Therapeutic Psilocybin?

December 7, 2010, 10:12 pm


Hallucinogens as Medicine. Griffiths RR, Grob CS. Scientific American Dec 2010;303:77-79.


Last April, TPR posted about an article in the New York Times reporting on the resurgence of medical research into possible therapeutic effects of hallucinogens, especially psilocybin. Two of the researchers cited in that article — Roland R. Griffiths from Johns Hopkins and Charles S. Grob from UCLA — have co-authored an piece in this month’s Scientific American describing their work.

In a study recently published online by Archives of General Psychiatry, Grob et al gave a relatively small dose (0.2 mg/kg) of psilocybin to patients with advanced-stage cancer and anxiety.  This was a double-blind, placebo-controlled study, with each subject serving as his or her own control. At least the authors call it a double-blind study.  In fact, as they admit in the discussion section, both subjects and researchers almost always knew whether the drug given was psilocybin or the placebo (niacin 250 mg).  So much for blinding.  Another detail not mentioned in the Scientific American article is that of Grob’s 12 subjects, 8 had previous experience with hallucinogens.  These were not your typical terminal cancer patients. Thus I have significant reservations regarding their conclusions: “This study established the feasibility and safety of administering moderate doses of psilocybin to patients with advanced-stage cancer and anxiety. Some of the data revealed a positive trend toward improved mood and anxiety”.

The study by Griffiths et al administered somewhat larger doses of psilocybin (30 mg/70 kg) to hallucinogen-naive adults active in religious or spiritual activities.  The control was methylphenidate (40 mg/70 kg).  Subjects completed a series of questionnaires including the “Mysticism Scale” and the APZ, which measures factors such as “oceanic boundlessness” and “dread of ego dissolution”. Some of these scales have not previously been used in drug research, and I have no way of knowing whether they are valid or not. Although the authors claim that psilocybin brought on positive mystical experiences in many subjects, 22% experienced “notable anxiety/dysphoria” and/or paranoia during the sessions.

Unfortunately, the Scientific American article presents an uncritical summary of this research, and often reads more like hype than science.  Perhaps this is not surprising when you have researchers presenting their own work.  But I get worried when they become so in love with their area of study that they claim, as they do here, that it “may ultimately be crucial to the survival of the human species”.


  1. Bradley Ball Says:

    Here’s the same comment I just sent to SCIAM:

    Mariette DiChristina, Editor in Chief, Scientific American

    In regard to “Hallucinogens as Medicine” from December 2010 issue by:
    Roland R. Griffiths
    Johns Hopkins University of Medicine

    Charles S. Grob
    David Geffen School of Medicine, UCLA
    Division of Child & Adolescent Psychiatry, Harbor-UCLA Medical Center

    I note that the report appears to be anecdotal accounts centered on culturally acquired beliefs and expectations of hand-picked participants without a control group for comparison. For a drug to be shown to be of benefit for psychiatric use a protocol including items similar to the following is widely thought to be necessary in order to minimize bias:

    1. Standardized selection of participants, i.e., the participants must all be selected according to the same checklist of criteria in which particular behavioral and/or emotional issues are indentified.

    2. A statistically significant number of participants must be used, particularly to minimize bias due to drop-outs as well as to be representative of the population at large.

    3. Participants must be randomly assigned to either the drug group or a placebo group while both groups receive the same scripted therapy, e.g., talk therapy, guided insight therapy, etc.

    4. A control group given no drug at all but undergoing a different standardized therapy must be included for comparison.

    5. Follow-up evaluations per original criteria should be
    conducted at timely intervals, e.g., 6 months, 1 year, 2 years.

    I did not see any of this mentioned in the article, but rather what appears to be selection of participants and analysis of outcome by a more or less personalized schema. Given that there are no alternate settings sans psilocybin to compare, the analysis is useless for science. It does not tell us anything other than researchers will tend to see only what they want to see when they set up a scenario to reveal only what they want to see.

    It is not surprising to me that SCIAM’s editorial take on this article is less than rigorous. I would hope that the same lack of rigor is not true of the authors. Currently the most widespread problem in psychiatry is that it has long ago become a subsidiary of the pharmaceutical industry. SCIAM in general is beholden to industry sources, and indeed so are many psychiatrists and other doctors who readily dispense so-called psychiatric medications like so many magic pills, even if the medications work no better than placebos, and even if the medications might cause brain damage (e.g., tardive dyskinesia resulting from use of neuroleptics).

    Bradley Ball

  2. Leon Says:


    Thanks for the great comments. I agree with much of what you say. I think the scientific laxity of the Scientific American article is the fault of both the magazine and the authors, whose related peer-reviewed studies — as discussed in my post — had major flaws. I agree that the pharmaceutical industry has an exaggerated and unhealthy influence on the psychiatric community, but am not clear that it has similar influence on Scientific American.