In my previous blog (1) I explained the background to the allegation against Jennie Brand-Miller and Alan Barclay by Rory Robertson. Essentially he argued that the so-called Australian Paradox which claimed that the increased incidence of obesity was not related to an increased consumption of sugar. Robertson concluded that the information showing that the sugar intake had fallen, over the period when obesity had increased was fundamentally flawed.
The University of Sydney commissioned Professor Robert Clark, Chair of Energy Strategy and Policy at the University of New South Wales and a former Chief Defence Scientist of Australia to conduct an Inquiry. He had just delivered his report, which is available on the internet although parts of it have been redacted (2).
The key statement in the Australian Paradox paper is that:
“This analysis of apparent consumption, national dietary surveys and food industry data indicates a consistent and substantial decline in total refined and added sugar consumption by Australians over the past 30 years”(3).
Robert Clark dismissed 6 of the 7 allegations which Rory Robertson had made. On the face of it, this might appear to be a satisfactory outcome as far as the academics were concerned. The only allegation which was substantiated referred to the decline in sugar intake from sugar sweetened soft drinks between 2002 and 2006. In the paper Brand-Miller and Barclay has used a value of 600g per person. This figure was wrong. It should have been only 150g.
Nevertheless a close reading of the report shows that all in the garden is not rosy and that there are many lessons to be learned. In particular Clark was not impressed by the quality of the original “Australian Paradox” paper. He concluded that the paper was not tightly written and contained a number of arithmetic errors. He expressed the view that important new findings would usually be published in a high-impact, rigorously peer-reviewed journal. Subsequently the results would be considered in special edition publications of conference journal format. Clearly this had not been done as the original was published in what was regarded as a ‘soft’ journal, where the quality controls are less stringent. As a consequence it was recommended that a new one be prepared which specifically addresses the key factual issues raised in the inquiry. Furthermore this paper should be written in a constructive manner which respects the issues relating to the quality and reliability of data, raised by Rory Robertson.
The crux of the dispute is the conclusion in the Australian Paradox Paper which claims that there has been a decline in the consumption of the sugar from 1980 to 2003 during which time there has been a substantial increase in the incidence of obesity. This is fundamental to the Paradox, which is that the obesity is not caused by sugar, a phenomenon which contrasts with that of many other countries, including the USA and the UK.
However as Robertson has pointed out repeatedly the sources used by Brand-Miller and Barclay are subject to many criticisms. This raises serious questions about the reliability of the data and therefore casts doubt on the conclusions drawn.
The body providing the relevant information is the Australian Bureau of Statistics (ABS). Clark was provided with data from the ABS which showed that there had been a decline in the per capita consumption of sugar between 1975/6 and 1998/9. So the critical issue is to determine the reliability of these data. In correspondence between the ABS and Clark it is accepted that some errors may not have been taken into consideration. More specifically, in order to compute a value for the amount of sugar in the food supply it is essential to have information on all the different sources of sugar. These include all the different foods, especially those which are manufactured, which contain sugar. So in order to work out the total supply, the sugar content of all these foods must be known. This is extremely complicated because there must be variability within products and from time to time. Product formulation is continually changing. A good example is the growth in products which are promoted as “low fat”. Very often these are devised by removing the fat and replacing it with sugar. After 1998/9 the ABS discontinued publication of the figures. One of the reasons was that the ABS believed these conversion values were no longer appropriate. The inevitable question, of course, is “Were they ever accurate and reliable?” On top of all this, all the evidence amassed by Robertson, an experienced economist well used to number-crunching, shows that the trend is UP not DOWN
In view of the absolutely critical importance of the conclusions inherent in the Australian Paradox, it would have been incumbent on the authors to provide a cast-iron case. Clearly they have not done so and it is regrettable that Clark did not reach that conclusion. Furthermore he made no attempt to place this issue in the context of the growing appreciation of the role of sugar in the development of a range of different diseases. Of much greater concern than obesity is Type 2 diabetes (T2D) which in Australia has increased more than 3-fold in the past 30 years or so (4).
It is also becoming clear that sugar intake is a key factor leading to the development of heart disease, many cancers and Alzheimer’s Disease. The evidence continues to accumulate (5). The presence of fructose may be especially damaging to health (6).Bodies such as the World Health Organisation and the American Heart Association now accept that current levels of sugar consumption are not compatible with good health. Hence official recommendations are being reduced significantly.
Despite the somewhat unequivocal findings of Robert Clark there is no doubt that Rory Robertson’s campaign has had a genuine impact and that the credibility of the Australian Paradox has been seriously undermined as shown by some of the local media coverage (7).
- A W Barclay & J Brand-Miller(2011) “The Australian Paradox: A Substantial Decline in Sugar Intake over the Same Timeframe that Overweight and Obesity Have Increased” 3. pp491-504