Personalized placebo effect

A new article in Nature explores the placebo effect–therapeutic benefit from taking medicine without active ingredients. On the one hand, minimizing the placebo effect is a principle of ethical medical practice, a bulwark against hype, oversell, and charlatanism. Without controlling for therapeutic effects not caused by the drug, one may overestimate the drug’s potency. On the other hand, if the goal is to make the patient well, one might consider it ethical to use every option available—including whatever mysterious mechanisms the body can muster to augment the pharmaceutical therapy. The oldest tradition in medicine is that of helping the body heal itself.

Most interestingly, the placebo effect is highly individualized. This is not surprising, given the complex psycho-somatic interactions involved. Yet in an age in which personalized medicine is the idea of the moment, it seems essential to open the black box of the placebo effect. The authors summarize not only psychosocial variables known to influence patients’ responses to drugs—expectations, anxiety states, hypnosis, and so forth—but also a range of genetic and anatomical correlates of placebo responses. The strongest data are in the area of drugs for anxiety disorders and depression. Documented examples include polymorphisms in serotonin loci and in modulators of monoaminergic tone, plasma noradrenaline levels in interleukin-2 release, and brain anatomy in placebo analgesia.

Such data are exciting in historical as well as therapeutic context. In 1902, at the dawn of Mendelism, Archibald Garrod suggested that humans were as variable and individualized at the biochemical level as they are at the phenotypic level. If this were true, he continued, the phenomena of obesity, the various tints of hair, skin, and eyes, and “idiosyncrasies as regards drugs and the various degrees of natural immunity against infections” could be amenable to biochemical-genetic analysis.

Rufus of Ephesus
Rufus of Ephesus

Going further back, the Hippocratic physician Rufus of Ephesus noted that drugs act differently on different people, and that to prescribe appropriately the physician must ask the patients about their habits, their diet, preferences, sleep patterns, familial diseases, degrees of pain, and numerous other facets of the patient’s history and context. For the Hippocratic physician the body’s own powers of healing, mysterious though they be, were a crucial component of therapeutics. The Nature article’s abstract concludes with a nice statement of individuality that sounds like updated Hippocratism:

Personalizing placebo responses — which involves considering an individual’s genetic predisposition, personality, past medical history and treatment experience — could also maximize therapeutic outcomes.

Personalized medicine is often portrayed as a return to ancient Hippocratic ideals. It is and it isn’t. Despite the hype, “one-size-fits-all” medicine is too cost-effective to abandon. Personalized medicine has always been reserved for those who can pay for it, and little in the age of high-tech genomic medicine seems likely to change that. And yet, this attention to the unique constellation of essential and historical factors that combine in a given patient’s “irrational” drug response does bode well to be anodyne against some of the dehumanizing forces in the recent history of medicine.


Anti-determinism on the march!

Nice piece today from SciCurious, guest blogging over at Scientific American. The post is an analysis of a recent article in Nature claiming that by knocking out serotonin in two different ways (both neurotransmitter production and receptors), they abolished sexual preference. The mice apparently mounted either sex with equal frequency.

SciCurious does a beautiful job dissecting the assumptions in the Nature article, analyzing the data, presenting alternative hypotheses, and looking at the history of the research. For example, the authors might have merely lowered the threshold of sexual activity–an extension of “all girls get prettier at closing time”. Or perhaps the researchers influenced the perception of other cues, for example olfactory cues. “So does this paper prove that there are drastic increases in sexual behavior associated with low serotonin?,” SciCurious writes. “Absolutely. Does it show that low levels of serotonin change sexual PREFERENCE? Well, that’s difficult to say.”

Also, Ed Yong looked at the article from a different but equally skeptical point of view. He points out how difficult it is to translate these kinds of behavioral findings from mice to humans. Further, he writes, “serotonin isn’t all about sex.” When I was a teaching assistant for the Neural Systems and Behavior course at Cornell back in the late 1980s, we used to drill in to the students’ heads the idea that neurotransmitters do not have behaviors. They act in many regions of the brain and influence all sorts of behaviors in ways that are very far from straightforward.

Yong worries (rightly) that anti-gay groups will use findings like this to argue for a simple biological basis for homosexuality, perhaps even proposing serotonin therapy as part of their effort to “cure” it. And SciCurious links to news stories soon appeared suggesting that the researchers had “turned mice gay.”

Such stories illustrate a fundamental fallacy that is one of the gravest dangers of popularizing science. For the sake of argument, let’s say the researchers did in fact eliminate sexual preference. In what sense is “no sexual preference” the same as “gay”? Ans: only in a world so normative that strict, unwavering heterosexuality is the only behavior considered normal. Of course, there are lots of people like that–I read about them all the time. But it is blinkered, naive, and deeply chauvinistic.

Biological and especially genetic explanations of behavior are a double-edged sword. The gay community has oscillated in its support for research to find “gay genes” and other traces of the biological basis of homosexuality. If homosexuality is innate, the reasoning goes, then it is cruel and pointless to try to “cure” gays, in the same way it was cruel to “cure” left-handed people.

But “gay” is a cultural construct. There were no “gays” in Ancient Rome or in 19th century Paris, and there are no gays in the Foré of Papua New Guinea. Here and now, in our culture, we need the term in order to protect human rights that are trampled on by people unreflectively absorbing an outdated cultural taboo on homosexual activity. But in the long run, the ideal should be to get rid of the concept–for us all, in short, to be “gay-blind.”

Good skeptical science writing helps that cause, because it exposes fallacies in the ways we think about science. I’m fine with describing a physiological mechanisms for a behavior, but we need to be careful not to equate mechanism with cause. I’m wary of science writing that talks about the “roots” or the “basis” of complex behavior or disease. It implies a hierarchy that blinds us to many biological mechanisms that work in the other direction. In biology, cause and effect go both directions: behavior changes gene activity as much as gene activity changes behavior. Studies purporting to examine the biological “basis” of behavior rely on cross-species analogies and make unsupportable assumptions about motivation.

In short, a “gay” mouse is a ludicrous concept.



%d bloggers like this: