|dc.description.abstract||Globally, noncommunicable diseases have increased rapidly. They represent the major
health burden in industrialised countries and are fast increasing in developing countries.
In South Africa, a culturally diverse developing country, the scenario is no different:
there is an increased prevalence of noncommunicable diseases amongst South
Africans, and South African Indians in particular. Indian migrants throughout the world
have a higher prevalence of diabetes, coronary heart disease and dyslipidaemia. Very
few statistics exist on the health and nutrition profile of South African Indians.
The aim of this cross-sectional, epidemiological descriptive study was to conduct a
survey on the prevalence of risk markers for noncommunicable diseases and on the
associations of dietary and lifestyle exposures with the risk of noncommunicable
diseases among apparently healthy adult Indians in KwaZulu- Natal, in order to make
realistic recommendations for a sustainable intervention strategy.
Two-hundred-and-fifty apparently healthy Indians, aged 35-55 years, living m
KwaDukuza, were randomly selected for this survey. Before the actual study, a
quantitative food frequency questionnaire (QFFQ) was adapted for Indian diets and
tested for comparative validity against three 24-hour food recalls as the reference
method, and for reproducibility by a second administration in a subsample of 50
respondents. The QFFQ was found to be relatively reproducible but showed moderate to-
poor validity. When testing for reproducibility, the Spearman rank correlation
coefficient was strongest for cholesterol (0.76) whereas total carbohydrate (0.153)
accounted as the weakest non-significant correlation coefficient With regards to the
differences between the reported intakes of the two administrations of the QFFQ, 13 of
the 19 nutrients tested fell within a 10% difference thus showing good agreement When
the Bland-Aitman procedure was used, there was significant proportional bias for all the
nutrients except for saturated fatty acids. Furthermore, the proportion of individuals
classified in the same or adjacent quartiles for energy and macronutrients intake ranged
from 64% (total energy) to 92% (polyunsaturated fatty acids), showing relatively good
agreement In contrast, the results of the Spearman rank correlation and Pearson
coefficients, paired t-tests, Bland-Aitman technique and quartile distribution showed
moderate to weak correlation for comparative validity between the two instruments. A
relatively low range for the Spearman rank correlation (-0.09 to 0.42) and the Pearson
correlation coefficient (0 .008 to 0.0556) was reported. Significant proportional bias was
present for eight nutrients using the Bland-Aitman technique and the proportion of
individuals classified in the same or adjacent quartiles was relatively high.
Thereafter, the main study followed, examining the socio-demographic and
anthropometric profile, diet, physical activity, blood pressure, fasting blood glucose,
triglycerides and total cholesterol of respondents.
In order to determine the prevalence of risk markers, Asian cut-points were used. About
92% of respondents recorded diastolic blood pressure levels >85 mmHG, 88.8% of
respondents recorded triglyceride levels >1 .69 mmol/L, and 38.8% of respondents
recorded fasting blood glucose levels >5.55 mmol/L. When using the cut-off points for
central obesity as defined by waist circumference for men at ≥ 90 cm and women 80
cm, 100% of women and 87.4% of men were classified as centrally obese. In terms of
physical activity, 62.5% of respondents were classified as inactive (<600 METS min).
However, when using algorithms such as the European SCORE and a modified doubled
score, respondents showed minimal risk for cardiovascular diseases.
The reported intake of fruit and vegetables were well below the recommended World
Health Organization population nutrient goals (≥400g) were men reported a mean daily
consumption of 221 .9 g and women 236.9 g. In addition, 94.4% of respondents
consumed below the estimated average requirement for fibre
Diet indices used to assess diet quality showed that the South African Indian diet
reflects good diet quality. Two principal components were identified, based on the
percentage of fat contributed to each food group: Factor 1 (Legumes, cereals and
cereal products, vegetables), which accounted for added fat, and Factor 2 (Fats and
oils, sugars and sweets and the milk group), which accounted for visible fat. According
to regression analyses, only fasting blood glucose was significantly inversely correlated
with both factor scores for the whole sample.
The following associations of nutrient intake with clinical risk markers were noted: there
were a number of significant correlations of risk score (the aggregate total of all risk
markers defined by Asian standards) and risk score 1 (the aggregate of all risk markers
with fasting blood glucose being doubled as compared to the other risk markers) with
nutrient intake, where the percentage energy from fat showed the strongest correlation .
Furthermore, waist circumference showed significant associations with most clinical risk
markers. It was correlated with systolic blood pressure, fasting blood glucose,
cholesterol, triglyceride and body mass index were body mass index showed the
strongest association amongst all the clinical risk markers.
Collectively, the prevalence of risk markers and the noted associations of risk markers
with diet and clinical parameters resonates into the call for dietary modification through
increased consumption of fruit and vegetables, weight reduction, decreased intake of fat
consumption and increased physical activity for a sustainable intervention strategy to
reduce and control the burden of noncommunicable diseases in the Indian population of