In fact, the only thing the financial markets economist has quit is sugar, something he credits with allowing him to lose a significant amount of weight.

And now, after nearly two years of campaigning from Robertson, the author of a key study used by the soft drink industry to contend that sugar is harmless has admitted it might be flawed.

Professor Jennie Brand-Miller made the admission in a Radio National Background Briefing program broadcast Sunday morning.

Known as 'GI Jennie' for her work developing the glycemic index and the so-called low GI diet Professor Brand-Miller is the co-author of a paper entitled The Australian Paradox that argues that per capita sugar consumption has been sliding during the three decades in which obesity has been climbing.

The Australian Beverages Council has used the data to support it's argument that health authorities trying to reduce obesity rates should not target Australians' consumption of sugar-sweetened drinks.

Robertson has maintained that the data used to support the argument is flawed, and has waged an ongoing campaign to have it corrected.

He is even offering $40,000 of his own money to anyone who can show there really has been a consistent and substantial decline in total refined or added sugar consumption over the past 30 years as claimed by the paper.

Sydney University has responded to complaints about the paper by appointing an external investigator.

In the program, Professor Brand-Miller continues to maintain that Australia “is unique”.

“Australia is actually bucking the trend with respect to added sugars there is good evidence that we are not increasing our intake,” she says.

But when told one of the graphs in her paper belied her claim that “per capita sales of sugar-sweetened beverages have decreased by 10 per cent” she says she might have meant to refer to market share.

“It might be that a key word came out, which is normally... a key word has come out,” she tells the program.

Robertson tells the program that he might not be a nutritionist but that he has spent his “whole professional career” examining data and charts.

The paper relies on Food and Agriculture Organisation figures which show total consumption of sugar fat since 1999. Robertson says this is the year the Australian Bureau of Statistics stopped publishing sugar consumption data because it was unreliable.

He says the FAO simply published a flat line after that because they used an algorithm which republished the last available number.

In the program Professor Brand-Miller said she had assumed that the FAO obtained its information from other sources.

The Beverages Council cites the paper on its website and praises it in the program saying its findings imply “efforts to reduce sugar intake may not reduce the prevalence of obesity”.