Professor finds flaws in obesity research

Defective information might prove obesity isn’t contagious after all.

After reading “The Spread of Obesity in a Large Social Network over 32 Years,” IU Mathematics Professor Russell Lyons determined its message was flawed.

The original paper was published in the New England Journal of Medicine in 2007 by Harvard University Professor Nicholas Christakis and University of California, San Diego Professor James Fowler.

The paper claimed to provide evidence of an influence rule within networks, like families and friends, where characteristics of obesity could be socially
transmitted.

Lyons said the results of the paper could be right, but it’s unlikely.

“If you look at the models they used, they contradicted their conclusions,” Lyons said.

The models imply no one affects each other, and the numbers are untrustworthy and meaningless, Lyons said.

“They simply misused statistics,” he said.

Lyons said he isn’t denying the existence of peer pressure, but he took issue with the research that supported the paper’s claim.

The way Christakis and Fowler did their research shouldn’t be repeated, he said.
In the future, Lyons said he believes people will begin to use different methods, as this issue brings attention to the correct research methods.

Lyons’ critique, “The Spread of Evidence-Poor Medicine via Flawed Social-Network Analysis,”  was published in Statistics, Politics and Policy.

The article was initially rejected by several journals.

“The two medical journals that I initially committed my paper to, rejected it without even doing a peer review,” Lyons said.

Other journals that turned him down said they have a policy to not publish critiques of papers they originally didn’t publish, giving his work no chance, he said.

Policies of journals prohibited publication, not discrepancy, Lyons said.

“It was never rejected, because people disagreed with the content of what I wrote,” he said.

­­— Jessica Williams