What's New in Neurofeedback

• Scientific American Frontiers had a 10 minute segment on neurofeedback March 27, 2001. - http://www.pbs.org/saf/previous.htm

• New Hampshire Daily Gazette ran NF story - www.gazettenet.com/03162001/gazette.shtml

Yet not everyone realizes the overlap between the two disorder groups. That distinction fuels some of the argument against neurofeedback training. In an upcoming paper (in a journal not cited by Medline) it is argued that although neurofeedback appears to have anticonvulsant properties, evidence for treating psychological disorders with neurofeedback is entirely lacking, due to placebo effects and design problems. Thank god for the placebo effect. Critics would be lost without it. Well, let's complete their logic: isn't is likely that those psychological disorders now treated with anticonvulsants (bipolar disorder, OCD, PTSD, social phobia, alcohol withdrawal and dependence, to name a few) would be ameliorated by neurofeedback? One would think.

In light of batting about the p-term, some background on this all-too-familiar topic seemed warranted.

When Freud gave a name to the darkness, calling it the unconscious, he set loose a process of distrust in ourselves and our self-knowledge that will likely never end. From that infernal creation descends the placebo effect, the idea that the mind can make itself better if only each of us allow it. Healing through distraction. It's a powerful idea -- because it works.

Or does it?

Henry K. Beecher's influential 1955 paper "The Powerful Placebo," published in the Journal of the American Medical Association, was the first study to attempt to quantify the magnitude of the placebo effect. This paper is the source for the ubiquitous citation of 35% improvement associated with placebos. Beecher claimed that 35% of 1082 patients were "satisfactorily relieved" by a placebo alone. This value was never meant to be carved in stone. It was a mathematical mean from only 15 clinical trials which involved a variety of diseases, with an range of improvement from 21 to 58 percent. From the beginning the reliability of this number (35%) should have been questioned.

Forty years later, it was. In the Journal of Clinical Epidemiology, Kienle and Kiene (1997) determined, after examining all of the initial studies, that "no evidence was found of any placebo effect in any of the studies cited by (Beecher)." One hundred percent off the mark! Setting a record in scientific endeavors that cannot never be broken, only tied. Instead of a placebo effect, Kienle and Kiene (1997) concluded that the improvements reported in the 15 studies could be accounted for by the following: spontaneous improvement or recovery, symptom fluctuation, regression to the mean, concomitant treatments, scaling bias, obliging reports by patients, irrelevant response variables, experimental subordination, conditioned answers, psychosomatic phenomena, and other factors including an "active placebo" also known as an (active) treatment. And not one factor listed by these authors had anything to do with the spooky unconscious (that is, psychological anticipation).

In a related paper the same authors wonder whether the placebo effect was perhaps largely illusory. In fact they muse that it may not even exist at all and be another vestige of Freudian thought still clinging to post-modern thought. (These were not their words, however). Some authors concurred. For instance, Gotzsche (1995) argued that the concept of placebo should be discarded altogether. But others disagreed and continued to argue that psychological mechanisms underlay nonspecific effects (Kirsch, 1997). [In other words, Freud would not die.]

As any physicist can tell you, studying a phenomenon, even when it does NOT exist, can be very interesting and fruitful. For a placebo effect to exist, what neurobiological mechanisms must control it, what neuroanatomical or functional systems must be involved? What are the duration and dosage curves like and why? Levine volunteered endorphins to answer some of these questions. Others have gone so far to say that the placebo effect may occasionally be toxic! (cf. Shapiro & Shapiro, 1997; Freud kills). While important questions have yet to be addressed, some researchers in mental health began to sling the term around like mud at anything they did not have patent rights to. Perhaps with hindsight, Kienle and Kiene (1997) warned that "the placebo topic seems to invite sloppy methodological thinking." And sloppy thinking is the hurdle facing any treatment modality which competes (effectively) against the current standard.

The Ethics of Placebo Controls: Should neurofeedback research use a placebo control?

In 1992, Russell Barkley argued that neurofeedback must be compared to the placebo condition before he would take any claim seriously. He went so far as to suggest using bogus feedback in his report (CHADDer Box). Human protection committees and researchers such as Michael Linden (1996) recognized how the "option" of false feedback was probably unethical; and others have understood how it is also impractical. False feedback "breaks the blind" of the patient. Subjects, particularly children, are quick to detect when information on the monitor is not associated with their own EEG. This fact itself points to darker workings in the mind.

Those who would recommend placebo controls as a necessary hurdle to acceptance are probably unaware of the Nuremberg Code which was formulated shortly after World War II in response to Nazi atrocities. This code limits the extent of future human experimentation and was the precursor to the Declaration of Helsinki accepted by the World Health Organization in 1964, a declaration to which our governmental and regulatory bodies are expected to prescribe. The Declaration of Helsinki elevates concern for the health and rights of individual subjects over concern for society, for future patients, or for science.. "In any medical study," it asserts "every patient -- including those of a control group, if any -- should be assured of the best proven diagnostic and therapeutic method." Re-read that statement. It ends the use of placebo controls when a proven therapeutic method exists. A study that violates this provision should not be accepted for publication and any application to a regulatory body for a treatment which unnecessarily involved placebo -controlled trials ought be rejected.

Ought to be.

The Code of Federal Regulations under which the FDA operates includes mention of the use of placebo controls, thus deeming them acceptable for US biomedical research. Some researchers complain (in letter to BMJ, etc.) that without a placebo control, their grant proposals will never be funded. Even when alternative treatments are available, a placebo control is an implicit requirement in competitive funding situations. This de facto requirement includes disorders of moderate severity and pain, clearly in violation of the Declaration of Helsinki.

Why are placebos used at all ?

I can think of three reasons why placebo controls are desirable:
1. To evaluate absolute efficacy: Placebo controls supposedly determine whether a new treatment is better than nothing.
2. To avoid difficult decisions about treatment effectiveness: Differences in cost, unintended (side) effects, drug interactions, and other factors may make comparisons between treatments of similar efficacy inexact and subjective.
3. To bolster statistical significance: As any psych grad student knows, it is much easier to detect statistical significance between a placebo and an active treatment than between two active treatments. (And it is less expensive to use a placebo condition than to increase the n, the other method of bolstering one's stats.)

Ignoring the ethics of requiring placebo controls, or even the phenomenological argument of whether placebo effects exist at all, I realize that one can define an active treatment by a set of criteria that placebos cannot meet. :

• Specificity: A single symptomalogical or functional domain is impacted by a treatment compared to nonspecific or multiple domains
• Toxicity: Use of the treatment can cause worsening of symptoms in some individuals. (Good thoughts rarely kill, though nocebos do appear to exist in unusual cases.)
• Long Duration: Changes associated with the treatment endure for weeks, months, or years.
• Dose-dependence: Extent of changes reflect an accumulation of doses (cumulation and carry-over effects)
• Time-dependence: Changes occur in response to a manipulatable time-line. Symptom changes do not emerge before or long after treatment is complete, for instance. (time-effect curves)

Neurofeedback meets each of the above criteria for ADHD/ADD, epilepsy , and other conditions. Unfortunately, this is an unsubstantiated claim at the moment as most of this information is in the clinical record only and currently unpublished. Perhaps if some agreement can be reach on active treatment criteria, studies can be designed to address each issue, one at a time.

Here's a thought: If a treatment only meets four or less of the five criteria, should we say this is an active treatment, a placebo, or something in-between? For instance, what is to be made of the fact that the long duration criteria is not met by stimulant therapy for ADHD (e.g., when most individuals stop taking stimulant meds, the symptoms return). And on a similar lines of thinking, if a treatment meets and exceeds all five criteria for a condition, must it be designated a cure?

DK

News & Reviews

NEW BOOKS

The Neuropsychology of Emotion

Summarizes recent research in the neurochemistry of emotion with regard to head injury and other neurological traumas. -www.amazon.com/exec/obidos/ASIN/0195114647/top100

Anxiety, Depression, and Emotion
by Richard Davidson

Proceedings of the First Wisconsin Symposium on Emotion; focuses on emotion and psychopathology. -www.amazon.com/exec/obidos/ASIN/0195133587/top100

An Odd Kind of Fame: Stories of Phineas Gage
by Malcolm Macmillan

Starting with Gage, neuropsychology has remained preoccupied with the function of the frontal lobes for 150 years. -www.amazon.com/exec/obidos/ASIN/0262133636/top100

Cortical Functions
by John D. Stirling

Discusses EEG variants and EEG correlates of psychological anomalies -www.amazon.com/exec/obidos/ASIN/3540655735/top100

Exiting Nirvana: My Daughter's Life with Autism
by Clara Claiborne Park

A mother tries to understand her autistic daughter's perceptions of the world. Well reviewed. -www.amazon.com/exec/obidos/ASIN/0316691178/top100

Asperger Syndrome
by Ami Klin, Fred R. Volkmar, Sara S. Sparrow