My current blog is epistem.ink. This one is here just for archival purposes.
Stress and Serotonin
Trigger warning: bad pop neuroscience
i - Briefly
I recently read "Serotonin and brain function: a tale of two receptors" and found it elucidating.
Why are "typical" serotonergic psychedelics (e.g. LSD, psilocybin), MDMA/MDA, and SSRIs (or even MAOIs) so different in terms of effect.
Different amphetamines, cocaine, modafinil and ephedrine are quite different, but they include a lot of unifying experiences, they could almost be graded on a few scales in terms of their "main" effects.
The same goes for opioids, or even for broader and fuzzier classes of drugs such as dissociatives.
I'd expect the difference between MAOIs and LSD to be closer to the difference between alcohol and benzos, rather than the difference between smoking a cigar and receiving a prophecy from God.
This is a paper that seemed to kind of answer this question, and it does it so fucking elegantly I to marry it. To top it all off you can basically fit the whole thing into predictive coding theory without even trying
The tl;dr is that psychedelics are principally agonists for 5HT2A receptors (dense in cortical areas, especially "newer" cortical areas, especially DMN associated areas). While SSRIs (and afaict MAOIs, though the paper doesn't say it) are agonists for 5HT1A receptors (mainly present in the limbic system, heavy association with emotions and impulses). Both of these receptors are involved in dealing with stress, but the 1A receptors "deal" with stress by emotional downregulation, while 2A receptors "deal" with stress by making you fitter to act in order to change the stressful stimuli in the environment.
ii - Mechanisms
I should mention I don't quite understand the mechanisms of action of either "typical" serotonergic psychedelics or SSRIs (nor MDMA or MAOIs). This paper seems to go with "psychedelics bind preferably to 5HT2AR and SSRIs inhibit reuptake preferably for synapses involving a lot of 5HT1AR".
My understanding was that various psychedelics are actually indirect in their effect (i.e. they bind to the 5HTR in such a way as to not activate it fully, the brain produces a lot more 5HT due to this, now the psychedelic unbinds and those areas are swimming in 5HT).
Also SSRI effects and how much the reuptake inhibition helps directly (increased 5HT levels) vs indirectly (stimulating 5HT production at the sites with inhibited reuptake and/or causes other downstream effects ala BDNF release) seems like a hotly debated topic that's especially confusing given that it takes weeks for them to take effect.
Finally, the model I have for MAOIs is that they basically should just lead to a global 5HT increase due to more substrate being available. Which should make them similar to MDMA (agonist for both 1A and 2A)... which I guess if you squint really hard it kind of is (?).
But I don't think the mechanism is relevant there, and one should keep in mind all of these drugs affect (to some extent) most or all of the 14+ serotonin receptor types. What does seem to be corroborated by a lot of evidence, including brain biopsies in rats&mice and fMRI in human is:
- Psychedelics are agonists for 5HT2AR and/or trigger more activity in 5HT2AR rich areas (more so than for all other receptor types, and importantly, much more than for 5HT1AR)
- SSRIs are agonists for 5HT1R (more so than for all other receptor types, and importantly, much more than for 5HT2AR)
- Both 2A and 1A receptors are involved in anxiety and responding to external stress.
- Both psychedelic and SSRIs tend to "treat" depression, especially the kind linked with anxiety.
Note: MAOIs... might be more agonistic for 5HT1AR, at least in small quantities, but large quantities will kill you. This goes into a whole other theory at the end of the paper about 5HT1AR and 5HT2AR basically being gradations on a scale of stress... and I'm kind of annoyed the authors don't mention MAOIs. MDMA is weird and acts on both, but for the sake of brevity, I won't cover it here.
Again, 5HT2AR activations are the cause of a cascade that leads to actions that help with changing the stressful environment, 5HT1AR are the cause of a cascade that helps in adapting to not being so stressed in a new/crappy/unexpected environment.
The long term effects of these can be oddly similar from a psychiatric well-being perspective, but the felt experience of this, as well as the resulting behaviour, is widely different.
iii - Dealing with stress
Say that you are a primordial human dealing with the primordial problem of, ahem, your bedroom window being smashed (or something to do with a sabertooth tiger, whatever).
You were about to go to sleep and this introduces an external stressor in the form of cold wind coming into your bedroom. You are internally stressed not because the event itself is so bad, but because you didn't predict it. Your brain had this "prediction" of "at around 10:30 PM I will go to sleep in my cosy bed", or even a more general prediction of what constitutes "your bed" and "your bedroom" which included the lack of felt sensations of coldness.
There are 2 ways of reducing stress (i.e. minimizing prediction error), one is to learn to predict your bedroom as windy and cold (accepting) the other to put together a plan to make your bedroom warm, e.g fix the window, call someone to fix the window, sleep in a friend's spare bedroom, add more blankets, a combination of these (addressing).
These are two sides of the same coin, you need some amount of "not freaking out about the window being broken" in order to concoct and enact a plan to fix it. Too little 1A activation combined with a lot of 2A activation and you'll be an extremely creative anxious mess (see bad trips). You don't want 1A activation without proportional 2A activation because then you will just decide it's fine if the window is never fixed and keep on sleeping like this (see depressed people).
Both are a mechanism for dealing with stress at a fairly "long term" rational and emotional level. This is not the cortisol release that makes you act milliseconds after the stimuli has hit the first layer in the audio/visual/gusta/whatever cortex. This is not the adrenaline pumping through your blood seconds after the stimuli. This is not the pathway that makes you "reflexively" move your limbs a certain way. These are ways to deal with "chronic" stress that affects you over hours, days, months or years.
The "just-so story" here is that psychedelics are the "make the world better" version of antidepressants. Which instead of letting yourself rot happily, will make you actively go out and do something about the problem.
We all agree that living with your parents, languishing in bed all day, and occasionally going to work at Walmart is not "ok". Antidepressants "cure" this by making you feel ok, and by adjusting your behaviour slightly as a result... so maybe now you go to work every day, and also your workout a bit, and in 2 years you get promoted to middle manager at Walmart or whatever. Psychedelics "cure" this by making you quit your job, go on a long trip panhandling around southeast Asia, find spiritual fulfilment and return to chair a successful save-the-children/whales/animal-cruelty/something-something-vegan-organic NGO.
Anything that's H1A agnostic in favour of H2A will make you more prone to feeling ok, while the reverse will make you more prone to address the issue. These get confounded over time since feeling better often has cascading effects that fix the underlying issue, and fixing the issue often has cascading effects that lead to feeling better.
iv - Adapting
The problem arises when you consider that both of these receptors don't guarantee a solution, they just make you more prone to find it.
I think for most people it's pretty obvious why H1A activity can be bad. But why can H2A activity harm? I point back to the diagram above:
- Sensitivity to stimuli
- Ease of learning
These in of themselves are not things that solve problems, they are things that put you in a problem-solving mindset, looking at the world with fresh eyes, openness, and renewed energy.
This actually seems to reinforce an old bias of mine that people take psychedelics for the wrong reasons, i.e. they take psychedelics to solve "existential" problems that have no real solution other than some form of "accepting", when instead psychedelics are particularly good for solving a hands-on problem like studying the behaviour of ants, figuring out how your mind works or learning to play an instrument.
I found it curious that people who take psychedelics for introspection usually end up as religious cranks or burnouts. While people that take psychedelics because "they are fun" don't seem to experience many negative side effects. Under this framework, it makes perfect sense, activate highly conceptual functions in the cortex and tell them to solve an intrinsic and emotional, you'll end up with some very weird conceptual scaffolding.
v - Other interesting notes
When discussing this paper I stumbled upon an interesting note. Which is that "happiness" is generally associated with emotional downregulation. Being less anxious, impulsive, and aggressive is often almost analogous to being contempt, happy even. Basing this on felt experience, on the effects of the drugs used for a lot of "sadness associated" conditions and on the fact that there are giant religious traditions built around escaping from (impulsive) desires.
Downregulate to an extreme and you won't be happy, you'll just be "meh" all the time. Indeed, SSRIs only work on a subset of depression, in general, the kind that's associated with anxiety and anger (think 4chan user stereotype).
I think this speaks to happiness as commonly spoken of as being a higher level cortical function that takes emotional and impulsive inputs, and can get overwhelmed by them, but is not entirely dependant on them. So, one can take a strong impulse to eat and turn it into "hunger" or into "appetite", the difference between those two is in part one of impulse strength and in part one of interpretation.
Another curious thing is how this makes creativity and stress so tightly associated. I always had this association and I thought it was in part a "bad" thing that I should unlearn. But at the same time, I find it very hard to be creative when I'm not a bit stressed. I think this speaks volume to that being a much more fundamental trait of the mind than I thought. It might be wiser to instead just mitigate the downstream effects of stress better.
Just in case it's not obvious, these are very much sky in the pie interpretations, and they helped my own mental models, but I doubt that the authors would endorse many of them.
Before reading this paper I skimmed this work about the function of all serotonin receptors. If you plan to head into the paper it might be useful as a primer. I suggest you just go ahead and read the paper, I expect you will find a lot of other interesting takeaways from the proposed model.
See this reply if you're curious about reading a smarter perspective on the topic.
See this comment chain if you think the mechanisms I propose for the drugs listed are inaccurate and want some clarification on my phrasing.
Published on: 2021-06-14