“A theory that explains everything, explains nothing.”
― Karl Popper1
column by Nick Kristof in the New York Times first alerted me to a newly
published book entitled Count Down: How Our Modern World Is
Threatening Sperm Counts, Altering Male and Female Reproductive Development,
and Imperiling the Future of the Human Race, co-authored by Shanna Swan, an epidemiologist, and Stacey
Colino, a journalist. True to its title,
the book tries to make the shocking case that environmental exposures to trace
levels of certain chemicals, particularly a diverse group called phthalates,
are responsible for causing severe declines in various measures of male and female
reproductive health and are placing the entire human race at threat of
extinction. These chemicals are purported to act by disrupting the endocrine
systems of humans and other species and hence are referred to as EDCs. However, this sensational claim is neither
new – Swan has dedicated at least the past 25 years of her career to making it,
and Kristof has devoted at least several of his opinion columns2
over the past decade shamelessly promoting it – nor is the scientific evidence
they offer to support it very convincing.
Sagan once stated, "extraordinary claims require extraordinary
evidence”, and by this test Swan and her book get failing grades.
I’m not alone in my criticisms, as other scientists have also panned the evidence. Responding to Kristof’s column, Geoffrey Kabat, an academic epidemiologist and author of the book “Getting Risk Right: Understanding the Science of Elusive Health Risks”, which includes an entire chapter devoted to the EDC hypothesis, wrote: “Kristof’s commentary, and in fact the research that he selectively cites, tell only part of a complex story about the impact of exposure to environmental chemicals. Rather than conveying the grays of science, he ends up promoting the most sensational, almost black and white, and least probable explanations of a serious environmental and health issue.”
Kabat has offered similar criticisms of a prior Kristof column on the topic, as has Dr. Paul Turek, an internationally renowned expert in men’s sexual health and reproductive urology, who has pointed out that the reported decline in sperm counts derives from flawed epidemiology studies, and that the reported level of decline still leaves the vast majority of men well within what is widely considered to be the normal and healthy range.
After reading Swan’s book, Tom Chivers, a book author himself and science editor, wrote “But the book does not seem sensible. Because for Swan, everything causes fertility problems, not just phthalates. Sugar, beef, cycling, stress, watching TV, doing too much exercise, doing too little, eating the wrong diet; you can barely get out of bed in the morning without sabotaging your chance of having a baby. Undergoing stress and trauma will not only make you less fertile, they’ll affect your children’s fertility (should you remain capable of having them). And phthalates and other EDCs don’t just damage your chance of having a baby – they seem to be the cause of everything wrong in the world.”
Reading Chivers review I was reminded of the quote from the noted science philosopher, Karl Popper, “A theory that explains everything, explains nothing.”
Furthermore, at no point in the book do Swan and Colino even try to explain in any detail their theory of how EDCs supposedly interact with other risk factors to cause the harm they allege. It is simply presented as established science, which is far from the truth. But the problems with the book run much broader and deeper, as I explain in greater detail below.
Strengths of the Book
Count Down is not wholly without its merits. Swan and Colino deserve some credit for bringing attention to infertility, a topic of great interest to prospective parents who are struggling to conceive. My wife and I started our family while we were quite young and were blessed to have children when we wanted them. We took our fertility for granted. By contrast, our children and their spouses chose to start their families in their 30’s, and while they were also ultimately successful, on average it took them quite a bit longer to conceive. This created lots stress for them and for us, as we were anxious to have grandchildren while we still had our health to enjoy them. Unfortunately, some other couples, including friends of ours, have not been as lucky. So this is a serious issue for some individuals. For broader society? Perhaps not so much!
Swan and Colino devote several chapters to explaining the underlying biology of male and female reproductive development, and do a decent job of it, although I’m almost certain that more than a few clinicians will object to some of their more fanciful characterizations and claims, particularly as they pertain to small changes in sperm count and morphology. The book also includes a chapter that is focused on the long list of non-chemical risk factors for adverse reproductive health, including: age, heat, obesity, smoking, radiation, marijuana use, vaping, alcohol use, a wide variety of medications (including opioid use), diet, exercise, psychological stress, other medical conditions, and other exposures. Their list, while not as exhaustive as others I have seen published, is reasonably complete and captures the most well-established and impactful of the known factors.
Major Problems with the Book
In addition to the problems that were noted at the outset of this article, namely:
n The claims are overly dramatic, which Chivers attributes to Swan not believing that her issue is getting enough public attention, so to compensate she has overhyped it by writing a “polemic” and has done so in the style of “the fallacy of reversed moderation” which “…ends up looking obsessive and weird” (Indeed, lending credence to this notion, late in the book the authors confess to wondering why the issue hasn’t generated more public outrage); and
n the theory that EDCs cause everything simply lacks biological plausibility;
the book suffers from numerous other problems that mislead readers who are not familiar with the underlying science.
The Authors Views Are Extremist and NOT in the Mainstream
The authors fail to provide any perspectives on the issue other than their own. The endocrine hypothesis was introduced more than 30 years ago and has been steeped in controversy ever since. For a fascinating tour de force outlining this controversy I highly recommend the reader look at a doctoral dissertation by Alexis Abboud. Abboud carefully chronicles the evolution of the hypothesis and discusses how its proponents quickly abandoned doing good science in favor of active lobbying for policy changes. None of that controversy is discussed in Swan’s book, which consequently leaves the reader to mistakenly believe that the scientific community has reached consensus on it and that Swan’s opinions are in the mainstream. Nothing could be further from the truth. At least Kristof included a paragraph devoted to the uncertainties and speculative nature of the EDC hypothesis, although he buried it late in his op-ed, only after making his most outrageous claims. Remarkably, Swan does nothing of the sort.
As Kabat writes “Missing from Kristof’s column [and notably also from Swan’s book] is any understanding that the scientists he quotes are not representative of views in the scientific community. A very different picture emerges when one listens to some of the foremost authorities in the field of reproductive health.” Kabat goes on to quote from Sir Richard Sharpe one of the world’s foremost endocrinologists. He is the research scientist who originated the notion and study of ‘endocrine disrupting chemicals’ in the 1990s. After participating in and reviewing hundreds of studies on EDCs, Sharpe is convinced the concept is wrongheaded—an ideological belief and not science based. Kabat makes a further important point “Unfortunately, by leveraging the media (and often partnering with tort lawyers), these ‘advocacy scientists’ generate massive reams of dubious-quality research, sucking up the bulk of public funding that should be devoted to the more difficult work of elucidating the complex pathways of healthy reproductive development.”
Near Exclusive Focus on Sperm and Egg Count and Quality and Fertility is Misplaced
Turek largely dismissed the importance of sperm and egg counts and quality and even fertility3 (the number of live births per unit of population) as measures of reproductive health and has instead emphasized the importance of measuring fecundity – the potential to conceive. Turek recounted the findings of a NIH sponsored scientific meeting in which he was a participant that examined the question of decreased fecundity and concluded there is no hard evidence to suggest that this has changed at all. Turek further explained “… what surfaced early on and pervaded the entire meeting were the massive complexities that surround this subject. Factors such as cultural norms, socioeconomics, family planning behavior and pregnancy intentions are all rather large elephants in the room when discussing human fecundity. And they are all huge influencers about which very little is known.”
No Context is Provided for Evaluating Epidemiological Evidence
Much of the research that Swan cites to support her claims derives from observational epidemiology studies rather than from controlled experiments. Yet, she never discusses the inherent limitations of such studies – notably how they are fundamentally less rigorous than experimental studies and thus are far more prone to yielding unreliable findings due to chance, confounding and bias4. As a rule, epidemiologists do not put much stock in the findings from single, small, un-replicated studies. They are too unreliable. Much of the work Swan cites on EDCs derives from such studies and particularly from cross-sectional studies which are considered among the weakest study designs. I have written previously on this blog about these topics and I refer the reader to those articles for the details (see Bond 2019, Bond 2017, Bond 2016).
Epidemiologists want to see consistent results generated across multiple studies of different designs, conducted by different investigators acting independently, before they are convinced that the findings are likely to be real rather than due to artifacts. In fact, Stanford epidemiology professor John Ioannidis wrote a seminal paper proving that most published research results are false. The consequence of Swan’s failure to provide this essential context is to leave the reader with the mistaken impression that the science supporting her claims is more robust than is actually the case.
Although earlier I gave credit to Swan and Colino for including a chapter devoted to non-chemical risk factors, there are several glaring problems with the evidence they presented which need to be highlighted. Firstly, much of that evidence also derives from observational epidemiology studies. However, Swan fails to make any attempt to characterize that evidence in terms of the number of studies, their design or size, consistency of findings, measures of the strength of association, etc. that would allow the reader to understand how compelling that evidence is, especially when compared and contrasted with the evidence she cites to support a link with EDCs. Which of the many non-chemical risk factors are most likely to negatively impact reproductive health and what is their relative contribution compared to one another and to EDCs?
Unlike the evidence she cites on industrial chemicals which comes from a few poorly designed studies based on small numbers and that yield weak associations, much of the evidence on these other factors derives is much more robust (i.e., more numerous, better designed and larger studies with stronger associations reported). Moreover, she does not discuss how these other factors may actually interfere with or confuse the studies done on environmental chemicals (a phenomenon which epidemiologists refer to as confounding) if the investigators do not measure them and control for their influence in their studies. Again, the uninformed reader will be misled.
By her silence on these important topics, most readers of the book will mistakenly assume that the evidence is equally scientifically robust and that all of the various risk factors are equally potent in increasing the risk of reproductive harm. Neither is true.
The Authors “Cherry Pick” the Science, Do Not Weigh the Evidence and Mistakenly Assume Correlation Equals Causation
The chapter which presents the evidence that Swan alleges shows exposures to purported EDCs are an important contributor to adverse reproductive health effects is surprisingly short – only 18 pages in length or about 8% of the total pages in the book (not counting the bibliography or index) -- and it is shockingly superficial. Separate sections are devoted to phthalates, Bisphenol A (BPA), polybrominated flame retardants (PBDE’s), pesticides and what Swan refers to as “under the radar EDCs” which she lists as perfluoroalkyl substances (PFAS), Polychlorinated Biphenyls and dioxin.
Early in the book, Swan describes phthalates as a large and diverse set of chemicals, but later she lumps them all together and claims they need to be considered as a single class without providing any scientific justification for doing so. She does the same for PFAS, PBDEs and most crazily for pesticides, a very diverse set of chemistry. The American Chemistry Council’s High Phthalates Panel, which represents US producers of these chemicals, has strenuously objected to treating the diverse chemicals as a single class, noting their widely divergent toxicology profiles. Other ACC chemical panels have expressed similar objections (PFAS, and Flame Retardants).
Swan does not reveal which literature sources she searched to identify pertinent studies or the criteria she used to include some and exclude others from consideration. Nor does she discuss how she evaluated the quality of the individual studies or reconciled or weighted their disparate results. Further she does not share her criteria for assessing the likelihood that mere statistical associations are likely to represent causal ones. This is a major departure from standard scientific practice which demands a high degree of transparency on these critical steps.
Swan often contradicts herself on the weight of the evidence5 linking EDCs to adverse health effects. Sometimes she refers to “associations” and other times she strongly states her personal belief that the relationships are causal. Yet, most scientists who have examined the same evidence have concluded that it falls far short of standards for establishing causation. So, Swan and Colino quite often are guilty of mistakenly assuming mere statistical correlation equates to causation. This is a common fallacy that the media unwittingly promotes; however, readers should expect better from an academic scientist.
Kabat noted many of the same points I observed when evaluating the evidence Swan relied upon. She cites isolated results from poor quality studies, which haven’t been reproducible, and selectively ignores the most well conducted studies and scientific discussion that provide no support for the claim that reproductive health is under threat. Epidemiology studies claiming declines in sperm quantity and quality rarely take into account some of the most important risk factors, such time since last ejaculation, scrotal temperature, prolonged sitting, season, smoking, drug use, stress, trauma, obesity, under-nutrition, medications, and diets and as a consequence they are prone to confounding and yielding misleading results. Swan also claims that malformations of the male genitals are on the rise. This is not supported by systematic reviews of the scientific evidence. Furthermore, she highlights the rising incidence of testicular cancer and strongly infers a role for EDCs in causing it, but there is strong evidence to support that rise actually started in the early twentieth century and had its roots in early life exposures in the late nineteenth or early twentieth century, which casts strong doubt on the hypothesis that the increase can be explained by exposure to chemicals in the post-World War II environment.
Finally, Swan includes a brief discussion of three examples she offers of “regrettable substitutions” for one type of phthalate (DEHP), PBDEs and BPA wherein the substitute chemicals, according to her, turned out to be equally bad or worse. However, she provides no real substantive evidence to support her claims. I’m personally astonished at how the proponents for the endocrine hypothesis are so quick to claim “regrettable substitution” on the flimsiest of evidence and take no accountability for their role in causing it. They produce substandard science indicting chemicals and then work hand in hand with activists to exert tremendous public and marketplace pressure on users of those chemicals to quickly deselect them from their products. Its no wonder that in their haste to escape the harsh spotlight that the users will jump to the next available chemical that provides the functionality they desperately need at a competitive price. Swan is quick to praise the large retail chains for their efforts to pressure their suppliers to deselect chemicals, but she fails to acknowledge that those same retail chains demand that the formulators not pass along any increase in the cost of the consumer products they peddle. It should be no surprise then, that some of the users might consequently cut some corners in choosing replacement chemicals.
Potency and Dose are Completely Ignored
It is a fundamental tenet of toxicology that the dose makes the poison. This is credited to Paracelsus who wrote "All things are poison, and nothing is without poison; the dosage alone makes it so a thing is not a poison." Yet, Swan never discusses the dose levels she suspects are required to cause the health effects she alleges. In reality, evidence from animal toxicology studies, including reproductive and developmental studies, are available for many of the purported EDCs, especially pesticides, and safe levels of exposure have been established to protect against the adverse effects Swan alleges. Endocrine pharmacology and medical toxicology establish that safe levels of exposure exist for endocrine active chemicals, both natural and man-made. Biomonitoring data on samples of the general population generally confirms that exposures are usually well below levels of concern. The US EPA perhaps states it best, “…limited evidence exists for the potential of chemicals to cause these effects in humans at environmental exposure levels.”
It’s not only the dose, but also the potency of a substance, that determines its effects. Once again, Swan does not discuss the relative potency of the chemicals she indicts. The reality is that synthetic EDCs are much less potent in interacting with the endocrine system than many naturally occurring endogenous and exogenous EDCs.
To illustrate that her hypothesis that EDCs are capable of causing multigenerational health effects, Swan cites the well documented example of Diethylstibestrol (DES), a drug once prescribed during pregnancy to prevent miscarriages or premature deliveries. It was later found to cause a rare form of vaginal cancer in the young adult daughters born to women who took the drug. Yet extrapolating from the DES example to industrial chemicals in the ambient environment is very precarious indeed, and unreasonable on several levels. DES is a very potent endocrine active agent – orders of magnitude more potent than the industrial chemicals Swan is indicting, and it was taken orally in therapeutic doses which greatly exceed the trace levels of exposure to those same industrial chemicals (Golden et al, 1998; and Borgert et al, 2012). Thus, the DES story, however tragic, cannot be reasonably extrapolated to the trace environmental levels of industrial chemicals. Swan also promotes some of the more controversial aspects of the “fashionable” field of epigenetics and posits that environmental exposures, lifestyle choices we make, and even emotions we feel can be passed on to future generations. It is presented as if there is scientific consensus about it, which is highly misleading.
Toward the end of the book Swan briefly mentions the controversial non-monotonic, low-dose hypothesis, but as she does with most of the science in her book, she treats it very superficially. Contrary to the well-established toxicology principle that “the dose makes the poison,” some scientists and advocacy groups allege that in certain circumstances, a different dose-response interaction can possibly occur with alleged EDCs. The researchers speculate that a more significant response could occur at a low dose, compared to that observed at higher doses. These “low-dose effects” are postulated to occur at doses well below those levels previously tested and determined to be safe by regulatory authorities. Although this non-monotonic, low-dose hypothesis is often discussed in relation to endocrine-related science, scientists at regulatory agencies across the globe, including the EPA and European Food Safety Authority (EFSA), have argued that the hypothesis could, in theory, apply to all chemicals, regardless of the mechanism by which they cause toxicity.
Some of those same regulatory agencies also have conducted rigorous reviews and have been unable to validate the non-monotonic, low-dose hypothesis using reproducible, relevant testing.
Furthermore, EPA led a work group of scientific experts that reviewed various studies, and the conclusion of the group’s draft report affirms what mainstream scientists have said for years: the purported scientific evidence for the nonmonotonic, low-dose hypothesis, even as it might apply to endocrine active chemicals, is, at best, very weak.
Evidence Supporting Similar Effects in Other Species
Swan also claims that EDCs are causing similar threats to the reproductive health of other environmental species, but once again she is guilty of selectively telling only the parts of the story that support her hypothesis and ignoring anything that contradicts it. Among the examples she cites is the work conducted by Louis Gillette on alligators which he sampled from a part of Lake Apopka, Florida that had been heavily contaminated with pesticides, many of which were eliminated from use decades ago. Hardly representative of the trace level exposures that are more likely to be endemic to their broader habitat. The current reality is that the alligator and crocodile populations in the southeastern U.S. have robustly rebounded. This has been the result of protections established since the late 1960’s to guard against the most well-established threats to their health, NOT chemicals as Swan implies, but rather due to excessive market harvesting, poaching and loss of wetland habitat.
Chivers rightfully takes Swan to task for yet another example she cites, the recently noted declines in insect and bird populations. Although the scientific community has speculated on a variety of potential causes for this phenomenon, Swan ignores any non-chemical related explanations such as climate change, loss of habitat, etc. and focuses solely on EDCs, but then presents scant scientific evidence to support it. Once again, this omission will mislead the uninformed reader to assume that the broader scientific community believes EDCs are the primary threat.
Swan’s Recommended Actions for Remedying the Problem Are Uniformly Ill-Considered
The final three chapters of the book are dedicated to Swan’s recommendations of steps that can be taken to (1) improve reproductive health through better personal lifestyle choices, (2) efforts to reduce personal exposures to purported EDCs, and (3) changes to public policy, respectively.
To their credit, the authors first recommend good lifestyle choices, presumably because these are the most well-established risk factors and are most impactful, but they never really come out and say so directly which is unfortunate. Avoid tobacco use, maintain a healthy weight, eat smartly, exercise, and avoid unhealthy stress. Of course, such actions will help prevent a myriad of other adverse health events as well. The authors miss an opportunity to prioritize even these factors according to which ones will deliver the biggest bang for the buck, which would have been helpful to readers especially since making even one major change in lifestyle choices at a time can be an exceptional personal burden for most individuals.
Next up they offer a room-by-room approach to reducing chemical exposures in your home, mostly focused on phthalates, BPA, PBDEs and pesticides. Much of this advice is neither practical to implement nor scientifically justified, and once again it is just thrown over the fence with very little prioritization as to what might be the most cost-effective steps for people to take. Those who are socioeconomically disadvantaged, and who are also most likely to have other risk factors for poor reproductive health, will find it nearly impossible to, for example, buy organic produce; buy and eat animals labeled with USDA organic seal; filter their drinking water; replace their couches and wall to wall carpeting, and so on. As such, the advice comes across as directed at a small subset of the population that can afford to obsess about such issues. She directs readers to the Environmental Working Group (EWG) website for lists of “safer” cleaning and personal products.
As Kabat points out some of the advice such as getting rid of your vinyl shower curtain, is just downright ridiculous because any off-gassing of EDCs from it has got to be infinitesimally small and of no health consequence. Kabat also takes pains to point out the lack of scientific rigor in EWG’s “safer” products assessments, writing “It’s also troubling that Kristof and Swan tout the Environmental Working Group, an environmental advocacy group funded in part by the organic industry [read the GLP Profile of the Environmental Working Group]. EWG is notorious for not following standard protocols for measurement of pesticide residues in produce. It is also notorious for putting out an annual “Dirty Dozen” list of ‘pesticide soaked’ fruits and vegetables that is widely decried as scientifically illiterate (here, here, here), including by the USDA, during the Obama Administration.”
The public policy changes Swan calls for are ill-conceived and are based on her misinformed and distorted views of current chemical industry practice and regulatory differences between the U.S. and Europe. She and Kristof both misrepresent the chemical industry’s efforts to respond to the endocrine hypothesis. I was there at the beginning in the early 1990’s and have continued to be engaged ever since, and can personally attest to the seriousness with which the industry has always taken the issue. Anything less would be a dereliction of fiduciary responsibilities. Working as individual companies and through ACC, and the International Council of Chemical Associations (ICCA), the industry has and continues to constructively contribute by helping to scientifically develop and validate proposed endocrine screens and tests, conducting those screens and tests on priority chemicals, and funding novel science and organizing scientific meetings through their Long-Range Research Initiatives. Individual companies are incorporating environmental, health, safety and sustainability concerns, including endocrine into the earliest stages of new product development, and are employing the latest predictive toxicology and exposure tools to do so. In today’s environment, it simply doesn’t pay to “rathole” money invested on prospective products that will fail to meet marketplace or regulatory hurdles.
Swan, and then by extension Kristof, also claim that while Europe and Canada are regulating EDCs, the US is not. Again, this is simply not true, and their description of what is actually happening is incomplete and distorted. For instance, she ignores the fact that the EPA has established a science-based endocrine disruption screening and testing program (EDSP)determining which substances have the potential to interact with the endocrine system. Substances with such potential may then be further evaluated by EPA to determine whether adverse effects can occur and what exposures might trigger such responses. Chemicals that are found to cause adverse effects are subjected by EPA to a comprehensive risk assessment, so researchers can understand the potential for exposure to the chemical and the likelihood of harm under real-life scenarios. This science-based risk assessment helps scientists determine the difference between the levels of exposure that can produce adverse effects, and the typical exposure levels experienced by humans and wildlife. EPA then determines if this safety standard is appropriate to protect public health and the environment, including groups that might be particularly sensitive, or if limiting certain uses of the chemical should be considered. Swan also completely ignores the considerable international efforts through OECD to advance approaches to endocrine screening and testing.
Finally, Swan also makes the case for implementing the Precautionary Principle (PP) to ban EDCs, and argues that the EU does this whereas the US does not. Such claims have been thoroughly investigated and found to be incorrect. A full discussion of the problems with the PP is beyond the scope of this review, but I refer the reader to the writings of Jonathon Adler and Cass Sunstein, who served in the Obama administration, both of whom argue the PP is flawed, incoherent and unnecessary since precaution is already adequately incorporated into existing US health and environmental laws and regulations.
n The book brings attention to infertility, a complex topic of great interest to prospective parents who are struggling to conceive.
n However, the authors dramatically overstate their case and fail to provide convincing scientific evidence to support their hypothesis that trace levels of synthetic chemicals (purported EDCs) in the environment are playing a meaningful role in causing infertility and threatening the human race with extinction.
n They undercut the biological credibility of their hypothesis by asserting that trace levels of EDCs cause a wide range of adverse health effects. They oversimplify a very complex situation that the scientific community is only now beginning to seriously investigate.
n They mislead the reader in a number of critical ways by:
o failing to acknowledge that the mainstream scientific community remains highly skeptical of their hypothesis;
o focusing on the wrong metrics (i.e., sperm, egg and fertility rather than fecundity);
o failing to provide essential context to the reader on the unreliability of much of the scientific evidence they cite;
o “cherry-picking” the evidence presented and failing to critically review it and weight it properly;
o misrepresenting mere correlation as evidence of causation; and
o ignoring the critical factors of dose and potency.
n The authors recommendations for improving fertility are not prioritized in any way, are often ill-conceived, impractical, elitist, and not scientifically justified.
n Furthermore, the authors either ignore or misrepresent current industry and regulatory agency practice which leads them to suggest unsound changes to public policy.
As a consequence of all of these problems, it should be no surprise that I cannot recommend the book. Those who may still decide to purchase and read it are urged to do so with a healthy dose of skepticism – caveat emptor!
1Although most sources I checked attribute this quote to Popper, evidently there is some dispute about its origins.
3The problem with using fertility as a measure is that there are so many other more important factors that impact it than trace levels of chemicals in the environment. For example, as Chivers also noted, the major reason scientists believe that fertility rates have dropped in developed countries are improvements in educational level, socioeconomic status, and reproductive rights (e.g., birth control) of women, and and then the resulting increase in choices they have about whether and when to have children, and ultimate family size.
4As Ioannidis notes, many researchers misapply or misinterpret statistical analysis and consequently underestimate the role of chance in producing the results they report. Confounding is a common problem in observational epidemiology studies and is a consequence of a third factor (e.g., smoking) being corelated with both the exposure agent under investigation (e.g., occupational exposure to a chemical) and the disease under study (e.g., lung cancer). If not controlled for in either the design or analysis, this third factor can mislead the investigator to believing there is a relationship between the exposure agent and the disease under study. Although investigator bias can indeed unduly influence all phases of the design, conduct, analysis and reporting of scientific studies, the use of the term bias in this context refers to the consequences of errors in measurement of exposures and disease, which actually occur much more commonly than the public might expect.
5Weight of evidence refers to a systematic approach that scientists use to evaluate the totality of scientific evidence to assess if the science supports a particular conclusion. Weight of evidence decisions do not involve a simple tally of positive and negative studies, but rely on expert scientific judgment to assess, review and integrate all of the results to form a meaningful conclusion.