Dr Rebecca Pradeilles, School of Health and Related Research, University of Sheffield; Dr Paula Griffiths and Dr Emily Rousham, Centre for Global Health and Human Development, Loughborough University.

Rebecca Pradeilles, SIID Fellow has led a recent study of adolescents in Johannesburg-Soweto, South Africa with colleagues from the Centre for Global Health and Human Development at Loughborough University that she summarises here.


Our study in urban South Africa was conducted as part of my PhD studies that I completed last year. We found that even within a small urban area, anthropometric patterns don’t change in the same way for everyone. There are patterns that emerge in these anthropometric changes, which are different for adolescent boys and girls and for people from different socio-economic backgrounds.

South Africa is experiencing a “triple burden” of disease

Changing diets are occurring simultaneously to the persistence of other health problems in South Africa. Like many urban settings in developing economies, South Africa is experiencing the so called “triple burden” of disease, characterised by continued high levels of infectious diseases and under-nutrition, alongside rapid rises in diseases that are historically associated with high income countries, such as obesity, cardiovascular diseases, cancer and diabetes. This triple burden has obvious health and economic consequences in these resource-poor settings. Of particular concern in South Africa are the high rates of HIV, with approximately 1 in 5 South African adults diagnosed as HIV positive. This huge burden is happening alongside persistent underweight (three times higher in men) and overweight/obesity (twice as high in women), reaching levels comparable with that in the UK for women, as almost two-thirds are affected. This triple burden of diseases can play out on a local level, as people living in the same community, household and even individuals can suffer with all three diseases.

soweto1soweto 2

Neighbourhoods in Soweto

Socio-economic background influences nutritional status

Studies have shown that the way socio-economic status and overweight are related depends on how urbanised and developed (economically) a country is. Many studies of women of child-bearing age in low-resource settings have shown that overweight and obesity is higher in the wealthiest women, there seems to be a similar pattern for men but there are fewer studies to conclude this from. However, it seems that as a country develops economically, overweight and obesity shift from being a problem of the wealthy to become a problem of the poor. This shift happens at an earlier point of economic development for women.

Our study of South African adolescents

The aim of our study was to investigate whether socio-economic status was linked to under-nutrition and overweight (including obesity) in city dwelling South African adolescents. We were interested in whether this differed between adolescent boys and girls so we could better understand how the nutrition transition was affecting this population. We think it is important to study adolescents as they will become the next generation of adults, and any risk factors for chronic diseases may continue when they become adults.

Our study was conducted in Johannesburg-Soweto, which is the largest urban area in South Africa with a population of around 4.3 million. South Africa is classed as an upper-middle-income country, but extremes of wealth and poverty exist, which explains partly why it is recognised as the most unequal society in the world. The sample of adolescents comes from an internationally renowned dataset called the Birth to Twenty Plus (Bt20+) cohort study. Bt20+ is a longitudinal cohort study (sample size of 3273) of births occurring between April-June 1990 to mothers who were permanent residents of Johannesburg-Soweto. The ambitious overall aim of the Bt20+ study is to “determine the biological, environmental, economic and psychosocial factors associated with the survival, health, well-being, growth and development of children living in an urban environment”.

For our study, adolescents in the cohort completed a questionnaire which assessed several things. Firstly, we were interested in classifying the socio-economic status of the households that children lived in, as accurately as we could. We did this by using different variables such as material living conditions (water/toilet facilities, and consumer durable ownership like car, fridge etc.) and the level of their caregiver’s education. We used international cut-off points to define nutritional status based on body mass index (thinness, normal weight, overweight and obese).


Birth to Twenty logo – The children enrolled in the cohort are known as Mandela’s Children as they were born in the seven weeks following Nelson Mandela’s release from prison on the 11th of February 1990.


What we found in our study

Underweight and overweight patterns differ between boys and girls

Our study revealed that weight status differs between boys and girls, as boys were twice as likely to be thin and girls were three times more likely to be overweight. The overweight: thinness ratio was 0.37 in males and 2.47 in females. This represents a clear shift from thinness to overweight in females and also highlights the existence of a dual burden of malnutrition within the area of Johannesburg/Soweto, with high prevalence of thinness in males and high prevalence of overweight in females. There is evidence that this ratio of overweight to thinness is even greater among adult women (Mendez et al. 2005). Our findings suggest that the nutrition transition starts before adulthood in South African females. Our earlier work with the cohort revealed little difference in mean Body Mass Indices between boys and girls aged 9/10 years.

The way that socioeconomic status influences underweight and overweight patterns is complex

The way that socio-economic status and weight are linked also showed clear differences between boys and girls, and the general pattern of association in our study is complex, with different trends observed for under- and over-nutrition. In boys, the prevalence of overweight (including obesity) was higher in those that were wealthier and in those with higher educated caregivers. We also saw a pattern of poor males being more likely to be thin. We observed a different dynamic in females as the prevalence of overweight and obesity was high regardless of the socio-economic level, probably because it is so widespread. We also found that poor females were less likely to be thin than wealthy ones, which is the opposite trend to that we observed in males. Our study overall implies that, in females adolescents, thinness is less prevalent in the poorest households and that the nutrition transition is more advanced than in males.



Our study sheds new light on the relationship between household socio-economic level and weight in adolescents living in this urban, transitioning society. We highlighted that even within a relatively small urban area, the nutrition transition manifests itself differently for boys and girls and in different socio-economic groups. It seems that in South African male adolescents, the shift we would expect to see of overweight spreading from wealthy to poor groups has not yet begun. The wealthiest male adolescents are more at risk of overweight and obesity and thus policy regarding promoting healthy dietary and physical activity patterns could focus on wealthy and well-educated households. In females, the results suggest that policy should target all female adolescents regardless of their household socio-economic level. The fact that there are different problems affecting boys and girls within the same area makes it more complex to develop appropriate health policies and public health services. Understanding the challenges for adolescent boys and girls at different ages is vital in helping to plan public health services.


More information

We presented the findings of our study in a recent joint publication in the journal of Public Health Nutrition. 

As part of a SIID funded project on changing urban diets in Africa, Rebecca will be analysing the Bt20+ data to explore dietary patterns in relation to socio-economic status. She is doing this in collaboration with Michelle Holdsworth (SIID Fellow, University of Sheffield); Paula Griffiths and Emily Rousham (Loughborough University); Alison Feeley and Shane Norris (University of the Witwatersrand).

No Comments have been posted yet.

Post a comment