We are very bad at assessing risk, often giving the most attention to the things that are least likely to harm us. Geoffrey Kabat’s new book teaches us how to think more clearly about scientific studies of environmental health risks.
Sharks get a lot of bad press and inspire a lot of fear, but in 2014 no one was killed by a shark in the US, while 36 people were killed by dogs, and 83 were killed by other mammals (including horses and cows). And tobacco kills a whopping 5.4 million people around the world every year.
Why do things that are unlikely to harm us get the most attention? In his new book Getting Risk Right, Geoffrey Kabat asks and answers that question. He says:
…we have been encouraged to worry about deadly toxins in baby bottles, food, and cosmetics; carcinogenic radiation from power lines and cell phones; and harm from vaccines and genetically modified foods… When looked at even the least bit critically, many of the scares that get high-profile attention turn out to be based on weak or erroneous findings that were hardly ready for prime time.
Kabat explains the cognitive processes that underlie the paradox. The scary studies are often flawed and biase
d, and authors frequently make leaps way beyond the actual data; but we block out those considerations because of the visceral impact of scary findings. They “sound” true because of a broader cultural narrative. Positive studies impress us more than negative studies.
He discusses the illusion of validity and the value of negative thinking (thinking about what data we might have overlooked), how a weak hypothesis can take on a life of its own due to social and psychological factors, how a scientific consensus of no risk can be outweighed in the public mind by a few vocal activists, the difference between correlation and causation, Hill’s criteria of causation, bad science and voodoo science, the complex relationships between environmental pollutants and clinical variables and the danger of focusing on one risk at a time, the fact that studies showing associations with cancer can be found for nearly every food in your cookbooks, why most research findings are false, poor reporting by the media, publication bias, the importance of dose and timing, slow thinking vs. fast thinking processes, availability cascades, and many, many other factors that may mislead us.
He makes this important distinction between “hazard” and “risk”: the drain cleaner under your sink is a hazard, but it only becomes a risk if you drink it or get it on your hands. He points out the danger of believing one’s own hypothesis. He shows that invoking the precautionary principle may not be wise.
Threats that are invisible and not under our control (like ionizing radiation) evoke a much stronger reaction than things like smoking and obesity, which have been “domesticated” and may seem to be under our control.
The dangers of focusing on unlikely harms
When people are desensitized by the latest threats, they pay less attention to known serious health risks such as smoking. When a topic is popularized by the media, scientists are more likely to keep studying it even when the results are negative. Research money is diverted to that topic and is less available for topics that are more likely to be productive. Biases and agendas have undue influence. The confusion of conflicting studies erodes public trust in science. We need to understand the interactions between researchers, the media, regulatory agencies, and advocacy groups.
Successful research vs. problematic research
…what does successful research in the area of health and health risks look like and how does it differ from the research that draws our attention to sensational but poorly supported or ambiguous findings that never seem to get confirmed but have great potential to inspire fear?
To illustrate the difference, he examines two examples of successful research (kidney damage from Aristolochiaand cancer due to HPV) and two examples of the other kind (cell phones and endocrine disruptors). He says:
If we have in mind a model of what a true advance in the area of public health looks like, this might provide a much-needed reference point for judging the many sensationalized findings that get so much attention.
Aristolochia: an herb that caused a mysterious disease and a worldwide public health threat
Kabat tells a fascinating detective story. Doctors in the Netherlands were seeing an epidemic of kidney disease, and they traced it to a mixture of Chinese herbs used at a local weight loss clinic, eventually determining that a toxic herb, Aristolochia, had been inadvertently substituted for one of the intended herbs. The pathology was reminiscent of that seen in a mysterious disease that spottily affected certain villages in the Balkans, and they eventually figured out that Balkan endemic nephropathy was caused by contamination of wheat supplies with a weed, Aristolochia. They found that aristolochic acid modifies DNA and causes a unique mutational signature, leading to kidney cancer many years after exposure. Also known as birthwort, Aristolochia is a widely used herbal remedy in China. Taiwan already has the highest rate of kidney disease and kidney cancer in the world, and it is estimated that 100,000 people have been exposed to the toxin and are at risk of developing kidney cancer down the road. More than 100 products containing Aristolochia can be purchased on the Internet. This is a cautionary tale: delayed toxic effects of herbs were missed by herbalists for centuries. So much for “natural” and “ancient wisdom!” Of the 51,000 new dietary supplements introduced since 1994 under DSHEA, only 0.3% have any documentation of their safety. There may be other Aristolochias waiting to be discovered.
HPV and cancer
Science was slow to recognize that viruses could cause cancer. The role of human papilloma virus (HPV) in cervical cancer was hard to detect, since the pathology progresses gradually over 15-20 years, and since many infections resolve on their own. We now have strong evidence that without HPV there would be no cervical cancer, and we have an effective vaccine that also protects against oropharyngeal cancers in both men and women. There is even the prospect that we might be able to virtually eradicate these cancers. The HPV story is a model for how basic science, molecular medicine, and technology can be effectively applied to real-world problems.
Cell phones and brain tumors
In 1993, David Reynard sued a mobile phone manufacturer for causing his wife’s fatal brain cancer, which happened to be on the side where she held her phone. Cell phones were a new technology that was not well understood, and this dramatic anecdote had enormous impact. Kabat describes the series of scientific studies that followed, and how the positive studies were plagued by bias and error. The best studies showed no association of cell phone use with brain tumors; in fact, regular use of cell phones was associated with reducedrisk of glioma and meningioma in one study. And despite widespread cell phone use, the incidence of brain tumors has not risen. While long-term effects can’t entirely be ruled out, evidence from many avenues of research converges to suggest that cell phone radiation is not carcinogenic.
BPA (Bisphenol A) has been banned from baby bottles; does it really pose a threat? Kabat provides the best, most thorough analysis I have seen of BPA and other so-called “endocrine disruptors.” He puts the evidence into perspective. The alarm was raised when the British Medical Journal reported a decline in sperm counts and speculated that it might reflect a decline in male fertility and contribute to testicular cancer and other abnormalities. Researchers tried to tie the phenomenon to environmental exposure to chemicals that disrupt endocrine function. It is exceedingly difficult to specify a causal factor when dealing with low- and very low-level exposures in food, air, and water. Levels of “estrogen equivalents” from pesticides are on the order of one-thousandth the levels found in red wine or beans, and one-ten thousandth that found in cabbage. The estrogenic effects of BPA are 37,000 times weaker than estradiol.
Research into environmental exposures is problematic:
- The effects may be subtle, transient, or nonexistent. Just because we can measure miniscule amounts in the blood doesn’t mean it is having a detectable effect.
- Lifestyle behaviors such as smoking, drinking, obesity, etc. may overwhelm any effects of environmental exposures.
- Genetic makeup may influence the ability to metabolize and detoxify these exposures.
- Exposures typically involve mixtures rather than single substances, and exposures change over time and are difficult to measure.
- If the effect is subtle or confined to a subgroup with particular vulnerability, few studies will have the ability to pick this up.
The BMJ researchers used poor methodology: they compared data from different countries at different points in time. Decline in sperm quality turned out not to be widespread, and in some locales, the quality actually increased. Sperm quality is known to vary with time since last ejaculation, scrotal temperature, prolonged sitting, season of the year, smoking, and drug use; these factors were not controlled for. The high levels of BPA reported in some studies may represent contamination of blood samples by BPA in medical devices in the hospital or clinic setting. Recent studies have shown that actual exposures are very low, and more than 99% of ingested BPA is efficiently metabolized.
Kabat suspects that scientists who favor the endocrine disruptor hypothesis have become too invested in their beliefs to do unbiased research. Enough research has been done to reassure most scientists that if there are any adverse effects, they must be very small in magnitude. Enough attention has been given to BPA and endocrine disruption, and it is time to direct research elsewhere.
Conclusion: Much valuable information in a small package
Kabat has packed a wealth of information into his 180 pages of text, and everything he says is copiously supported by references. There is much to learn here, not only about the four subjects covered in depth, but about how science works and the factors that can lead us astray. This book will provide you with defensive armor against alarmist headlines and it will help you judge the credibility of new studies. Highly recommended.
This article was originally published in the Science-Based Medicine Blog.