Experts struggling to explain a new study that finds little harm from saturated fat in the diet have found another clue: Errors were discovered in the new paper.
"A new version of the publication had to be posted shortly after it appeared on the website of the Annals of Internal Medicine to correct several errors. And although the study's first author stands by the conclusions, a number of scientists are criticizing the paper and even calling on the authors to retract it," writes Kai Kupferschmidt at Science.
Harvard's Walter Willett, whose decades of research was flatly contradicted by the new study, told Kupferschmidt that the study's authors "have done a huge amount of damage. I think a retraction with similar press promotion should be considered."
In a post on this controversy yesterday, I wrote that "nobody is quite sure how to reconcile the old advice with the new finding."
The errors might make it a good deal easier to sort out the confusion.
The study, led by Rajiv Chowdhury of the University of Cambridge, was a meta-analysis of 72 individual studies. Such analyses are intended to combine data from separate studies to reveal insights that can't be gleaned from the individual studies themselves. The legitimacy of meta-analyses depends upon the care with which the data is combined and the validity of the assumptions made during the analysis of the pooled data.
Kupferschmidt writes that questions about the paper arose even before it was published. "For instance, the authors took one study on omega-3 fats, one type of unsaturated fats, to show a slightly negative effect while, in fact, it had shown a strong positive effect. The correction means that the meta-analysis now says people who report eating lots of this particular fat have significantly less heart disease; previously, it said there was no significant effect."
If that's true, it explains a lot.
"Critics also pointed out two important studies on omega-6 fatty acids that the authors had missed," Kupferschmidt continues. He quotes an email from Jim Mann of the University of Otago in New Zealand who said the errors "demonstrate shoddy research and make one wonder whether there are more that haven't been detected…If I had been the referee I would have recommended rejection" of the paper.
Chowdhury says the new study's conclusions are still valid, despite the corrections. One of the paper's authors, Dariush Mozaffarian, said he was not happy with the conclusion that there was no evidence for the benefit of polyunsaturated fats in the diet, but he couldn't persuade his co-authors of that. "Science isn't a dictatorship," he said.
Right. But neither is it supposed to bury minority views.
Kupferschmidt gives us a very clear and well reported analysis of the problems with the new study. The discovery of the errors by no means invalidates the new findings, but it does make us think. So does Kupferschmidt.