Conclusion

In a pithy little book published in 2014 titled Are We All Scientific Experts Now?, the sociologist of science Harry Collins explored public attitudes toward science in an attempt to explain recent phenomena like the antivaccine movement and climate skepticism.¹ Collins noted that the public’s trust in science has declined from its apogee in the decades following World War II, and he attributed this decline to a simplistic reading of Thomas Kuhn and the rise of social relativism, starting in the 1970s. He went on to provide a useful inventory of different types of expertise to distinguish what is special about scientific expertise. Collins’s aim was to explain what it is about scientific activity that sets it apart from other activities we are familiar with, in order to restore science to the position it deserves. He held up as the epitome of real science physicists working on gravitation waves, who delayed publishing important results because their confidence in their measurements was not absolute.

However, both the circumstances and the culture of science in the area of biomedicine and public health differ in important respects from those of experimental physics. Physicists are not in the habit of lobbying the public to gain support for their particular point of view, and they are not given to claiming that their findings are politically correct. Furthermore, the public is not clamoring for answers to the mysteries of black holes and dark matter. Because the problems physicists study are so far removed from everyday life, they are able to pursue their work in relative isolation from external influences.

When it comes to the study of factors that may affect our health, the situation is quite different. Because research in this area focuses on clinical diseases affecting real people, the results appear to have a direct relevance to our lives. We are all eager for solid information that would permit us and our loved ones to avoid, or survive, the chronic diseases that are the major causes of death and disease in advanced societies. For this reason, findings from studies of disease have a particular power and mystique, which can influence what results are reported, how they are reported, and what is made of them in the wider society.

If research in public health were conducted out of the spotlight, one could leave it to internal mechanisms of the discipline and to time to weed out what stands up and what is important. But, as we have seen, the landscape in which health risks are studied and in which findings are disseminated is pervaded by false claims, oversold results, biases operating at the level of observational studies as well as psychological and cognitive biases, and professional and political agendas. On certain topics, as we have seen, scientists holding conflicting views cannot find common ground and are polarized into opposed camps, or “silos.” Needless to say, this type of behavior is not in line with scientists’ view of their profession nor with Collins’s view of science.

If scientists disagree on many questions regarding health, nutrition, biotechnology, and the environment, it is small wonder that nonscientists are confused about what are arcane and difficult issues that require special expertise to begin to understand, no less to assess critically. Such questions—and the research that addresses them—only become more confused when they are catapulted into the public arena. Rather than being assessed on strictly scientific grounds, they are refracted through many different lenses according to the outlook or agenda of different groups or individuals. In many cases the resulting versions of the relevant science involve serious simplifications or distortions. Often the goal is to distill the results of a particular study, or of the totality of evidence on a question, to a simple yes/no dichotomy that either confirms or contradicts what we would like to believe. But there is no reason for us to expect that all studies on a difficult question will deliver a clear-cut answer, and much less that they will all line up “on the right side.” Furthermore, as we have seen, all studies are not equal.

When difficult scientific questions are drastically simplified to fit a specific purpose or agenda—journalistic, regulatory, ideological, political, or personal—often the most basic distinctions and considerations are lost sight of. The mental fog produced by so much misinformation and partisan spinning of the science can be dispelled only by keeping in view certain fundamental facts that rarely get attention. These are worth reprising below.

There are real problems and there are false problems, that is, problems that, to the best of our knowledge, are not problems at all. Vaccines, genetically modified crops and foods, and cell phones are not threats to our well-being. Rather, they are among the greatest advances contributing to human welfare. We need to get better at distinguishing false problems from real problems.

Biology is complex, and we should not underestimate the difficulty of the problems we want to see solved. This difficulty helps explain why progress in understanding diseases such as pancreatic cancer, Alzheimer’s, and many others has been so slow. It also explains how dramatic advances can be made on some fronts, but, in spite of concerted efforts, progress on other fronts can be disappointing. Often we recite the meager knowledge we have on a question and fail to acknowledge just how little we actually know. Being open about our ignorance would both serve as an incentive to fill in the knowledge void and, at the same time, serve to highlight the real progress that has been made in answering other questions.

The quality of research and the rate of progress in uncovering new knowledge vary dramatically between different areas. For example, in the area of genetics and genomics there has been impressive progress, and high standards have been established for the replication of findings, allowing the field to move forward. In contrast, findings concerning environmental exposures and their effects on health are much weaker and are subject to controversy.² The fact is that there are fields where the methods and the hypotheses are more robust than those in other fields.

As we have seen, it is now widely recognized that much of what is published is either wrong or exaggerated and that there is an epidemic of false claims that gain wide circulation and are not easily dispelled, even when more solid contradictory evidence becomes available. False claims are fueled and reinforced by the many biases that affect both the scientific work and how it gets presented to the public. The true extent of false claims and misinformation in biomedicine has only recently begun to receive systematic study. The prevalence of error in the published literature points up the difficulty of the problems studied and the need for improved standards in research. It turns out that a lot of what we think we know is wrong.

How one approaches a question can have a decisive effect on how productive one’s efforts will be. If one frames the issue in a way that screens out relevant considerations, it stands to reason that one is reducing one’s chances of finding something new and important. This is the lesson that Richard Sharpe drew from twenty years of high-profile but fruitless efforts to find evidence supporting the endocrine disruption hypothesis. Rather than asking how exposure to endocrine disrupting chemicals in the environment causes reproductive disorders, he posed the question, what causes these disorders? As he commented, “Such a simple difference, but it takes your thought processes in a very different direction.” We should note that Sharpe’s restatement of the problem meant going against the tide, since it meant rejecting a fashionable idea that had wide support from the public and that improved one’s chances of obtaining funding.

Sharpe also stresses that “getting it wrong is alright” and that failure is an essential part of the research process—that is, if it prompts one to take a fresh look at one’s framing of the problem rather than cling to one’s hypothesis.³ Failure is an essential part of the research process because it forces one to go back and ask where one went wrong. If the work was done right, being led to a dead-end forces one to redirect one’s attention to another, possibly more promising question.

In addition to framing the question in such a way as to maximize the possibility of finding a meaningful answer, there are other “simple” distinctions that can increase the chances of identifying a fruitful path. First and foremost is a consideration of the characteristics of the agent one is interested in. If the focus is “endocrine disrupting chemicals,” one has to start by documenting the relative dose and potency of human exposures in different environments to these chemicals, which are hypothesized to be having detectable effects on the population. Since all researchers agree that the DES experience of pregnant women in the middle of the last century provides the cornerstone for studies of environmental estrogens, one would expect researchers to acknowledge the enormous difference in dose and potency between DES administered as a drug to these women and typical exposure to trace amounts of chemicals in the environment. However, this is rarely done. To recognize that the environmental exposure is many orders of magnitude weaker than the pharmacologic doses does not rule out that the former is worthy of study, but it does mean that any effects are likely to be much harder to detect, and also that research would not incite the kind of fear and certainty that it tends to in the public.

By our nature, we are disposed to want to find external causes to account for diseases we don’t understand. These are things that are beyond our control, and this helps explain the enormous appetite for stories about what are extremely low-level exposures in the environment. Such low-level exposures may be having real effects and may well merit study. But we should keep in mind that the major causes of chronic disease that have been identified in the past sixty or so years are smoking, heavy alcohol consumption, heavy sun exposure, excess body weight, a poor diet, lack of physical activity, exposure to certain micro-organisms, and socioeconomic inequality. These are factors that have large effects. There are few things that are studied in the realm of public health that influence one’s risk of disease by a factor of five, or ten, or more, but these factors do just that. And yet, curiously, these factors, which are mundane and appear to be under our control, do not inspire anywhere near the kind of fear that is inspired by trace exposures to “chemicals” and “radiation” in the environment.

As regards the future, we hear a great deal about the many exciting developments that have the potential to yield undreamed of advances—“precision medicine,” “targeted therapies,” “gene therapy,” “regenerative medicine,” “tissue engineering,” and the use of “Big Data” to uncover new relationships in unprecedentedly rich datasets. Here too, however, we need to keep in mind how difficult and slow real progress on these fronts is likely to be and that, while these approaches may revolutionize the treatment of specific illnesses, they are less likely to transform our lives. Big Data, or “data mining,” represents a powerful tool that can supplement or, in some cases, replace hypothesis-driven research, as in the search for genes linked to disease. As has been pointed out, however, the use of Big Data to solve meaningful problems will require much better data than are currently available, as well as new methods to analyze the data and to avoid spurious findings.⁴ Although the project of understanding the role of genetics in complex diseases is certain to lead to fundamental change in how we prevent and treat disease, it is significant that Eric Lander, the head of the federal Human Genome Project, has cautioned that the real payoff from this work is generations away.

* * *

The exciting and beautiful thing is that an astute observation and the determination to solve a mystery can, in fortunate circumstances, lead to the formulation of a hypothesis that can transform our understanding of a problem and lead to new strategies to prevent or cure disease. This is what happened when Denis Burkitt, observing the swollen jaws of Ugandan children with a mysterious disease, thought to undertake a continent-wide survey that led to the first linkage between a virus and a human cancer. This is what happened when a Belgian pathologist noted the similarity between the type of kidney damage observed in women on a weight-loss regimen, which included Chinese herbs, in a Brussels clinic in the 1990s and that seen in patients with Balkan nephropathy. This is what happened when Arthur Grollman decided to undertake molecular studies of upper urothelial cancer in the Balkans and in East Asia—studies that demonstrated a unique type of genetic damage to the urinary tract caused by exposure to aristolochic acid. And this is what happened when Harald zur Hausen questioned the dogma that herpes simplex virus must be the cause of cervical cancer and made a connection between the observation of virus-induced lesions in cottontail rabbits in the 1930s and the pathology of human cervical cancer.

Much promising work is going on that will undoubtedly lead to new breakthroughs, although we cannot say where these will occur. When a breakthrough does come about, it is more likely to come from the persistent work of different groups pursuing a strong hypothesis that has been refined as a result of challenges and self-criticism than as a result of focusing on a culprit that appeals to our ill-defined fears and ignoring contradictory evidence and competing hypotheses. We need to promote a model of what science is and how it operates at its best, based on examples like those presented in the preceding chapters and described by many others. The achievements that we are surrounded by and that we often take for granted, for the most part, did not come about as the result of one individual’s stroke of insight and certainly didn’t come from following some fashionable but ill-defined idea. Rather they are the result of trying to answer an important question by building on and extending existing knowledge. Such achievements required persistence and collaborative work conducted out of the spotlight by many different groups, each contributing a piece to the larger puzzle, with no guarantee that the work would turn out to be important. It is these real accomplishments that should serve as models of what science can achieve and, at the same time, provide a standard for judging overstated claims, implausible findings, and appeals to irrational fear.