Retraction Watch

Tracking retractions as a window into the scientific process

Lancet retracts 24-year-old paper by “father of nutritional immunology” after reopening inquiry

with 9 comments

lancetFollowing questions from outside experts, a retraction of a related paper, a university investigation and a court case, The Lancet has decided to retract a 1992 paper by Ranjit Kumar Chandra, the self-proclaimed “father of nutritional immunology.

In a lengthy retraction note included in the January 30 issue, the journal explains that:

the balance of probabilities in our judgment is that the reliability of the 1992 Lancet paper by Chandra can no longer be assured.

Chandra is objecting to the retraction.

This retraction was a long time coming, so sit back and relax as we fill in the backstory.

For many years, Chandra, formerly of Memorial University of Newfoundland (MUN), was at the top of his field. But faith in his research began to falter when BMJ editors questioned one of his submissions, according to a recent news article in the journal:

Chandra’s career began to unravel in 2000 when he submitted a paper to The BMJ that seemed to show that his own patented vitamin supplement could improve memory in healthy elderly patients. The study, which followed up one published in the Lancet in 1992, immediately raised concern among BMJ editors, including the then editor in chief, Richard Smith, partly because the huge amount of psychometric testing involved seemed beyond the capabilities of a single author whose expertise was nutrition.

BMJ did not publish the study, which made its way into Nutrition in 2001. In 2005, the journal retracted it.

That Nutrition paper included the same subjects as the 1992 Lancet paper, which suggested supplements helped reduce the risk of infection in the elderly. It has been cited 390 times, according to Thomson Scientific’s Web of Knowledge.

In 2006, the Canadian Broadcasting Corporation aired a three-part documentary series which examined allegations of fraud against the former MUN scientist. Chandra sued the CBC, arguing the series constituted libel and invasion of privacy, and asking for more than $100 million.

Chandra lost his libel suit against the CBC in July, 2015. After the trial, MUN — where Chandra conducted the research included in the Lancet and Nutrition papers — released the report of a 2009 investigation conducted by  William Pryse-Phillips, a professor emeritus at MUN, which concluded Chandra had committed misconduct in the Nutrition paper.

When we covered the court decision in July, we asked the Lancet about the 1992 paper. A spokesperson for the journal said it had no plans to retract it:

That 1992 paper was the subject of further editorial and external review in 2003. We found no evidence of fraud, fabrication, or falsification of the data published in 1992, so no grounds for retraction. We are not aware of any new information, so there are no plans to further investigate, or to retract the paper.

That plan changed, according to the retraction note:

On Oct 6, 2015, we asked the Dean of Medicine of Memorial University, St John’s, Newfoundland and Labrador, Canada, to conduct a formal investigation into the research by Dr Ranjit Kumar Chandra that culminated in his publication in The Lancet in 1992. Chandra gives Memorial University as his address in the 1992 paper, and Memorial University is also acknowledged as supporting the research through a University Research Professorship Award.

We reopened our investigations into this paper because of documentation released into the public domain during a libel trial in Canada involving Chandra and the Canadian Broadcasting Corporation (CBC), which Chandra lost in July, 2015. Chandra has informed us that he is appealing against the judgment. The libel trial was about three programmes made by the CBC which raised concerns about Chandra’s research, including the 1992 paper. Previous concerns about the 1992 publication were raised by Kenneth J Carpenter, Seth Roberts, and Saul Sternberg as documented in Correspondence in The Lancet in 2003, with a reply by Chandra; and when Nutrition retracted a paper by Chandra in 2005, which involved the same subjects as the 1992 Lancet paper.

After consulting MUN and reviewing a copy of the Pryse-Phillips report, the journal determined it could no longer trust the data from the 1992 paper. According to the retraction note:

On Nov 27, 2015, Memorial University sent us a report by Dr William Pryse-Phillips, dated Oct 23, 2009, with supplementary comments dated Nov 3, 2015. The Pryse-Phillips report into a paper by Chandra that was submitted to the BMJ in 2000, published in Nutrition in 2001, and retracted in 2005, concluded that this paper “was not in full compliance with the scientific, ethical and/or integrity standards of Memorial University at the time”. On Dec 11, 2015, after further correspondence with Memorial University, we received a letter from Dr Gary Kachanoski, President and Vice-Chancellor of Memorial University, stating that it is “our view that the concerns documented by the Pryse-Phillips report in relation to ‘subjects and methods’ in combination with concerns identified by other commentators, provide confirmation that there are serious problems with the veracity of the 1992 Lancet publication”.

In view of the concerns raised, together with the conclusions drawn by Memorial University, which was Chandra’s institution in 1992, the balance of probabilities in our judgment is that the reliability of the 1992 Lancet paper by Chandra can no longer be assured. Chandra disputes these concerns and Memorial’s conclusions, and does not agree with the need to retract the 1992 paper. Nevertheless, we retract the 1992 Lancet paper from the scientific record.

A spokesperson for the Lancet declined to comment further on the retraction.

In October, The BMJ retracted Chandra’s 1989 paper about the role of breastfeeding and formula in infant eczema. That retraction cited yet another investigation by MUN that was completed in 1995, which the university did not send to the journal at the time. It came into the public domain as a result of the court case.

The Lancet retraction note does not cite that 1995 report, so it’s not clear if it played any role in the journal’s decision to pull the 1992 paper.

We should note that, in the 1990s, journals didn’t issue retractions as frequently as they do now. In 1995, there were roughly 35 retractions; in 2010, the number rose to at least 400. Now, of course, they keep us busy every day.

On a related note, these BMJ and Lancet retractions, 26 and 24 years in the making, would normally be contenders for record holders as the longest time to retract, but we recently discovered a retraction that occurred 80 years after the paper was published.

Like Retraction Watch? Consider making a tax-deductible contribution to support our growth. You can also follow us on Twitter, like us on Facebook, add us to your RSS reader, sign up on our homepage for an email every time there’s a new post, or subscribe to our new daily digest. Click here to review our Comments Policy.

    • Ivan Oransky January 29, 2016 at 9:54 pm

      Fixed, thanks.

  • JF Luc January 29, 2016 at 8:44 pm

    Interesting development… So I wonder if he will have to surrender his numerous awards, including the Order of Canada (received in 1989).

  • Richard Smith January 30, 2016 at 12:24 pm

    There are, I’m sure, many more retractions to come. Indeed, when Chandra was producing fraudulent work as far back as 1989 there must be severe doubts about all his work. Why once you had started inventing data would you start generating it for real?

    I wonder if there isn’t a case for retracting everything by him.

    As with so many other cases, many people either knew his work was fraudulent or strongly suspected so for a long time but nobody did anything. How can we move away from such a culture?

    • Dave Fernig January 30, 2016 at 12:36 pm

      PubPeer is a step in the right direction, though the all too common lack author engagement demonstrates that cultural change takes time.

  • MannyHM February 1, 2016 at 12:39 pm

    It should be taught in med school on how to spot fraud in research or clinical practice like detecting disease as part of pathology. Let me add – On how to deal with it if reported and ignored.

  • Post a comment

    Threaded commenting powered by interconnect/it code.