Finkel and Bieniek CTE article as published 4-25-18.pdf - Adam M. Finkel

Please enable JavaScript to view the full PDF

Human and Ecological Risk Assessment: An International Journal ISSN: 1080-7039 (Print) 1549-7860 (Online) Journal homepage: http://www.tandfonline.com/loi/bher20 A quantitative risk assessment for chronic traumatic encephalopathy (CTE) in football: How public health science evaluates evidence Adam M. Finkel & Kevin F. Bieniek To cite this article: Adam M. Finkel & Kevin F. Bieniek (2018): A quantitative risk assessment for chronic traumatic encephalopathy (CTE) in football: How public health science evaluates evidence, Human and Ecological Risk Assessment: An International Journal To link to this article: https://doi.org/10.1080/10807039.2018.1456899 Published online: 25 Apr 2018. Submit your article to this journal View related articles View Crossmark data Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=bher20 HUMAN AND ECOLOGICAL RISK ASSESSMENT https://doi.org/10.1080/10807039.2018.1456899 A quantitative risk assessment for chronic traumatic encephalopathy (CTE) in football: How public health science evaluates evidence a b Adam M. Finkel and Kevin F. Bieniek a Environmental Health Sciences, University of Michigan School of Public Health Ann Arbor, Michigan, USA and Wharton Risk Management and Decision Processes Center, University of Pennsylvania, Philadelphia, Pennsylvania, USA; bNeuroscience, Mayo Clinic College of Medicine and Science, Jacksonville, Florida, USA ABSTRACT ARTICLE HISTORY How should science and policy interpret the recent finding that 110 of Received 5 February 2018 111 former National Football League (NFL) players had brain pathology Revised manuscript known as chronic traumatic encephalopathy (CTE) at autopsy? Some accepted 21 March 2018 physicians view this (and related epidemiologic and mechanistic KEYWORDS evidence) skeptically, emphasizing that the association between epidemiology; OSHA; repeated head trauma (RHT) and CTE may be artifactual, that this quantitative risk assessment; “incidence” is biased by self-selection of players with cognitive or head trauma; science-policy; emotional symptoms, and that even if RHT causes CTE, the lesions chronic traumatic themselves may be inconsequential. Public health scientists look at this encephalopathy (CTE) emerging evidence quite differently; in particular, they tend not to fall prey to certain illogical arguments justifying inaction. We present a quantitative risk assessment showing that even accounting for the non- representativeness of the 110 cases, the risk of CTE in the NFL workforce amply meets both parts of the test for “a significant risk of material impairment of health” that would permit the U.S. Occupational Safety and Health Administration to intervene to reduce RHT exposure. We further conclude that according to available evidence, CTE is a public health problem, and that lawyers and physicians need to understand that this conclusion is based on standards of evidence at least as long- standing and robust as their own. Introduction Scientific consensus about the nature and magnitude of environmental and occupational health risks is often elusive, but accruing evidence generally helps experts converge if not agree. This is not so in the case of chronic traumatic encephalopathy (CTE) as a possible grave health risk of occupations and avocations that involve repeated head trauma (RHT) – occupations that include professional contact sports (notably American football) and jobs in the logging, commercial driving, firefighting, and military sectors (Tiesman et al. 2011). In recent years, as several case series of former National Football League (NFL) players with CONTACT Adam M. Finkel afi[email protected] Clinical Professor of Environmental Health Sciences, University of Michigan School of Public Health, Ann Arbor, MI, USA and Senior Fellow, Wharton Risk Management and Decision Processes Center, University of Pennsylvania, Philadelphia, PA, USA. © 2018 Taylor & Francis Group, LLC 2 A. M. FINKEL AND K. F. BIENIEK CTE, epidemiologic studies of CTE incidence in persons with and without history of RHT, and mechanistic studies relating RHT and CTE have appeared in the literature, scientists, clinicians, and advocates have been expressing increasingly polar and irreconcilable posi- tions with respect to the three most basic risk-related questions: (1) is CTE causally related to RHT, or are the two related by non-causal association or by mere coincidence?; (2) if RHT contributes to CTE, how likely is a given amount of exposure to RHT to cause such harm? (and more specifically, what is the average CTE risk among NFL players given their exposures, and hence what is the expected number of CTE cases in this cohort?); and (3) how severe are the neurological and other consequences of CTE to those afflicted with these lesions?1 The authors – one a pioneer in methods of quantitative risk assessment (QRA) and cost– benefit analysis (and a former chief regulatory official at the U.S. Occupational Safety and Health Administration (OSHA)) and the other a neuroscientist with a special focus on CTE – observe that there is a large and growing disconnect between experts who believe that the evidence is ample to justify individual caution and possible regulatory action or “nudges” (Sunstein and Thaler 2008) to limit RHT, versus those who believe the “science remains in its infancy” (Carey 2016) and that no such conclusions can be drawn. Outside the peer- reviewed literature, diametrically clashing opinions on these questions are on prominent dis- play, and feature poignant stories by next-of-kin (Carpenter 2017) and by recently retired players dealing with what they consider to be early symptoms of disease (Utecht and Tabb 2017; Kelly 2018), contrasting with indirect denials from the League pointing out to a group of reporters, for example, that “the average NFL player lives five years longer than you” (Gantt 2017)2. This article focuses on the nearly-as-wide spread of opinion within the pub- lished literature. With respect to the first of the three basic questions above, current conclusions about the relationship between RHT and CTE run the wide gamut from “[repeated head trauma], with or without symptomatic concussions, can alter the structure and function of the brain to potentially underpin cognitive, behavior and mood deficits observed in some former ama- teur and professional football players” (Alosco et al. 2017) to “Beyond identifying the exis- tence of abnormal tau protein in a person’s brain, researchers know very little about CTE” (In re Nat’l Football League Players Concussion Litigation 2015). With respect to the second question, assertions about the eventual excess incidence of CTE among retired NFL players range from the prediction that “a shockingly high percentage” of NFL players will develop CTE (Fainaru-Wada and Fainaru 2013) to “the reporting of CTE in former professional American football players has led to widespread speculation far beyond the conclusions that can be drawn…; there is no credible data with which to establish the incidence or prevalence of CTE” in the NFL cohort (Maroon et al. 2015). And with respect to the clinical severity of CTE lesions to the individual affected, conclusions range from “The clinical features [of 1 Note that we array these questions according to the time-honored paradigm first codified in the early 1980s (National Research Council 1983), in which quantitative risk assessment begins with the “hazard identification” stage (what health effect(s) can the substance or agent induce?), proceeds in parallel through exposure assessment (how much exposure to the hazard do humans undergo?) and dose–response assessment (what is the function relating exposure to probability-of- harm?), and concludes with the “risk characterization” stage (how large and how severe are the consequences of the popula- tion’s exposures?). 2 As a reviewer pointed out, this observation is both unremarkable (NFL players surely benefit from the “healthy worker effect” and are expected to live longer than average merely because they were once fit enough to make an NFL team) and incom- plete (it says nothing about any quality-of-life differences between NFL players and the general population). HUMAN AND ECOLOGICAL RISK ASSESSMENT 3 CTE] include impairments in mood (for example, depression and hopelessness), behavior (for example, explosivity and violence), cognition (for example, impaired memory, executive functioning, attention, and dementia), and, less commonly, motor functioning (for example, parkinsonism, ataxia, and dysarthria)”(Montenigro et al. 2014), to the diametrically opposite conclusion that CTE lesions are “tiny abnormalities [that] might not have any specific clini- cal significance” (Noy et al. 2016) or that they are “merely an immuno-histochemical curios- ity” (Schnabel 2016).3 Some of these chasms are doubtlessly widened by confirmation bias – the selective or preferential marshaling of evidence to bolster preconceptions (Kahan 2013) – and of course it is possible that financial considerations, tribal loyalties, or personal agendas can have a profound influence on those preconceptions (Haidt 2012). It is also possible that we are exaggerating the size and characteristics of the cadre of CTE skeptics – that for this particu- lar disease and possible cause, only a small subgroup of the physician/legal observers see things very differently from the public health orientation.4 Nevertheless, we believe that much of this and other controversies stems from a fundamental disconnect between how most public health scientists, as opposed to some clinicians or tort lawyers, amass and interpret evidence of association, of risk, and of harm. In particular, the type and quality of evidence sought and deemed useful to avoid errors of one type only (false positives, or Type 1 errors) in clinical diagnosis and legal compensation for specific causation may be misleading and counterproductive in public health evaluation. This article therefore uses the disparate interpretations about CTE as a real-time, evolving example of the need for better communication across the range of scientific and other special- ties. We suggest that some clinicians and lawyers need a much better understanding of how public health professionals look at evidence of causality, probability, and severity, not because the public health burdens of proof are “lower,” but because they are fit for different purposes. Most generally, the science can indeed “be in its infancy” for the purpose of estab- lishing specific causation at the individual level, but at the very same time it can be more than ample to establish a public health concern and a worthy target of regulatory attention.5 And specifically in the case of CTE, as we will show, a large case series (despite being derived from a likely nonrepresentative convenience sample) can give a skewed picture of the pre- vailing risk of a disease, but at the same time it can be more than ample to establish a public health concern and a risk whose central estimate is clearly “significant.” On the other hand, we also urge public health professionals to consider more seriously what their orientation cannot offer society in terms of our understanding of specific causa- tion, mechanisms of action, and perhaps even the ability of public health interventions to 3 For a recent review article discussing the evidence linking RHT to CTE as well as some of the possible weaknesses in the cur- rent knowledge base, see Asken et al. (2017). 4 In this case, it may be that the most parsimonious division lies between physicians whose specialty is not neuropathology (e.g., concussion specialists, psychiatrists, orthopedists) and those who specialize in neuropathology and whose approach to evaluating evidence may tend to be similar to the public health orientation. 5 For a prominent example of this, consider the case of smoking and lung cancer: 54 years after the landmark Surgeon Gener- al’s Report (US Dept of HEW 1964), we still do not know with certainty which constituent(s) of tobacco smoke cause lung cancer, which target(s) in the genome are affected, the molecular mechanism by which any constituent exerts its toxic effect, whether genetic correlates place specific individuals at much higher or lower risk per unit of exposure, or whether any indi- vidual smoker, no matter how intense her exposure, got her disease there or from something else. All science can say is that it’s reasonable to believe that widespread reductions in smoking have been and would be associated with widespread reductions in lung cancer, even though one could also rightly say that “the science of tobacco and lung cancer is still in its infancy.” 4 A. M. FINKEL AND K. F. BIENIEK provide unequivocal benefits at the population level (Cox 2016; but see also Finkel 2018). With respect to specific causation, we caution that we will never know, for example, whether CTE caused Aaron Hernandez to commit violent acts (Nierenberg 2017), but we also cau- tion that it would be folly to deny that reducing RHT in football and elsewhere would be plausibly likely to reduce the incidence of CTE and its grave consequences. Although the fields of both chemical risk assessment and evidence-based medical guidelines are beginning to converge towards reliance on systematic review protocols to assess the weight of evidence supporting causality (Sterne et al. 2016), controversies remain as to whether conclusions that drive public-health decisions should strive to weigh the evidence oblivious to the conse- quences of an erroneous conclusion (Neutra et al. 2018). This article is a companion piece to a forthcoming article in the Arizona Law Review (Fin- kel et al. 2018) that discusses the legal and policy aspects of CTE and other health/safety haz- ards in the NFL. That article concluded that OSHA definitely has jurisdiction over the NFL workforce (the League’s fans call the employees “players,” but they are clearly salaried work- ers covered by the OSH Act), and that a hypothetical regulation reducing CTE risks, while politically unlikely, would likely be permissible as meeting the various evidentiary hurdles OSHA must cross in order to impose salutary and “feasible” controls.6 While that article did not go into detail on the disciplinary differences that explain the widespread skepticism that RHT is a hazard sufficiently understood and sufficiently serious for OSHA regulation, this article confronts the divergence directly. However, this article does not discuss any of the possible interventions that might reduce CTE in professional football, either in terms of their likely effect on reducing RHT or in terms of the theory and evidence suggesting that any par- ticular diminution in RHT in football might affect CTE risk. As part of this article, we provide (for the first time in the literature, to our knowledge) a QRA estimating a lower bound for the risk that an NFL player will develop CTE given current RHT exposures in the sport. This assessment does not profess to offer a central estimate of this risk, but the estimation of a lower bound does counter the statement of Maroon et al. (2015) that data do not exist to make such a calculation. We will show that this lower bound amply exceeds the 10¡3 (1 chance per 1000) lifetime-risk level that the U.S. Supreme Court opined was the boundary line below which a grave health risk might become something other than unambiguously “significant” (Industrial Union Dept. AFL-CIO v. American Petroleum Institute 1980, hereinafter referred to as “the Benzene case”). A risk of 1/1000, of course, is much higher than the level (generally on the order of 1/105 or 1/106) that Congress has instructed the US Environmental Protection Agency (USEPA) to regard as possibly de minimus, and much higher than levels usually regarded as “acceptable” for involuntary risks (Fischhoff et al. 1983). The remainder of this article has four major sections: (1) a general exploration of how public health science evaluates accruing evidence for, in turn, hazard, risk, and severity; (2) a brief review of the current state of knowledge about the neurological hazards of RHT, the risk of CTE in professional football, and the severity of CTE lesions; (3) the basis for our ten- tative conclusions that public health experts in general, and OSHA regulatory scientists in particular, would look at the current CTE evidence and conclude there is reason for 6 The article ultimately recommended, however, that various “soft law” mechanisms would likely be more useful and construc- tive here, particularly the model of an “enforceable partnership” wherein OSHA, the NFL, and the players’ union would jointly craft a code of conduct to reduce RHT. In such a model, OSHA would disclaim any ambition to write and enforce a regulation, but the private-sector parties would in turn agree that adhering to the code of conduct would constitute the employers’ “general duty” under the OSHA statute and that OSHA could issue citations for failure to perform that duty. HUMAN AND ECOLOGICAL RISK ASSESSMENT 5 concern—but in each of the three steps of the assessment we follow these appraisals with various caveats and counterarguments; and (4) a general summary about how public health and medical/legal experts might reconcile their disparate views about the adequacy of the evidence base here. In what follows, we often try to explain features of the appraisal of CTE and RHT by drawing analogies to how the well-developed methods used in chemical risk assessment han- dle hazard identification, dose–response assessment, and severity appraisal. Such analogies may seem strained, since RHT is not a toxicant and CTE is not a consequence of DNA dam- age. But readers of this journal should appreciate that our field of risk assessment has much to offer to the appraisal of certain physical stressors that may cause long-term harm. The human, animal, and mechanistic evidence linking RHT to CTE can feed into the kind of weight-of-evidence classification methods used in chemical risk assessment; more practi- cally, if voluntary or mandatory limits are ever to be set for exposure to head trauma, ana- lysts will need to develop and evaluate indices of exposure and assess their fit to various physiologically plausible dose–response models, just as we do routinely in cancer risk assessment. How regulatory science assesses emerging evidence In the courtroom or at the bedside, interested parties must grapple both with general causa- tion (is it more likely than not that particular exposures can cause a particular disease?) and specific causation (is it more likely than not that this did occur to the particular plaintiff or patient?) Public health, by contrast, is concerned with whether scientifically sound reasoning can support the key premise that exposure reductions among a population will be associated with incidence reductions among that population: for this to be true, the exposure does not have to be the sole or even the proximate cause of the disease, and we certainly do not need to know how or why the exposures are linked to adverse effects. However, we believe the dif- ference in orientation goes far beyond this, and has more to do with whether and how uncer- tainties are both acknowledged and handled, as we will discuss. The evidence public health professionals consider invariably emerges from some combi- nation of toxicologic findings (controlled studies on experimental animals), human evidence (ranging in ascending order of robustness from isolated case reports, to epidemiology studies7, to full-scale clinical trials, which obviously are rare when environmental toxicants are involved), and mechanistic evidence (e.g., in vitro studies of how a particular chemical effects its harm at the cellular level). Because this is an article about occupational risks where OSHA has primary jurisdiction, we will focus our discussion of risk assessment and science-policy practice on that agency; however, most of this discussion would apply equally to USEPA, the Food and Drug Administration (FDA), the Consumer Product Safety Com- mission (CPSC), the National Highway Traffic Safety Administration (NHTSA), and the other public health regulatory agencies, as they all developed their approaches to evaluating 7 There are two fundamentally complementary types of epidemiologic studies. “Cohort studies” compare persons exposed versus unexposed to a putative hazard, to see if the incidence of a disease is greater in the former group. “Case-control stud- ies” look at persons with a disease and a matched control group of healthy persons, to see if the former group is made up of persons more likely to have been exposed to the hazard. Since CTE can only be diagnosed definitively at autopsy, not enough time has elapsed for any true cohort studies to have been mounted – the best epidemiologic evidence at present comes from studies like Bieniek et al. (2015) (CTE incidence in subgroups with and without prior RHT exposure). 6 A. M. FINKEL AND K. F. BIENIEK epidemiologic and toxicologic evidence at roughly the same time and with very similar results (Rhomberg 1997). If (and such cases are uncommon) we had direct observational evidence of a “natural experiment” wherein small reductions (and subsequent increases) in population exposure reliably led directly to reductions (increases) in population disease incidence (Li et al 2010), or to reductions/increases in upstream biomarkers in turn known to be causally related to disease (Rich et al. 2012), there would be little controversy about the sign and magnitude of population risk as a function of exposure. But controversy would likely still remain about the probability that any identified individual would be harmed by a given amount of expo- sure (due to interindividual variability in susceptibility; that is, the risk per unit of exposure). More importantly for this article, even absent any controversy about the magnitude of an individual’s risk, doubt will always remain about whether the risk will manifest itself in harm – that is indeed the very definition of risk (a probability that either will or will not result in harm), just as the 0.50000… “risk” of a coin landing on “tails” can describe only the behavior of a large number of coin flips, never any specific flip before it takes place. But even when observers from different fields can agree to discuss population risk, not individ- ual causation, they may place very different emphasis on the twin human faculties of observation and inference. One of us (Finkel 2009) has referred to human health risk assessment as sur- rounded by a “fundamental conundrum”: namely, that risks too large to tolerate are also often too small to detect reliably. Recognizing this fact helps explain why the kinds of “proof” sometimes demanded in clinical medicine and law are discordant with what public health and regulatory sci- ence seek. To draw an analogy, when Galileo saw four moons of Jupiter (four more than anyone else had ever known were there) circa 1610, he set in motion the development of more and more powerful telescopes through which scientists detected many more moons. But even before we sent probes to observe the Jovian satellites close at hand, we also became able to reliably con- clude that additional moons existed beyond our limits of Earthbound observation, by tracking the motions of those moons we could see and inferring that perturbations in their orbits had to be caused by other satellites we could not yet see. So public health and regulatory science are comfort- able using, even relying on, indirect methods of inference. Regulatory agencies generally call these inferences “default assumptions,” and ideally they express the scientific consensus about how to interpret evidence in the absence of compelling (often case- or chemical-specific) reasons to the contrary (NRC 1994; NRC 2008). For exam- ple, OSHA and USEPA typically require that a chemical must cause a significant excess of tumors in two different species of laboratory rodent before it will treat it as a “presumed ani- mal carcinogen” (USEPA 2005). Some agencies are left to their own devices as to whether to favor “conservative” defaults that tend to infer (greater) hazard, risk, or severity in the absence of contrary information (NRC 1994, Appendices N-1 and N-2) – but OSHA has been given explicit guidance by the Supreme Court to use “conservative” assumptions, “risk- ing error on the side of overprotection rather than underprotection” (Industrial Union Dept. AFL-CIO v. American Petroleum Institute 1980). The National Academy of Sciences has twice (NRC 1994; NRC 2008) urged USEPA and other agencies to articulate all their defaults, and to use clear and predictable logic if and when they decide to depart from a default, such as a policy that alternative evidence would need to be “clearly superior” to the evidence supporting the default. The NRC has also recommended strongly that agencies explicitly state their criteria for what general quantity and quality of evidence (for each of the major defaults) will likely be needed to make a “clearly superior” or other showing. HUMAN AND ECOLOGICAL RISK ASSESSMENT 7 A brief summary of the inferences OSHA (and USEPA) generally employs is as follows: Hazard identification USEPA and other agencies often focus their concern based solely on animal toxicology data, adopting the reasonable presumption (amply validated in general terms from past experi- ence) that exposures capable of producing significant excesses of disease in other mammals are likely also to do so in humans (Allen, Crump, and Shipp 1988). Although OSHA has on several occasions regulated an occupational toxicant based solely on evidence in experimen- tal animals8, in the case of CTE the existing evidence is largely from human experience, so this article will not further discuss toxicologic “defaults.” When evaluating case reports and epidemiology studies, OSHA generally follows the mainstream of scientific opinion about how to evaluate such evidence in these ways, among others:  For an epidemiologic result to be considered “positive,” the lower bound of the relative risk estimate must exceed 1.0 (at a relative risk of 1.0, there would be no difference at all between disease or injury rates in the exposed versus the unexposed subpopulations).In other words, there needs to be (usually) a less than 5% probability that any observed excess of disease among the exposed cohort (or an excess of exposure among those with the disease, in a case-control design) could have arisen due to chance when the expo- sure truly had no effect on disease (Mitts 1998).  Other things being equal, a well-conducted positive epidemiology study outweighs one or more “negative” ones. It is, regrettably, fairly easy to conduct a human study with insuf- ficient power (that is, too few subjects, too little incremental exposure, too little time during which exposures occurred, insufficient latency between exposure and assess- ment, etc.) to detect a true positive effect. Studies that show a significant relative risk (see, e.g., the discussion infra about Bieniek et al. 2015) must still be carefully evaluated to make sure that confounding or other factors are not creating spurious results.9 Case reports alone are infrequently the basis for an OSHA regulation, primarily because by def- inition a case series cannot itself yield a quantitative estimate of risk. To estimate risk, in addition to an estimate of the numerator (the number of exposed persons whose risk has been manifest), one also needs the denominator (the number of persons in the occupation or the number exposed to the hazard), as well as the background rate – the number of persons in the numerator who would be expected to exhibit the impairment or contract the disease in the absence of the specific exposure. However, OSHA (and the courts) has deemed various case series ample for rulemaking, especially when a case cluster had appeared in a narrow geographic or occupational niche. For example, in 1977 OSHA moved quickly (nine months from initiation of rulemaking to final promulgation) to set a strict exposure limit for the nematocide DBCP (dibromochloropropane) upon learning that seven workers who handled DBCP in one plant in California had lowered or 8 Several final OSHA health standards, including those for methylene chloride (62 Federal Register, 1494–1619), methylenedi- aniline (57 Federal Register, 35,631), and ethylene oxide (49 Federal Register, 25,734), are based solely or primarily on animal bioassay data, with little or no human evidence. 9 Confounding is probably the most important flaw affecting epidemiologic studies. However, true confounding is not as com- mon as laypeople may assume, based on how frequently this criticism is invoked. For a factor to be a true confounder, it must be associated both with exposure and with effect. For example, smoking can only create a false-positive relationship between a chemical exposure and lung cancer if smoking is associated with lung cancer (which it certainly is), and if for some reason smoking is significantly more common in those exposed to the chemical compared to those unexposed (or those more highly versus those less highly exposed). 8 A. M. FINKEL AND K. F. BIENIEK absent sperm counts (Bingham and Monforton 2013).10 It took more than 10 additional years for careful epidemiologic studies to be completed on these populations, but the regulation had long since been in place by that time – so the parallel to the case series (see below) of more than one hundred former NFL players with CTE is especially instructive here.11 One relevant lesson from the instances where case reports led to correct inferences is that even if the investigator(s) offering the reports provide no information about quantitative risk, it is often possible to reliably estimate risk based on exogenous information; the denominator can often be estimated with little addi- tional effort, and the background rate can sometimes reliably be estimated to be zero or near-zero. In these situations, case series can be (nearly) as useful as full-fledged epidemiologic investigations. Dose–response assessment When prevailing occupational or environmental exposures to a hazard are much smaller than those observed in epidemiologic or toxicologic studies (or when proposed rulemaking has reached a point where an agency must select and justify a particular lower limit of expo- sure as appropriately “safe”), it is necessary to extrapolate below the region of observation using some kind of model for dose–response. In the case of CTE, of course, the very cohort (NFL workers) whose risks are being estimated is the group from whom all the case reports are derived, so there is no need to extrapolate to lower levels of RHT – until some future date when OSHA or the NFL might be contemplating specific rules or suggestions to lower the number and/or the intensity of head impacts. But in general, where extrapolation is required, agencies generally assume that carcino- genic substances exhibit approximately linear dose–responses, from the exposure level at which “epidemic” levels of tumor incidence are seen in the laboratory or in human subpopu- lations, down to levels perhaps 1–3 orders of magnitude lower. For non-carcinogenic expo- sures, public health regulatory agencies generally assume that after one finds a “no observable adverse effect level” (NOAEL) in an animal bioassay, the agency can “adjust” this level downward by a factor of 10 (to account for the fact that the typical human may be 10 times more sensitive to the effects of this substance than rodents are) and then by a sec- ond factor of 10 (to account for the fact that some humans may be 10 times more sensitive than the typical human) – although many scientists have urged that this “bright line” approach be replaced by a unified approach that models both cancer and non-cancer effects as phenomena with a continuous dose–response relationship (NRC 2008). Severity of harm Even assuming a strong and dose-related association between and exposure and an occupa- tional disease, OSHA does not have sufficient license to move towards rulemaking without a reasoned belief that the adverse health outcome at issue truly constitutes “material 10 Note that the seven workers were exposed to DBCP at approximately 0.3 parts per million; the OSHA standard required employers to lower concentrations to below one part per billion. 11 In one of the most renowned examples of vindicated precautionary action outside the workplace, the FDA delayed approval of the sedative thalidomide circa 1960, largely on the basis of a very limited case report from an Australian obstetrician who delivered three babies with phocomelia (a rare constellation of birth defects) and recognized maternal ingestion of thalido- mide as the common factor. Very few U.S. babies were affected, in contrast to tens of thousands in Germany and the U.K. where the case reports were not heeded. HUMAN AND ECOLOGICAL RISK ASSESSMENT 9 impairment of health or functional capacity”12. Occupational (and non-occupational) dis- eases all fall along a continuum of adversity, and most of the diseases OSHA has regulated to date, particularly cancer, obstructive or restrictive airway disease, and infectious hepatitis, are near the upper end of this continuum and clearly constitute “material impairment.” This is so even though some types of cancer can readily be cured, and even though some individ- ual cancers may be indolent, cause few symptoms until the final weeks of life, or even can be found at autopsy within a person whose cause of death was wholly unrelated; in each case, OSHA would consider this tumor type a “material impairment of health.” In some cases, OSHA has defined “material impairment” fairly expansively and been supported in this by the judiciary. For example, the Eleventh Circuit in 1992 upheld OSHA’s construal of even transient sensory irritation (e.g., itching of the eyes caused by chemical exposure) as material impairment, and further stated that even “purely subjective responses” could be so construed (AFL-CIO v. OSHA 1992). In general, OSHA has even construed biochemical changes that cause no symptoms as material impairment, especially when it is plausible that these are early warning signs of a latent or prodromal disease state and that future exacerbation would result in clinically salient symptomatology. OSHA’s rationale has always been that if it is possible that identifying a physiological change could trigger additional medical monitoring (or employee-specific reduction in further exposures) that could conceivably slow the pro- gression towards overt disease, then the perturbation itself should count as material impairment.13 Applying the public health paradigm to CTE Background information on CTE According to the 2017 Berlin Consensus Statement on Concussion in Sports by 36 neuro- scientists, “chronic traumatic encephalopathy (CTE)… appears to represent a distinct tauop- athy with an unknown incidence in athletic populations” (McCrory et al. 2017). The term “tauopathy” refers to one of several neurodegenerative diseases, including frontotemporal lobar degeneration (FTLD) and progressive supranuclear palsy (PSP), in which the protein tau (found in normal human brain) becomes abnormally aggregated and forms inclusion bodies within nerve and glial cells in the brain. The abnormal tau becomes burdened by an excess of phosphate groups (hyperphosphorylation). A frequent inclusion body containing abnormal and insoluble tau is the “neurofibrillary tangle,” a lesion most often seen in the Alzheimer’s disease (AD) brain.14 In CTE, these tangles accumulate in a characteristic pat- tern that is different from AD and other tauopathies. A consensus conference of 12 Section 6(b)(5) of the Occupational Safety and Health Act of 1970 (P.L. 91–596) gives OSHA the authority to set standards only when employees are in danger of suffering “material impairment of health or functional capacity.” 13 See, e.g., the newly promulgated final rule for exposure to respirable crystalline silica, where OSHA determined that the mildest possible abnormality on a chest X-ray (a divided grade of “1/0” for small opacities) would be counted as silicosis, even though at this point in the disease process the employee would have no symptoms whatsoever (81 Fed. Reg. 16826, 16837 (March 25, 2016). See also 82 Fed. Reg. 2470, 2547 (Jan. 9, 2017) promulgating a final rule for exposure to beryllium, in which OSHA concludes that early-stage chronic beryllium disease (CBD), an asymptomatic period during which small lesions and inflammation appear in the lungs, is material impairment because of “evidence and expert testimony that early- stage CBD is a measurable change in an individual’s state of health that, with and sometimes without continued exposure, can progress to symptomatic disease” (emphasis added). 14 AD also involves aggregation of hyperphosphorylated tau, but because its primary hallmark is the accumulation of fibrils of amyloid protein within plaques, it is considered a “secondary tauopathy.” CTE per se (when not co-existing with AD) does not display amyloid accumulation. 10 A. M. FINKEL AND K. F. BIENIEK neuropathologists, sponsored by the National Institutes of Health, defined the pathogno- monic lesion of CTE as hyperphosphorylated tau aggregates in neurons, astrocytes, and cell processes around small vessels in an irregular pattern at the depths of the cortical sulci (McKee et al. 2016). There appear to be two different clinical presentations of CTE (Stern et al. 2013); some cases begin relatively early in life (ages 30–50) and involve behavioral and mood disturbances with minimal cognitive impairment, while others begin later in life and display marked cognitive impairment and often motor disturbances as well. Evidence about the RHT-CTE association The evidence supporting an association between either repetitive head trauma or football (or both) and CTE, one that suggests but does not establish causality, includes case reports, epi- demiologic studies, and mechanistic evidence and hypotheses: Case reports As of this writing, there are several case series in which survivors of NFL players have sought post-mortem brain examination. The largest of these series involves the team at Boston Uni- versity, which to date has examined 111 former players; in 110 of these cases (99%), CTE pathology was found, either in isolation or co-existing with other proteinopathies, notably beta-amyloidosis (Mez et al. 2017). This very high incidence rate (see the “risk estimation and characterization” discussion below) is certainly biased by the fact that only some next- of-kin were motivated to seek a pathological explanation for symptoms they were observing or experiencing. But although the case series consists preferentially of players whose symp- toms led to a (postmortem) diagnosis of CTE, the obverse relationship (a very high preva- lence of prior symptoms among those found with CTE) is also in evidence. Notably, Stern et al. examined medical records and conducted next-of-kin interviews (in a blinded fashion) regarding 36 deceased athletes with confirmed CTE, and found that 33 of them had either the early-onset behavior/mood variant of symptoms, or the later-onset cognitive variant (Stern et al. 2013). Of the only three subjects who were judged as asymptomatic, one died at age 17, and the other two had advanced graduate degrees, possibly suggesting that “cognitive reserve” can delay or mask CTE symptoms (to paraphrase, a highly intelligent subject whose cognition has declined relative to his own baseline may still pass all available tests). Epidemiologic studies Bieniek et al. (2015) studied the brains of 1721 male subjects (from the Mayo Clinic Brain Bank in Jacksonville, Florida) who had a variety of neurodegenerative disorders. The researchers identified 66 men with a documented history of contact-sport exposure, and matched them by age and comorbidities with 132 men without such history, and 66 addi- tional controls (age- and comorbidity-matched females). When each brain was assessed for the presence or absence of CTE pathology, the results were striking: 21 of the 66 (32%) of the contact-sport group had CTE, but zero of the 198 controls did. Notably, this was so even though 33 of the controls had a documented history of a single head trauma (e.g., a fall or a motor vehicle accident) unrelated to contact-sport participation. Findings such as these have led one research group to claim (citing McKee et al. 2013, Mez et al. 2013) that “all cases of neuropathologically confirmed cases of CTE to date have had one thing in common: a history of repetitive head impacts” (Montenigro et al. 2015, emphasis added).15A statement HUMAN AND ECOLOGICAL RISK ASSESSMENT 11 this strong simply cannot be made about most other diseases, even where an overwhelming consensus exists that a particular risk factor is unequivocally associated with a particular dis- ease. For example, the two statements “all known cases of non-small-cell lung cancer (NSCLC) have one thing in common: a history of smoking” and “no case of NSCLC has ever been found in anyone without a smoking history” are both clearly false, and yet this does not shake our evidence-based belief about the causal link between one and the other. It is, of course, still possible that the CTE-trauma association is not a causal one, or can be attributed to some other factor we fail to discern (see below). Mechanistic evidence Although the science of how CTE might be associated with cranial impacts is at an early stage, there are several plausible pathologic mechanisms that could explain much of what has been observed. In particular, various research groups have hypothesized that rapid movement and deceleration of the brain within the cranium transmits shearing forces (Ghajari, Hellyer and Sharp 2017), which can disrupt the permeability of axons (individual nerve fibers); this in turn is thought to allow marked influx of calcium ions into the neurons, and concomitant release of caspases (proteins that cleave or degrade other proteins) (Stein, Alvarez, and McKee 2014). Misfolding of tau, leading to its aggregation in tangles, is one plausible consequence of this cascade of molecular events (Montenigro et al. 2014). Iso- meric conformational changes in tau are another plausible molecular consequence (Kondo et al. 2015). More recently, a study in mice (Tagge et al. 2018) suggested that head trauma may cause microvascular leakage of proteins, particularly albumin, that are known to cause inflammation; these investigators found phosphorylated tau (though not mature neurofibril- lary tangles) in mouse brains following a single head impact (see note 19 below for an analo- gous finding in humans). The plausibility of this or any alternative mechanism would be enhanced if experimental studies in animals could reliably reproduce CTE in vivo, but this evidence is decidedly mixed at present; see, e.g., Ojo et al. (2016), who report that of 12 stud- ies in rodents exposed to repetitive concussive head trauma, seven showed evidence of hyperphosphorylated tau post-exposure, while five did not.16 And although (see above) there are several plausible mechanisms explaining how repeated impacts could cause neurofibril- lary tangles, none of these mechanisms has been demonstrated reproducibly in a lissen- cephalic animal model (Goldstein et al. 2012; Kondo et al. 2015). And although the lesions said to be the hallmark of CTE seem to occur in a similar pattern in those diagnosed post- mortem, the entire human experience of roughly 200 brains studied may be insufficient to establish that this pattern is indeed necessary and sufficient to define CTE.17 Also, although researchers have developed a staging system for CTE, it is of course not possible in any indi- vidual case to compare CTE pathology at two different points in time to document the 15 But see the discussion below of the subsequent paper by Gao et al. (2017), which may provide the first case report of CTE in one individual with no history of head trauma. Also see the discussion below of the findings by Tagge et al. (2018) suggest- ing that a single severe impact can lead to pathologic changes consistent with early stages along the pathway to CTE. 16 However, it would not be unprecedented even if none of the rodent studies eliciting head trauma could reproduce the changes seen in human brains; although rodents are considered excellent models for chemical carcinogenesis, for example, one of the most notorious human carcinogens (tobacco smoke) was considered a “negative rodent carcinogen” for many decades. See this quote from Coggins 2007 (at p. 331): “Until recently, the published literature on inhalation studies with lab- oratory animals and cigarette smoke consisted entirely of negative findings, as far as neoplastic disease is concerned.” 17 In re Nat’l Football League 2015: “Beyond identifying the existence of abnormal tau protein in a person’s brain, researchers know very little about CTE.” 12 A. M. FINKEL AND K. F. BIENIEK progression from stage to stage. Therefore, it is possible that the mechanism by which impact and force increases the risk for CTE lesions may not yield an association between these factors and adverse health outcomes (Turner et al. 2015). As discussed above, public health agencies do often regulate risk factors without any specific knowledge of the mecha- nism(s) linking exposure to disease, though occasionally such information is brought to bear either to bolster the case for regulation or to allow the agency to assert with some confidence that the risk factor is irrelevant to humans or is of concern only at exposures above some threshold. Evidence about dose–response A recent study (Montenigro et al. 2016) provides further evidence that among living former football players (not necessarily players who went on to the NFL, however), a retrospective exposure index that estimates the cumulative number of impacts the player likely sustained during his participation was strongly associated with the probability of later-life cognitive and mood disturbances. Specifically, these researchers assert that the data support a “linear above threshold” dose–response function relating various adverse outcomes to cumulative impacts – below about 2000 impacts there is no apparent excess risk, but 5000 additional impacts increases the incidence of symptoms by nine-fold.18 Studies such as this one suggest that it is possible that a large number of sub-concussive impacts over a career can be as dan- gerous, or more so, than a small number of serious concussions; such a conclusion is bol- stered by the observation of Bailes et al. that football linemen (a position where sub- concussive jolts are routine but where high-speed hits are rare compared to those among receivers and defensive backs) may be over-represented among CTE diagnoses (Bailes et al. 2015)19. Additional but indirect evidence of dose–response comes from, among other stud- ies, the case series reported by Mez et al. (2017); they found a monotonic relationship between frequency of CTE and a proxy for length of participation in football (CTE found in 21% of subjects who had only played in high school, 91% who went on to play in college, and 99% of NFL players). Evidence about CTE’s severity There is an association between CTE stage and symptoms, with more severe cognitive (though not mood-related) symptoms associated with former players found to have had more advanced stages of CTE (Stern et al. 2013; Mez et al. 2015). In addition, CTE has been shown to cause injury to axons, ranging from focal axonal injury in the cerebral cortex and white matter in CTE stages I and II, to more extensive, diffuse axonal loss in the cortex and 18 Indeed, Montenigro et al. found a strictly monotonic dose-response relationship for six of six different adverse outcomes, as a function of six different levels of estimated lifetime G-forces sustained in sport impacts. A public health regulatory agency would consider the consistency of these demonstrations to be further evidence of a true association between impacts and symptoms. 19 On the other hand, at least one recent study (Hay et al. 2016) has suggested that although neurocognitive deficits following a single moderate or severe traumatic brain injury have generally been assumed to represent Alzheimer’s disease, the pathology in some of these patients may in fact more closely resemble CTE. Therefore, the “linear above threshold” function posited by Montenigro et al. may not apply when a single impact with substantial force is involved (there may be a thresh- old in terms of cumulative G-force, but perhaps not if the exposure is measured in number of impacts). HUMAN AND ECOLOGICAL RISK ASSESSMENT 13 white matter in stages III and IV (Stern et al. 2013); causal injury to neural pathways would presumably have CTE represent more than an “immunohistochemical curiosity.” Tentative conclusions about CTE as an occupational disease worthy of attention Hazard identification In evaluating human data, OSHA and other public health agencies recognize three related and very basic truisms about diseases whose risk is increased by specific exposure(s). We bring these up largely because of comments on a section of the companion article (Finkel et al. 2018) made by physicians asked to review it; they objected to expressions of regulatory concern about RHT and CTE (and to assertions that the CTE lesions are harmful), using all three of these incorrect arguments. First, for virtually every disease causally related to a particular exposure, it is always the case that some individuals will be found who have sustained the exposure(s) but have escaped the disease, as well as others who have succumbed to the disease but have never been exposed. This follows directly from what “risk” most fundamentally means: exposure increases the proba- bility that disease will occur, but increases it from some non-zero background level upwards to some non-inevitable new level. Put more colloquially, even for an association like that between tobacco and lung cancer (one now universally accepted as scientifically sound), it is completely unremarkable, and in no way exculpatory to the dangers of exposure, that there must and there will be found persons who have smoked and never developed lung cancer, and others who have developed lung cancer and never smoked. With regard to CTE, we actually are not sure even of a single case of CTE in someone without RHT, although one case report in one individual (an autopsy on a patient with ALS (Gehrig’s Disease) that found CTE lesions) may be the first such instance (Gao et al. 2017). But in any event, the study of an unexposed person (or an unexposed population) cannot by definition change our views about the qualitative association between the exposure and the disease.20 Therefore, the statement by Castellani (2015) that “athletes… may find solace in the fact that one can survive a career in the NFL neurologically intact and with no significant proteinopathy” (based on one autopsy of a middle-aged NFL lineman who died of nat- ural causes) may provide evidence that the human brain can “endure protracted physical punish- ment in the setting of high-energy collision sports” – but it does not, we hope needless to say, provide evidence about the proportion of a population of human brains that will escape such pun- ishment unscathed. The second truism involves a similar pair of logical fallacies when considering the rela- tionship between a health effect and its consequences, and so we will mention it in the sub- section on “severity of impairment” below. Third, it is always possible that there is something special (usually, a genetic factor) about the “subset” of individuals with both exposure and disease. But this does not mean that “only certain people will succumb” – again, risk is a probabilistic increase in a process that is 20 Gao et al. (2017) wrote that “To date, repetitive traumatic brain injury has been shown to be associated with no neuropatho- logical changes, with CTE alone, with CTE and another neurodegenerative disease, or with non-CTE neurodegeneration.” But this is no different from saying “To date, smoking has been shown to be associated with perfect health, with NSCLC alone, with NSCLC plus bladder cancer, and with bladder cancer alone.” In neither case does the existence of individuals who fit any of these three patterns cast any doubt on the association at the population level between exposure (to RHT or tobacco) and a specific lesion (CTE or NSCLC). 14 A. M. FINKEL AND K. F. BIENIEK at least partly stochastic. So it is also completely unremarkable to see persons with similar exposures, some with and some without the disease of concern, whether or not there is any genetic factor at work.21 And so, a public health agency like OSHA would clearly regard RHT-CTE as an associa- tion of concern. It could, however, recognize these possible counter-explanations and caveats (that is, keeping the door open that contrary evidence might rise to the level of supplanting the default conclusion):  Some other factor, unrelated to football and cranial impacts, may be partially or wholly responsible for causing CTE. It is possible that some of the reports (by survivors of play- ers found to have CTE) of a past history of cranial impacts were artifacts of recall bias (the tendency to preferentially remember non-existent or rare exposures when prompted to confirm or deny these after a diagnosis) (Silverberg et al. 2015). However, it is unclear whether recall bias matters at all for a disease that might be more related to subconcussive impacts (by their nature not subject to recall, biased or not) than to memorable impacts. But if impacts are spuriously associated with CTE, then a search for alternative factors may come into play. Various researchers have posited that CTE may be related to normal aging (McCrory et al. 2017) or to opioid abuse (Ban et al. 2016). However, the former explanation is undercut by the greater incidence of CTE in younger persons exposed to contact sports as compared to older and unexposed per- sons. The latter explanation is plausible given that opiates are known to increase tau concentrations and lead to neurofibrillary tangles (Stein et al. 2014), and opiate use may plausibly be a true confounder if it also turns out to be more prevalent among for- mer NFL players than among the general population, but in order for OSHA to dis- claim interest in head trauma as a risk factor for occupational disease, research would need to show that opiates alone could explain the bulk of the excess incidence. Further- more, there is a wealth of evidence, including sound epidemiological findings, that dementia and neurodegenerative pathology consistent with CTE, previously referred to as “dementia pugilistica” in boxers, can also be observed in other professional athletes whom have experienced numerous repetitive head injuries, as well as military personnel exposed to RHT, in which blood levels of tau have been found correlated with the sever- ity of neurologic symptoms (Olivera et al. 2015). Thus while CTE may not be specific to football, the cranial impacts sustained by its participants are intrinsic to the sport and may serve as a prominent cause of CTE.  Only “genetically susceptible” players are at risk of CTE. For Alzheimer’s disease, we know that one of the major alleles (epsilon 4) of the apolipoprotein E (ApoE) gene is associated with higher risk. However, Maroon et al. reported that their analysis of all 21 More importantly, even if there is a genetic component, public health agencies never abandon their interest in exposure while awaiting word about the genetic contribution. Public health regulatory agencies like OSHA cannot influence people’s genetics, but sometimes can influence their exposures. Therefore, they bear in mind the tenet that “genetics may load the gun, but environment pulls the trigger” (this tenet was apparently originally coined by Prof. Judith Stern of the University of California at Davis, circa 2000) – their mission is to explore cost-effective ways to reduce the population prevalence of dis- ease by controlling one or more factors that will do so, regardless of whether other factors might also do so. Even where sci- ence strongly suspects that persons with a genetic variant are at much higher risk per unit of exposure than others are (e.g., for chronic beryllium disease), regulatory agencies concentrate on controlling the exposure, which of course is the most direct means of reducing for risk for both those who may be susceptible and those who may be “resistant.” And as a policy matter, OSHA and other agencies even regulate the exposures of both women and men to toxicants suspected of causing damage primarily to the developing fetus (cf. lead); male workers are presumably wholly “resistant” to these effects, but the employer is not permitted to establish male-only areas of the establishment where exposures can be higher. HUMAN AND ECOLOGICAL RISK ASSESSMENT 15 confirmed cases of CTE (football and non-football) “revealed no significant group dif- ference between ApoE4 carriers in the CTE population and the general population (p D .26), suggesting that ApoE may not be a significant risk factor for the development of CTE” (Maroon et al. 2015). Even if such an association is later found, as discussed above, this would not change the regulatory response, if any, to the situation. Interindi- vidual variability in risk per unit of exposure is the rule, not the exception, so this repre- sentative statement by Bailes et al. (2015) misunderstands the nature of exposure and risk: “Despite the large number of people exposed to concussive and subconcussive injury, CTE appears problematic for only a small subset of the population exposed to neurotrauma.” Again, the “subset” of exposed persons who manifest a disease are sim- ply the ones who were unfortunate enough to have their risks (potential for harm) borne out as harm – one would never sensibly say that “only a subset of skiers caught in avalanches find them problematic.”22 Indeed, unless OSHA could devise some kind of regulatory or other controls that would only apply to “susceptible” individuals (and yet not run afoul of their equal-opportunity rights23), the existence of genetic suscepti- bility would probably increase the stringency of regulation to reduce CTE risks. Dose–response assessment The confidence we can have in any risk estimate for CTE in a particular occupation would be significantly increased if we had multiple data points for CTE incidence as a function of RHT exposure that could be fit to some kind of dose–response relationship. At this time, however, there are no studies (or combinations of studies) that provide the basis for observing changes in incidence or prevalence of CTE as a function of RHT exposure. There is at least one study (Montenigro et al. 2016) that demonstrates a monotonic dose–response (their Table 6) over six strata of a cumulative exposure index (estimated number of head impacts in youth, high school, and college football, by position and number of years played), for each of six different outcome measures – but all of these measures involve symptoms as the outcome variable, not the CTE lesions themselves. However, while a well-fitting dose–response model is useful for risk assessment, it is by no means essential for the purposes of considering an exposure to be a public health concern and for proposing a safer exposure level. Because cancer risk (in the “high risk” region) is generally believed to obey a roughly linear dose–response relationship in the absence of information to the contrary, regulatory agencies can extrapolate from one point where expo- sure and cancer incidence are known, and posit a lower exposure limit based on linearity and on consideration of where risks become “acceptable” or controls become economically 22 This point was made particularly well by eminent epidemiologist Sir Richard Peto in 1977: “It is a common misconception that because many smokers do not develop lung cancer there must be constitutional or environmental reasons why they do not… Even in A.D. 3000 when all the details of cellular susceptibility and environmental and metabolic peculiarity have been elucidated, a complete and full description of the process of cancer induction will still require that good and bad luck be invoked to explain why my brother got cancer and I did not… There will still be unexplained stochastic variation which it will be scientifically worthless to investigate, in exactly the sense that it would be scientifically worthless to ask, once the molecular basis of Mendel’s laws is properly understood, what was the reason why a particular child born to two parents each with one recessive gene for red hair did not, in fact, have red hair.” (emphasis added) 23 See the 1991 decision in UAW, et al., v. Johnson Controls Inc. (499 U.S.187), which upheld OSHA’s right to require lead expo- sure reductions that protected pregnant women as well as all other workers, on the grounds that excluding women from high-lead areas was not an acceptable means of control. 16 A. M. FINKEL AND K. F. BIENIEK infeasible (this procedure is, of course, less precarious when the frank-effect exposure level and the regulatory level are not too far apart). And for non-neoplastic diseases, the accepted paradigm involves finding a single NOAEL or “lowest observed adverse effect level” and applying one or more adjustment factors to derive a lower exposure level thought to be below a population threshold exposure and to have an ample margin of safety associated with it. Such a method could be used with respect to CTE (see below).24 Risk estimation and characterization In order to meet its legal requirements to consider regulating, OSHA would have to show that the status quo risk of CTE in the NFL is “significant” per the Benzene decision – that is, above one chance per thousand over a working lifetime (although lower risks can still be considered significant per this decision). But how could OSHA possibly estimate the risk of developing a condition during life that at present can only be diagnosed post-mortem?25 Risk (the probability of harm) can reliably be extrapolated from controlled bioassay experi- ments on animals, because these studies are designed to estimate the fractional incidence of health effects in a known population – but no such data are available for CTE. Case series (see above) are also insufficient to yield a quantitative risk estimate: for that we need a numerator (the number of cases of CTE, over and above what would be expected as the background risk in an unexposed population) and a denominator (the number of persons, or person-years of exposure, from which those cases in the numerator emerged). As discussed above, we have every reason to believe that a random sample (as opposed to the biased initial sample) of players would show less, perhaps far less, than the 99% incidence rate documented by Mez et al. (2017). Yet we can use the 110/111 statistic to show that the ultimately correct estimate of CTE incidence in the NFL must in fact be greater than 1/1000, as follows. Suppose that, contrary to any reasonable expectation, there will not be another NFL player active in the relevant time period (see below) found with CTE (either at autopsy or if a reliable in vivo test is finally developed). Then the question for OSHA would be how the “lowest bound” incidence of 110 cases out of N total players compares to the 1/1000 bench- mark. The easiest (and most favorable to the possibility that the CTE risk is not significant per Benzene) approximation would be to simply note that about 26,000 players have ever appeared in a league game during the past 50 years (Nixon 2014). One hundred and ten cases out of a largest-possible universe of 26,000 equals a risk of 4.2 per thousand, well above the Benzene benchmark. But this estimate of the denominator is clearly biased upward, since many of these 26,000 players appeared in only one or several NFL games before being cut from their teams. A 24 In cases where genetic susceptibility may play a major role, as exemplified by chronic beryllium disease, observational stud- ies sometimes find higher incidences of disease among subgroups exposed to lower exposures, presumably because these subgroups were systematically (or randomly) “enriched” with more susceptible individuals. So applying the margin-of-safety concept to a subgroup with a measurable increase in disease may fail to protect those exposed to lower “safe” levels, because these levels may in fact still be capable of yielding unacceptably high incidences among susceptible individuals. 25 Although a definitive diagnosis of CTE, based on confirming the characteristic pathology, can only occur post-mortem, this does not mean that the disease cannot be “diagnosed” via the standard judgments and heuristics that define the art of dif- ferential diagnosis. Indeed, since along with signs, symptoms, and medical history, there exist in vivo tests for several of the diseases that could explain a patient’s condition, negative results on these tests along with clinical judgment can allow for a reasonable diagnosis of CTE during life. HUMAN AND ECOLOGICAL RISK ASSESSMENT 17 public health regulatory agency like OSHA would never estimate lifetime risk by counting all exposed persons, but would estimate the number of person-years of exposure. So a more rea- sonable approach here is to consider the number of player-careers that accrued over the rele- vant time period of 1963–2008.26 There exists some controversy about the average length of an NFL career (NFllabor.wordpress.com 2011): the league uses a figure of 7.1 years, while the NFL Players Association commonly cites a shorter career of 3.5 years. Either way, for an average career of X years, an average of (1/X) players enter or leave the league each year (the league is a “closed system” with a constant total number of players). Using a complete dataset from the NFL on snaps per player (NFL undated), in the 2016 season, 1334 of the 2,275 players who appeared on an NFL field for at least one snap played more than a trivial number (10%) of their team’s offensive or defensive snaps. So 1334 play- ers per year with a “churn” of (1/3.5) players per year would yield 17,150 player-careers in the past 45 years; assuming a longer career length of 7.1 years, there would have been 8,455 player-careers in the same period. Therefore, the range of “lowest possible risk estimates” for this cohort would go from 110/17,150 (6.4 per thousand) to 110/8,455 (13 per thousand).27 Both these figures are far above the 1/1000 benchmark. But even if this estimated risk of roughly 6 to 13 per thousand was not a lower bound, but a central estimate, it would nevertheless be biased low for OSHA’s pur- poses. OSHA seeks (and this science-policy choice has never faced judicial challenge) to reduce significant risks not merely over an average working lifetime, but for a num- ber of years that an employee could reasonably choose to work; in other words, if a dis- ease risk would become only significant after (say) 20 years of exposure, OSHA does not ignore that risk because “on average, people don’t work that long,” but seeks to empower workers to have both long and healthy careers if they so choose. OSHA has always used 45 years as its standard estimate of a working lifetime, on the grounds that although this figure is exceeded by some (anyone who works from age 18 to 63 or lon- ger exceeds it), it is a “reasonable upper bound.” OSHA would certainly not assume the annual CTE risk in the NFL should be multiplied by 45, but would likely choose some “reasonable upper bound” such as 10 seasons as its “working lifetime” in the league. Therefore, the best way to estimate the denominator would be to tally up the number of player-seasons accrued during 1974–2009, and then divide that number by 10 to yield an estimate of “player-working-lifetimes.” We do not have access to a com- plete dataset of all NFL players by the number of seasons each played, but presumably if OSHA was to initiate pre-regulatory analysis of CTE, it would produce such an 26 Using fully 45 years of player experience is a conservative (biases low) choice, but it is reasonable in light of the fact that the player within the CTE numerator who was probably the first to appear in the league (John Mackey) began his career in 1963, and perhaps the youngest and most recently diagnosed player with CTE (Paul Oliver) entered the League in 2008. In other words, the 110 cases in the numerator all occurred among players who were in the league between 1963 and 2008, although virtually all of the 110 presumably involve players who were active during a shorter window within this 45-year period. 27 We also estimated risk more simply, by observing that at any given time, there are approximately 1000 “regular players” active in the NFL: 22 players on the offensive plus defensive squads, plus perhaps 10 others from the 24 remaining players (46 players are dressed for each game) who garner significant playing time, for each of the 32 teams (32 £ 32 D 1024). So if the average career is 7.1 years, there are (1024/7.1) players starting or ending their careers each year, or about 150 players, for a total of 6,490 player-careers of experience over the 45-year period; if there is more “churn” in the NFL (average career of 3.5 years), the total number of player-careers would instead be 13,166. Even using the larger denominator, the incidence of CTE could be no lower than 110/13,166, or 0.0084, which is more than eight times higher than the 1/1000 benchmark (and using the league’s estimate of churn, the incidence would be 16.9 per thousand). These figures are very similar to those we derived using player-snap data. 18 A. M. FINKEL AND K. F. BIENIEK estimate, which would certainly be smaller than the denominators of 8,455 to 17,150 above. So however one construes the denominator, the risk of (110/N) is clearly substan- tially greater than 1/1000, even if no additional CTE cases will ever be found in the 1963–2008 cohort.28 We offer two caveats to the calculations above. First, strictly speaking, they do not esti- mate the excess risk of CTE above the natural background rate, if any, in the majority popu- lation unexposed to repeated head trauma, but rather the rate including the background. We have little scientific basis to estimate the background rate at present. The only article that bears directly on this issue, to our knowledge, is a 2015 letter report from the U.K. that found CTE pathology (though exclusively the two earliest stages of the disease, not stages III or IV) in 32 of 268 brains (12%) from subjects with neurodegenerative diseases (Ling et al. 2015). This rate is clearly biased upward, as many of the positive cases had precisely the same his- tory of RHT that contributes to the “exposed” rate estimated above.29 However, Ling et al. also found lesions that might correspond to CTE in 6 of 47 (13%) of the “control” subjects – but it is not at all clear how these controls came to the brain bank and whether they were unexposed, asymptomatic, or disease-free (and they were all older than 60 at time of death, so they would not contribute to the proper comparison rate for those NFL players who died with CTE at younger ages).30 Nevertheless, it is conceivable that if few or no additional cases of CTE are found in former NFL players, the incidence in that population will eventually be found not to be as highly significantly elevated above the true background rate. In addition, there is an additional (a legal/regulatory) issue about the background rate that OSHA might have to consider. Strictly speaking, the excess CTE incidence due to employment under OSHA’s jurisdiction would be that above the specific background rate of CTE including persons exposed to RHT before joining the workforce. There certainly exists evidence that playing college-level football is associated with CTE: in the Mez et al. series (2017), 48 of 53 subjects who played football in college but not in the NFL had CTE, a lower percentage than the 110/111 who did play professionally, but still very high. However, it should be noted that the association with “severe” (as opposed to “mild”) CTE pathology was much stronger among the 110 professional players; 86% of them had severe CTE, versus 28 A reviewer suggested a very worthwhile alternative calculation: that we try to calculate the proportion of deaths in which CTE was found at autopsy versus the total number of deaths from all causes among the relevant subgroup of NFL players (namely, all players who entered the League between 1963 and 2008). The calculations in the main text estimate CTE deaths among all players, the vast majority of whom are still living and who may therefore add to the numerator, so the alternative presented in this note assumes that when all of these players eventually die, the proportion of them who will have CTE at autopsy would be the same as the proportion found to date. This likely introduces some unknown amount of upward bias, in that CTE very likely caused many of the 110 documented cases to die before their normal life expectancy, “enriching” this subgroup with CTE cases. According to Nixon (2014), 14,658 players entered the League between 1963 and 2008; of these, 711 had died as of June 2014. Therefore, 15.5% (110/711) of the deaths in this cohort involved CTE cases. This corresponds to a lifetime risk (in the sense of “what will your cause of death eventually be?”) of 155 per 1000, which is many-fold higher even than the substantial working-lifetime estimates presented in the main text. Note that the larger figure of 1300 deceased NFL players cited in Ward, Williams, and Manchester (2017) is not the proper reference group, since many of these players were active long before 1963 and were never suspected of having CTE or autopsied – they are part of the denomina- tor but not the numerator. 29 In addition, Ling et al. note that even some of the cases without documented head trauma may have experienced trauma, since one of the major diseases selected for in this brain bank (progressive supranuclear palsy) leads to “early and frequent falls” that themselves could have caused CTE pathology. 30 In addition, the pathology Ling et al. depict in their manuscript is not characterized according to consensus standards for describing CTE, in that the authors only illustrate perivascular glial pathology that in and of itself is not specific to CTE and can be found in the elderly (in the elderly it is known as “aging-related tau astrogliopathy,” or “ARTAG”; see Kovacs et al. 2016). The mean age of death for the “CTE” cases in Ling et al. is 81, which is much older than other CTE studies and raises the issue of whether they are not measuring CTE but rather ARTAG. HUMAN AND ECOLOGICAL RISK ASSESSMENT 19 56% of the 48 college players with CTE lesions (Mez et al. 2017, Table 1). It may also be the case that the additional RHT experienced during professional football may be more damag- ing (per unit of exposure) to players who began tackle football before age 12 (Stamm et al. 2015). So this situation contrasts with most chronic exposures of concern to OSHA, where the agency can safely assume that exposures began coincident with joining the workforce (few if any students or young volunteers are exposed to concentrations of carcinogenic chemicals like those found in the workplace). Therefore, OSHA might find it necessary to subtract out an estimate of the “pre-work” risk from the numerator of its excess risk calcula- tions, as any OSHA regulation covering employees would not benefit some who came to the league with incipient disease. Again, it is unclear whether such a refinement could possibly move the estimate of excess OSHA-covered risk below 1/1000, given by how much the face- value estimate exceeds that benchmark.31 We also emphasize that the consideration of “pre- work” risk only applies to OSHA; general public health science-policy would of course con- cern itself with all RHT exposure regardless of age or employment relationships. Severity of impairment Compared to the many and varied conditions OSHA has in the past regarded as “material impairment,” OSHA would clearly regard cerebral pathology as material impairment, and reject any notion that the CTE lesions are “tiny abnormalities” or “an immunohistochemical curiosity” (see above). In recent years, evidence has accumulated that pathologic protein, including tau in CTE, at a single site may spread or propagate to other areas resulting in widespread damage to the brain (Meng et al. 2017). Such evidence may be bolstered by the recent finding that extracts of phosphorylated tau from CTE patients can infect human embryonic kidney cells (Woerman et al. 2016). Moreover, CTE is at least statistically associ- ated with a constellation of symptoms that are in several ways more grave than any disease OSHA has yet regulated; compared to cancer, for example, progressive cognitive and behav- ioral impairment is incurable, generally with a prolonged period of time between debility and death, and can result in grave collateral harm to the victim’s family or bystanders32. In any event, the emphasis of some physicians that there is not a one-to-one concordance between CTE and symptoms is a fallacy similar to the two we mentioned above. It is always possible to find someone without a particular disease or lesion who evidences the same symp- toms as someone with the disease/lesion, just as it is unremarkable to find persons with the lesion who experience no (detectable) symptoms. Again, this follows directly from the defini- tion of “severity,” which is rarely binary, and from the co-existence of multiple factors that can lead to any particular symptom. To return to the NSCLC analogy, the fact that many non- smokers can develop nonmalignant diseases that also cause shortness of breath in no way casts doubt on the ability of lung tumors to cause this symptom – and the fact that some lung 31 It is also not completely clear whether OSHA can, or in the future could, consider collegiate student-athletes as “employees” for the purposes of OSHA jurisdiction (see Willborn 2014–2015, who makes a case that they are employees for the purposes of the National Labor Relations Act). In that case, the proper excess risk estimate might construe risk as above the back- ground rate among those unexposed to RHT before college. 32 OSHA might well be especially concerned that behavioral symptoms (especially violence towards oneself or others) possibly associated with CTE may put others at risk. OSHA has in the past paid special attention to risk factors that can leave the workplace and be “taken home,” particularly certain toxic substances like lead, beryllium, and asbestos that can increase risks among cohabitants (OSHA 2014). 20 A. M. FINKEL AND K. F. BIENIEK tumors can follow a moderately indolent course in no way detracts from their ability to be quite grave in other patients. So to say, as Solomon recently did (2018) that “it is incumbent on us [clinicians] to focus on the relevant, multivariate factors in the long-term outcome from sport-related concussions, and not to reinforce the linear thinking of concussions or subcon- cussive impacts as the cause of all neuropsychiatric ills” seems to us a strawman. Just as no one has made the claim that RHT is the cause of all neuropsychiatric symptoms, no one has said that benzene is the cause of all leukocytosis (elevated white blood cell count). Nevertheless, while leukocytosis in a patient with known exposure to harmful bacteria would normally lead to the diagnosis of an infection, a high white count in a patient with known exposure to ben- zene would instead increase suspicion for leukemia – and in neither case would the fact that there are (at least) two independent possible causes lead to dismissing the health effect itself. There are two more reasonable counter-arguments to the conclusion that CTE lesions are consequential:  It is possible that the only tangible manifestations of CTE – the neurofibrillary tangles depos- ited in particular patterns – are not in fact capable of causing symptoms or deficits. However, it is well recognized that Alzheimer’s disease lesions are present in the brain many years before the onset of clinical symptoms, and so a finding of one or more asymptomatic persons with CTE lesions would be unremarkable (see above) as well as perhaps premature. It is true that not all pathologic findings are harmful (e.g., some cysts, some anatomical variants like the “empty sella turcica,” and goiters are compatible with normal lifespan and function). But few if any truly benign lesions exist – many “benign” tumors, of course, are life-threatening even though incapable of metastasis. Clinical evidence that CTE might be, akin to certain thyroid nodules or skin lesions, truly inconsequential, would be of great interest to OSHA but would be an exception to a long-standing clinical presumption (Dwass 2015). And we emphasize that the claim that the observed symptoms are not caused by the CTE lesions is logically incompatible with the other argument (that some of the same critics use) – that the Mez and other case series are biased high due to self-selection. It is only the presence of symptoms that would impel any self-selection, so either the 110/111 is unbiased (which we doubt), or the symptoms are real and associated with CTE.  The association between reported symptoms and CTE may be artifactual. Although there is as yet no specific evidence supporting this claim, it is possible that some unknown factor, perhaps atrophy of the pituitary gland (Meehan et al. 2015), might co-exist with CTE and be the actual cause of symptoms that only appear to be caused by the CTE lesions. Note, however, that for this sort of explanation to rebut a presumption that CTE lesions constitute material impairment, the hitherto-unknown factor must be plausibly uncorrelated with repeated head trauma.Otherwise, the explanation would amount to “correcting the incor- rect name we were using for this football-related disease.” If head trauma causes pituitary atrophy, and this pathology is the source of players’ symptoms, then OSHA might reason- ably seek to reduce that impairment. In that case, we would be taken back in time several years to 2014, when the NFL hired an actuarial firm who issued a report estimating (albeit for the purpose of trying to demonstrate that the monetary amount set aside to pay claims was adequate) that an upper bound of 28% of all former NFL players would develop Alz- heimer’s, Parkinson’s, dementia, or ALS (but not including CTE) (Segal Group 2014). The figure of 28% is, of course, 280 times higher than OSHA’s 1/1000 benchmark — so even if CTE itself is somehow not “material impairment,” the League itself has suggested that a large percentage of its players may develop conditions that clearly are. HUMAN AND ECOLOGICAL RISK ASSESSMENT 21 Conclusions Many books and articles (Michaels 2008; Oreskes and Conway 2010) have chronicled the system- atic attempt to undercut public health inference by: (1) casting unwarranted doubt on existing evi- dence; (2) offering exculpatory theories without ample data to support them; or (3) fabricating data. In the case of CTE and football, some skeptics may be motivated by the same impulses as purveyors of doubt in the tobacco and like industries – protecting an industry over protecting those harmed by it – but it is important to distinguish sowing of doubt within the public health paradigm from what we have here, which largely involves incomprehension of or discomfort with the paradigm itself. As in the classic parable involving blind men examining an elephant, none of whom are entirely wrong, CTE in football can resemble a rope (if all you are examining is its tail) or a tree trunk (if you encounter only its leg), because it is both, and more. We believe that the authors and advocates who view CTE as not clearly related to football and/ or not clearly a grave impairment of health may be espousing one or more of these four coherent perspectives that are different from, or incompatible with, the public health risk paradigm:  Instead of relying on science-based default assumptions that tend to be “conservative” and precautionary, they adopt the obverse stance and view associations as artifactual unless compelling evidence emerges to “prove” causality;  Adopting the perspective of an individual patient they might be treating, they may view deferring a diagnosis as prudent and superior to offering a tentative and perhaps erro- neous one – whereas in population health, we are more attuned to the social costs of delaying actions that may be preventive;  Similarly, they may regard an incorrect intervention based on a false-positive diagnosis as a violation of the “first, do no harm” credo, whereas public health considers that in specific cases the untoward costs of risk-reduction measures may well be smaller than the expected benefits of acting even absent perfect information; and  They may fall prey to a different “availability heuristic” (Kuran and Sunstein 1999) than the one that normally affects risk assessment – rather than seeing patterns of harm whose repetition makes the cases more salient than they perhaps should be, they instead fixate on patterns among persons who do not develop the disease or the symptoms their exposure “should” result in, clouding their ability to see the entire body of evidence. The dichotomy between the fundamental public health risk question (“what is the proba- bility that exposure X at level Y will cause disease in people?”) and the clinical/legal question (“did XjY cause disease in this person?”) should be easy to understand, although one recent development unrelated to CTE has muddied these waters.33 RHT clearly is associated with CTE lesions – so arguments that we do not know enough to be sure that RHT causes CTE are very important at the bedside and in court, but are simply inappropriate for discussion of whether to consider population-based interventions and rights to know. In contrast, the arguments over whether CTE lesions are truly consequential to the person(s) afflicted seem to violate both public health logic and most norms of medicine and law. The 33 In 2012, OSHA revised its 20-year-old Hazard Communication Standard to align with the Globally Harmonized System (GHS) developed by the United Nations. The GHS requires that containers of known human carcinogens have warning labels on them stating “Danger: May Cause Cancer.” This is a true but inane statement, as it replaces information about something that definitely causes cancer in people with the reassuring information that it only “may” cause cancer in you, the individual reader of the label. There are virtually no substances, even at extremely high levels of exposure, that “will” cause cancer in every single person so exposed, so the “may” information abruptly changes a playing field that U.S. companies had under- stood for decades from a public health warning to a clinical warning. 22 A. M. FINKEL AND K. F. BIENIEK simple question for those who offer CTE as inconsequential is “would you be indifferent between having the lesions and not having them?” and we have heard no response from any skeptics that they themselves would be indifferent. Not every harm, of course, is equally grave, but while hav- ing a lesion or disease that may be inconsequential is preferable to one known to be grave, it is still worse than having no “immunohistochemical curiosity” at all. Most societies in the developed world strive to reduce risks, when feasible and cost-effec- tive, because a probability of something irreversible is of concern to individuals and popula- tions. Along the pathway from exposure to risk to consequence, intervening steps may mitigate its seriousness. An exposure may confer a small excess probability of harm, the probability may be uncertain and may even be zero (if the relationship is artifactual rather than causal), or the severity of the harm may be assumed rather than known. These factors may make it less important to control the exposure, but not necessarily unimportant. In par- ticular, knowing that a risk is “much smaller than a biased sample would suggest” may be of little comfort if this in fact means “much smaller than an apocalyptically high incidence.”34 Similarly, it may be true, as NFL Commissioner Roger Goodell has said (Belson 2016), that “There’s risk in life; there’s risk in sitting on the couch,” but there are of course a myriad of ways to reduce the risks associated with sedentary activity without courting different and possibly more grave risks. What matters is what we do while we await (and, one would hope, accelerate) the arrival of science that is no longer “in its infancy.” There is value in emphasizing the optimistic pos- sibilities as to why concern over RHT may turn out to be needless, but urging others to tem- per their concern with skepticism can lapse into hectoring them to withdraw their concern until there is unanimity. Clinicians and public health professionals both strive to “do no harm,” but both fields (should) realize that while any intervention, at the personal or societal levels, may cause some harm, greater harm still may come if uncertainty (real or manufac- tured) makes us paralyzed and unable to act on what we know and on what we reasonably believe. Acknowledgments The first author thanks Chris Deubert and I. Glenn Cohen (both formerly with the Harvard Football Players Study), John Evans, and Sidney Shapiro for helpful comments during the drafting of this arti- cle; we also appreciate the improvements suggested by two anonymous reviewers. Funding Dr. Bieniek is supported by the Florida Department of Health, Ed and Ethel Moore Alzheimer’s Dis- ease Biomedical Research Program (7AZ08) and the Mayo Clinic Younkin Scholars Program on Syn- aptic Biology and Memory. This publication was in part made possible by CTSA Grant Number UL1 TR002377 from the National Center for Advancing Translational Sciences (NCATS), a component of the NIH. Its contents are solely the responsibility of the authors and do not necessarily represent the official view of NIH. 34 For example, a person might have to drink the proverbial “case of diet soda per day” to match the exposure of rodents in toxicology studies, but if the cancer incidence among the rodents is 50%, an exposure 1/1000 of that is still a very large risk (assuming the dose-response can be modeled as linear in this range) (Finkel 1994). HUMAN AND ECOLOGICAL RISK ASSESSMENT 23 ORCID Adam M. Finkel http://orcid.org/0000-0001-6259-9387 Kevin F. Bieniek http://orcid.org/0000-0003-4922-864X References AFL-CIO v. OSHA. 1992. 965 F. 2d 962, at 974–5 Allen BC, Crump KS, and Shipp AM. 1988. Correlation between carcinogenic potency of chemicals in animals and humans. Risk Anal 8(4): 531–44. doi:10.1111/j.1539-6924.1988.tb01193.x Alosco ML, Kasimis AB, Stamm JM, et al. 2017. Age of first exposure to American football and long- term neuropsychiatric and cognitive outcomes. Transl Psychiatry 7. doi:10.1038/tp.2017.197 Asken BM, Sullan MJ, Deosky ST, et al.. 2017. Research gaps and controversies in chronic traumatic encephalopathy: A review. JAMA Neurol 74(10):1255–62. doi:10.1001/jamaneurol.2017.2396 Bailes JE, Turner RC, Lucke-Wold BP, et al. 2015. Chronic traumatic encephalopathy: is it real? The relationship between neurotrauma and neurodegeneration. Neurosurgery 62(Suppl. 1):15–24. doi:10.1227/NEU.0000000000000811 Ban VS, Madden CJ, Bailes JE, et al. 2016. The science and questions surrounding chronic traumatic encephalopathy. Neurosurg Focus 40(4):1–10. doi:10.3171/2016.2.FOCUS15609 Belson K. 2016. Roger Goodell insists football is safe: “There’s risk in life.” New York Times, Feb. 5, p. D2 Bieniek KF, Ross OA, Cormier KA, et al. 2015. Chronic traumatic encephalopathy pathology in a neuro- degenerative disorders brain bank. ActaNeuropathol 130:877–89. doi:10.1007/s00401-015-1502-4 Bingham E and Monforton C. 2013. The pesticide DBCP and male infertility. Chapter 9 in Late Les- sons from Early Warnings: Science, Precaution, Innovation, pp. 235–44. European Environment Agency, Copenhagen, Denmark Carey B. 2016. On C.T.E. and Athletes, Science Remains in its Infancy. New York Times, March 27. Avail- able at: https://www.nytimes.com/2016/03/28/health/cte-brain-disease-nfl-football-research.html Carpenter R. 2017. Requiem for a running back. Documentary film. Available at: https://www.requiem forarunningback.com/press/ Castellani RJ. 2015. Chronic traumatic encephalopathy: A paradigm in search of evidence? Lab Invest 95:576–84. doi:10.1038/labinvest.2015.54 Coggins CRE. 2007. An updated review of inhalation studies with cigarette smoke in laboratory ani- mals. Intl J Toxicol 26:331–38. doi:10.1080/10915810701490190 Cox LA Jr. 2016. Concentration-response associations used to estimate public health benefits of less pollution are not valid causal predictive models. Annals Amer Thoracic Soc 13(12):2280–81. doi:10.1513/AnnalsATS.201608-619LE Dwass E. 2015. The brain tumor is benign, but threats remain. New York Times, April 28, 2015, p. D4 Fainaru-Wada M and Fainaru S. 2013. League of Denial: The NFL, Concussions, and the Battle for Truth. Crown Archetype, New York, NY, USA Finkel AM. 1994. Rodent tests continue to save human lives. Washington Times (magazine section) 50:20–22, December 12 Finkel AM. 2009. Rejecting cost-benefit analysis is an anti-environmentalist position. Invited presenta- tion at the Institute for Policy Integrity, NYU Law School, New York, NY, USA. November 10. Slide presentation. Available at: http://policyintegrity.org/documents/Finkel-RejectingCBAisAnti- Environmentalist.pdf Finkel AM. 2018. Doubting the health harms of air pollution—Not a ‘mistake,’ but not ‘just science’ either. Essay. Available at: https://medium.com/@adamfinkel/doubting-the-health-harms-of-air- pollution-not-a-mistake-but-not-just-science-either-e56189178f4f Finkel AM, Deubert CR, Lobel O, Cohen IG, and Fernandez Lynch H. 2018. The NFL as a workplace: The prospect of applying occupational health and safety law to protect NFL workers. Ariz Law R 60:291–368. Available at: www.tinyurl.com/finkel-OSHA-NFL Fischhoff B, Lichtenstein S, Slovic P, et al. 1983. Acceptable Risk. Cambridge University Press, Cam- bridge, UK 24 A. M. FINKEL AND K. F. BIENIEK Gantt D. 2017. Roger Goodell: Football players live longer than non-football players. August 1. Avail- able at: http://profootballtalk.nbcsports.com/2017/08/01/roger-goodell-football-players-live-lon ger-than-non-football-players/ Gao AF, Ramsay D, Twose R, et al. 2017. Chronic traumatic encephalopathy-like neuropathological findings without a history of trauma. Int J Pathol Clin Res 3(1). doi:10.23937/2469-5807/1510050 Ghajari M, Hellyer PJ, and Sharp DJ. 2017. Computational modelling of traumatic brain injury pre- dicts the location of chronic traumatic encephalopathy pathology. Brain 140(2):333–43. doi:10.1093/brain/aww317 Goldstein LE, Fisher AM, Tagge CA, et al. 2012. Chronic traumatic encephalopathy in blast-exposed mili- tary veterans and a blast neurotrauma mouse model. Sci Translational Med 4(134):134ra60. doi:10.1126/scitranslmed.3003716 Haidt J. 2012. The Righteous Mind: Why Good People are Divided by Politics and Religion. Pantheon Books, New York, NY, USA Hay J, Johnson VE, Smith DH, et al. 2016. Chronic traumatic encephalopathy: the neuropathological legacy of traumatic brain injury. Ann Rev Pathol Mech Dis 11:21–45. doi:10.1146/annurev-pathol- 012615-044116 In re Nat’l Football League Players’ Concussion Injury Litigation, 307 F.R.D. 351, 397–401. (E.D. Pa. 2015) Industrial Union Dep’t, AFL-CIO v. American Petroleum Inst, 448 U.S. 607. 1980. Kahan D. 2013. Ideology, motivated reasoning, and cognitive reflection: An experimental study. Judg- ment Decis Making 8:407–24 Kelly E. 2018. I’m the wife of a former NFL player: Football destroyed his mind. New York Times, Feb. 2, p. SR1 Kondo A, Shahpasand K, Mannix R, et al. 2015. Antibody against early driver of neurodegeneration cis P-tau blocks brain injury and tauopathy. Nature 523:431–36. doi:10.1038/nature14658 Kovacs GG, Ferrer I, Grinberg LT, et al. 2016. Aging-related tau astrogliopathy (ARTAG): Harmo- nized evaluation strategy. Acta Neuropathol 131(1):87–102. doi:10.1007/s00401-015-1509-x Kuran T and Sunstein CR. 1999. Availability cascades and risk regulation. Stanford Law Rev 51 (4):683–768. doi:10.2307/1229439 Li Y, Wang W, Kan H, et al. 2010. Air quality and outpatient visits for asthma in adults during the 2008 Summer Olympic Games in Beijing. Sci Total Envt 408(5):1226–7. doi:10.1016/j. scitotenv.2009.11.035 Ling H, Holton JL, Shaw K, et al. 2015. Histological evidence of chronic traumatic encephalopathy in a large series of neurodegenerative diseases. Acta Neuropathol 130:891–93. doi:10.1007/s00401-015- 1496-y Maroon JC, Winkelman R, Bost J, et al. 2015. Chronic traumatic encephalopathy in contact sports: A systematic review of all reported pathological cases. Plos One 10(2). doi:10.1371/journal. pone.0130507 McCrory P, Meeuwisse WH, Dvorak J, et al. 2017. Consensus statement on concussion in sport: The 5th international conference on concussion in sport held in Berlin, October 2016. Br J Sports Med 51(11). doi:10.1136/bjsports-2017-097699 McKee AC, Stein TD, Nowinski CJ, et al. 2013. The spectrum of disease in chronic traumatic enceph- alopathy. Brain 136:43–64. doi:10.1093/brain/aws307 McKee AC, Cairns NJ, Dickson DW, et al. 2016. The first NINDS/NIBIB consensus meeting to define neuropathological criteria for the diagnosis of chronic traumatic encephalopathy. Acta Neuropa- thol 131(1):75–86. doi:10.1007/s00401-015-1515-z Meehan W, Mannix R, Zafonte R, et al. 2015. Chronic traumatic encephalopathy and athletes. Neurol- ogy 85(17):1504–11. doi:10.1212/WNL.0000000000001893 Meng Q, Zhuang Y, Ying Z, et al. 2017. Traumatic brain injury induces genome-wide transcriptomic, methylomic, and network perturbations in brain and blood predicting neurological disorders. EBio Medicine 16:184–94. doi:10.1016/j.ebiom.2017.01.046 Mez J, Stern RA, and McKee AC. 2013. Chronic traumatic encephalopathy: Where are we and where are we going? Curr Neurol Neurosci Rep 13:1–12. doi:10.1007/s11910-013-0407-7 HUMAN AND ECOLOGICAL RISK ASSESSMENT 25 Mez J, Solomon TM, Daneshvar DH, et al. 2015. Assessing clinicopathological correlation in chronic traumatic encephalopathy: Rationale and methods for the UNITE study. Alzheimer’s Res Ther 7:62–76. doi:10.1186/s13195-015-0148-8 Mez J, Daneshvar DH, Kiernan PT, et al. 2017. Clinicopathological evaluation of chronic traumatic encephalopathy in players of American football. J Am Med Assn 318(4): 360–70 Michaels D. 2008. Doubt Is Their Product: How Industry’s Assault on Science Threatens your Health. Oxford University Press, Oxford, UK Mitts DL.1998. Epidemiological evidence as a basis for causation: Implications for a suspected pesti- cide-induced cancer. San Joaquin Agric Law Rev 8:187–210 Montenigro PH, Baugh CM, Daneshvar DH, et al. 2014. Clinical subtypes of chronic traumatic encepha- lopathy: Literature review and proposed research diagnostic criteria for traumatic encephalopathy syndrome. Alzheimer’s Res Ther 6:68–84. doi:10.1186/s13195-014-0068-z Montenigro PH, Corp DT, Stein TD, et al. 2015. Chronic traumatic encephalopathy: Historical Ori- gins and Current Perspective. Ann Rev Clin Psychol 11:309–30. doi:10.1146/annurev-clinpsy- 032814-112814 Montenigro PH, Alosco ML, Martin BM, et al. 2016. Cumulative head impact exposure predicts later- life depression, apathy, executive dysfunction, and cognitive impairment in former high school and college football players. J Neurotrauma 33:1–13 National Football League. Undated. Excel spreadsheet giving number of snaps played during the 2016 NFL season for each player who appeared on the field. On file with first author National Research Council, Committee on the Institutional Means for Assessment of Risks to Public Health. 1983. Risk Assessment in the Federal Government: Managing the Process. National Acad- emy Press, Washington, DC, USA National Research Council, Committee on Risk Assessment of Hazardous Air Pollutants. 1994. Sci- ence and Judgment in Risk Assessment. National Academy Press, Washington, DC, USA National Research Council, Committee on Improving Risk Analysis Approaches used by the U.S. Environmental Protection Agency. 2008. Science and Decisions: Improving Risk Assessment. National Academy Press, Washington, DC, USA Neutra RR, Cranor CF, and Gee D. 2018. The use and misuse of Bradford Hill in U.S. tort law. Juri- metrics 58:127–62 NFLlabor.wordpress.com. 2011. What is average NFL player’s career length? Longer than you might think, Commissioner Goodellsays. April 18. Available at: https://nfllabor.wordpress.com/2011/04/ 18/what-is-average-nfl-player%E2%80%99s-career-length-longer-than-you-might-think-commis sioner-goodell-says/ Nierenberg AA. 2017. Defensible or not? Psychiatr Ann 47(12):571–72. doi:10.3928/00485713- 20171108-01 Nixon J. 2014. List of all pro football players since 1920. Available at: https://www.sportsblog.com/jeff nixon/list_of_all_pro_football-2/. Excel file available at: https://www.scribd.com/doc/231126190/ list-of-all-pro-football-players-since-1920 Noy S, Krawitz S, and Del Bigio MR. 2016. Chronic traumatic encephalopathy-like abnormalities in a routine neuropathology service. J Neuropathol Exp Neurol 75:1145–54. doi:10.1093/jnen/nlw092 Occupational Safety and Health Administration, US. 2014. If you work around lead, don’t take it home! OSHA bulletin. Available at: https://www.osha.gov/Publications/OSHA3680.pdf Ojo JO, Mouzon BC, and Crawford F. 2016. Repetitive head trauma, chronic traumatic encephalopa- thy, and tau: Challenges in translating from mice to men. Exptl Neurol 275:389–404. doi:10.1016/j. expneurol.2015.06.003 Olivera A, Lejbman N, Jeromin A, et al. 2015. Peripheral total tau in military personnel who sustain traumatic brain injuries during deployment. JAMA Neurol 72(10):1109–16. doi:10.1001/ jamaneurol.2015.1383 Oreskes N and Conway EM. 2010. Merchants of Doubt: How a Handful of Scientists Obscured the Truth on Issues from Tobacco Smoke to Global Warming. Bloomsbury Press, New York, NY, USA Peto R.1977. Cancer risk (letter). New Sci 73:480–81 Rhomberg LR. 1997. A survey of methods for chemical risk assessment among federal regulatory agencies. Human Ecol Risk Assess 6:1029–196. doi:10.1080/10807039709383746 26 A. M. FINKEL AND K. F. BIENIEK Rich DQ, Kipen HM, Huang W, et al. 2012. Association between changes in air pollution levels during the Beijing Olympics and biomarkers of inflammation and thrombosis in healthy young adults. JAMA 307(19):2068–78. doi:10.1001/jama.2012.3488 Schnabel J, quoting Univ. of Maryland neuropathologist Rudolph Castellani. 2016. How Real is CTE? Apr. 2, 2016, Available at: http://www.schnablog.com/how-real-is-cte [https://perma.cc/JW64-JGY5 Segal Group to Special Master Perry Golkin. Sept. 12, 2014. In re: Nat’l Football League Players’ Con- cussion Injury Litigation, 12-md-2323. ECF No, 6168 Silverberg ND, Iverson GL, Brubacher JR, et al. 2015. The nature and significance of preinjury recall bias following mild traumatic brain injury. J Head Trauma Rehab 31(6):388–96. doi:10.1097/ HTR.0000000000000198 Stamm JM, Koerte IK, Muehlmann M, et al. 2015. Age at first exposure to football is associated with altered corpus callosum white matter microstructure in former professional football players. J Neu- rotrauma 32:1–9. doi:10.1089/neu.2014.3822 Stein TD, Alvarez VE, and McKee AC. 2014. Chronic traumatic encephalopathy: A spectrum of neu- ropathological changes following repetitive brain trauma in athletes and military personnel. Alz- heimer’s Res Ther 6(4):1–11 Stern RA, Daneshvar DH, Baugh CM, et al. 2013. Clinical presentation of chronic traumatic encepha- lopathy. Neurology 81:1122–29. doi:10.1212/WNL.0b013e3182a55f7f Sterne JAC, Hernan MA, Reeves BC, et al. 2016. ROBINS-I: A tool for assessing risk of bias in non- randomised studies of interventions. BMJ 355. doi:10.1136/bmj.i4919 Sunstein CR and Thaler RH. 2008. Nudge: Improving Decisions about Health, Wealth, and Happi- ness. Yale University Press, New Haven, CT, USA Tagge CA, Fisher AM, Minaeva OV, et al. 2018. Concussion, microvascular injury, and early tauop- athy in young athletes after impact head injury and an impact concussion mouse model. Brain 141 (2):422–58. doi:10.1093/brain/awx350 Tiesman HM, Konda S, and Bell JL. 2011. The epidemiology of fatal occupational traumatic brain injury in the U.S. Am J Prev Med 41(1):61–67. doi:10.1016/j.amepre.2011.03.007 Turner RC, Lucke-Wold BP, Logsdon AF, et al. 2015. Modeling chronic traumatic encephalopathy: The way forward for future discovery. Front Neurol 6:1–18. doi:10.3389/fneur.2015.00223 U.S. Department of Health, Education, and Welfare. 1964. Smoking and Health: Report of the Advi- sory Committee to the Surgeon General of the Public Health Service, PHS Publication No. 1103, U.S. Government Printing Office, Washington, DC, USA U.S. Environmental Protection Agency. 2005. Guidelines for Carcinogen Risk Assessment. EPA/630/ P-03/001F, Risk Assessment Forum, 166 pp. Available at: https://www.epa.gov/sites/production/ files/2013-09/documents/cancer_guidelines_final_3-25-05.pdf Utecht B and Tabb M. 2017. Counting the Days While my Mind Slips Away: A Story of Perseverance and Hope. Howard Books, Brentwood, TN, USA Ward J, Williams J, and Manchester C. 2017. 110 NFL brains. New York Times, July 25. Available at: https://www.nytimes.com/interactive/2017/07/25/sports/football/nfl-cte.html Willborn SL 2014–2015. College athletes as employees: an overflowing quiver. U. Miami Law Rev. 69:65–115. Woerman A, Aoyagi A, Patel S, et al. 2016. Tau prions from Alzheimer’s disease and chronic trau- matic encephalopathy patients propagate in cultured cells. Proc Natl Acad Sci USA 113(50): E8187–96. doi:10.1073/pnas.1616344113