Science writer Gary Taubes has been writing columns everywhere promoting his new book The Case Against Sugar. He has written columns in The Guardian, and The New York TImes among other places, and has been reviewed somewhat critically in The Guardian, Food Insight and The Atlantic.
Taubes’ central argument is that calories from sugar are not the only reason for obesity, but argues that sugar itself is uniquely toxic.
Taubes: “If the research community had been doing its job and not assuming since the 1920s that a calorie is a calorie, perhaps we would have found such evidence long ago.”
In a nutshell, the flaw in his argument is revealed in the above statement in the Times article. There must be more to sugar’s causing obesity than just calories, but researchers haven’t been doing their job!
And, in fact, despite Taubes’ persuasive writing, this is most of his argument. He cites no research in his articles (I have not read his actual book) or even mentions researchers who agree with him.
His thesis echoes that of Dr Robert Lustig, who makes much the same arguments in his book Fat Chance, and in the movie Fed Up but both Lustig’s and Taubes’ similar ideas have been debunked in articles, such as this one in Science Based Medicine. And Food Insight called this “blind fealty to correlation as causation.” Scientific American pointed out the fallacies in this argument in 2013.
In fact, while obesity continues to increase, sugar intake in the US actually decreased from 1999-2008, mainly because of decreased consumption of sugary soft drinks.
Taubes’ other somewhat distressing argument is that the sugar industry has been influencing research outcomes for years by sponsoring research. This suggests that not only that scientists are unethical but that the journal peer-review process itself is corrupt, and that is hard to swallow. The idea that research funding influences outcomes had been thoroughly debunked in this article by van Eenenaam, who notes that such corrupt research is a sure path to a short academic career.
He cites this PLoS One paper which reviews papers for their findings, correlating them with the source of their support. The authors suggested that papers with no declared “conflict of interest” are more likely(83%) to find that sugar sweetened beverages could be a risk for weight gain, but for those “disclosing some financial conflict of interest” 83% found that there was no such correlation.
The trouble with that paper is that there are only a few such studies: there were only 12 in the first category and 6 in the second category, and only 10/12 and 5/6 supported the author’s conclusions.
There are other reviews of sugar consumption that we need to consider. For example, Weed et al. studied reviews of health outcomes from sugar sweetene beverage (SSB) consumption, and rated the review quality using the AMSTAR review rating scheme, and found that most of them received moderately low quality scores, regardless of the conclusions of the paper. This would mean that the conclusions of these reviews are probably not entirely convincing, and basing Taubes’ sugar conspiracy theory on such weak data is not fully substantiated.
Moreover, this recent paper by Keller et al. reviews papers on sugar sweetened beverage consumption among children and adolescents, reporting that 9 reviews found a correlation between obesity and SSB consumption, while 4 did not. But that the quality scores of the reviews was low to moderate and that the two papers with highest quality scores reported discrepant (inconclusive) results.
The most important conclusion we can draw from reading Taubes’ many opinion pieces is we eat too much sugar, but that studies so far have not shown that sugar is more to blame than calories from any other source. No such research seems yet to exist.
Last night, 
Food Science Institute 