Researcher admits faking data

'Egregiousness' of conduct over almost 15 years leads to first-ever lifetime ban on US grants

Mar 21, 2005
Doug Payne(dougpayne@islandtelecom.com)

A well known obesity researcher with more than 200 articles to his name will plead guilty to making material false statements in a 1999 grant application worth $542,000 from the US National Institutes of Health, according to a plea bargain announced Thursday (March 17) in Vermont. Eric Poehlman said he fabricated data in 17 applications for US federal grants and agreed to be barred for life "from seeking or receiving funding from any federal agency in the future, including all components of the Public Health Service."

Poehlman is expected to be arraigned either this week or next. The criminal charge—an unusual step for investigations of scientists—of fraud to which he agreed to plead guilty could see the researcher go to jail for up to 5 years.

However, it is possible he will serve little or no jail time. Stephen Kelly, a prosecutor in the US Attorney's office in Vermont, told The Scientist that his office's policy has been to ensure "that the scientific record is as accurate as possible." In addition, he said, "we do feel, and will report to the court, that Dr. Poehlman has been cooperative and will try to redress the serious wrongs he has committed." According to a joint release from the Office of Research Integrity and the Justice Department in Vermont, Poehlman also has "to submit numerous letters of retraction and correction to scientific journals related to his scientific misconduct."

Poehlman held various research positions at the University of Vermont (UVM) College of Medicine in Burlington from 1987 to 1993 and again from 1996 to 2001. He was at the University of Maryland in Baltimore between 1993 and 1996. He conducted research related to exercise physiology and metabolism funded primarily by grants from federal public health agencies and departments.

UVM began to investigate Poehlman's work after one of his research assistants, Walter DeNino, questioned study data. DeNino became suspicious in the fall of 2000 after being asked to analyze data in a study on aging in Vermont. He subsequently reported Poehlman to the university, which began an investigation in December 2000 that ran until April 2002. Poehlman resigned his tenured professorship in 2001, taking a Canada Research Chairs endowed chair worth CDN $1 million (USD $825,000) over 5 years, at the Université de Montréal, which knew nothing about the UVM investigation and was eager to highlight his work.

The Research Chairs' acting director of operations, Julie Dompierre, told The Scientist that the agency "was not aware of any concerns with the Poehlman file when it was [first] submitted to the secretariat by the Université de Montréal." Poehlman resigned from that post on January 22, the Université de Montréal told The Scientist, adding that "privacy laws prevent the Université de Montréal from divulging the reasons for his resignation."

The list of the papers affected by Poehlman's misconduct includes publications in 10 journals. Among the best known studies cast into doubt is Poehlman's longitudinal menopause study, published in the Annals of Internal Medicine in 1995, in which he claimed that he had tested 35 healthy women for the effects of menopause on resting metabolic rate and other variables and that he had retested them six years later. He fabricated the results for 32 of the women. As a result of the university investigations, the article was retracted in 2003.

Fran Carr, the vice president for research at University of Vermont, characterized Poehlman's actions as "betraying the trust of his colleagues, his subjects, the university, and the entire scientific community." When asked whether the university would change anything in its overview process as a result of what she called "an incident of grave importance," she said no.

"The university clearly has a system in place to respond to allegations of scientific misconduct, and I think that the process and the people that were involved in the entire investigation are validated," Carr said, adding, "We carried through to the end, in spite of the fact that the person involved had already left the institution."

In the meantime, the university said the basic patient information originally obtained by Poehlman is still in their files, and it is likely the results will be published in peer-reviewed journals in due course. The college is also setting up a hotline for volunteers who took part in the Poehlman studies.

Poehlman's lawyer, Robert Hemley, declined comment, and Poehlman could not be reached for comment.