Killed by something that doesn't exist

Placebos have no repeatable physical effect that can be broadly demonstrated to exist. But, if people believe the placebo can help them, it often does—especially for inherently subjective issues like pain relief.

Nocebos are what happens when a placebo (again, something that technically has no physical effect on the body) causes a negative side-effect, simply because the person believes that such side-effects are likely to happen to them.

There is a lot we don't understand about both of these effects. After all, running really detailed tests would inherently involve unethical behavior—intentionally not treating patients or intentionally trying to induce a negative reaction in them. But that doesn't mean you can ignore these phenomena.

A great example comes in a recent column by Alexis Madrigal on The Atlantic. You're probably familiar with the idea of sleep paralysis—the experience of waking up, being mentally awake, but still physically paralyzed. This happens to people all over the world. And, all over the world, it's long been explained in folklore as the work of demons and evil spirits. (The fact that sleep paralysis is often accompanied by feelings of terror, and the sensation of something sitting on your chest doesn't hurt in that regard.) Normally, sleep paralysis brings a few minutes of terror, but no lasting harm. In the mid-1980s, however, it suddenly became capable of killing. The catch, the men it killed were all recent Hmong immigrants, living in the United States. Researcher Shelley Adler thinks it was actually a nocebo effect that killed these men—they believed themselves into an early grave.

[In America] some Hmong felt that they had not properly honored the memories of their ancestors, which was a known risk factor among the Hmong for being visited by the tsog tsuam. Once the night-mare visitations began, a shaman was often needed to set things right. And in the scattered communities of Hmong across the country, they might not have access to the right person. Without access to traditional rituals, shamans, and geographies, the Hmong were unable to provide themselves psychic protection from the spirits of their sleep.

Drawing on all this evidence, Adler makes the provocative claim that the Laotian immigrants of the 1980s were in some sense killed by their powerful cultural belief in night spirits. It was not a simple process.

"It is my contention that in the context of severe and ongoing stress related to cultural disruption and national resettlement (exacerbated by intense feelings of powerlessness about existence in the United States), and from the perspective of a belief system in which evil spirits have the power to kill men who do not fulfill their religious obligations," Adler writes, "the solitary Hmong man confronted by the numinous terror of the night-mare (and aware of its murderous intent) can die of SUNDS."

Via Christopher Ryan


  1. Since there was just a post about Wade Davis, it’s worth mentioning that he discusses this in depth as a possible cause of Vodou magical killings (and zombies) in The Serpent and The Rainbow — the belief that a powerful sorcerer had cast a spell to kill you could be strong enough to actually kill you.

    I don’t think he uses the term “nocebo,” though, as it probably hadn’t been coined.

  2. Fascinating post.  I’m pretty sure those Hmong immigrant deaths (or some case very similar to them) was the inspiration for  Wes Craven’s first Nightmare on Elm Street movie.

  3. If you’re interested in placebo studies, check out Harvard’s Program in Placebo Studies. It’s hosted at Beth Israel! The director is Ted J. Kaptchuk, who wrote The Web that Has No Weaver, a well-known text on classical Chinese medicine. Really fascinating to see a (former?) acupuncture proponent heading a placebo program at what is arguably the most prestigious hospital in America.

  4. I’m saying these men were literally killed by nightmares.  After all you’ve seen, Scully…why is it so hard to believe that the mind has powers that can’t be explained by science?

  5. Isn’t it pretty much demonstratable that psychological stress and torment can cause physical illness? As an anecdote I definitely know that emotional abuse/psychological torment has lead to physical illness and destructive behavior due to mental problems in myself. It doesn’t seem far fetched that stress and anxiety could exacerbate the sleep problems, causing more stress and anxiety, and the whole thing could become destructive for the person.  Also, night terrors suck. I had them all the time for a long time and it really *does* seem like *something* is doing that to you. The worst is when you can actually see the room, but have the sensation of something in it holding you down while you… try to scream. 

  6. Had a terrifying, first experience with sleep paralysis after reading

    “The Horla” by Guy de Maupassant in college. Dreamt someone came

    into the room and was suffocating me with a pillow.  I woke up and couldn’t move.

    Scared the crap out of my roommate when my scream came out.

    “Killed by something that doesn’t exist”  Next up, national boundaries.

  7. In p2 Maggie calls a placebo “something that technically has no physical effect on the body”, but isn’t that inaccurate?  A placebo or nocebo definitely has a physical effect — in the Hmong immigrant case, death — right?  The mechanism is just not understood.

    I think in the Hmong immigrant case it’s actually the Hmong immigrants’ explanation that is the more technically and scientifically accurate one — it’s more technically and scientifically accurate to say they were killed by tsog tsuam spirits than to say they were killed by a nocebo effect.  Both are true in some epistemological sense, but the spirit description is more useful and thorough.

    The spirit model has predictive power – it apparently prescribes a visit to a shaman and rituals that can fix the condition, and although I guess we don’t know if that actually would have worked in their specific case, I don’t have any reason to doubt that it would have… it worked for the Hmong who did have access to the shamen and rituals, right?

    It’s also a more successful scientific model by Occam’s Razor standards – nocebo has to invoke some kind of nebulous psychosomatic process that cannot be observed or interrogated whereby the belief in the ghosts causes the effect the ghosts are believed to seek.  Why say “the Hmong believe in X that want to do Y and because of some not-understood psychological process Z, this belief causes Y to actually come about” when you already have the much simpler option of “X causes Y”, which explains the same set of facts much more succinctly?

    The Hmong model of the affliction is better developed than the nocebo or SUNDS models, even though it’s just a superstitious ghost story.

    On the other hand, I guess the nocebo/SUNDS models predict you can avoid death by not believing in the spirits, which also probably works… so maybe those are better after all.

    1. You would have to demonstrate that somebody — let’s say a Hmong at least — raised without contact with Hmong culture and never told about the spirits also experiences sleep paralysis and eventual unexplained death during sleep before you can just say “X causes Y.” Otherwise you are gripping Occam’s Razor by the blade, as many people do, if your idea of a “simple” explanation is that “X causes Y if you are a Q and somebody has told you about the X, and may or may not if you are a Q who does not believe in X, or if you are not a Q but have been told about the X but believe it.” Because clearly, ancestors can haunt people to death, but not without proper permission. That would just be absurd.

      What they’re saying is, you’re told the pill is sugar, you eat it and nothing happens. You’re told the pill is poison, you eat it and you die. You can describe the conclusion any number of ways, but I’m pretty sure it is not, “The pill kills you.”

      (Oh, and if further clarification is needed, I have experienced sleep paralysis a number of times but my ancestors have not killed me, even a little bit.)

      In any case we’re overlooking the really important lesson here, that “Nocebo” would be a great name for a band.

  8. As fascinating as the topic is, the veneer of institutional ‘scientific’ condescension is rather offputting. Like, really? Unexplained and experimentally uncontrollable negative effects of the, at best, vaguely understood or classified entities ‘beliefs,’ themselves about a poorly explained phenomenon (sleep paralysis) occurring at the edge of a much bigger and at least as poorly explained phenomenon (sleep) – this is the ‘explanation’ that is so clearly distinct from and preferable to muddled, superstitious notions like spirits and shaman rituals? Get real.

    Generally, I honestly can’t think of a more useful definition of ‘real’ than ‘that which can kill you.’

  9. If I put proper citations in this I’m pretty sure I’ll be blocked as spam, because there are tons of respectable, peer reviewed studies chemically measuring the placebo affect.  Being told you’re taking a painkiller raises the level of endorphins in your blood, which could account for lessened sensations of pain.  Delivering an opiate blocker along with the placebo pain killer results in no pain relief.  People told they’re about to get something delicious experience a larger uptick in ghrelin levels than people told they’re getting a diet shake, and the two groups respond different after consuming the same thing.  And that’s what I can think of off the top of my head.

    1. So does that mean Science-Based Medicine’s frequent rants about objective pacebo effects in general, and Kapchuk in particular, are off base?

      1. There’s a lot here and this is the first time I’ve read SBM so I can’t claim authority on it, but my reading of this particular post is that, in an attempt to fight some very harmful ideas, they’re overlooking some helpful nuggets.   For example, you can’t cure cancer by wishing really hard, but depression increases mortality from cancer.  Given that antidepressants don’t seem more effective than placebos in many people, it’s not ridiculous to say that placebos could be a useful treatment in context, especially for patients whose bodies are already under tremendous strain

  10. All of a sudden, hundreds of thousands of people have Tourettes and other faked tweaks and loose nuts.  Most of it’s B.S. I have proof. 

    I worked in a mental health facility for about a year. When I started, none of the patients bit their fingernails. But I did. Staff noticed some of the residence started biting their nails and couldn’t figure out why, the first group was  mainly  autistic.  A few months later, and a few more patients gnawing their nails, the staff  figured out maybe one or two learned from me, then it snowballed.  We all wondered what else they “copied”. 

    The strange thing was, I only bit my nails when I charted at the end of the day, usually at the front desk at 11 p, not many of them could have seen me going down on myself. 

    1. Right, and you don’t think copying a mild and typically not distracting habit to an extreme extent is a bit… I don’t know… imbalanced? 

      There is a lot of interesting research on the “shape” of mental illness, its manifestations. These vary an amazing amount culturally, but what doesn’t vary is the existence of behaviors that a culture might refer to as “crazy.”

      Also, that’s not evidence. It’s a nice story. We all have nice stories, but it’s never a good idea to broach them as more than anecdote, especially when making a massively wild assertion like “Most of [mental illness?] is B.S.”

  11. Also the sensation of being levitated out of your bed to a place with instruments and people with big heads and large eyes and the famous anal probe.

  12. These simple minded explanations kill me.  Schizophrenia was originally caused by a cold, withholding mother, look it up in your history books kids, ulcers by stress, and the spanish flu by those unclean carriers of plague, the spanish(no offense meant to our hispanic brethren).
    No autopsies, no serious medical inquiry and we’re good with the easiest, quickest explanation possible, hysteria, stress and a good glossing over.
    I recommend Charles Forte’s books which discuss supernatural and unexplained phenomena but are most fascinating for the demonstration of how explanations are formed by “experts” with no factual information to quickly close discussion.
    Given that for many serious diseases we don’t know the cause and don’t know the cure we would be better off approaching medical mysteries with humility rather than hubris.

  13. The book The Spirit Catches You and You Fall Down:A Hmong Child, Her American Doctors, and the Collision of Two Cultures by Anne Fadiman explains the power of “spirits” or the internal/mental side of health that Hmong people feel. As Westerners we have ignored, or more accurately forgotten about that side of our medicinal practices. While I do think we can explain many of our ailments with our medicine and science, our brains are already wired to accept treatment in the form of shamans and other spiritual leadership so it might be reasonable to augment modern treatment of disease with more subtle and non-physical treatments. 

Comments are closed.