Bitter feud: Sweetener industry says review accusing it of bias is itself biased

Just one week after the sugar industry was left defending itself against accusations it tried to downplay the link between sugar and coronary heart disease for decades, the sweetener industry has found itself doing the same.

A review published in PLOS ONE journal, which analysed 31 studies into artificial sweeteners conducted between 1978 and 2014 found that none of the nine reviews performed by scientists without conflicts of interest reported results favourable to sweeteners, while four reviews with authors declaring conflicts of interest reported favourable results.

The researchers, led by Daniele Mandrioli of the Ramazzini Institute in Italy and Cristin E Kearns who also contributed to the sugar article published in JAMA, urged industry-sponsored studies to be greeted with a degree of scepticism as a result.

But the ISA challenged these conclusions in a statement today saying there were important biases and limitations in their systematic review, such as the small number of studies used, which raised serious concerns about the quality of the study and the accuracy of its outcomes. 

“It is based on a highly heterogeneous sample of studies including both reviews that summarised data quantitatively (with meta-analysis) as well as only qualitatively reviews,” it said.

“It is important to highlight that two out of the four industry-supported reviews include also a meta-analysis, and therefore the comparison conducted

in this systematic review between industry-supported and non-industry funded studies is simply biased.”

Co-author Dr Daniele Mandrioli denied this was the case. “In our review we complied with the highest standard of practice for systematic reviews, including PRISMA guidelines, and we adhered to Cochrane Collaboration standards for inclusion and exclusion,” he told this publication.

 A less-than-perfect system

According to Jack Winkler, emeritus professor of nutrition policy at London Metropolitan University, it is “entirely credible” that industry funded research produces biased results.  

“More than most, companies have a lot riding on research outcomes, not just reputations but lots of money. So, understandably, they want to produce supportive evidence - skew the research design, select supportive data, write biased interpretations, hide results they do not like.“

But the problems into go far beyond scientists that are on industry's payroll, he says.

“Much other research, without industry funding, is also tendentious. Academics are not angels either. Not least, some of them are for hire to companies to produce the results desired. Some have strong, genuinely-held beliefs about how nutrition works, built up over their careers, and so design their own research to support them, and denigrate contrary research. Others are opportunists, spotting hot topics and producing ‘controversial’ results to gain attention, or funding, or promotion.  Dissent can be a smart career move.”

Added to this are limitations of peer review journals – where academics review each other’s work leaving the system open to personal prejudice – and publication bias where journals give priority to strong, affirmative results while scientists with non-confirming results struggle to find a publisher, even though their findings may be just as relevant.

Interpreting the results of nutrition and public health science therefore becomes a minefield for consumers, who are confronted with sensationalist headlines from tabloid newspapers and serious news outlets reporting scientists’ back-and-forth flow of findings, says Winkler.

“It means that there is some truth in the widespread public complaint that ‘they can never make up their minds, so we can safely ignore nutrition science’. And this is reflected in the declining respect for scientists in public opinion polls.”

It also becomes difficult for governments and public bodies that draw up national dietary advice, resulting in countries with differing guidelines, he added.