Human beliefs activate neural circuits in a dose-dependent manner
A brain imaging study demonstrates that nicotine-related beliefs are so powerful that they can induce dose-dependent neural responses
Happy Friday! Welcome to the new subscribers of GWAS Stories. Every Friday I pick one of my past Twitter posts and share it in the GWAS Stories. This is an effort to transfer all my past writings from Twitter to Substack.
For this week, I picked a post that I posted on Twitter last year October while I was attending the World Congress of Psychiatric Genetics (WCPG) in Montreal. I remember receiving a message about a preprint from my colleague Jonathan Marchini with the comment “mind-blowing” (I don’t think I have heard those words from him before). When I read the abstract I felt the same and dropped everything else and started drafting a Twitter post about the paper.
From the Twitter archives
Something I read recently and thought very relevant in the current political state. This paper will probably blow your minds off and will remind you how profoundly powerful are human beliefs.
A preprint (Perl et al. 2022) posted last year shows that beliefs about drug intake can be so powerful that they influence not just the behaviour but also the molecular processes driving the behaviour (i.e., the neural responses) in a dose-dependent manner, something that was thought only possible with the actual drugs.
In a fascinating experiment conducted by Xiaosi Gu and colleagues from Mount Sinai, twenty nicotine-dependent human participants were asked to smoke e-cigarettes and then, perform a monetary decision-making task (known to be influenced by nicotine) while undergoing an fMRI scan. The participants were made to believe that their e-cigarettes contained either low, medium, or high levels of nicotine, while in reality, all the cigarettes had the same amount of nicotine. What did the scans show?
As the participants performed the decision-making task under the scanner, a brain region called the thalamus (which is a key binding site of nicotine and is densely populated with nicotine receptors) lit up at an intensity proportional to the dosage of nicotine that the participants believed to have smoked.
The dose-dependent effect between drug intake and brain activity has been well documented in the context of tobacco smoking, alcohol intake, etc. The underlying molecular mechanisms, too, are well understood. For example, when a person smokes, the nicotine from the cigarette enters the brain, binds to neuronal acetylcholine receptors (nAChR) in brain regions housing the reward neural circuits (striatum, thalamus, etc.), releases dopamine and activates the neurons projecting to the higher brain centres (cortex). This manifests as rewarding outcomes—a relaxed state, elevated mood, increased concentration, etc. The higher the amount of nicotine a person smokes, the higher the brain activation and the higher the rewarding effects (up to a certain limit, of course, beyond which the aversion neural circuits take over, preventing the person from smoking further). This kind of dose-response relationship requires a pharmacological agent such as nicotine to activate the neural circuits.
A Pavlovian type of conditioned response between smoking cues and brain reward centre activation has been shown before. However, this is the first time we are learning that the brains of chronic smokers can be conditioned to the level that they can autonomously activate the neural circuits causing a physiological response in proportion to the dosage of nicotine that a person merely imagines. That is mind-blowing. If this turns out to be true, then this has major implications for addiction treatment.
Apart from addiction, the findings are also perhaps relevant in the context of other types of human beliefs (e.g. radicalization) and their associated behaviours. The findings suggest that it is not only possible to easily implant beliefs in a conditioned human brain, but one can fine-tune the beliefs to precisely produce the desired behavioural outcomes. Changing one’s beliefs (e.g. using cognitive behavioural therapy) would often mean completely rewiring one’s brain at the molecular level which is, as you can guess, not a trivial task.
This work was published in Nature Mental Health a few days ago if you’d like to read it in detail.
I'm going to the ice down some of the claims here, with the caveat that I can only read the abstract and access the open data of the study.
Foremost, this is a behavioral fMRI study with n = 20, which inherently merits some skepticism until it's replicated with a larger sample.
Second, belief-dependent activation was only seen in the thalamus. The area of the brain most directly implicated in addiction (striatum) did not show a change in response, which is what I'd wager the authors were expecting/hoping for instead. Given that only the thalamus (a very busy chunk of tissue) showed a difference, there's no way to know whether this reflects some kind of auto-simulation of nicotine consumption, or another process. The same applies to the thalamic-vmpfc connection.
Third, no comparison is made to the effects of actually vaping low, medium, and high amounts of nicotine. A difference is observed between the groups, but it could be a trivial one by comparison to the actual effects of variable levels of nicotine, or have a different function over dose.
Fourth, the activation was dependent on receiving nicotine to begin with.
It's therefore not accurate to say that "the brains of chronic smokers can be conditioned to the level that they can autonomously activate the neural circuits causing a physiological response in proportion to the dosage of nicotine that a person merely imagines." Activation was neither autonomous nor (as far as we know) in proportion to the dosage the person believed they received, nor are we sure these are the relevant neural circuits. Whether you want to call a change in thalamic activity 'physiological' might be a thing to quibble over, but most readers would think you meant something like heart rate, not how much blood is flowing to a specific spot in the brain. And even if that thalamic activation perfectly matched expected nicotine response, it's one small piece of what happens when nicotine is used. When your conditions are "low, medium, and high" all you can really see is if the effects are in the expected direction, and as mentioned, there is no comparison group which did receive variable levels of nicotine. As subjects in this study did not generate changes in activation patterns in the most addiction-relevant brain structures, and had to consume nicotine to generate the effect, I don't see the "major implications for addiction treatment" myself.
I do not see at all how the findings suggest "that it is not only possible to easily implant beliefs in a conditioned human brain, but one can fine-tune the beliefs to precisely produce the desired behavioural outcomes". There's no such thing as a 'conditioned human brain'--subjects here were addicted to nicotine, and the findings, such as they are, are limited to this population. In the sense that you mean, every brain is 'conditioned' by all kinds of things in the environment. There is no suggestion in this study that one can "precisely produce the desired behavioral outcomes" by "implant[ing] beliefs". We're talking about small changes in performance on tasks people voluntarily signed up to do. And the only sense in which beliefs were "implanted" is that people were told things and (probably) believed them. There is an ocean of literature in the vein of "changing task performance by lying to the participants," and I'm sure that loads of those studies, if conducted in fMRI, would turn up associated activation differences.
"Changing one’s beliefs (e.g. using cognitive behavioural therapy) would often mean completely rewiring one’s brain at the molecular level which is, as you can guess, not a trivial task." This, I'm sorry to say, is either far afield or vacuous. Of course changing your beliefs involves changing your brain--it would be shocking if it were otherwise. People who haven't integrated that all thoughts and behaviors ultimately flow from the brain will be surprised again and again when "neural correlate of x" is discovered (not what happened in this study). But as far as "completely rewiring one's brain at the molecular level," I have no idea what that would mean. No brains were 'rewired' in this study; regions were activated more or less at different timepoints. And there's a sense in which everything in the brain is happening at the molecular level (though we don't always operate at that level of analysis). But if you meant something like "changing beliefs can involve changing the strength of connections between different regions of the brain," that has both meaning and evidential support. If you're interested, there are some studies that look at the effects of CBT specifically, as seen in the fMRI scanner:
https://doi.org/10.1016/j.biopsych.2012.07.026
https://doi.org/10.1093/scan/nst009
https://doi.org/10.1176/appi.ajp.2013.12111484
It is still very exciting! Thanks for sharing it. And yes, the sample size is small and the statements from the paper might seem bold, but we all started somewhere. I hope this study will pave a way to larger studies of this kind. Can’t wait to see their results!