By Marika Sboros
Did South African scientists really do something scientifically dodgy to silence their colleague, Prof Tim Noakes? Did they draw a UK scientist in to help them? Were their actions close to scientific fraud? The case against them appears to be building.
The US peer-reviewed journal PLoS ONE is now investigating the Naudé Review, published in July 2014. The Health Professions Council of SA used the review as key evidence to charge Noakes with unprofessional conduct.
The HPCSA found Noakes not guilty on all points of the charge on April 21, 2017. Thereafter, Noakes’s instructing attorney, Adam Pike, of Pike Law, wrote to PLoS ONE. Senior editor Dr Renee Hoch replied to say that the journal was “conducting a full reassessment” of the review.
All the universities involved have refused to investigate or “reassess” their academics’ role in it. UCT Faculty of Health Sciences deputy dean of research Prof Karen Sliwa said that only one out of the six researchers (Prof Marjanne Senekal) is from UCT. Four are from Stellenbosch University, one from the Liverpool School of Tropical Medicine. Thus, it would “seem preferable”, Sliwa said, for Noakes to complain to Stellenbosch.
Noakes says that’s like the Anglican Church saying it won’t investigate paedophilia claims against its priests because the Catholic Church has more claims against its priests.
The Naudé Review is a meta-analysis named after lead author and Stellenbosch nutrition academic Dr Celeste Naudé. The authors have concluded that low-carb diets are “no more effective for producing weight loss than are high-carbohydrate, or so-called isoenergetic, ‘balanced’ diets”.
They have also concluded that there is “probably little or no difference in weight loss and changes in cardiovascular risk factors up to two years of follow-up”.
As well as Senekal, Naudé’s co-authors include Prof Jimmy Volmink, Stellenbosch Dean of Faculty of Health Sciences and Liverpool School of Tropical Medicine professor Paul Garner.
All the Naudé reviewers have received funding from food and/or soft-drink industries for their research. US investigative journalist Russ Greene showed links between Senekal and others opposed to Noakes with the Coca-Cola front, International Life Sciences Institute. Senekal also a consultant to the HPCSA against Noakes, her colleague at the time in UCT’s Health Sciences faculty. She was also one of the four authors of what has become known as the UCT Academics Letter that a Cape Town newspaper published, attacking Noakes for irresponsible in his scientific opinions on nutrition.
Click here to read: Noakes vs ILSI ‘queenpins’ trying to nail him
However, in their declaration of competing interests, the Naudé researchers paint a very different COI portrait.They say: “No authors currently receive or have received funds from commercial organisations that could directly or indirectly benefit from the question addressed by this research or its findings”.
Why so many mistakes?
That’s debatable – at best.
Noakes is a UCT emeritus professor. He realised the pivotal role of the Naudé Review in the case against him during the November 2015 hearing session. He, therefore, asked British obesity and public health researcher Dr Zoë Harcombe to investigate.
Harcombe found the review to be ‘riddled with errors’, 14 of them major, enough to make the review fatally flawed, she says. She became an expert witness for Noakes at the October 2016 hearing session.
She and Noakes published their findings in the SAMJ in December 2016. They asked whether UCT and Stellenbosch scientists had made a mistake or mischief?
The authors’ first response was a media statement. They claimed that evidence at the October 2016 hearing session had addressed all the errors.
They also claimed that Harcombe had “conceded more than seven times” that the errors were not material to their findings. Both statements are not correct, as hearing transcripts show.
Senekal was present throughout the hearing. Thus, she would – or should – have conveyed that to her co-authors.
The review authors responded in a letter to the SAMJ in December 2016. Again, they claimed to have addressed all errors but did not, in fact, do so. Instead, and remarkably, they pointed Noakes and Harcombe to new errors.
They also claimed that Noakes and Harcombe lacked “understanding of current methods in evidence synthesis”.
In a rebuttal letter to the SAMJ in April 2017, Noakes and Harcombe described that as a “cheap” shot. “We understand the Naudé et al protocol only too well,” they wrote. “Indeed, we appear to understand it rather better than do its authors.”
Instead of addressing the numerous errors in detail, the Naudé Review authors admitted to only one error – the inclusion of a duplicate study. That alone is grounds for retraction of the article, Noakes and Harcombe said.
The researchers also admitted to using data from ITT ( intention-to-treat) analyses instead of completers for some studies. As well, they only reported analysing the data that way once Noakes and Harcombe forced a response to the errors.
As Noakes explained it to me: “When researchers set up a trial, they may follow the subjects during the trial and then report what they find. For example, they may start with 100 patients, of which 80 comply with the trial and do everything they’re expected to do. Thereafter, the researchers report the data for those 80. Or, if they do an ITT, they don’t bother to check if everyone complied. They just report the outcomes.”
Click here to read: ADSA desperately seeking to nail Noakes?
In analysing the data as ITT in some trials only, the Naudé researchers again favoured the low-fat group, Noakes said.
Harcombe said that ITT might be more ethical and beneficial in drug trials. “It captures the important issue of non-compliance in ways that completers don’t,” she said.
Noakes has often speculated on whether the Naudé Review is dodgy enough possibly to constitute scientific fraud. He and Harcombe have also suggested that were it not for the Naudé Review the HPCSA hearing might not have happened.
For example, in the SAMJ, they referred to the “disproportionate prominence” the HPCSA gave to the review. They quote verbatim from the testimony of HPCSA’s witnesses, starting with dietitian Claire Julsing-Strydom.
Strydom referred to the importance of a meta-analysis generally and/or the Naudé meta-analysis specifically on 12 separate occasions during her testimony. Also under oath on November 24, 2015, she said of the Naudé Review: “So before any media statements could be made we had to get that information and all these associations were waiting on that…It is not like everybody joined together to now make a statement against Prof Noakes.
“We were all waiting for the evidence to be published.”
Another prosecution witness, North-West University Prof Hester “Este” Vorster, referred to the review five times and quoted verbatim from it once. A third prosecution witness, paediatric professor Ali Dhansay, referenced it twice. He also used the term “Cochrane” when referencing. That would have afforded the review “the appropriate esteem”, Noakes and Harcombe said.
Given all that, it’s not surprising that Noakes and Harcombe see the review as “dogdgy”. It is relevant that the Heart and Stroke Foundation of SA commissioned the research when there was already a robust Brazilian meta-analysis on the same topic. That review came to the opposite conclusion from the Naudé reviewers.
Why repeat robust research?
That raised the question, says Noakes, why the HSFSA wanted Naudé to duplicate robust research. It’s also common knowledge, he says, that the HSFSA actively promotes low-fat, high-carb diets and opposes his opinions on LCHF.
The SA Medical Research Council also funded the Naudé review. Dhansay is an MRC staffer. In his evidence, he made clear his opposition to Noakes and LCHF. Greene reported that the MRC is investigating Dhansay’s ILSI links.
The responses of UCT and Stellenbosch Universities to concerns over the Naudé review are significant.
Noakes wrote to UCT pharmacology professor Marc Blockman to request an investigation. Blockman is chair of the Faculty of Health Sciences Human Research and Ethics Committee.
He said that in the Committee’s opinion, “potential issues” related to the Naudé review were not “issues of non-compliance”. Somplaints of research misconduct also fell “outside the HREC’s mandate”. Blockman, therefore, passed the buck to Sliwa.
However, UCT’s own rules for claims of scientific misconduct require Sliwa to form an interim committee to consider what action the university should take. Thus, in declining to investigate, Sliwa appears to be in conflict with UCT guidelines.
She would also presumably have had to involve cardiology professor Bongani Mayosi in the case of such serious allegations against a UCT academic. Sliwa and Mayosi did not reply to questions via email. Nor have UCT’s vice-chancellor Dr Max Price and the university’s management at the highest levels responded.
On the UCT Academics’ letter and involvement of UCT’s marketing and communications chief, Linda Rhoda, registrar Royston Pillay responded with a broad brushstroke. “UCT does not intend responding to all your queries or allegations in your e-mail,” Pillay said. “Failure to do so should not be construed as an admission of the allegations contained therein.”
An appeal to UCT’s head of Council Sipho Pityana elicited similar closing of the ranks. He dismissed it all as “robust and rigorous scholarly engagement” that Council “is not equipped to pronounce on”. UCT chancellor Graca Machel also ignored emailed requests for comment.
Garner and Liverpool School of Tropical Medicine researcher Prof Paul Garner were likewise reticent. Through its media department, the LSTM said that it “fully supports scientific rigour and the peer review process”.
The LSTM also oddly claimed that PLoS ONE was considering “minor corrections” only to the Naudé Review. It refused to clarify why it believed that, despite the PLoS ONE announcement.
In author Daryl Ilbury’s book on Noakes, The Quiet Maverick, Naude claimed “surprise” at the HPCSA’s prominent use of her review. After all, it was on adult, not infant, nutrition, she said.
Noakes, his lawyers and expert witnesses were just as surprised. After all, the HPCSA insisted throughout that its case was on infant, not adult, nutrition. Yet Strydom and all the HPCSA’s nutrition “experts” gave length testimony on adult nutrition.
In early May 2017, the HPCSA announced that it will appeal the not guilty verdict for Noakes a month earlier. Sixty days later it has yet to give its grounds. The Naudé Review may well prove one of its many scientific Achilles heels in its ongoing quest to find Noakes guilty.
- Click here for full coverage of The Noakes Trial
- Follow me on Twitter @MarikaSboros
- Like my Facebook Page
- Subscribe for email notifications of Foodmed.net postings
- I am co-authoring a book on the trial with Prof Tim Noakes. Penguin Random House will publish in November.