Changes in dietary intakes and cardiovascular disease risk factors in black South Africans from 2005 to 2015 : the PURE-SA-NW study

Abstract

Background: The diets of populations are constantly changing, and the South African population is no exception. Over the last couple of decades, the diets of rural and urban South Africans have been undergoing a rapid nutrition transition, changing from prudent traditional diets that have been typically low in fat and energy, to Westernised diets containing foods high in fat and energy. This shift in dietary patterns and nutrient intakes is referred to as the nutrition transition. Dietary changes associated with a nutrition transition are linked to an increase in non-communicable diseases (NCDs) globally, but also in South Africa in particular. The World Health Organisation (WHO) has identified NCDs as the leading cause of death worldwide. The 2018 NCD report of the WHO revealed that the majority (19%) of NCD deaths in South Africa could be attributed to cardiovascular disease (CVD). The emergence of cardiovascular diseases (CVDs) has been linked to several preventable risk factors, including hypertension, overweight, obesity, diabetes and high cholesterol levels. Lowering the prevalence of these preventable risk factors would be key to addressing the ongoing public health dilemma. Research evidence has shown that dietary and nutrient intake is associated with CVD risk. A growing body of evidence also suggests that the type of dietary fat rather than the total amount of dietary fat consumed drives CVD risk. The primary aim of this thesis is to examine the changes in dietary intake from 2005 to 2015, as well as the associations between the changes in dietary fat intake and the changes in CVD risk from 2005 to 2010, of rural and urban South African groups amid a nutrition transition. Methods: This study is a sub-study of the South African leg in the North-West province of the multinational Prospective Urban and Rural Epidemiological (PURE) study (PURE-SA-NWP). Data for PURE-SA-NWP were collected from two communities, one urban and one rural. Baseline data were collected in 2005, with two follow-up collections conducted in 2010 and 2015. Dietary data were collected using a 145-item culturally sensitive quantified food frequency questionnaire (QFFQ). Only participants with complete dietary data were included in the analysis, resulting in a total of 1154 participants in Chapters 3 and 4, and 722 participants in Chapter 5. In article 1 (Chapter 3) Wilcoxon’s two-sample test was used to test for differences between the rural and urban participants (by gender) in terms of their nutrient intakes, as well as the changes in nutrient intakes between 2005 and 2010. Wilcoxon’s signed-rank test was used to test for changes in nutrient intakes between 2005 and 2010, for each of the four groups (by gender and

locality). In article 2 (Chapter 4), stepwise regression models were estimated to test for associations between the change in fat intake and the change in CVD risk factors. In article 3 (Chapter 5), a linear mixed-model approach was used to assess the trends in energy and nutrient intake of urban and rural men and women over time (2005-2010-2015). Results: In 2010, rural participants consumed the amount of energy (men 9·7 MJ/d; women 9·1 MJ/d) that urban participants had consumed in 2005 (men 9·9 MJ/d; women 9·0 MJ/d). Despite the higher energy intake, not all the participants met total micronutrient needs in 2010. Among rural men, an increase in the change in waist circumference from 2005 to 2010 was associated with a decrease in the change in poly-unsaturated fat (β=-0.280; p=0.002) and an increase in the change in mono-unsaturated fat (β=0.200; p=0.027) intake. For rural men and women, trend analysis (2005, 2010, 2015) indicated significant upward trends in total energy (p < 0.001) (men: 7.60 MJ to 13.26 MJ, women: 6.60 MJ in 2005 to 11.89 MJ), and most macro- (p < 0.001) and micronutrients (p < 0.001) from 2005 to 2015. Upward trends were also observed in the proportion of total energy intake from all macronutrients (except alcohol) (p < 0.001). The observed trends for rural and urban participants differed significantly (p < 0.001 for interaction effect) for total energy intake as well as intakes of most micro- and macronutrients. Discussion and conclusion: Findings from the 2005 to 2010 data clearly showed that the nutrition transition was occurring at a rapid pace in rural and urban groups. Despite the significant increases in energy and fat intake from 2005 to 2010, changes in fat intake were, overall, not strongly associated with the changes in CVD risk factors. From the in-depth trend analysis of energy and nutrient intakes of urban and rural participants of the PURE-SA-NWP study (2005-2015), it is evident that rural and urban groups differ significantly in terms of dietary changes occurring during the nutrition transition. The rural groups were experiencing a nutrition transition so rapid that they not only caught up with the urban groups in terms of energy and nutrient intake but had surpassed them by 2015. Energy and nutrient intakes in the urban areas, on the other hand, seemed to have flattened down by 2015. Our results highlight the importance of timeous monitoring of the various drivers of the nutrition transition and the urgent need to steer the nutrition transition in a more positive direction to prevent the progressive worsening of our NCD burden.