Sunday, May 11, 2014

The Australian Paradox: Part 2

In the last post, I showed how in the Australian Paradox, the authors own data suggests it more likely that sugar and SSB consumption has increased and that the only way to conclude there has been a “consistent and substantial decline in total refined or added sugar consumption by Australians over the past 30 years” is to look at the (faulty) FAO data and ignore all the other results. 

RCTs > Ecological Studies 

Once the authors looked the FAO data and concluded the “consistent and substantial decline”, they turned their attention towards the implication of their ‘results’.  Essentially that reducing sugar or taxing SSBs may not reduce obesity: 

“The implication is that efforts to reduce sugar intake may reduce consumption but may not reduce the prevalence of obesity” 

“The findings challenge the implicit assumption that taxes and other measures to reduce intake of soft drinks will be an effective strategy in global efforts to reduce obesity” 

The problem is that even if there was the “consistent and substantial decline” the authors would be overstating their ‘results’.  The Australian Paradox is an ecological study, a type of observational study that looks at populations rather than individuals, and ecological studies shouldn’t be used to make inferences about the biological effects on individuals [1] (doing so is referred to as the ecological fallacy), a point which is made briefly by the authors 

“A limitation common to all ecological studies is that relationships observed for groups do not necessarily hold for individuals” 

The Australian Paradox can’t determine whether sugar and/or SSBs are obesogenic or whether taxes will reduce obesity.  RCTs will answer the first and other studies (most likely case studies) will inform the second 

Fortunately, at least for SSBs, there’s a number of RCTs investigating their relationship with weight.  Prior to 2011 (the Australian Paradox was published in 2011), there were six systematic reviews without a conflict of interest with the food industry, which examined the relationship between SSB and excess weight gain in observational studies and/or RCTs. Five out of those six systematic reviews found a positive association [2] 

However, in the discussion when the authors compared their findings to similar research (ecological, other observational and RCTs) they tended to cite studies that found a negative or no association between sugar/SSBs and excess weight gain, and only one study that found a positive association 

Who are the Authors? 

Alan Barclay, PhD (AWB) is an Accredited Practicing Dietician, the Head of Research at the Australian Diabetes Council and did his PhD thesis on the relationship between glycemic index (GI) and chronic disease (Australian Diabetes Council 2013). Jennie Brand-Miller, PhD (JBM) is a Professor of Biochemistry in the School of Molecular Bioscience at Sydney University and her research interests include the role of carbohydrates and GI on health

The authors disclosed some conflicts of interest: AWB is a co-author of one of the books in The New Glucose Revolution (NGR) book series and the chief scientific officer of the Glycemic Index Foundation (GIF); JBM is a co-author of the NGR book series and the director of the GIF 

What isn’t mentioned with these conflicts of interest is that: (1) The NRG book series makes claims of high GI bad, low GI good, sugar ok; and statements like: “There is an absolute consensus that sugar in food does not cause diabetes” [3] [4].  An absolute consensus means 100%.  You won’t need to look too hard to prove this statement wrong.  (2) The GIF promotes low GI whole foods (like fruit), lower GI alternatives (lower GI bread and potatoes), but also lower GI sugar and junk food [5] 

The authors did not disclose that some members of the food industry are corporate governors of the nutrition research foundation at Sydney University’s School of Molecular Bioscience (where JBM works) [6] 

Also, JBM was the guest editor of a special issue of Nutrients (the journal, only started in 2009) called Carbohydrates, which was the issue that the Australian Paradox was published in.  With all that’s been said, it’s likely the authors choose to publish in Nutrients, with JBM as guest editor, because under normal circumstances the paper would likely have not passed peer-review 

* I was fortunate (being familiar with Rory Robertson and Background Briefing) that the Australian Paradox was one of the papers we could use for an assignment called ‘critical analysis of the literature’, hence the subject of the blog posts.  From my experience, the papers they select for these types of assignments tend to be poorer quality.  The assessor liked the assignment, noted a few other issues with the paper that I hadn’t discussed and wondered whether the paper had previously gone under review at a different journal.  I discussed the paper with a few people and they had similar views on it.  If the Australian Paradox was presented in an academic setting I wonder what the audience response would be.  I certainly wouldn’t want to be the one presenting

No comments:

Post a Comment