The End of Doom (19 page)

Testicular cancer does appear to be increasing in many developed nations. Proponents of the endocrine disruption thesis have conducted epidemiological studies that weakly suggest that they do correlate with greater risk for testicular cancer. On the other hand, lots of other physiological phenomena also correlate with higher testicular cancer rates. A 2012 study reports that one of the stronger correlations uncovered by several studies is that the risk of testicular cancer increases with adult height. The researchers in that study find that men over six feet tall have a higher risk of testicular cancer, and they suggest that “the trend of increasing adult height and the increasing TC [testicular cancer] incidence are biologically interconnected with improved nutrition in early life.” They note that while testicular cancer rates in Europe increased throughout the twentieth century, they stalled for the generation that suffered nutritional deprivation as a result of World War II. Interestingly, being overweight reduces the chances of testicular cancer, so one might think that as the obesity rate rises, the testicular cancer rate should fall.

The
State of the Science
consensus report notes, “The prevalence of obesity and type 2 diabetes has dramatically increased worldwide over the last 40 years.” As noted above, the consensus report, however, admits there is very little epidemiological evidence linking endocrine disrupting chemicals to increasing obesity rates, but that does not stop researchers from trying to do so. A typical study, “Association of Endocrine Disruptors and Obesity: Perspectives from Epidemiologic Studies,” published in 2010, found a correlation between excess weight and endocrine disrupting chemicals. However, that study begins by acknowledging that “changes in diet and physical activity are undoubtedly key causal factors related to the increase in obesity.” Well, yes.

A far more plausible explanation for the rise in obesity can be found in a 2013 study that tracked the changes in the number of calories Americans consumed daily between 1971 and 2010. The researchers found that in 1971, Americans ate an average of 1,955 calories daily. That average rose to 2,269 per day by 2003 and has recently dropped a bit, to 2,195 calories daily. While Americans are eating more, they are also exercising less. A 2011 study reported, “Over the last 50 years in the U.S. we estimate that daily occupation-related energy expenditure has decreased by more than 100 calories, and this reduction in energy expenditure accounts for a significant portion of the increase in mean U.S. body weights for women and men.” It is well established that being overweight escalates considerably the risk of type 2 diabetes. A 2012 workshop organized by the US National Institute of Environmental Health Sciences reviewed seventy-five studies relating exposures to synthetic endocrine disruptors to type 2 diabetes and concluded, “In no case was the body of data considered sufficient to establish causality.” It's hard not to conclude that in comparison to eating more calories and increasingly sedentary lifestyles, the effects of endocrine disrupting chemicals on rising obesity and the prevalence of type 2 diabetes, if such effects exist, would be negligible.

The
State of the Science
report also asserts that there is a “trend towards earlier onset of breast development in young girls.” Again, some epidemiologists have sought to identify synthetic endocrine disruptors as the possible villains behind this trend. However, considerable research persuasively suggests that earlier breast development and puberty onset in girls is linked to increases in body fat. In other words, rising obesity rates strongly correlate with earlier onset of puberty in girls. There does appear to be an epidemic of
diagnoses
of attention deficit hyperactivity disorder (ADHD), but many researchers question the claim that actual cases are increasing. A 2014 comprehensive review in the
International Journal of Epidemiology
analyzed 135 studies dealing with trends in ADHD and reports, “In the past three decades, there has been no evidence to suggest an increase in the number of children in the community who meet criteria for ADHD when standardized diagnostic procedures are followed.”

A devastating article, “Endocrine Disruption: Fact or Urban Legend,” by a team of European and American toxicologists, published in the journal
Toxicology Letters
in December 2013, reviews hundreds of studies on the effects of alleged endocrine disrupting chemicals. Their review of the science finds no effects on sperm counts, no effects on the rate of penile deformations, no effects on any cancer, no effects on diabetes, no effects on breast development, and … simply no effects whatsoever from exposures to chemicals, natural and synthetic, that emit a weak hormonal signal in very sensitive lab tests. Among many other things, the researchers point out that voluntary human exposures to powerful hormone modifying substances such as the estrogens in birth control pills have resulted in practically no significant health effects on either those who take them or their progeny. The effective potency of contraceptive pills is more than a million times greater than the potencies of the weak endocrine disrupting chemicals on which environmental activists focus. Taking this and much other evidence into account, the researchers assert that “the hypothesis that the negligible exposure of humans to chemicals of negligible hormonal potency could have an effect on human fertility is absurd, defying a scientific basis as well as common sense.” They conclude that the man-made environmental disruptor hypothesis “has now been evaluated experimentally and epidemiologically for nearly 20 years and no convincing evidence has been found of an actual decline in human fertility, and even less of a causal relation with synthetic hormonally active substances.”

The Problem of Epidemiology

Epidemiologists are hard at work sincerely trying to uncover relationships between all sorts of exposures and health outcomes. Many of the assertions made about the possible effects of endocrine disrupting chemicals in the
State of the Science
report are based on the results of observational studies by epidemiologists. The number of observational studies has increased from 80,000 in the 1990s to more than 260,000 in the first decade of the twenty-first century. Most observational studies are case-control studies in which epidemiologists identify a population that suffers from a disease or other condition and then attempt to match them with a similar population free of the disease or condition. They then look for differences in lifestyle, diet, or the environment that might account for the disease. This should work in theory, but the problems with controlling for biases in the data and for confounding factors are well known to epidemiologists. Confounding factors are variables that have been overlooked by researchers. Confounders can easily generate spurious associations.

For example, one famous study found a link between heavy coffee drinking and pancreatic cancer that disappeared once the smoking and alcohol consumption habits of coffee drinkers were taken into account. In addition, biases can creep in because it turns out that the control population differs in significant but unrecognized ways. For example, a finding that exposure to electromagnetic fields caused leukemia disappeared when differences in the incomes of the case population and the control population were taken into account. There is a well-known epidemiological relationship between poverty and cancer.

The problem is, epidemiologists generally find far more false positives than they do true positives—that is, they identify far more associations between phenomena than eventually are found to be the case. How do we know that there are far more false positives than true positives? Because the vast majority of epidemiological studies are not replicated. In other words, other, later researchers do not find that the risk factor identified in the initial observational study is in fact associated with a disease. S. Stanley Young of the US National Institute of Statistical Sciences estimates that only 5 to 10 percent of observational studies can be replicated.

In addition, there is a strong tendency among epidemiologists to publish only studies with positive results. Reporting only positive results (that is, a finding that some risk factor is associated with disease) skews the literature toward implying that various risk factors are more dangerous than is really the case. Others worry that researchers may try to please their sponsors, especially the government regulatory agencies that fund their research. As one anonymous researcher at the National Institute of Environmental Health Sciences told
Science,
“Investigators who find an effect get support, and investigators who don't find an effect don't get support. When times are rough it becomes extremely difficult for investigators to be objective.”

Even with the best of scientific intentions, it is not easy to sort actual risk factors from the statistical background noise of confounders and researcher biases. “With epidemiology you can tell a little thing from a big thing. What's very hard to do is to tell a little thing from nothing at all,” said Michael Thun, an American Cancer Society epidemiologist, in 1995. Former Boston University epidemiologist Samuel Shapiro agrees: “Epidemiologists have only primitive tools, which for small relative risks are too crude to enable us to distinguish between bias, confounding, and causation.”

So, most epidemiologists will agree that one study that identifies a small effect means very little. However, if a number of studies consistently find a similar small relative risk for a factor, then perhaps the factor is causal. But consistency among studies can go only so far. If all of the studies have the same design, they could all be implementing the same biases and missing the same confounders and thus producing the same spurious positive results. That brings up the issue of pathological science.

Possibly Pathological Science?

In 1953, Nobel Prize–winning chemist Irving Langmuir identified what he called “pathological science,” or “the science of things that aren't so.” In his 1992 summary of Langmuir's insights, Denis Rousseau noted that “the first characteristic of pathological science is that the effect being studied is often at the limits of detectability or has very low statistical significance. Thus it can be difficult to do experiments that reliably test the effect.” In such scientific situations “unconscious personal bias may affect the results.”

Experimental researchers into the effects of endocrine disrupting chemicals have devised a set of exquisitely sensitive cell-based and animal model tests to detect their subtle influences. They then turn around and justify the relevance of the barely detectable responses conjured from their hypersensitive assays by citing supposed endocrine disruptor effects found in human epidemiological studies. Amusingly, some epidemiologists return the favor and suggest that the plausibility of their very weak findings is justified by the experiments using the sensitive assays. Neither bothers to explain why the results of these highly sensitive tests have any relevance to human health.

Rousseau adds, “Because the effect is weak or of such low statistical significance, there may be no consistent relationship between the magnitude of the effect and the causative agent. Increasing the strength of the causative agent may not increase the size of the effect.” Another hallmark of pathological science is that its effects can be elicited by only some researchers, who are then unable to communicate how they achieve them to other researchers. Proponents of endocrine disruption generally respond that outside researchers who can't replicate their findings are simply not careful enough. Naturally, practitioners of pathological science are impervious to critiques and accusations of sloppiness on their part from other scientists.

In addition, practitioners of pathological science propose eccentric theories, positing mechanisms that appear nowhere else in related sciences. In this case, champions of the environmental endocrine disruption hypothesis reject the standard thinking in toxicology, according to which the biological effects of a substance increase as the dose increases, often summarized as “the dose makes the poison.” Instead, they propose the concept of nonmonotonic dose response, in which exposures to allegedly endocrine disrupting compounds exhibit a U-shaped dose response curve—that is to say, barely detectable doses produce a big effect that falls off as the dose increases to a certain level, at which point bigger doses also produce big effects.

Perhaps environmental endocrine disruptor researchers have uncovered a real phenomenon, but most toxicologists doubt it. “Although this hypothesis is consistent with the ideas of homoeopathy, it contradicts centuries of toxicological and pharmacological experience demonstrating that active substances produce a specific dose-response in the affected organism,” tartly asserted one group of critical toxicologists. Homeopathy is a medical pseudoscience in which the alleged remedies are so diluted that they often do not contain a single molecule of any supposed therapeutic substance. Low-dose effects indeed.

To explain how researchers and whole fields of science can end up studying phenomena that don't actually exist, Stanford University biostatistician John Ioannidis fancifully describes the highly active areas of scientific investigation on Planet F345 in the Andromeda Galaxy. The Andromedean researchers are hard at work on such null fields of study as “nutribogus epidemiology, pompompomics, social psychojunkology, and all the multifarious disciplines of brown cockroach research—brown cockroaches are considered to provide adequate models that can be readily extended to humanoids.”

The problem is that the Andromedean scientists don't know that their data dredging and highly sensitive nonreplicated tests are massively producing false positives. In fact, the Andromedean researchers have every incentive—publication pressure, tenure, and funding—to find effects, the more extravagant the better. But in fact, the manufactured discoveries are just estimating the net bias operating in each of these “null fields.”