Dr. Magda Havas, PhD Environmental Studies Research Papers

Breast Cancer and EMFs: a response

Posted on October 10, 2009

Havas, M. 2009. Breast Cancer and Occupational Exposure to Electromagnetic Fields. Response to Request from Heidi Evelyn, Tribunal Counsel Office, Workplace Safety and Insurance Appeals Tribunal, Dated January 7 & 9, 2009. Toronto, Ontario, Canada, Response submitted February 9, 2009, 42 pp.

excerpt from testimony, from page 7 . . .

Threshold magnetic field to protect workers and public

Threshold magnetic field to protect workers and public

Breast Cancer and EMFs

Posted on October 10, 2009

Havas, M. Breast Cancer and Occupational Exposure to Electromagnetic Fields. Report to the Workplace Safety and Insurance Appeals Tribunal, November 18, 2008, 20 pp.

On October 28, 2008, Mr. Gary Newhouse asked me to address the following questions:

1. What is the current level of evidence that EMF and/or ELF is cancer causing or promoting, with particular reference to breast cancer? Please explain.

2. What is your opinion on this comment from Dr. M. Bitran found at page 14 of Exhibit 20:

Comprehensive reviews of epidemiological and laboratory studies carried out by authoritative organizations have consistently concluded that the evidence does not substantiate a cause-effect link between ELF magnetic fields and cancer. Recent epidemiological studies on breast cancer and occupational and residential exposure to ELF magnetic fields are negative on balance. A meta-analysis found a small risk increase that may be due to artifacts. Recent reviews of epidemiological and animal data conclude that ELF magnetic fields are most likely not a risk factor in breast cancer.

3. What is your opinion on this comment from Dr. M. Bitran found at page 23 of Exhibit 20:

The link between ELF magnetic field exposure and female breast cancer has become more tenuous as newer, larger epidemiological studies with better exposure assessment have become available.

4. Please comment on the elements of average exposure, transient peak impact exposures, and cumulative dose estimates of exposure for each of the three workers in terms of the relationship between those exposures and whether or not exposure to EMF would have been a significant contributing factor in the development or onset of breast cancer in each case.

My expert testimony regarding each of these questions follows:

Excerpt from page 5 . . .

To address this issue I would like to quote from a document written by Dr. Martin Blank at Columbia University (Blank 2007).

We should be reminded that ‘scientific proof’ is not symmetric (Popper, 1959).

One cannot prove that EMF is harmless no matter how many negative results one presents. One single reproducible ([statistically] significant) harmful effect would outweigh all the negative results.

Scientific method is not democratic. The word ‘proof’ in ‘scientific proof’ is best understood in terms of its older meaning of ‘test’. It does not rely on an adversarial ‘weight of the evidence’, where opposing results and arguments are presented and compared. Answers do not come from keeping a scoreboard of positive versus negative results [note: positive and negative results refers to studies with and without statistically significant effects] and merely tallying the numbers to get a score.

The above characteristics of science are generally acknowledged to be valid as abstract principles, but in EMF research, it has been quite common to list positive and negative findings and thereby imply equal weights. . . .

Negative studies play an important role in science, and there is good reason to publish them when they are failures to replicate earlier positive results. This can often lead to important clarifications of the effect, the technique, etc.

However, negative studies are being used in another way. Although they cannot prove there is no positive effect, they do have an influence in the unscientific ‘weight of evidence approach’. In epidemiology, where it is difficult to compare studies done under different conditions, it is common to make a table of the positive and negative results. The simple listing has the effect of a tally, and the overall score substitutes for an evaluation. In any case, one can write that the evidence is ‘not consistent’, ‘not convincing’ or claims are ‘unsubstantiated’ and therefore ‘unproven’. The same is true in experimental studies . . . the contradictory results are juxtaposed and a draw is implied. This is a relatively cheap but effective way to neutralize or negate a positive study (Blank 2007).