ANALYSIS: Is SA the 'fattest nation' in sub-Saharan Africa?
Researched by Kate Wilkinson
South Africans' weight made headlines last month. " SA is fattest nation in sub-Saharan Africa," declared Business Day online. The Port Elizabeth Herald told readers: " One in three South Africans now deemed obese". The Times' print edition went with the snappier " SA pigs out in fat ranking".
The newspapers were quoting the CEO of South African medical aid company Profmed, Graham Anderson. According to Business Day online, he said: "One in three South Africans are now deemed obese‚ making the country's population the most obese nation in sub-Saharan Africa".
Is this true? Do South Africans tip the scales as the heaviest nation in sub-Saharan Africa?
A quarter - not a third - is obese
Profmed's public relations company, Epic Communications, sent us the press release they issued on behalf of Anderson. It references and links to the World Health Organisation's (WHO) 2014 Global Status Report on Noncommunicable Diseases. This report presents progress made around the world in preventing and controlling noncommunicable diseases, such as harmful use of alcohol and tobacco, high blood pressure, diabetes and obesity.
However, the report doesn't show that 31.3% of South Africans are obese. This was the WHO's 2008 estimate.
The WHO's ' crude estimate' for the number of South Africans who were obese in 2014 is 25.6%. So-called 'crude estimates' are based on local studies and surveys, the team leader in the WHO's prevention of noncommunicable diseases department, Leanne Riley, told Africa Check.
These included the South African National Income Dynamics Studies, South Africa's Demographic and Health Surveys and the WHO's study on global ageing and adult health. The aggregated data was then adjusted to be correct for the year of reporting and representative of the population.
So is South Africa the 'fattest'?
But countries' crude estimates should not be compared because the age and sex breakdown of populations differ. Comparing the incidence of obesity in a country with a very young population against that in a country with an older population would be misleading, for example.
To allow for better comparison, the crude obesity estimates are weighted by the WHO to reflect the "age and sex structure of most low and middle-income countries". This removes any population differences that could skew comparisons.
Their weighted estimate suggests that 26.8% of South Africans were obese in 2014. This is the figure that should be used when comparing South Africa with other countries.
With an average estimate of 26.8%, South Africa topped the WHO's rankings. It was followed by Seychelles, where 26.3% of the population are considered obese.
Country % obese (with 95% confidence intervals)
South Africa 26.8 [22.8-31.3]
Seychelles 26.3 [20.5-31.7]
Botswana 22.4 [17.3-27.5]
Namibia 18.9 [14.4-23.5]
Mauritius 17.9 [12.9-22.5]
Ranking not that straightforward
A senior researcher at Stellenbosch University's Centre for Evidence-based Health Care, Dr Celeste Naude, cautioned Africa Check that 'statistical confidence intervals' need to be taken into account when making comparisons between averages.
The WHO's confidence interval for South Africa ranges between 22.8% and 31.3%. This means that while their best average estimate is that 26.8% of South Africans are obese, they are 95% sure that the figure falls somewhere between 22.8% and 31.3%. The WHO provides this range because sampling errors and the statistical model used can introduce uncertainty.
The confidence intervals of the top four countries (South Africa, Seychelles, Botswana and Namibia) overlap.
"The fact that South Africa's confidence intervals overlap considerably with Seychelles in this WHO report means that it's highly likely that there is no difference between the estimates for the two countries and, strictly speaking, one cannot make the statement that South Africa is worse off than Seychelles," Naude said. The WHO's Riley confirmed this.
Likewise, when comparing South Africa with Botswana and Namibia, it must be noted that their statistical confidence intervals overlap to varying degrees.
Conclusion: 1 in 4 South Africans obese, country tops average sub-Sahara African ranking
The report that appeared in newspapers was based on a press release. The WHO report this was based on estimated that roughly a quarter of South Africans were obese in 2014, not 1 in 3 as the press release claimed. That was the situation in 2008.
While the situation today has improved, 1 in 4 is nevertheless a worrying statistic. According to the WHO obese people are more likely to suffer from diabetes, hypertension, coronary heart disease, stroke and certain types of cancer.
When it comes to declaring South Africa the 'fattest nation' in sub-Saharan Africa, care must also be taken. While South Africa tops the average rankings, reporting should take into account uncertainty about the estimates. This qualification doesn't fit neatly in a news headline, but it is more accurate.
Press releases issued on behalf of companies are frequently drafted entirely by public relations firms. When Profmed's PR company released comments they said were made by the medical aid's CEO, the claims weren't checked for accuracy.
The key lesson? Always check the original source.