Did Diet Politics Corrupt World Cancer Research Fund Recommendations?

It is the most important barometer of cancer research - a survey of over 7000 recent studies on cancer that took five years to complete - but one of the World Cancer Research Fund's key recommendations on how to avoid cancer may be flawed because of what was not included in the survey.

Among ten recommendations on how to avoid cancer, the report argues that there is "convincing" evidence that red meat and processed meats increase the risk of colorectal cancer, and that as a result people should limit the intake of such food products to 18 ounces a week. As the Los Angeles Times reported:

Once an individual reaches the 18-ounce weekly limit for red meat, every additional 1.7 ounces consumed a day increases cancer risk by 15%, the report said. Every 1.7 ounces of processed meat consumed a day increases cancer risk by 21%, it added.

However, the largest ever study examining the link between colorectal cancer and red and processed meat consumption did not find any association. The study "Meat and fat intake and colorectal cancer risk: A pooled analysis of 14 prospective studies," by Eunyoung Cho, and Stephanie A. Smith-Warner for Harvard's Pooling Project of Prospect Studies of Diet and Cancer Investigators, was abstracted in 2004. But it was never been published - even as 19 other studies on cancer and diet were published by the Pooling Project.

When contacted by STATS.org, Smith-Warner said they wanted to add a few more studies before publishing their results next year. But the fact is that their colorectal cancer study had more subjects than many of the other studies published by the Pooling Project - and the four-year delay in publication cannot but raise the question of whether their results just didn't fit in with the nutritional beliefs of Harvard's School of Public Health, one of whose senior figures - Dr. Walter Willett - has long recommended limiting red meat and who, coincidentally, is a board member of the World Cancer Research Fund.

Perhaps the additional data mined by Cho and Smith Warner will find a statistically significant link to support the WCRF's recommendation. Or perhaps not. Either way, the decision to withhold the results of what appears to be a statistically robust study of enormous scope taints the report's recommendations with the unhealthy appearance of ideology.