Much of the JAMA “complementary, alternative, and integrative medicine” series was a plea for more investigations of “alternative” methods. The overarching theme was that these methods are not taken seriously because they lack scientific evidence, but that this lacking is due to cultural, political, and economic forces, and not to any lack of real scientific potential1 Can this be true? Does fairness require that it is necessary to subject all claims, even those that conflict with established knowledge, to experimental studies?
A good example of a claim not worth studying is homeopathy. Why not study it? It is widely practiced, it has been the subject of many conflicting studies, and it is therefore cited by proponents and even some skeptics as needing more. Why should the facts suggest the cessation of studies when they seem to point to the need to find out, once and for all, if homeopathy really works?
The explanation is counter to the position “there exists no scientific explanation for how it could possibly work.”2 It can be shown, with no further studies, that homeopathy’s status is readily explained by established knowledge, that this explanation is overwhelmingly likely to be correct, and that more studies are therefore unjustified.
For a hypothesis to be judged useful, it must be subjected to several “criteria of adequacy”: testability, fruitfulness, scope, simplicity, and conservatism.* These criteria are generally accepted by scientists, including those who would, in principle, support the study of “alternative” methods. They are also implicitly used in many circumstances, even if they are not consciously articulated. Not every criterion, save the first, must be fulfilled to avoid discarding the hypothesis; but if not, there should be compelling reasons.
The two remaining “laws” of homeopathy together can be thought of as the “homeopathy hypothesis.” The first is the “law of similars,” which posits that materials that have effects similar to the patient’s symptoms cure those symptoms (like cures like). Second is the “law of infinitesimals,” which posits that after multiple dilutions, a dose of a remedy becomes more and more “potentized,” or powerful. Even when the preparation contains nary a molecule of the original substance, it is claimed to have a therapeutic effect attributable to that substance. Homeopaths grant that there is none of the original substance, but perceive that their patients get better and insist that there must be an explanation that supports the hypothesis. Some suggest novel hypotheses, e.g., “water memory,” while others don’t bother, arguing, “If it works, who cares why?” Granted, if the premise holds; but does it? Let us scrutinize the hypothesis:
- Is it testable? The homeopathy hypothesis is certainly testable. It has been tested, but it has not fared well in tests (see below).
- Is it fruitful, i.e., does it make novel predictions of phenomena besides those claimed by the original hypothesis? First, there is no example in nature of water, previously exposed to some relevant substance, exerting an action upon any system, biological or otherwise, attributable to that substance. One can imagine situations in which such a phenomenon should, if real, occur, but clearly does not. A survivalist in the woods should be immune to the bite of a poisonous snake, since the stream water he drinks must, at some point, have been in contact with the venom of any species indigenous to the region. All noninfant humans must be resistant to cholera, since all drinking water must have previously been in contact with its exotoxin. One need not fill a car radiator with antifreeze, since any water must, at some point, have been in contact with ice. Any substance whatsoever, including air, would be expected to arouse a patient from a coma, since volatile hydrocarbons must have been in contact with such a substance. Condemned prisoners can’t be expected to succumb to electrocution because the water that they drink must, at some point, have been exposed to lightning or the emissions of an electric eel. The possibilities are infinite.Homeopathy, moreover, predicts the opposite of the most relevant experimental data: the pharmacologic dose-response curves.
- Does the homeopathy hypothesis have scope? That is, does it explain more than competing hypotheses based on the law of mass action? No. Standard pharmacology explains much of what we observe about the relations between drugs and their substrates, and there is no reason to expect that new discoveries, e.g., elucidation of the mechanism of action of volatile anesthetics, will require any change in basic pharmacological concepts. Homeopathy adds nothing to the discussion. The homeopathy hypothesis doesn’t even explain homeopathy particularly well, as will be discussed below.
- Does homeopathy have simplicity? In a way it does, as suggested by its simplistic laws, but this is not what is meant by “simplicity” here. Simplicity means that the hypothesis need not require much rhetorical wrangling in order to justify itself. When subjected to this kind of scrutiny, homeopathy is found wanting. Since there is nothing but sucrose or water or alcohol in the remedy, novel hypotheses such as “spiritlike essence” or “water memory” must be invented to explain the supposed effect. Such hypotheses mostly or completely fail these same “criteria of adequacy.” This lack of rigor inevitably leads to even greater absurdities, such as the remedy known as “luna” (water exposed to moonlight), “potentized noise,”3 or Jacques Benveniste’s claim to be able to “digitize” biological messages from homeopathic dilutions and transmit them over the Internet.4
- Is the homeopathy hypothesis conservative? That is, does homeopathy avoid conflicts with well-established theories? We have seen that it does not. One such conflict is with Avogadro’s number. Another is the competing hypothesis of biological response based on the law of mass action, which has performed admirably during an era of immense discovery.5 Mass action has been particularly fruitful, e.g., in predicting the existence of molecular receptors.6 Still another conflict is with the physics of water. In fact, we would have to discard most of what we know about chemistry, biochemistry, physiology, molecular biology, and pharmacology in order to accommodate the homeopathy hypothesis.
In summary, the homeopathy hypothesis, when subjected to “criteria of adequacy,” fares dismally. How, then, to explain the outcomes of homeopathy itself? The problem must be approached by asking if there might be an alternative hypothesis that can both explain the outcomes of homeopathy and survive the scrutiny of the “criteria of adequacy.”
Here’s one: any apparent physical, chemical, or biological action of a preparation that has been diluted beyond Avogrado’s number is attributable to the carrier itself and the context in which such action takes place (the “carrier hypothesis”). The scrutiny:
- It is testable.
- It is somewhat fruitful. It predicts the action of a small amount of sucrose, water, or whatever else the carrier might be, on a substrate. In addition, it accurately predicts not only the outcomes of all the thought experiments described above, but also all the claims of the field known as “homeopathy,” both anecdotal and experimental. Clinical improvements apparent to practitioners and patients would be expected of any largely inert treatment because of myriad well-established phenomena, including but not limited to the natural course of most diseases, the placebo effect, willingness of the patient to please the practitioner, practitioner expectation, selective attention, misdiagnosis, and variable follow-up. The carrier hypothesis also predicts conflicting experimental outcomes. Less well-designed studies, particularly those without good controls against investigator bias, would be expected to show positive effects. The better ones would be less likely to show such effects, but nevertheless would include some that do, both because of chance and because of the extreme difficulty of controlling for such confounding factors as “cueing.” How does this prediction compare to actual data from homeopathy investigations? Here is the conclusion of a recent review of homeopathy trials:
There is some evidence that homeopathic treatments are more effective than placebo; however, the strength of this evidence is low because of the low methodological quality of the trials. Studies of high methodological quality were more likely to be negative than the lower quality studies. Further high quality studies are needed to confirm these results.7
- It has scope. It explains not only all the phenomena associated with homeopathy, but it also explains them better than does the homeopathy hypothesis itself. In the case of experimental investigations, for example, the homeopathy hypothesis predicts strong, consistently positive treatment effects, whereas the carrier hypothesis predicts the conflicting results that have actually been found.
- It has simplicity. Its acceptance requires no additional novel hypotheses.
- It is conservative. It poses no conflict whatsoever to existing, well-established knowledge.
And there you have it. The homeopathy hypothesis fails miserably when subjected to the “criteria of adequacy”; the most obvious competing hypothesis passes with flying colors. Further studies of homeopathy, moreover, are certain to yield one of two types of outcomes: the conflicting ones of the past or completely negative ones. Either of these conflicts with the “homeopathy hypothesis” but is entirely consistent with the “carrier hypothesis.” Any honest comparison between the two would overwhelmingly favor the latter. The conclusion is inescapable: further studies of homeopathy are unjustified.
* The format for this argument is adapted from chap. 7 of Schick T, Jr, Vaughn L. How to Think about Weird Things; Critical Thinking for a New Age. 2d ed. Mountain View, Calif: Mayfield Publishing Company; 1999. These are standard tools of the critical-thinking trade.
- Atwood KC. Factors that shape critical thinking about “alternative” medicine. Sci Rev Alt Med. 2001; 5(3): 160–164.
- Eskinazi DP. Factors that shape alternative medicine. JAMA. 1998; 280(18): 1621–1623.
- Ransom S. Homeopathy: What Are We Swallowing? Uckfield, England: Credence Publications; 1999: 64–68.
- DigiBio Research Laboratory home page. Available at: www.digibio.com.
- Ross EM. Pharmacodynamics: mechanisms of drug action and the relationship between drug concentration and effect. In: Hardman JG, Gilman AG, Limbird LE, eds. Goodman & Gilman’s The Pharmacological Basis of Therapeutics. 9th ed. New York, NY: McGraw-Hill; 1996: 29–41.
- Clark AJ. The Mode of Action of Drugs on Cells. London, England: E. Arnold & Co.; 1933.
- Cucherat M, Haugh MC, Gooch M, Boissel JP. Evidence of clinical efficacy of homeopathy: a meta-analysis of clinical trials. Eur J Clin Pharmacol. 2000; 56(1): 27–33.