Cell phones, DNA damage, and questionable data.

While other ScienceBlogs bloggers (notably Revere and Orac) post periodically on the state of the scientific evidence with regard to whether cell phones have biological effects on those using them, I've mostly followed the discussion from the sidelines. Possibly this is because I'm a tremendous Luddite who got a cell phone under protest (and who uses a phone with maybe three functions -- place a call, receive a call, and store phone numbers). Possibly it's because in my estimation the biggest health risk posed by cell phones is that they shift the attention of the maniac driver shifting across four lanes of freeway without signaling.

What has me jumping into the fray now is a news report in Science about fraud charges that have been raised against a group of scientists whose papers offered evidence of the potential for biological harm from cell phone use. From the Science article:

The only two peer-reviewed scientific papers showing that electromagnetic fields (EMFs) from cell phones can cause DNA breakage are at the center of a misconduct controversy at the Medical University of Vienna (MUV). Critics had argued that the data looked too good to be real, and in May a university investigation agreed, concluding that data in both studies had been fabricated and that the papers should be retracted.

The technician who worked on the studies [Elisabeth Kratochvil] has resigned, and the senior author on both papers [Hugo Rüdiger] initially agreed with the rector of the university to retract them. But since then, the case has become murkier as the senior author has changed his mind, saying that the technician denies wrongdoing. He will now agree to retract only one paper, and he also says his critics have been funded by the cell phone industry, which has an obvious interest in discrediting any evidence of harm from its products.

It's worth noting that the findings in question here were already contentious when they were reported, due in large part to the fact that there is not an obvious mechanism by which EMFs could cause DNA breakage. This does not rule out findings of a strong correlation between EMF exposure and DNA breakage -- after all, plenty of the biological mechanisms that seem obvious today weren't obvious in the past. But, the lack of an obvious mechanism is likely to make scientists examine the results much more critically.

Obvious mechanism or not, the Rüdiger lab was not the only research group reporting that the EMFs from cell phones seemed to have biological effects:

Other teams have reported only cellular effects of EMFs that are more subtle than DNA breakage, such as changes in gene activation or expression. "If this work isn't solid, then one really has to give up the hypothesis that these fields cause genotoxic effects," says Anna Wobus, a developmental biologist at the Leibniz Institute of Plant Genetics and Crop Plant Research in Gatersleben, Germany, who has studied the effects of EMFs on stem cells.

In any case, perhaps because the results reported by the Rüdiger group seemed so surprising, other scientists scrutinized the data very closely. One of the scientists mounting this scrutiny, Alexander Lerchl, professor of biology at Jacobs University Bremen, concluded that the variation in the data from a 2005 paper in Mutation Research was too low to be consistent with "data from biological experiments".

Lerchl alerted the editors of Mutation Research, who took his concerns under advisement. However, they noted that with a properly blinded experimental setup, "it would have been impossible to make up data that produced a desired result".

The ethics commission at MUV, however, launched an investigation of the 2005 paper and a 2008 paper in nternational Archives of Occupational and Environmental Health. That investigation concluded:

The data were not measured experimentally but fabricated.

Since the full report of the MUV ethics investigation wasn't made public, it's hard to know on what basis the review panel made this judgment. But it may have been something along the following lines:

[I]n April, unaware of the university's investigations, Christian Wolf, the interim head of Rüdiger's former department, was taking an independent look at the data after hearing they were under dispute. Wolf told Science that he and a colleague examined the lab notebook of technician Elisabeth Kratochvil, first author of the 2005 paper and a co-author of the 2008 study. Wolf says that they noticed a column of numbers corresponding to a code from the instrument designed to expose cell lines to EMFs. The code revealed which chamber was exposed to EMFs and which was the control. Rüdiger's team was supposed to receive the key only after sending their observational data to the device's manufacturer in Zürich, but Wolf found that the code could be observed by the turn of a knob to an "unused" channel. After being confronted with the notebook, Wolf says, Kratochvil resigned. Later, Wolf says, they found code entries in laboratory notebooks going back to the fall of 2005.

Data that were supposed to be blinded were not. Experimenters thus had biased expectations about what would happen to the cell lines under study. Even if there was no conscious attempt to fabricate or falsify data, unblinding the experiment compromised the objectivity of the results.

Principal investigator Rüdiger argues that only the 2008 paper (done on the lab's own EMF exposure device -- the one that Wolf discovered revealed the secret code) is problematic. However, he defends the 2005 paper. That paper, he says, was based on data collected using an EMF exposure device in a lab in Berlin for which "there is no evidence that she [Kratochvil] knew" the code.

In light of the too-good-to-be-true data from the 2005 paper and the lab notebooks indicating that Kratochvil was doing unblinded experiments that were reported in the literature as being blinded, is the lack of direct evidence of cheating enough here?

There may not be enough direct evidence to prove misconduct in the research that was reported in the 2005 paper (or there may be -- again, we don't know what the MUV ethics investigation turned up. But the serious problems with the 2008 paper raise serious questions about the credibility of the particular scientists involved. If there are good grounds for doubting your integrity and objectivity on a later paper, it's natural to wonder whether such problems could have affected your earlier published work.

It might not be the sort of paper you'd want to trust as the basis for your own further research. It might not be the sort of paper you'd want to trust as the basis for government regulations on cell phones.

Of course, Rüdiger raises a concern of his own: the conclusions of the MUV ethics investigation may have been biased by the fact that "the chair of the ethics committee was a lawyer who had worked for a telecom company".

Clearly, objectivity is hard work all the way around.