Third retraction for Indiana University scientist who altered figures in NIH-funded research

Another shoe has dropped in the case of Emily M. Horvath, the Indiana researcher whose tinkering with figures while on a $369,000 federal grant ended in sanctions by government officials.

Biochemical and Biophysical Research Communications, an Elsevier title, has retracted another paper on which Horvath was an author, bringing to three the number of her articles tainted in the scandal. The paper, “A novel membrane-based anti-diabetic action of atorvastatin,” was published online in June 2008, and cited four times since, according to Thomson Scientific’s Web of Knowledge. (Atorvastatin is sold as Lipitor.) According to the journal:

The editors would like to confirm the retraction of this paper at the request of the authors. One of the authors, Emily M. Horvath, admitted to altering data in two of the figures. Specifically in Figs. 2 and 4A, some of the values were incorrectly reported for the radioactive glucose uptake assays. None of the other authors associated with this publication were aware of the data manipulation. The authors sincerely regret if this has caused problems with

Earlier this year, The Scientist reported that the Office of Research Integrity (ORI) announced that it had reached a settlement with Horvath, a former biochemist at Indiana University. According to ORI, a division of the Department of Health and Human Services, Horvath

admitted to falsifying the original research data when entering values into computer programs for statistical analysis with the goal of reducing the magnitude of errors within groups, thereby gaining greater statistical power.

The bad data made it into two grant applications to the National Institutes of Health and even Horvath’s doctoral thesis, which she finished at the end of 2008 and which was supported by a federal predoctoral fellowship.

The case has led to the retraction of a paper in Molecular Endocrinology and one in Endocrinology, in which the forged data were published as figures. The articles involved research into the way the body processes insulin. According to the journals, none of Horvath’s co-authors, nor the researchers who relied on her fabricated data, was aware of the deceit.

The retraction notice in Molecular Endocrinology stated that the authors

are retracting this manuscript because one of the authors, Emily M. Horvath, admitted to altering data in four of the figures. Specifically in Figs. 2C, 5, 6D, and 11, some of the values were incorrectly reported for the radioactive glucose uptake assays. None of the other authors associated with this publication were aware of the data manipulation. The authors sincerely regret if this has caused problems with investigators that have used this information experimentally.

The notice in Endocrinology makes for similar reading.

The senior author of all three articles, Jeffrey Elmendorf, professor of cellular and integrative physiology at Indiana University, served as principal investigator on the NIH grant applications implicated in the fraud case. Elmendorf did not return repeated requests for comment.

Horvath agreed to a three-year agreement that severely curtails the scope of future work with the Public Health Service, of which the NIH is a part. Under the terms of the voluntary agreement, she is barred from serving on advisory boards for the agency. Any research group in which she participates must create a plan to supervise her if it wants to solicit grant money from the service. The agreement also called for Horvath to notify the journals in which she published her tampered data about the fraud.

Horvath had an outwardly promising career as a graduate student in biochemistry at Indiana, winning a department poster award in 2005-6.

For another Endocrinology retraction, see this post.