Lloyd Morgan's Column

[Return to column index]

Industry-Funded versus Independently-Funded Cellphone Research: A Scatter Plot Comparison of Outcomes

[Click here to open the scatter plots in a new window]

Introduction

The facts are in. Cellphone exposure is clearly causing health issues, most especially, the danger from brain tumors. As illustration, the latest epidemiological study by a Swedish team found not only a 290% increase in the risk of developing a high-grade astrocytoma (the nastiest type brain cancer) from cellphone use (as opposed to non-users), but also found this risk after only 10 or more years of exposure (30 years is the typical length of exposure for tobacco users developing lung cancer).

So the question must be asked: Why do these facts continue to be hidden from voice and view? And why do we continue to be lulled into complacency with report after report assuring us of the unequivocal safety of cellphone use?

The answer is easy: the cellphone industry provides the bulk of the funding for cellphone research, and the information distributed to the media concerning cellphone health and safety. Added to this cellphone industry domination is an inherent conflict-of-interest— because the very cellphone industry that funds the research has an all too real economic stake in the research outcomes. It should therefore come as no surprise that the domination of cellphone industry generated research would have not only a profound impact on the research, but also an impact on the public's knowledge concerning cellphone health and safety.

Evidence: Impact of Funding Sources on Research

The best evidence of the cellphone industry's impact on cellphone research is a side-by-side examination of the research outcomes of both cellphone industry-funded and independently funded epidemiological research.

The cellphone industry funnels money into cellphone research via their individual corporations and via their trade associations. They have substantially funded the "Interphone" studies. The Interphone studies, using a standard protocol (method), have performed studies in 13 countries. As of late-July 2006, seven Interphone studies have been published on the risk of brain tumors from cellphone use. It is interesting to note that, unlike the independent researchers, the Interphone studies will not disclose their studies' standard protocol.

The independent researchers are from the Swedish Government's National Institute for Working Life and from Örebro University in Örebro, Sweden. The National Institute for Working Life is equivalent to the U.S. Government's National Institute for Occupational Safety and Health, or, NIOSH. This "Swedish Team" has studied all brain tumors diagnosed in three regions of Sweden between 1997 and 2003.

When the cellphone industry-financed Interphone studies are compared to the independently-financed Swedish Team studies, three telling results come into view:

1. Risk of Brain Tumors
   The preponderance of the cellphone-industry funded research winds up in the "null findings" (i.e., no risk) category for the risk of brain tumor from cellphone use; in sharp contrast the preponderance of the independently-funded research finds an alarming risk of brain tumors from cellphone use with this risk being too early with respect to known latency times.
   
   Surprisingly, for cellphone use of 10 or more years of use, the majority of the cellphone industry-funded studies also show an excess risk of brain tumors.
   
   
2. Reporting Issues
   The cellphone-industry funded research—even the research that actually finds causes for concern!—has been reported by the media, usually with banner headlines, as further affirmation of cellphone safety. In contrast, the independently-funded research that continues to find, with each updated study, harmful outcomes from cellphone exposures receives, if any, scant media attention.
   
   
3. Cellphone Exposure Protects the User from Brain Tumors
   Six of the seven epidemiological research studies funded by the cellphone industry actually present "statistically significant" findings so out-of-whack to all other research that it quizzically implies that cellphone exposure is actually protective for brain tumors! (Is this perhaps an indication of some research bias or error in the Interphone studies?)

Reading the Scatter Plot Comparison Charts [View scatter plots (new window)]

1. The charts depict a comparison of the epidemiological research as of late-July, 2006 from both the Interphone and Swedish Team studies.
   
   
2. The 3 charts compare the epidemiological findings of cellphone exposures and the increased risk of 3 types of brain tumors: brain cancer, acoustic neuroma and meningioma.
   
   
3. The epidemiological studies of the two compared research groups are represented by the geometric shapes on the chart: red triangles indicate individual findings from the cellphone industry-funded studies; red squares indicate two findings at the same point on the charts from cellphone industry-funded studies; dark diamonds indicate individual findings from independently-funded studies. A black box on the chart indicates multiple findings too close together to select individually. Clicking the box on the chart will bring up every finding and its corresponding citation to the research within the box.
   
   
4. The numbers on the bottom of the chart (the horizontal axis) represent the decreasing or increasing risk of developing a brain tumor-with the "0%" in the middle representing the 50-50 chance of a coin toss, equivalent to no risk.
   
   When examining each chart, note whether the findings (red or black) are about equally distributed to the left and right of the 0% midline of the chart. If the distribution is to the left, it implies that the weight of the findings indicates that use of a cellphone protects the user from brain tumors. If the distribution is to the right, it implies that the weight of the findings indicates that use of a cellphone is a risk for brain tumors.
   
   
5. The vertical axis (with "% Confidence" on the left side of the chart) indicates the statistical confidence of the finding. In terms of scientific "weight" (or how seriously most scientists adjudge a study's findings), the most important part of the chart is this left-hand column of numbers that illustrates the "percentage confidence" (or its complement, the "p-value", the probability of a chance finding) for the particular data point. These numbers tell the story as to the strength of the finding (as opposed to what percentage of chance one has) in the risk of developing a brain tumor. In other words, the percentage confidence tells us the percentage of confidence that a finding is a true result.
   
   
6. The strength of the finding increases upwards from the bottom of the chart (light yellow-green equals "no statistical meaning"; light pink equals "possible statistical meaning"; light green equals "probable statistical meaning") towards the green line that equals the "line of statistical significance". This line crosses the chart at 95% confidence and is seen by the scientific community as the point at which findings need to be taken seriously.(However, it is important to note that within the public health community some consider this too high a bar-kind of like legal "certainty" rather than "preponderance of the evidence".)
   
   
7. Findings in the gray area indicate studies that show an excess risk-but a risk that is less than 100% in excess. Epidemiologists hold a certain skepticism about these findings because they may be the result of study flaws due to bias or confounding. An example of bias (recall bias) is the case of studies whose results depend on the accuracy of the study's subjects' memory of past events. An example of confounding can be found in risk-of-heart-attacks-per-age studies that have failed to adjust for the "confounding" effect that long-time smokers may have on the population studied.

Summary: What the Comparison Scatter Plots Tell Us [View scatter plots (new window)]

Reading the comparison scatter-plot charts, what emerges is the very picture of why the public has so little knowledge concerning the hazards of cellphone use. Consistently, the data points (findings) representing the research has the Interphone studies' findings pretty much showing no harm or even protection (as in the meningioma chart) or very little evidence of harm along with seeming protection (as in the brain cancer and acoustic neuroma charts). In contrast, the Swedish Team's studies show statistically strong evidence of an excess risk of brain tumors (except perhaps in the case of meningiomas where there is weaker evidence of an excess risk).

Yet, the majority of Interphone studies have also found an elevated risk of brain tumors, when the study examines that risk after using a cellphone for 10 or more years (e.g., see the acoustic neuroma and brain cancer charts). However, the substance of these results is not what is disseminated to the public.

Finally, as we view these charts, we are seeing before our eyes two distinct tragedies being played out: the first tragedy is that of the trumping of science for corporate economic needs. The second tragedy is much worse, and the direct outcome of the first: when independent scientific inquiry is thusly manipulated and marginalized, we all lose as humanity is deprived of a working warning system to which attention must be paid. And, in the case of cellphone exposure and human health, with the cellphone industry's increasing domination in research funding matched by the shriveling (soon, non-existent) of independent research funding—and early research findings as troubling as early tobacco studies—we may already be too late to head off a health avalanche coming our way.

Lloyd Morgan and Diana Bilovsky