49 / 62
CARDIOVASCULAR JOURNAL OF AFRICA • Volume 31, No 1, January/February 2020
AFRICA
47
Geographical influence on the distribution of the
prevalence of hypertension in South Africa:
a multilevel analysis
Muchiri E Wandai, Shane A Norris, Jens Aagaard-Hansen, Samuel OM Manda
Abstract
Background:
As a response to the growing burden of non-
communicable diseases, the South African government has set
targets to reduce the prevalence of people with raised blood
pressure, through lifestyle changes and medication, by 20% by
the year 2020. It has also recognised that the prevalence varies
at local administrative level. The study aim was to determine
the geographical variation by district of the prevalence of
hypertension among South African adults aged 15 years and
above.
Methods:
Data from all five waves of the National income
Dynamics Study, a panel survey, were used for estimation
by both design-based and multilevel analysis methods. In
the multilevel analysis, a three-level hierarchy was used with
panel participants in the first level, repeated measurements
on patients in the second level, and districts in the third level.
Results:
After accounting for demographic, behavioural,
socio-economic and environmental factors, significant vari-
ation remained in the prevalence of hypertension at the
district level. Districts with higher-than-average prevalence
were found mostly in the south-western part of the country,
while those with a prevalence below average were found in the
northern area. Age, body mass index and race were the indi-
vidual factors found to have a strong effect on hypertension
prevalence for this sample.
Conclusions:
There were significant differences in hyperten-
sion prevalence between districts and therefore the method
of analysis and the results could be useful for more targeted
preventative and control programmes.
Keywords:
hypertension prevalence, district variability, random
effects, multilevel analysis
Submitted 5/2/19, accepted 31/7/19
Published online 20/9/19
Cardiovasc J Afr 2020; 31: 47–54
www.cvja.co.zaDOI: 10.5830/CVJA-2019-047
Hypertension is a major risk factor and consistent predictor for
cardiovascular diseases, such as coronary heart disease, stroke,
transient ischaemic attack and congestive heart failure.
1,2
A
study based on data from the 36-year follow-up Framingham
study pointed out the urgent need for primary prevention of
hypertension by addressing associated risk factors through
weight control, exercise and reduced salt and alcohol intake.
3
In 2015, global age-standardised prevalence of raised blood
pressure was estimated to be 24.1% (21.4–27.1) of men and 20.1%
(17.8–22.5) of women. The number of adults with raised blood
pressure has increased from 594 million in 1975 to 1.13 billion
in 2015, with the increase largely in low- and middle-income
countries (LMICs).
4
According to the 2012 South African
National Health and Nutritional and Health Examination
Survey (SANHANES), the prevalence of hypertension was
approximately 26.0%,
5
and the 2016 Demographic Health Survey
estimated the prevalence to be 46.0 and 44.0% for women and
men, respectively.
6
A number of studies have reported higher-
than-global average prevalence in LMICs,
7-9
and this has been
attributed to non-compliance with treatment, urbanisation,
population ageing and behavioural risk factors, including
tobacco and alcohol use, poor diet and physical inactivity.
7,9,10
In 2013, the South African National Department of Health
developed a strategic plan for the prevention and control
of non-communicable diseases, which targets reducing the
prevalence of people with raised blood pressure by 20% by the
year 2020, through lifestyle change and medication.
11
While
prevalence has been estimated at both provincial and national
levels, little is known on the prevalence of hypertension at levels
below the province due to limited data that can reliably be used
for estimation.
In South Africa, existing surveillance and estimation of
hypertension and other non-communicable disease (NCD)-
related risk factors are overwhelmingly focused at the first
(national) or second (provincial) level geographies,
5,12-14
but
gaining a better understanding of variations at the finer
resolutions (district level in particular) could be important in
decision making for improving the effectiveness and efficiency in
the response to hypertension.
While efforts have been made to estimate hypertension
prevalence at the district level, the method used has fallen short
as it does not account for factors that are known to be associated
with prevalence. In one study, district-based prevalence of
MRC Developmental Pathways for Health Research Unit
(DPHRU), Department of Paediatrics, School of Clinical
Medicine, Faculty of Health Sciences, University of the
Witwatersrand, Johannesburg, South Africa
Muchiri E Wandai, MSc,
muchiriwandai@gmail.comShane A Norris, PhD
Jens Aagaard-Hansen, MD, MPH
Health Promotion, Steno Diabetes Centre, Copenhagen,
Gentofte, Denmark
Jens Aagaard-Hansen, MD, MPH
Biostatistics Research Unit, Medical Research Council,
Pretoria, South Africa
Samuel OM Manda, PhD
Department of Statistics, University of Pretoria, Pretoria,
South Africa
Samuel OM Manda, PhD