Group for Cardiovascular Disease and Epidemiological Transition : G.C.T.

Hypertension in countries in an early stage of epidemiological transition: epidemiology of hypertension and other risk factors; knowledge, attitudes and practices on hypertension, and utilization of health services for hypertension in Dar es Salaam, Tanzania, 1998-2001

Executive summary

This research program has been initiated in collaboration with the Tanzanian Health Authorities since 1998 and with the support of the Swiss National Foundation for Science. As cardiovascular disease is rapidly emerging in developing countries, this research project aims at investigating the epidemiological characteristics of hypertension and relevant other risk factors and the therapeutic response of hypertensive patients an urban population in early epidemiologic transition (Dar es Salaam, Tanzania). Research includes prevalence and determinants of hypertension, strategies for efficient diagnosis of hypertension, and utilization of health services by hypertensive patients. These data will help inform strategies to prevent, detect, and treat hypertension in low-resource settings. In total 6 specific studies are being conducted.

Investigators

Investigator: Pascal Bovet (1,2), Jean-Pierre Gervasoni (1), Christian Lengeler (3), Fred Paccaud (1)
Main and deputy study coordinators in Dar es Salaam: Allen Ross (1), Mashombo Mkamba (4) and Anne Rwebogora (4).

(1) Institute for Social and Preventive Medicine, Lausanne, Switzerland
(2) Ministry of Health, Seychelles
(3) Swiss Tropical Institute, Basel, Switzerland
(4) Ministry of Health, Dar es Salaam, Tanzania

Funding agency and ethic clearance

The study is funded by the Swiss National Science Foundation, Grant No 32-51189.97 (1998-2001). The study was approved by the Ministry of Health in Tanzania and by the ethics committee of the National Institute of Medical Research (NIMR) and the Tanzania Commission for Science and Technology Dar es Salaam, Tanzania. Participants are free to participate and give informed consent before participating to the study.

Objectives of the study

Study 1: Cardiovascular risk factor levels in the population of Dar es Salaam assessed through cross-sectional examination of the entire population aged 25-64 of 5 branches (14 866 eligible participants; 9254 participants),
(see abstract 1 below).

This study was conducted between November 1998 and August 1999. The main findings have relevance to various areas of health. High prevalence of hypertension and obesity in middle-aged adults stresses the need for early public health interventions to limit the burden of cardiovascular disease. The direct association of BMI with socio-economic status found in this population (in contrast to the inverse relationships observed in most industrialized countries) may drive further increase in body weight and blood pressure in the population while this low-income country becomes more affluent. The findings emphasize the need for early public preventive programs to foster healthy lifestyle through health education and the need to adapt health care infrastructure (that is currently almost entirely geared towards infectious diseases control) to meet the emerging challenge of cardiovascular disease.

Study 2: Study on non-participants to Study 1 to examine migration rates in the study area.

This study was conducted between October and December 1999 on 500 randomly selected non-participants, i.e. 12-14 months after a census where these persons were identified. Up to 4 visits were made to encounter all these non-participants with the assistance of local administrative authorities. The main finding is that the migration rate within the study area is very high, with up to 25% of persons in the area who moved or died within the 1-year period under study. These results are similar to the yet unpublished data from a study by another group in Dar es Salaam (AMMP). This has implication for the analysis of data that rely on follow-up in urban settings (like Study 3 and 4). In addition levels of risk factors and other relevant socio-economic variables were similar in this population compared to those measured in study 1, when adjusted for sex and age.

Study 3: Follow-up of all cases with high blood pressure at the first visit and controls matched for sex to identify best strategies to define truly hypertensive cases.

Triplicate measurements of blood pressure were performed during 4 separate visits in 839 persons with elevated BP (³160/95 mmHg at the first measurement of the first visit) and/or under treatment (cases) and 839 persons with normal BP (controls) to examine the natural evolution of BP over time. In addition, the 4 visits were planned according to two time schedules, a short one (one week) and a long one (one month). The allocation of hypertensive/normotensive persons (cases/controls) to the two schedules was done randomly. Preliminary analysis suggests that the largest difference in BP readings over time occurs from first to second visits rather than from second to third visits. There is not much difference in BP readings whether the second visit takes place within 3-4 days or within 2 weeks. The second reading is regularly much lower than the first one and the third reading is also substantially lower than the second reading, irrespective of the visits and the intervals between visits. These findings suggest that a strategy for the diagnosis of definite hypertension in this setting could be one based on triplicate measurements on only two visits taking place at intervals of no more than a week. Final analysis and write-up of an article on this study part will be completed by the end of 2001. These findings aiming at minimizing the number of false positive hypertensive cases while also minimizing the procedures (hence, the resources) will have high relevance in a setting with extremely limited resources (~12 US $ health expenditures per capita and per year).

Study 4: Utilization of health services in identified hypertensive cases during the 12-month period following diagnosis of hypertension.

Data for this part of the study have been collected. Some verification of the data is ongoing and the final database is expected to be ready for analysis by the second quarter of 2001. Analysis and a scientific publication on this part of the study are expected before the end of 2001.

Study 5: Validity of measurements with automatic sphygmomanometers (as used in studies 1 to 3) compared to mercury sphygmomanometers, by comparing measurements with the 2 devices in 400 other persons.
(see abstract 2 below).

The main finding is that readings with automated sphygmomanometers give lower values that mercury sphygmomanometers, particularly for persons with large arms and high blood pressure. This phenomenon is consistent with some formal studies to validate these devices. These findings have important relevance as automatic devices will be used increasingly often in view of the negative impact of mercury on the environment. The mentioned abstract is attached to this report and a formal publication on this issue is being prepared (the paper is expected before the end of 2001)

Study 6: Survey of all dispensaries, pharmacies available in the study area to provide more insight in the interpretation of data on health service utilization.

The simple and small study is desirable to provide more insight in the interpretation of data on health service utilization. The collection of such data, according to a simple protocol, is ongoing and it is expected to be completed the by the first quarter of 2001.

Administration of the project

The study involved 17 clinical officers who were hired to collect the data and 2 officers for data entry. Since October 1999, the study team has been reduced to 3 clinical officers and 2 officers for data management, as the later parts of the project involved fewer patients.

Key human resources (in addition to the investigators)

Dr. Deo Mtasiwa, City Medical Officer and Deputy Director of DUHP
Mr. Pierre Pichette, Director DUHP
Prof. Marcel Tanner, Director and Chairman of STI
Dr. Nigel Unwin, Senior Lecturer, University of Newcastle
Prof. Bernard Waeber, Director of DPP

Host institution

IUMSP: Institute for social and preventive medicine (Director: Fred Paccaud), http://www.iumsp.ch/

Collaborating institutions

AMMP: Adult Morbidity & Mortality Project, PO Box 65243, Dar es Salaam, Tanzania, tel:+255 051 71956, fax: +255 051 46034, email: ammp.dar@twiga.com
DCC: Dar es Salaam City Commission, PO Box 9084, Dar es Salaam, Tanzania
DPP: Division of Physiopathology, University Hospital, 1011 Lausanne, Switzerland, tel ; fax: ; e-mail: bernard.waeber@dmed.unil.ch, web: http://www.hospvd.ch
DHUP: Dar es Salaam Urban Health Project, PO Box 63320, Dar es Salaam, Tanzania, tel: +255 051 113535; +255 (051) 325 179; email:duhp@twiga.com
MOH: Ministry of Health, PO Box 9083, Dar es Salaam, Tanzania, tel: +255 051 27191; fax: +255 05139951
STI: Swiss Tropical Institute, Socinstrasse 57, 4002 Basel, Switzerland, tel: +41 61 284 81 11; fax: +41 61 271 86 54; http://www.wb.unibas.ch/sti

Prevalence of high blood pressure in a population in early stage of epidemiological transition (Forum Med Suisse, Suppl2, 102S, 2001)

Gervasoni JP(1), Bovet P(1,2), Ross A(1), Mkamba M(3), Rwebogora A(3), Lengeler C(4), Paccaud F(1).

(1) University Institute of Social and Preventive Medicine, Lausanne, Switzerland
(2) Unit of Prevention and Control of Cardiovascular Disease, Ministry of Health, Victoria, Seychelles
(3) Ministry of Health, Dar es Salaam, Tanzania
(4) Swiss Tropical Institute, Basel, Switzerland

Objective: To assess prevalence of high blood pressure in a population in early stage of epidemiological transition.
Methods : All houses in 5 branches of Dar es Salaam (Tanzania) were visited and the encountered adults aged 25-64 were examined. Blood pressure (BP) was measured with automatic devices (Visomat 2) and the average of the 2nd and 3rd readings was considered.
Results: 9254 adults were examined, which corresponds to 55% of the adults registered in the branches. Prevalence (%) of BP ³ 140/90 mmHg (hypertension grade I+), ³ 160/100 mmHg (grade II+) and body mass index ³ 30kg/m2 (obesity) appear in the table:

   Men  Women
   n  I+  II+  Ob  n  I+  II+  Ob
25-34 1930     6.4      0.9      2.5      3601     5.2      1.4      6.1     
35-44 910     14.6      4.2      4.8      1191     15.5      4.6      14.8     
45-54 4810     28.3      10.6      8.9      540     34.1      16.7      20.4     
55-64 278     45.9      17.9      8.2      322     15.5      22.7      18.9     

In total, 143 participants reported treatment for hypertension (97 with BP ³ 140/90 mmHg). Systolic BP was significantly associated with age, male sex and body mass index (1.0 mmHg per kg/m2), and was lower in persons with skilled non-manual occupation (vs. unskilled manual) in multivariate linear regression.

Conclusions: High prevalence of high BP was found in middle-aged adults (e.g. ³ 45 years) although the total number of cases was proportionally limited in this demographically young population. Programs to prevent and control hypertension should be conducted promptly as the number of hypertensive adults is expected to further increase due to demographic and epidemiological transition.

SYSTEMATIC BIAS BETWEEN BLOOD PRESSURE READINGS MEASURED WITH MERCURY AND AUTOMATICA SPHYGMOMANOMETERS (Forum Med Suisse, Suppl2, 103S, 2001)

Gervasoni JP(1), Bovet P(1,2), Rwebogora A(3), Mkamba M(3), Paccaud F(1)
(1) University Institute of Social and Preventive Medicine, Lausanne, Switzerland
(2) Unit of Prevention and Control of Cardiovascular Disease, Ministry of Health, Victoria, Seychelles
(3) Ministry of Health, Dar es Salaam, Tanzania

Objective: To assess if there is a systematic difference in blood pressure (BP) readings measured with a mercury sphygmomanometer (MS) and electronic sphygmomanometers used at humeral level (ASH: Visomatâ 2, Hestia) and wrist level (ASW: Visomatâ Watch Handy)
Methods : BP was measured by 4 clinical officers in voluntary men (n=201) and women (n=203) living in Dar es Salaam, aged 24 to 65. Readings were obtained at 3-minute intervals, successively with each type of device. The sequence of BP readings was initiated randomly with any of the 3 types of devices and was conducted three times (so triplicate readings were available for all subjects and with each device). Several devices of each type were used.
Results: Average systolic/diastolic BP readings were as follows (AC = arm circumference):

   N  MS  ASH  ASW
All patients 404 133.8/87.7  129.1/82.8  129.5/81.6 
BP (ASH) ³160 or 100 72 172.0/110.1  168.7/111.0  161.6/103.3 
BP (ASH) ³140 or 90 173 154.8/100.5  150.2/97.2  146.4/93.7 
BP (ASH) <140 and 90 231 118.1/78.0  113.2/72.0  116.9/72.6 
AC ³ 30 cm 122 136.6/88.7  134.5/85.5  132.6/83.4 
AC < 30 cm 282 132.6/87.2  126.7/81.6  128.2/80.9 

Overall, systolic and diastolic BP readings were lower with electronic sphygmomanometers than with MS (P<0.001), particularly for diastolic BP. Compared to readings with MS, systolic/diastolic BP was particularly lower in persons with high BP using ASW and in persons with small AC using ASH and ASW
Conclusions: BP readings were systematically lower with electronic sphygmomanometers than with a MS and differences tended to vary according to arm size and BP. These findings have important relevance as automatic devices are used increasingly often and considering that currently available data and recommendations on BP are mainly based on readings with MS.

BACK TO PROJECTS     BACK TO HOME