CARDIOVASCULAR JOURNAL OF AFRICA • Vol 23, No 7, August 2012
376
AFRICA
Discussion
The purpose of this study was to establish the association
between SES and CVD risk factors (including dietary intakes)
in an African population in the North West province of South
Africa. SES was differentiated by use of residential strata (urban
or rural), employed or not employed, and education levels
(none, primary schooling, or secondary schooling and/or higher
education/training). The salient results are summarised briefly
and discussed, after which a possible socio-economic drift in the
CVD risk factors is critically evaluated.
It should be noted that in the North West province of South
Africa, a region taken over by rural-to-urban migration, the
majority of those moving are in search of a better life in the
cities, where they can possibly get jobs and help those that have
stayed behind in the rural setting, usually children, women and
the elderly. Using this as the trend in the North West province,
one’s setting can be used as a proxy for socio-economic status.
Although this might not hold true for everyone, it does for more
than 80% of the population, as seen in our 1996 to 1998 results
from the THUSA study.
4
It appears that more processed foods are
reaching rural areas, unlike in earlier years.
Dietary intakes
The increased intakes in total energy, fat and protein,
characterised by increases in animal protein and saturated fat,
with a concomitant decrease in total carbohydrates taken in
during urbanisation confirm the changes in dietary patterns
observed during the nutritional transition in other developing
countries,
22
as well as in South Africa.
3
The observed increases
in dietary fibre and micronutrient intake in the urban subjects
are probably related to the increased energy and thus food
intake. The increases were less than expected for dietary fibre
in women, which shows change in the types of foods consumed.
Although the mean increases in selected micronutrients
were substantial, they did not reach recommended values,
a phenomenon also observed in other urban black South
Africans.
23
The observation that intakes of macronutrients by the
more highly educated women compared to those with little or no
education were actually more prudent (lower in energy and fat)
suggests that the nutritional transition may have reached a point
in these women where healthier diets are now being followed.
CVD risk factors
The only significant difference between rural and urban subjects
regarding serum lipid levels was the increased triglyceride levels
in urban women. The same group was significantly older than
the rural women, had a higher fasting mean glucose level, and
higher BMI. It is therefore possible that the higher triglyceride
and glucose levels in urban women could be related to the older
age and higher BMI, both known to influence these variables.
17
Both urban men and women had higher mean blood pressures,
while the rural subjects had higher mean plasma fibrinogen
levels. Plasma fibrinogen is accepted as a risk factor for CVD.
24
In addition, fibrinogen is an acute-phase reactant.
25
The higher
fibrinogen levels in the rural subjects could therefore also reflect
chronic (perhaps low-grade) infection despite the fact that
apparently healthy subjects were recruited for the PURE study.
However, the mean values of the highly sensitive C-reactive
protein did not differ significantly between urban and rural men
and women. In the THUSA study,
26
urban subjects tended to have
higher plasma fibrinogen levels, although values were also raised
in rural subjects, especially in those living on commercial farms.
Effects of educational level
The salient observations regarding the effects of educational level
on CVD risk factors were that with increased education there
were increases in BMI and in energy and macronutrient intakes
of both men and women when those with primary education
were compared to those with no education. This change to a
Westernised diet was sustained in men with secondary education
but not in women.
The other CVD risk factors did not show significant
differences in uneducated and educated subjects, except for
lower serum triglyceride levels and blood pressures in the more
educated women. However, these women were also slightly
but significantly younger than the uneducated women and it
seems that they followed a more prudent diet than women
with only primary schooling. The diet of the educated subjects
resembled that of the urban subjects, indicative of the changes
observed in the nutritional transition, with indications that in the
highly educated women, energy and macronutrient intakes were
changing back to more prudent intakes.
Effects of employment
Of the 1 833 subjects for whom data on employment were
available, 84.2% of the men and 90.0% of the women were
unemployed, which may indicate a bias in the sample selection.
This bias could be the result of recruiting volunteers who were
available for a very long weekday of measurements, for which
employed people would have to take leave. The official figure of
the unemployment rate in South Africa between 2000 and 2006
averaged at 26.38%, with the highest in March 2003, at 31%.
27
However, the PURE questionnaires used to assess employment
status have a category described as homemaker, which is
interpreted as unemployment. It is possible that employed
domestic workers, especially in the urban areas, indicated
this category, which may be partially responsible for the high
unemployment figure. The employed subjects were significantly
younger than the unemployed. This may explain why, despite
the higher fat intake and more Westernised diet, the employed
men and women had significantly lower fasting glucose and
triglyceride (women) levels.
Is there a social drift in CVD risk factors in this
population?
To evaluate this question, one should consider the limitations of
the study. Firstly, as already mentioned, the reported employment
rates may suggest that the sample was biased. However, potential
bias would be similar for all four sites where subjects were
recruited. A second limitation was the absence of reliable data on
personal income. In many African families, household income
from different sources is shared by a varying number of extended
family members.
28
As proxy for socio-economic status, educational level,
employment and residing in urban or rural areas were used.
The data showed that many subjects with higher education were
