A 2001 paper following up on a study by a prominent Canadian scientist showing striking cognitive benefits for people over 65 who took a daily multivitamin mineral supplement—which has been cited more than 350 times since its publication in 1992—has become the center of an increasingly vituperative exchange of claims and counterclaims about research data.

Ranjit Kumar Chandra—who worked for 27 years as a nutritionist at Canada's Memorial University and the Janeway Child Health Center in St. John's, Newfoundland, and Labrador before retiring in August 2002—published the original study in the Lancet. The British Medical Journal rejected a follow-up to the Lancet study submitted in late 2000. It was accepted by Nutrition, where it was published in September 2001.

The BMJ's editor, Richard Smith, then wrote the editor of Nutrition, Michael Meguid, to inform him of the BMJ's rejection and that the BMJ's unidentified reviewer had concluded that the data had “all the hallmarks of being entirely invented.” Smith also wrote to Memorial University president Axel Meisen asking him to investigate if there was any wrongdoing on Chandra's part.

An editorial and exchange of commentaries eventually appeared in Nutrition in which Meguid and his European editor Alan Shenkin said there were serious statistical flaws in Chandra's paper and expressed regret “that our peer review process failed to identify these problems before publication.” Critics have called for Chandra to produce his data.

“After Richard Smith wrote to us,” Jack Strawbridge, the director of faculty relations at Memorial University, told The Scientist, “we asked Chandra for the raw data. He never, ever, gave us a thing.”

Chandra, however, by this time had resigned, which he had said he would do as he was approaching mandatory retirement age. Chandra told The Scientist this week that he stands by his work.

Strawbridge told The Scientist that “what is most vexatious to us is the claim by Smith that Memorial and only Memorial has an obligation to investigate all of his work—even that which is not under dispute. We say not. We are not in the business of vetting research that goes forward to a publication. It is up to the journals and the peer review process to say whether something is worthy of being published” and that have the ability to attach queries, comments, or even withdrawals against citations, “not us.”

Memorial had also investigated Chandra in 1994 after a research nurse complained about a study on the effects of maternal diet during lactation on allergy in infants. She said that the number of mothers studied was less than what had been reported. But Chandra said that “the investigation concluded… that there was no wrongdoing on my part and the file was closed.”

Stawbridge—who still calls Chandra “a good clinician”—admitted the 1994 committee investigating Chandra made some procedural errors, resulting in “a draw.” Chandra was, however, given “a very explicit letter” in which he was “told to keep data after he'd published it,” Strawbridge said.

Chandra, who has published more than 140 papers, said his critics should simply replicate his work on the study under question. “They should not hide behind the argument that this will waste time and money,” he told The Scientist.

Chandra holds a patent on the multivitamin mineral supplement described in his papers. The rights to it were licensed to a company founded by his daughter, the Javaan Corporation, in Somerville, Mass., which features several of his papers to support their products. However, Chandra told The Scientist, he has not received any royalty from sales. “When that does happen,” he said, “the marketing company has been given the authority to donate my royalty to any educational institution of their choice without further permission from me.”

Meguid is on jury duty this week and could not be reached for comment.