Two years ago this month, Nutrition retracted a study by former Memorial University nutrition researcher Ranjit Chandra, but since then no other journals have retracted or issued warnings about the other papers alleged to contain fraud, if not outright fabrication. And some experts following the case are getting frustrated at the journals' decisions to let the data stand.
"It's ridiculous," Saul Sternberg, a psychologist at the University of Pennsylvania, told The Scientist. Sternberg, along with Seth Roberts from the University of California, Berkeley, published a paper in Nutrition Journal last November about suspicious patterns in Chandra's data. The papers are a "permanent piece of the scientific establishment. It's irresponsible on the part of the journal editors" to avoid taking action on questionable data, Sternberg noted.
Editors at the journals that published the accused papers told The Scientist they have taken the allegations against Chandra seriously, but so far...
Statistical reviewers for Smith said the 2001 paper had all the hallmarks of being fabricated. Since its publication, scientists have raised concerns about the Lancet paper, along with two replication studies (1, 2) Chandra and another researcher, Amrit Jain, published in 2002 in the journal Nutrition Research, which supported the Lancet findings. Chandra founded Nutrition Research and was editor at the time. A meta-analysis paper published in the BMJ in 2005 identified the three studies as outliers, compared with other similar studies on vitamins and immune response. Besides the retraction in 2005, none of the editors at other journals that published Chandra's data have issued warnings that his data may not be trustworthy.
Chandra retired from Memorial University in 2002 and has since worked in India and L'Université Internationale des Sciences de la Santé in Switzerland that he claimed to have established, according to Sternberg. Sternberg's research focuses on human experimental and mathematical psychology, but he said he decided to weigh in on the case against Chandra "for the sake of science." Chandra declined to comment on the situation, citing an ongoing court case in Canada.
"It's shameful that the scientific community doesn't have a way of responding to [the alleged misconduct]," former BMJ editor Richard Smith told The Scientist. "Most of the [investigatory] work has been done by mass media, and we're left not knowing how much of [Chandra's] stuff can be believed." Bruce Watkins, now editor of Nutrition Research, said he would be willing to consider a retraction if he could see Chandra's data, and has so far not seen any strong evidence the 2002 papers Chandra published in Nutrition Research were fraudulent. "These are charges that I have no evidence for. I've never seen the data," he said. "Without evidence or allowing Chandra to comment on his publication, it's like taking someone to court without facts and evidence."
While he admitted that the data in the 2002 papers was "weak," he said issuing a retraction or expression of concern would do little to correct the scientific record. Watkins said he had tried to contact Chandra in Switzerland but had never heard back. Astrid James, deputy editor of the Lancet, told The Scientist in an Email regarding Chandra's 1992 paper, "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."
James added that the editors also saw no grounds to issue an expression of concern regarding the suspicion over Chandra's work, even after Nutrition retracted a related paper. No one has publicly accused the highly-cited 1984 JAMA paper of containing suspicious data, but editor-in-chief Catherine DeAngelis told The Scientist in an Email that it would be virtually impossible to investigate a 23-year-old paper. Similarly, American Journal of Clinical Nutrition editor-in-chief Charles Halsted said he is aware of Chandra's "curious" past, but feels that re-examining AJCN Chandra papers would be fruitless.
"This issue has been investigated," he told The Scientist in an Email. "Anything further would be bringing up water under the bridge." Some journal editors simply aren't equipped to fully guard against misconduct in the literature, Pritpal S. Tamber, secretary of the World Association of Medical Editors told The Scientist. "There's no real training for this," and properly investigating claims of misconduct is "incredibly time consuming," expensive, and requires a certain expertise, said Tamber, who is also the managing director at the Faculty of 1000 Medicine, a sister company of The Scientist.
Indeed, the long, drawn-out process of uncovering fraud in science likely keeps most researchers quiet, David Klurfield, national program leader on human nutrition for the USDA, told The Scientist. "Chandra's gone away in retirement and most people are happy to have this go away," he said. "There's no collective stomach for a fight about this."
Correction (posted March 4): When originally posted, the story said Saul Sternberg and Seth Roberts co-authored a paper in Nutrition Research describing suspicious patterns in Chandra's data. That paper is published in Nutrition Journal. While correcting the error, other references to papers in Nutrition Research were mistakenly changed to Nutrition Journal. The references have been corrected, and The Scientist regrets the errors.