Excerpts from The Natural Mind

by Andrew Weil

Retrospective studies are risky ways of framing hypotheses; they are fraught with logical traps known to the ancients, and it is remarkable that men of science still fall for them.

The saga of LSD and chromosomes is a case in point, for much of the evidence was of this retrospective sort. The initial hypothesis, first reported in 1967, was based on the observation that LSD users seemed to have a higher frequency of broken chromosomes in certain white blood cells (lymphocytes) than "normal" persons (1). The _New England Journal of Medicine_ gave this observation great prominence in an editorial titled, "Radiomimetic Effects of LSD," suggesting that the drug mimicked radiation in its damaging effects on genetic material. Evidence that was more circumstantial then appeared: LSD was shown to affect chromosomes of cells growing in test tubes; a few mothers who had used LSD gave birth to deformed babies. The scientific and lay press gave all these findings front-page attention. The National Institute of Mental Health eagerly seized upon and disseminated the new information in a propaganda campaign against LSD. And, for a few months, use of the drug appeared to decline.

But throughout this campaign, a number of facts were overlooked. First was the total absence of any prospective studies supporting the hypothesis. No one had tested the hypothesis in a legitimate way -- by looking at chromosomes before exposure to the drug, giving the drug in a controlled fashion, and then keeping watch on chromosomes. Second was the known fact that many things affect chromosomal integrity, among them such common drugs as aspirin and chlorpromazine (Thorazine) and recent viral infections. No effort was made to control for these other factors in the clinical cases. Third was the general problem of tissue-culture studies: cells growing in test tubes do not behave the way cells do in the body. In addition, the doses of LSD that caused visible changes in chromosomes of tissue-culture cells were far higher than the doses living cells get when a person takes an acid trip. Fourth, chromosomal breaks are seen in cells of all people; the arguments turned on a statistical difference in frequency, not an all-or-nothing difference, and the frequency of chromosomal breaks in lymphocytes seems to correlate more directly with laboratory technique than with other variables. (The technique of preparing lymphocytes to make chromosomes visible is complicated and likely to produce factitious changes.) Fifth, the lymphocyte is one of the only cells in which human chromosomes can ever be seen under the microscope. Even if the changes were real, they said nothing about the state of chromosomes in other cells (such as reproductive cells). In fact, through the whole controversy no one showed _why_ it was bad to have broken chromosomes in your lymphocytes. It sounds bad, certainly, but one cannot say that it is bad without making a number of shaky assumptions.

All of these logical flaws in the medical arguments against LSD were obvious in 1967. They do not mean that the hypothesis should never have been published, but surely it should not have been promoted by the medical profession, the press, and the National Institute of Mental Health without more thought. And it is significant that these logical flaws were first pointed out in the _Berkeley Barb_ and other underground newspapers at least eight months before the _New England Journal of Medicine_ voiced similar doubts. The necessary prospective studies were not published until the end of 1969 (2). Not surprisingly, they failed to demonstrate any relationship between LSD use and chromosomal changes. They generated very little national publicity.

This episode ought to be profoundly embarassing to journal editors and government scientists. At one stroke it created an irreparable gap between users of drugs and drug experts. Since 1968 I have not met a single user of hallucinogens who will believe any reports of medical damage associated with drugs, and the use of hallucinogens has never been higher.

(1) M. M. Cohen, K. Hirshhorn, W. A. Frosch, "In Vivo and in Vitro Chromosomal Damage Induced by LSD-25," _New England Journal of Medicine_ 227 (1967), p. 1043.

(2) J. H. Tjio, W. N. Pahnke, A. A. Kurland, "LSD and Chromosomes: A Controlled Experiment," _Journal of the American Medical Association_ 210 (1969), p. 849. For a recent review of the whole field, see N. I. Dishotsky, W. D. Loughman, R. E. Mogar, W. R. Lipscomb, "LSD and Genetic Damage," _Science_ 172 (30 April 1971), p. 431.