Scientific Absolutism

by Siegfried Othmer | February 15th, 2006

The other day it was 87 degrees in our neighborhood in Los Angeles. Unusual? Yes, but the previous record was 86 degrees, and that was twenty years ago. One can’t base much of a case on such isolated extrema in weather, but viscerally they do give one pause. I have yet to close the sunroof in my car this winter, or to put away my short-sleeved shirts. The heater has hardly gone on in our house at all. Last year was the warmest year on record around the world, the continuation of a consistent trend over the past few years. So far this winter is turning out to be the warmest as well, the winter storm in the East notwithstanding.

To get a better handle on global warming, one should look at those phenomena that have built-in smoothing functions–like arctic ice. Greenland ice sheets have been thinning at accelerating rates, measurably larger every year than the one before, and glaciers are going into accelerating retreat. Evidence is starting to accumulate on the slow-down of the Atlantic Conveyor, with grave prospects for European farmers north of the Alps as climate goes unstable and temperatures plunge in Northern Europe.

Ostensibly the Administration is simply waiting for the “science” to get more definitive on global warming. More likely, however, the Administration is trying to postpone inconvenient measures and using the scientific “duty to doubt” as the unimpeachable rationale. Scientific certainty may not seem too much to ask when high economic costs and dislocations may be at stake. Yet such certainty is not likely to be forthcoming any time soon. Each scientific study sets limited objectives, and results are usually presented in nuanced fashion. In a matter as complex as climate, the results of different studies need to be integrated at a higher level into a coherent picture. So if policy demands it, then the formal status of scientific uncertainty can be spun out into the indefinite future.

The smoothing function we are looking for in arctic ice is also in play to moderate the near-term effects of global warming: the ocean has enormous heat capacity. Even if we were to stop making things worse on the spot the die would already be cast on further global warming for many years and even decades to come. The same factor is operative in shifting the depth of winter far past the solar minimum in December. That delayed, cumulative, and unavoidable liability should be taken into account in policy-making. Ideally we would be matching some kind of cost-benefit analysis with the likelihood of various climate outcomes. To put such kinds of calculations “beyond use” in the anticipation of the immaculate scientific study is folly, not wisdom. We call such an attitude “scientific absolutism.”

A second instance of this kind of scientific absolutism relates to the hazard of smoking. The matter is once again relevant in that second-hand smoke has just been declared a toxic constituent of the atmosphere. Pardon me if I repeat reference to a 1958 prospective study published in the American Scientist on the hazards of smoking. This was a thoroughly exhaustive study that not only established (statistically) the hazards of smoking, but also could have been taken to establish the hazards of second-hand smoke. The relevant data were already there. Yet the tobacco companies and their biostitute scientists managed to keep the ball in the air on scientific uncertainty for another several decades–a half-century in the case of second-hand smoke.

Even now, the communicable aspect of cigarette smoking is being played down–i.e., the greater likelihood that someone routinely exposed to cigarette smoking will also become addicted to nicotine. Only a few years ago, tobacco company executives still felt they could announce to a Congressional Committee under oath that they personally did not believe nicotine was addictive. The proposition had not yet been absolutely, irrevocably proved, and until that was done, the execs were free to nurture their own favorite (and conveniently self-serving) personal opinions.

Another issue of the same kind is the biological effects of radiation. For decades the issue raged as to whether a threshold existed below which ionizing radiation was biologically benign. It is at low radiation levels that uncertainty is most difficult to expunge. The zone of uncertainty was large enough to drive a whole policy agenda through for decades. Only now, some sixty years after the atomic bomb, is the window finally closing on this hypothesis. In the meantime, the government has been able to resist claims by whole hosts of cancer victims, including those of American soldiers directly exposed to atomic blasts in the early tests, until most of them have died.

The issue has surfaced here in Los Angeles recently, as it was admitted that cancer death rates were elevated in the neighborhood of the Santa Susana Field laboratory, operated by Rockwell International for rocket testing and nuclear reactor research since the time of WWII. This issue has been researched for decades, yet these findings surface now, when most of the victims are beyond help. The cloud of obfuscation in public briefings by scientists could be cut with a knife. Elevated levels of perchlorate (a constituent of rocket fuel) were found in wells only a few miles from where the rocket testing had been conducted for decades. A billion dollar development proposal was at stake at this point, so scientists could not bring themselves to make the connection of this mysterious toxin in the wilderness with its obvious nearby source. Scientific absolutism was running rampant. Since the speaker could not be absolutely certain, he preferred to make no judgment at all. His paymaster would be pleased.

The same person ventured that the “toxic plume” emanating underground from the facility “would have stabilized by now.” Such a proposition has essentially no evidence behind it, or even a good model. There is no “plume,” as in smokestack emissions, and hence no defined “front” to the progress of toxins. These move with water through fissures in our highly fractured geology. And they don’t just stop after a while. But the scientist was unlikely to be proved wrong in his lifetime, so it was safe to go out on a limb on this one. He was signaling, in effect, that if the property to be developed is clean now, then it will likely remain clean in the future. There was an essentially complete match of the “scientific” utterances to the political demands of the moment. The paymaster would be pleased once again.

The best example is perhaps that of the Holocaust deniers. I am referring here to those who (like David Irving) claim to be adhering to the best standards of historical research, insisting that all the pieces must fit into place before any credence at all can be given to the Holocaust story. As long as some contrary evidence exists, as long as some pieces of the story don’t fit together, judgment must be held in abeyance.

In all of the above instances, science is asked to function in a contentious policy environment, or in matters of strong political conviction. Under those circumstances, scientific integrity is likely to suffer. Scientific absolutism is characterized by black-and-white, all-or-nothing thinking, by rigidity with respect to the kind of evidence that is allowed, and by a haughtily authoritarian attitude all around. We are seeing all of this with respect to neurofeedback. Our critics, such as Russell Barkley, John Ratey, Peter Jensen, and Larry Silver show tendencies toward such scientific absolutism. The evidence for neurofeedback must be utterly unassailable and pristine before they will give it any credence. This is not how science should proceed.

We should recognize that there is likely to be an agenda here, and we should not accept the imposed strictures. Most likely, this is not about the science of neurofeedback at all, but rather about the care and feeding of drug companies, and about the maintenance of the status of these individuals as authorities on ADHD. There may be personal animus here as well. As authorities on ADHD, they have been blind-sided, and they probably resent it deeply.

In the meantime, they absent themselves into the spectator seats. Their utterances with respect to neurofeedback–or the lack thereof–don’t appear to be entirely intellectually honest. There is no doubt that they have created a climate in which other scientists have held back from investigating neurofeedback, lest their reputations be singed. An adversarial climate disturbs the conditions that must exist for good science to proceed, and these folks seem to promote such an adversarial climate.

Some of our principal mainstream critics behave like scientific fundamentalists, frustrating the healthy unfolding of the scientific process. How else to explain that John Ratey, when asked to speak on the topic of alternative remedies for ADHD at the “Learning and the Brain” Conference in Boston in 2000, failed to even mention neurofeedback. He was not restricting himself to those alternative approaches that met some kind of criterion of acceptability. He covered the whole waterfront–including even projecting forward to technologies of the future such as implanted chips. Yet he could not bring himself to utter the word neurofeedback. I think this is just scandalous, and the man should not be allowed to wriggle off the hook.

These people wrap themselves in the mantle of scientific probity, and raise skepticism to the level of overriding virtue, when in fact they are functioning as agitators and propagandists against neurofeedback. They insist on the immaculate experiment as a kind of scientific purification rite, when in fact nothing else in their mental universe is being held to such standards. They should be exposed and disgraced for their conduct. History will probably see to that over time, but it would be nice if it happened while it could still have a positive impact. In the meantime, they should at least not be catered to.

Now the scientific integrity of these people is not the real issue here. More to the point is that they seem to have succeeded in panicking people in our field to accept their standards. Somehow the scientific elite in our field has been intimidated into accepting absolutist criteria as being determinative. Thus, for example, the definitive journal article–and position paper–on ADHD, published last year in the AAPB Journal, proclaims lamely in its abstract: “On the basis of these scientific principles, EEG biofeedback is determined to be ‘probably efficacious’ for the treatment of ADHD.”

So the mountain labored and brought forth a mouse. The insurance companies have been handed all the evidence they need to declare the technique to be “experimental.” And the ethics of neurofeedback practitioners have been rendered questionable. There is a disconnect here between what the authors of this paper (Monastra, Lynn, Linden, Lubar, Gruzelier, and LaVaque) themselves presumably hold to be true on this issue, and what they have chosen to say or claim “officially.” The claim they make is conditioned on some assumptions, namely the efficacy guidelines. In real life, some of the same authors conduct themselves as if neurofeedback had been proved efficacious and treat the matter as no longer in question.

These authors seem perfectly capable of partitioning the world of scientific claims from the world of policy and of actual practice, where decisions have to be made even in the absence of firm knowledge. So let us do the same. If those little games have to be played with the efficacy guidelines, let’s at least quarantine that perspective to the world of the journals and not let it contaminate everything else that we do.

Just as in the cases of global warming, and of tobacco and radiation hazards, we face policy choices in which science should simply be the handmaiden. With the new hazards that have been exposed in stimulant medication, surely it is time for our society at large to be informed of a remedy for ADHD that does not entail a fatality risk. These decisions need to be made even though we don’t have perfect knowledge about either the hazards of stimulants or of the benefits of neurofeedback. A policy declaration is needed that is not tethered to the efficacy guidelines. I happen to believe that these guidelines represent an absurd construction when applied to a behavioral technique such as neurofeedback, but that is not the point at this moment. Rather, the declaration should simply be based on what we actually do, and on what we actually believe, quite independently of that other world of formal efficacy studies. The science must inform such a declaration; it shall not determine it absolutely. Rigorous science is always in catch-up mode with respect to what one actually believes, and to how one actually practices.

This discussion has been rendered acute because of the recent FDA call for warnings on stimulants. Here is a moment when we could have the nation’s attention. We should take advantage and not squander it in internecine squabbles. Some 51 people, including 19 children, have apparently died, with stimulant medication being a suspected contributory cause, in the 1999 to 2003 timeframe. Some 54 others suffered significant cardiovascular complications.

I was relieved to hear that David Graham was the FDA’s spokesperson on this matter. He was the one who raised the alarms about Vioxx, where his concerns have been tragically vindicated, and he was the one who testified forthrightly to Congress about the FDA’s neglect of post-approval monitoring. If we assume, as former FDA head David Kessler did, that only about one percent of such adverse drug reactions are ever reported to the FDA, then there are indications here of significant problems with stimulant medication. At a minimum, a subset of patients needs to be identified that may be at enhanced risk from stimulant therapy. This might include children with congenital heart defects or with tendencies toward arrhythmia. And these folks–at minimum–should be offered a non-drug remedy. (In that context, is it ethical to remain silent? Should penicillin not have been used during the war until controlled studies were done?)

Did the FDA in fact maintain due diligence in this matter, promptly raising the issue when the data warranted? Consider that the relevant publication on adverse reactions to stimulants in particular, and multiple drug cocktails in general, was published back in 1997. I am referring to the following publication:

Case Study: Adverse Response to Clonidine.
Journal of the American Academy of Child & Adolescent Psychiatry. 36(4):539-544, April 1997.
Cantwell, Dennis P. MD; Swanson, James PhD; Connor, Daniel F. MD

The use of clonidine alone and in combination to treat a variety of problems has increased in child and adolescent patients. Four cases of adverse experiences with clonidine are described. Clinical guidelines for the use of clonidine in particular and the use of polypharmacy in general are presented. J. Am. Acad. Child Adolesc. Psychiatry, 1997, 36(4):539-544.

The typical combination in which clonidine is conventionally prescribed to ADHD children is of course with stimulants. The FDA excluded the complications of “poly-pharmacy toxicity” from its analysis, even though if anything risks would be enhanced with such prescribing. If the agency were truly focused on risk, it should of course focus on where the risk is greatest.

Dennis Cantwell was among the most vocal critics of neurofeedback until he became aware of the above results. At that point, his perspective shifted, and it was entirely out of ethical considerations. Said he: “ADHD may be a very bad thing; but it doesn’t kill you, and neither should the remedy.” So he set out on a search for a non-drug alternative, and he recognized the need for a formal study of neurofeedback. “We have an obligation to test what is being offered to the public as a remedy.” That brought him to our door back some years ago. Unfortunately, Cantwell died suddenly as our proposal was going through the NIH approval process, and the UCLA Child Psychiatry Department has not seen fit to take up the baton over all these succeeding years.

The above is to illustrate that a fair-minded person—even one thoroughly steeped in the field of psychopharmacology—appreciated the ethical burden clearly back in 1997. Yet the FDA did not suggest monitoring stimulants for cardiac events until 2004. One again sees scientific absolutism at work. The evidence must be compelling before it is attended to, and it must be overwhelming before a warning is issued. The bias is in favor of drug sales rather than in favor of safety.

The story was much the same years ago with Rezulin, a medication for diabetes. People were dropping dead from liver toxicity in significant numbers, and the FDA kept recommending more extensive monitoring. Yet there was no evidence of any predictive power of such monitoring. In fact the evidence went the other way. The liver toxicity had rapid onset–often death occurred over a day or two, and it was not predictable. But more monitoring was ordered anyway, just to keep the Rezulin market viable. I recall that the Los Angeles Times won a Pulitzer for the exposure of this scandal.

It goes without saying that the availability of a non-drug alternative presents a thorough-going threat to this drug hegemony–with ethics as a lever. Hence the need for such complete denial about neurofeedback–just as we have seen with respect to the risk of vaccines in the autism spectrum.

What if the pristine experiment that Barkley calls for were actually done? Why then Barkley could finally get out of the corner into which he has painted himself. He would proclaim that the evidence he was looking for was now in hand, and he could now recommend neurofeedback positively. Attention would once again be focused on Barkley, and not on either the AAPB or the ISNR. In a similar manner, an OpEd in the Los Angeles Times started out with “President Bush was right to say that the United States was addicted to oil…,” when in fact he was wrong to be in denial about that fact for so long. No matter. His former apostasy< is forgiven. Just why is it that we are so anxious to hand Barkley the key to his own jail cell, when the consequence will only be that he can once again become the self-appointed arbiter, in this case of what constitutes good and bad neurofeedback? Right now, he has condemned himself to virtual silence on neurofeedback, since his neurofeedback denials are no longer treated as news. After all these years we are better off keeping things that way. In the recently published article on neurofeedback in Scientific American “Mind,” after due recognition of Barry Sterman’s work attention focused almost entirely on what is now happening in Europe. Elsa Baehr was quoted on the asymmetry protocol for depression to the effect that “This is an experimental protocol. Until there are controlled studies, we won’t know how effective the therapy is.” Such a charmingly self-effacing modesty is not all that helpful–or even correct. The truth is rather the opposite: controlled studies will not be done until people are convinced that the treatment is actually effective. Controlled studies are always about subsidiary hypotheses such as mechanisms, rather than about clinical effectiveness. The latter can be observed directly by the clinician. It amounts to the mere quantification and statistical assessment of outcome data. Now the self-effacing wall-flower strategy that is being adopted by our organizational leadership will not help the field to grow. At the very moment that the country is awakened to major hazards with stimulant drugs (Surprise! Surprise!), we remain engaged on the matter of efficacy proofs and placebo designs. We don’t need a bunch of eunuchs tending the temple, defenders of the faith of firmly established doctrine. In the words of the late Abba Eban, we seem to “never miss an opportunity to miss an opportunity.” The bottom line is that “science” has been contaminated by policy considerations on every hand in our society, having been made subservient to a corporatist agenda. In the face of this, we should forthrightly separate the domains of policy declaration and of rigorous research, and we should do justice to both by their respective standards.

Leave a Reply