The History of the “Allergy” Myth of Alcohol Addiction

    In 2022, it is still not uncommon to hear people claim that an addiction to alcohol is an “allergy.” Little wonder, when the Big Book of Alcoholics Anonymous, the abstinence-based fellowship of around 2 million members, contains a section titled “The Doctor’s Opinion,” which reads, in part:

    “We believe … that the action of alcohol on these chronic alcoholics is a manifestation of an allergy; that the phenomenon of craving is limited to this class and never occurs in the average temperate drinker. These allergic types can never safely use alcohol in any form.”

    The Big Book was published in 1939, but the internet has only made this misinformation more widely available. Googling “the doctor’s opinion” today will turn up countless articles on sites hosted by 12-step treatment centers, declaring the absolute truth of this debunked allergy theory.

    Yet the theory preceded AA, and it wasn’t initially applied to alcohol. We’ll take a look at its history. First, we should make clear what we’re talking about.


    What Is an Allergy?

    Our immune system protects us from diseases by sending out antibodies which can neutralize toxins or destroy or help other parts of the immune system to destroy viruses, bacteria or other parasites. Antibodies only attack proteins, typically large molecules between 5,000 and 50,000 times the mass of a hydrogen atom.

    In the process of defending our bodies against these invaders, antibodies cause the release of chemicals such as histamines, which can lead to runny noses, sneezing, rashes, etc.

    In some people, harmless substances such as cat dander or ragweed pollen can activate antibodies in the immune system, leading to the release of these chemicals. These people are said to have an allergic reaction—an “allergy” to a particular substance.

    Any substance which triggers an allergic reaction is called an allergen. Typically, allergens are large protein molecules. However, sometimes small molecules can also trigger an allergic reaction. A good example of this is the allergy some people have to nickel. In this case, a nickel atom binds to a protein in the human body and turns it into an allergen, which antibodies then attack.

    Antibodies were discovered in 1890. By 1891 they were being used to create a serum known as diphtheria antitoxin, which could neutralize the toxin released by diphtheria bacteria in the human body. Serums are created by introducing a small amount of a toxin into an animal, such as a horse. The animal then produces many antibodies to neutralize the toxin. A small amount of blood plasma from the animal can be injected into a human to neutralize the toxin. The German physiologist Emil von Behring won the Nobel Prize in 1901 for his discovery of diphtheria antitoxin.



    The idea that morphine addiction was some sort of allergic reaction was first introduced by Dr. Carlo Gioffredi in 1897. The Italian published five papers on the topic, in 1897, 1897, 1898, 1899 and 1900. Antibody theory was new at the time and very little was known about how antibodies worked, so Gioffredi’s hypothesis was not seen as outlandish.

    Based on experiments he conducted with cats and dogs, he claimed that a serum extracted from an animal which had developed a tolerance to morphine could confer morphine tolerance when injected into another animal. In 1902, Dr. Leo Hirschlaff also published a paper, claiming he had replicated Gioffredi’s results by using rabbits to create the serum and mice as test animals.

    Pushback soon arrived, however. In 1903, Dr. Julius Morgenroth published a paper which pointed out that Hirschlaff’s estimate of the fatal dosage of morphine was incorrect. Moreover, Morgenroth had attempted to replicate Hirschlaff’s mouse experiment and found that the serum had no effect at all on the rodents. Dr. Max Cloetta published another 1903 paper finding the serum had no effect on mice.

    The definitive study disproving the morphine allergy theory was published in 1925.

    The argument went back and forth. In 1907, Dr. Georg von Marikovszky published a paper claiming that he had found the serum to be effective on guinea pigs. A 1916 paper by Dr. Johannes Biberfeld showed no effect. In 1920, Dr. Emil J. Pellini and Arthur D. Greenfield published a paper showing that the serum had no effect on cats or mice. In 1914, Dr. Adriano Valenti published a paper which claimed that if serum from a dog undergoing morphine withdrawal were injected into a normal dog, the normal dog would also show signs of morphine withdrawal. However, in 1924, Pellini and Greenfield published a paper showing that this serum did not produce withdrawal symptoms in either dogs or cats, and attributed Valenti’s results to poor experimental design.

    But the definitive study disproving the morphine allergy theory was published in 1925 by pharmacist Andrew Grover DuMez and Dr. Lawrence Kolb, who worked together at the Hygienic Laboratory in Washington, DC (the forerunner of the National Institutes of Health).

    DuMez and Kolb spent two years conducting a series of meticulous experiments on white mice, and proved that the serum had no effect. Since the morphine molecule is rather small, with a mass about 285 times that of a hydrogen atom, and since morphine is an alkaloid rather than a protein, it is not surprising that morphine fails to act as an allergen.

    A careful clinical trial of Narcosan showed that it actually made withdrawal symptoms worse.

    However, a few cranks continued to posit the theory even after 1925. In 1926 Narcosan, a creation of biochemist Alexander Samuel Horovitz, was touted as a miracle cure for morphine withdrawal; it was stated that the lipoids comprising Narcosan neutralized the antibodies created by morphine, thus relieving withdrawal symptoms.

    A careful clinical trial of Narcosan showed that it actually made withdrawal symptoms worse. (Horovitz had previously tried to sell his lipoid concoction as a cancer cure.)

    Chemist Ivan Ivanovich Ostromislensky published a 1935 paper claiming that morphine withdrawal was a form of anaphylactic shock caused by antibodies formed by morphine administration, and that a coal-tar derivative known as rossium could relieve this shock and eliminate withdrawal symptoms. A clinical trial proved rossium to be worthless for treating opiate withdrawal.



    The first discussion of a serum for alcohol occurs in an 1896 paper, the first of two that year, by Dr. Édouard Toulouse. Toulouse reported creating an anti-alcohol (aka anti-ethylic) serum by giving alcohol to dogs and using the serum to successfully treat a case of delirium tremens.

    Drs. Emmanuel Joseph César Sapelier and Gabriel Dromard expanded on Toulouse’s work and published their research in a book in 1903. The pair created their serum by giving alcohol to horses, and using it to treat people for heavy alcohol use. Sapelier and Dromard claimed that humans injected with the serum developed an aversion to alcohol and quit drinking.

    Dr. William Silkworth seems to have been the first person in the United States to use the term “allergy” in reference to alcohol addiction.

    Anti-alcohol serum treatment became very popular in France, Spain and Latin America; however, it was never used in the United States and is almost completely absent from the US medical literature. A good discussion of the European and Latin American literature is found in Margaret W. Robinson’s 1950 paper, “The Immunological Properties of Alcohol: A Survey of the Literature.”

    Dr. William Duncan Silkworth seems to have been the first person in the United States to use the term “allergy” in reference to alcohol addiction—in his 1937 paper “Alcoholism as a Manifestation of Allergy,” published in the Medical Record.

    By 1937, the Medical Record had become pretty obscure; however, Silkworth was unable to publish in a quality medical journal since his paper contained no experimental evidence. Instead, it expounded his hypothesis that “alcoholics” and “social drinkers” were as categorically different from each other as Mendel’s green and yellow peas, and postulated an allergy as an explanation for that supposed difference.

    Silkworth’s paper was completely ignored by the medical community, and would almost certainly have been forgotten had not AA founder Bill Wilson prevailed upon him to write “The Doctor’s Opinion” as a part of the frontmatter of the Big Book. (There was limited prior discussion of the allergy theory of alcoholism in the US, although Dr. Walter Clark Ashworth had published articles discussing alcohol antibodies in 1929, 1931, and 1932.)

    Silkworth, although arguably a good medical practitioner, wasn’t much of a scientist. Prior to his allergy theory, Silkworth had only published two much earlier papers in the New York Medical Journal: “Notes on the Jungle Plant Combretum Sundiacum (1908) and “A Further Report on the Jungle Plant Combretum Sundiacum in Morphine Addiction” (1909). Although Silkworth claimed that Combretum sundiacum was an effective treatment for morphine addiction, a 1915 item in the Journal of the American Medical Association declared it worthless.

    Silkworth had worked at the ambulatory morphine reduction clinic in Newark, New Jersey in 1919 and 1920. Around 1932, he became medical director of the Towns Hospital in New York. He met Bill Wilson there in 1934, when Wilson checked in to detoxify.

    Thanks to press exposure, AA grew rapidly in the 1940s and 1950s, and many people, including doctors and medical researchers, were exposed to Silkworth’s allergy theory through the Big Book. But since the alcohol molecule is even smaller than the morphine molecule, with a mass about 46 times that of a hydrogen atom, it is an unlikely candidate for an allergen, and most doctors remained skeptical.

    The definitive study of the allergy theory of alcoholism was published by Robinson and Dr. Walter L. Voegtlin in 1952: “Investigations of an Allergic Factor in Alcohol Addiction” in the Quarterly Journal of Studies on Alcohol.

    Robinson and Voegtlin used pure ethyl alcohol diluted with neutral saline solution in their tests to make certain that there were no impurities which could cause an allergic reaction. They worked with over 50 rabbits and 17 humans (two abstainers, five social drinkers, and 10 people with alcohol addiction), to ensure that effects were not due to random chance.

    Robinson and Voegtlin first performed a precipitins test, to see if there was any difference between the anti-alcohol serum taken from the blood of rabbits treated with alcohol and the normal serum taken from rabbits which had never been given alcohol. In the precipitins test, if antibodies are present, they should bind with antigens to form large clumps. However, there were no differences between the reactions to the serums from the two groups of rabbits; the one showed no more clumping than the other. The results with human serum were the same; there was no difference.

    Robinson and Voegtlin also performed an anaphylactic reaction test. One group of rabbits was given an injection of purported anti-alcohol serum before being dosed with alcohol; another group was given no treatment before being dosed with alcohol. If antibodies were present, the first group should have had a much greater reaction to the alcohol than the second. However, there was no significant difference.

    AA has continued to endorse the idea of “an allergy of the body and an obsession of the mind,” ignoring science in the process.

    Two other tests were performed which I shall not detail here; neither showed the presence of any alcohol antibody. Robinson and Voegtlin’s experiments were definitive: Alcohol addiction is not the result of an allergy or of alcohol antibodies.

    In his 1960 book The Disease Concept of Alcoholism, E.M. Jellinek declared the allergy theory as dead as a doornail. However, AA has continued to endorse the idea of “an allergy of the body and an obsession of the mind,” ignoring science in the process.

    There is no categorical difference between so-called “alcoholics” and “non-alcoholics.” Alcohol use disorders exist on a continuum from mild to severe, and a mild alcohol use disorder can remit without ever becoming severe.

    But for some people, a black-and-white biological explanation for addiction has always seemed more attractive than the psychological and societal factors that require far more work to address.



    Photograph by Rebecca Bollwitt via Flickr/Creative Commons 2.0

    • Kenneth is the founder and CEO of HAMS (Harm Reduction, Abstinence, and Moderation Support), the world’s only lay-led, free-of-charge support group for harm reduction for alcohol. He is also the author of How to Change Your Drinking: a Harm Reduction Guide to Alcohol; BETTER IS BETTER!: Stories of Alcohol Harm Reduction; and the book series The Untold History of Addiction Treatment in the United States—which currently consists of Strychnine & Gold (Part One), Strychnine & Gold (Part Two), and From Inebriate Asylums to Narcotic Farms, with more to come.

    • Show Comments

    You May Also Like

    Five Harmful Anti-Alcohol Myths and the Evidence Against Them

    In Temperance America and beyond, it seems no amount of evidence will be accepted ...

    Drug Reporters Know This Is a War―So Why Don’t We Cover It Like One?

    [This article contains graphic images of injecting drug use.] A picture may be worth ...

    With the Focus on Opioids, Don’t Forget About Meth and Cocaine

    The “opioid crisis” has dominated drug conversations for at least the past decade, while ...