff 79% of European and 51.4% in women for those older than 60 years arms,27 using a smaller bladder might have led us to of age. overestimate the prevalence rates in our study. Age-specific mean SBP and DBP levels for the 35– A further study limitation concerns the sampling 64 age group in the present study are almost similar design. Although a random sampling based on prob- to those of the TEKHARF study in both women ability has not been used, (a) the appropriateness of (133.0/82.3 mm Hg) and men (125.1/80.0 mm Hg). rural-urban distribution of the study population Compared with the mean BP levels of the median with that of general population, and (b) a slight centre of the population samples enrolled in the depression of the histogram of age from right to left, WHO MONICA study (134/83 mm Hg in women and reflecting the predominance of young people in Tur- 135/86 mm Hg in men), men in our population have key, increase the likelihood of prevalence estimates much lower systolic and diastolic BPs (128.8/78.6 reported here to be close to the true population mm Hg) but the results are nearly the same in values. For all these reasons, our results should be women (134.2/81.4 mm Hg).25 cautiously interpreted. By assessing the data derived from the follow-up Because different criteria to define HT can have screening of the original TEKHARF study cohort in important effects on the estimates of prevalence and 1995, Onat et al have concluded that the mean sys- control26 and for the purpose of comparison with tolic and diastolic BP levels have not changed other studies, we computed the estimates of preva- between 1990 and 1995 and have estimated that lence of HT and its control using two different BP there are 10.2 million people (5.8 million women cut-off points: as recommended by the JNC V (140 and 4.4 million men) having BP у140/90 mm Hg or and/or 90 mm Hg) and according to the WHO/ISH taking antihypertensive medication in Turkey. criteria mostly used in Europe (160 and/or 95 Comparing their estimates from the TEKHARF mm Hg).1,4 study cohort with the data from NHANES III, Onat Although few national or regional surveys have et al have suggested that age-specific prevalence been undertaken on the epidemiology of HT in Tur- rates of HT are greater among Turkish people than key, the data from these studies indicate some simi- American people. Whereas, a reverse inference 14–16,18 larities with· our results. As we know, appears to be true with European populations, AYDINHIP is the third major study on HT epidemi- especially for men. ology in Turkey. In the early 1990s, two large-scale Not unexpectedly, the present study has indicated population screenings have been carried out: the that housewives, unemployed and the less educated TEKHARF study by the Turkish Society of Cardi- individuals have higher BP levels. Age-specific BP ology and Ministry of Health collaboratively in 1990 levels of the housewives are greater than those of and Hypertension Screening Study by the Turkish other female participants (P = 0.000). This might be Association of Hypertension and Atherosclerosis explained through obesity and lack of activity. Onat (TAHA) in 1993. Evidence from these country-wide et al have found that women over 40 years of age surveys have suggested that HT is relatively preva- have a body mass index (BMI) about 3 kg/m2 greater lent among Turkish people and women have higher than men and have suggested that this difference in prevalence estimates than men. BMI meets two-thirds of the difference of 10–11/4 In the present study, we found a higher preva- mm Hg between women and men.18,30 The results of lence rate in women than in men, which was con- the TAHA study have revealed an inverted associ- sistent with the results of other studies from Tur- ation between the level of education and BP levels. key.14,18,28 On the other hand, the WHO MONICA But our results of unemployed individuals are not study has also shown that almost universally the consistent with those of the TAHA study.14 problem of elevated BP is more prevalent in women Considering awareness and management of HT, than in men.29 our findings are worrisome. One of the most striking In the TAHA study, Erdine14 has defined the sys- observations of the present study is that therapy tolic and diastolic components of HT separately and does not necessarily signify control of the elevated found higher systolic and diastolic HT prevalence BP. Although overall awareness rates appear to be estimates in women than in men. In the TEKHARF good, which indeed are better than those of several study, a cardiovascular risk factor survey of adults countries,31–34 and the majority of aware hyperten- over 20 years of age, Onat et al have found that sives take antihypertensive medications, only one- 16.5% of women and 11% of men are hypertensive third of those under treatment (or 23.9% of all according to the threshold of 160/95 mm Hg for hypertensive individuals) are adequately controlled. defining HT.18 Age-specific rates for the 40–59 age Furthermore, it has not found any significant differ- group are 24.2% and 18.6% respectively. The esti- ence between BP levels of aware hypertensives on mates from the TEKHARF study are relatively lower treatment and of those taking no medications than those of East and South European countries (P Ͼ 0.05). The reason for this might be that nearly included in the ERICA project.24 However, our age- half of the hypertensive individuals take antihyper- specific estimates for the same age group are higher tensive medications irregularly. In addition, we in women and lower in men than the results of the think that there is poor therapeutic compliance even ERICA project. among the hypertensives taking regular treatment. In another Turkish study which includes subjects Onat et al have found a lower percentage of older than 59 years of age, 50.8% of men and 52.5% receiving treatment (one-third) and a higher percent- of women have been detected as hypertensive.16 In age of control (42% of treated women and 60% of our population these estimates are 30.1% in men treated men).15 Awareness, treatment and control of The epidemiology of hypertension in Aydın, Turkey HM So¨nmez et al 404 HT is still an important problem even in developed kan basıncının bazı dig˘er risk fakto¨rlerinden countries.35–38 bag˘ımlılıg˘ının nicelig˘i. Tu¨ rk Kardiyol Dern Ars¸ 1996; In conclusion, we have found a high prevalence 24: 328–336. of HT with a low degree of control. Specific inter- 18 Onat A et al.Tu¨ rkiye’de eris¸kinlerde kalp hastalıg˘ıve vention and control programmes, including detec- risk fakto¨rleri sıklıg˘ı taraması: 5. Hipertansiyon ve tion strategies, are needed to increase the extent of sigara içimi. Tu¨ rk Kardiyol Dern Ars¸ 1991; 19: 169– 177. control of HT in our region and country-wide. · 19 Devlet Istatistik· Enstitu¨ su¨ (State Statistics Institute, Turkey). ‘Istatistik Go¨stergeler 1923–1992’, 1994, p 14. ¨ Acknowledgement 20 Sag˘lık Bakanlıg˘ı (Ministry of Health, Turkey). H. U. Nu¨ fus Etu¨ tleri Enstitu¨ su¨ ve Macro International Inc. We thank Prof. Dr. Bilgin Timuralp and Associate 1994. Tu¨ rkiye nu¨ fus ve sag˘lık aras¸tırması 1993, Ank- Prof. Dr. Necmi Ata from the Cardiology Department ara, Turkey, p 5. of Osmangazi University School of Medicine for 21 Petrie JC, O’Brien ET, Littler WA, de Swiet M. Rec- their advice on this paper. ommendations on blood pressure measurement. British Hypertension Society. Br Med J 1986; 293: 611–615. 22 Kirkendall WM et al. Recommendations for human References blood pressure determination by sphygmomanometers. Circulation 1980; 62: 1145A–1155A. 1 The Fifth Report of the Joint National Committee on 23 Marija JN. SPSS for Windows Base System User’s Detection, Evaluation, and Treatment of High Blood Guide Release 6.0. Chicago, USA, SPSS Inc. 1994. Pressure (JNC V). Arch Intern Med 1993; 153: 154–183. 24 Lamm G (on behalf of WHO ERICA Research Group). 2 Al-Roomi K et al. The importance of hypertension in The risk-map of Europe. Ann Med 1989; 21: 189–192. the aetiology of infarctive and haemorrhagic stroke. 25 The WHO MONICA Project: Geographical variation in Med J Aus 1992; 157: 452–455. the major risk factors of coronary heart disease in men 3 Rudnick KV, Sackett DL, Hirst S, Holmes C. Hyperten- and women aged 35–64 years. Wld Hlth Statist Quart sion in a family practice. J Can Med Assoc 1977; 177: 1988; 41: 115–140. 492–497. 26 Birkett NJ. The effect of alternative criteria for hyper- 4 Zancheti A et al. Guidelines for the management of tension on estimates of prevalence and control. J Hyp- mild hypertension. Special ed. Memorandum from a ertens 1997; 15: 237–244. WHO/ISH meeting: ISH Hypertension News 1993; 27 O’Brien ET et al. Blood pressure measurement. Rec- pp 1–20. ommendations of the British Hypertension Society. 5 Poulter NR et al. Concomitant risk factors in hyperten- BMJ Publishing Group: Plymouth, 1997, pp 1–27. sives: a survey of risk factors for cardiovascular disease 28 Baykan N. Tu¨ rkiye’de kalp hastalıkları prevalansı amongst hypertensives in English general practices. u¨ zerine¸ calıs¸malar. Ankara U¨ niversitesi Yayını: Ank- Blood Press 1996; 5: 209–215. ara, 1973, pp 1–57. 6 Bannan LT, Beevers DG, Jackson SHD. ABC of blood 29 Wolf HK et al. Blood pressure levels in the 41 popu- pressure management: detecting hypertensive patients. lations of the WHO MONICA Project. J Hum Hypertens Br Med J 1982; 282: 1211–1213. 1997; 11: 733–742. 7 Beevers DG. Do we need a hypertensionologist? Br 30 Onat A et al. TEKHARF: 6. Obesite ve diyabet. Tu¨ rk Med J 1983; 287: 1597–1600. Kardiyol Dern Ars¸ 1991; 19: 178–185. 8 Final Report of the Subcommittee on Definition and 31 Colhoun HM, Dong W, Poulter NR. Blood pressure Prevalence of the 1984 Joint National Committee. screening, management and control in England: results Hypertension prevalence and the status of awareness, from the health survey for England 1994. J Hypertens treatment, and control in the United States. Hyperten- 1998; 16: 747–752. sion 1985; 7: 457–468. 32 Jones DW, Kim JS, Kim SJ, Hong YP. Hypertension 9 Casper M et al. Antihypertensive treatment and U.S. awareness, treatment and control rates for an Asian trends in stroke mortality, 1962 to 1980. Am J Pub population: results from a national survey in Korea. Health 1992; 82: 1600–1606. Ethn Health 1996; 1: 269–273. 10 Julius S et al. The association of borderline hyperten- 33 Ibrahim MM et al. Hypertension prevalence, aware- sion with target organ changes and higher coronary ness, treatment, and control in Egypt. Results from the risk. JAMA 1990; 264: 354–358. Egyptian National Hypertension Project (NHP). NHP 11 Pick WM et al. The epidemiology of hypertension in Investigation Team. Hypertension 1995; 26(6 Pt 1): family practice in Cape Town. S Afr Med J 1990; 78: 7–10. 886–890. 12 Nan L et al. Prevalence and medical care of hyperten- 34 Smith WC, Lee AJ, Crombie IK, Tunstall-Pedoe H. sion in four ethnic groups in the newly-industrialised Control of blood pressure in Scotland: the rule of nation of Mauritius. J Hypertens 1991; 9: 859–866. halves. BMJ 1990; 300: 981–983. 13 Manfred SG, Israela P. Prevalence and control of 35 Joffres MR et al. Awareness, treatment, and control of hypertension in a large cohort of occupationally active hypertension in Canada. Am J Hypertens 1997; 10: Israelis examined during 1985–1987: The Cordis 1097–1102. Study. Int J Epidem 1992; 21: 676–682. 36 Chockalingam A, Fodor JG. Treatment of raised blood 14 Erdine S. Hypertension Map Of Turkey. MD: Hyper- pressure in the population: the Canadian experience. tension and Atherosclerosis Association of Turkey, Am J Hypertens 1998; 11(6 Pt 1): 747–749. 1993, pp 1–101. 37 Chamontin et al. Prevalence, treatment, and control of 15 Onat A et al.Tu¨ rk eris¸kinlerinde kan basıncında yeni hypertension in the French population: data from a eg˘ilimler: TEKHARF çalıs¸ması 1990 ve 1995 verileri- survey on high blood pressure in general practice, nin analizi. Tu¨ rk Kardiyol Dern Ars¸ 1996; 24: 73–81. 1994. Am J Hypertens 1998; 11(6 Pt 1): 759–762. 16 Çetinarslan B et al. The hypertension prevalence and 38 Puras A, Sanchis C, Artigao LM, Division JA. Preva- some electrocardiographic findings in elderly people. lence. awareness, treatment, and control of hyperten- Archives of the Turkish Heart Society 1993; 21: 17–21. sion in a Spanish population. Eur J Epidemiol 1998; 17 Onat A et al.Tu¨ rk eris¸kinlerinde sistolik ve diyastolik 14: 31–36.