Source
In a nutshell
Using statistical learning techniques, scientists have correlated distinct types of psychedelic experiences to specific receptor subtypes and their distribution in the brain. This is a major step forward in understanding what neurochemical systems in the brain bring about distinct types of conscious experiences and how they are affected by psychedelics.
In depth
A team of scientists at SUNY, MIT, and Mcgill universities went over thousands of “trip reports” (collected from Erowid.com) linked to 27 distinct psychedelic compounds (such as LSD, psilocybin, Ketamine, and more) and used computational tools to analyze the subjective content of these experiences. In other words, they wanted to see if specific words and phrases kept popping up in relation to certain substances. They then looked at which drug type the report was about, and used pre-existing databases that list what receptors these drugs bind to, in order to try and sort out whether certain experiences are linked with specific receptor activities. Lastly, they mapped out known brain areas that express the highest density of these receptors, since it stands to reason that if a specific substance binds to a certain receptor type more than others, that substance would have the highest effect on those brain areas that express that receptor to the greatest degree.
Let’s try to get a clearer intuition about how this works. So, imagine that the trip reports are broken into their constituent words and phrases, and then those words and phrases which show up several times are all listed one after another, in a specific order. What order? Well, according to how likelier they are to appear in a report about a substance that is linked to receptor type X than a report about a substance that is linked to receptor type Y. So for example, take the relatively new psychedelic compound DipT (N,N-diisopropyltryptamine), it is known to bind the serotonin receptor 2A subtype most strongly, but with significant interaction also with the serotonin transporter (SERT) among others (source). Turns out that trip reports with this compound were quite likely to contain words such as “visuals”, “nausea”, and “sleep”, which the researchers group together as sensory phenomena. However these same descriptors were far less likely to occur in reports about salvinorin A, for example, a substance that binds the Kappa opioid receptor, but not the serotonin receptor (source). Indeed, drugs like Salvinorin A were more associated with words such as “universe”, “world”, “consciousness” etc. which the researchers classified as “mind expanding” descriptions, possibly associated with the mystical experience side of psychedelics.
Using this method they found several “factors” (a statistical term that basically means- which variables change together and which don’t) that together explain a large proportion of the observed data. The strongest factor was the axis mentioned above, where one pole (one end of the word list) represented sensory experiences and was linked to drugs that activate the serotonin 2A receptor subtype, but also other serotonin receptors (1A, 2C and 2B) as well as adrenergic receptors (receptors for the neurotransmitter adrenaline), and the dopamine D2 receptor, while the opposite pole represented mind expanding or mystical descriptions and was linked to substances that bind to the dopamine D1 receptor, Kappa-opioid receptor, and serotonin 5A receptor. The next strongest factor had specifically auditory effects (“auditory”, “pitch”, “sounds”, etc.) on one pole, and was again very highly correlated with the substance DipT, with the opposite pole composed of social-emotional effects (“friends”, “love”, “depression”, etc.) and was most strongly correlated with MDMA.
The researchers then went one step further, and looked at how these different receptors were spread out in the brain, asking what regions should then be most strongly affected. Turns out that their results agree with existing literature about the functions of various brain regions. So for example, for the first factor (the one whose poles were “sensory” on the one end and “mystical” on the other), they found that the areas which were most likely to be affected on the sensory sides were indeed “low level” sensory processing areas, meaning those areas in the brain that are directly connected to our sensory organs and are the “first stop” in processing incoming sensory signals, while the areas associated with the second pole were very high level areas (so-called “association cortices”). This finding is especially significant as the anatomical areas were not a part of the original statistical method, but rather came out of their own accord, so to speak, when they tried to map the receptor densities to anatomical brain maps.
All in all, the researchers found 8 such factors that passed the threshold for statistical significance. When looking at the extreme poles of the experiences associated with each factor, we get: (1) sensory <-> mystical ; (2) emotional <-> auditory ; (3) visual <-> emotional ; (4) body sensations <-> therapeutic outcomes ; (5) discomfort <-> environmental descriptions ; (6) euphoria <-> dysphoria (negative feelings) ; (7) physical location or context <-> therapeutic outcomes ; (8) bodily relaxation <-> emesis (vomiting, purging, and otherwise relieving oneself). It should be noted that this is a very brief summary, and the factors are significantly more complex than expressed here. As one example, most experience types were also correlated with specific time words so for example, “seconds” and “minutes”, would appear in other places than “days” or “years”, indicating that the experienced time-horizon was a strong part of the effect. Another interesting aspect of note is that while some factors are organized in a way that matches our intuitions (see factor 6- it stands to reason that euphoria and dysphoria are at opposite ends of the spectrum), other factors are organized in a manner that we would perhaps not expect at a first guess (see factor 1- why should sensory descriptions be missing from reports with mystical experiences). Such results could, over time, offer compelling new insights to understanding how our conscious experience itself is constructed from seemingly unrelated lower level processes.
Summary &
Caveats
Naturally, this study also has some limitations, which should be stated and taken into account. First, the results are purely correlational, and there are no direct causal experiments. Second, the experiential dataset was a collection of anonymous trip reports, obviously this is not a well controlled source of information, and serious confounds could exist, from the obvious fact that this is not a blinded report (meaning the report writer knows what substance they are taking and has expectations about it), to other issues such as mislabelling of the compound (the person writing the reports usually has no way of knowing if what was sold to them as MDMA is in fact that, for example), to the fact that there is likely a strong selection bias in terms of who writes and submits trip reports to Erowid (likely very experienced psychonauts with well-defined expectations and beliefs about psychedelics), which may not faithfully represent experiences in the wider population. However all in all, this work goes a long way towards providing a more graded and nuanced understanding of how psychoactive compounds bring about their specific effects, and does so by looking at the brain in a holistic manner, focusing on large sample sizes to see recurring themes, and tying these to large scale networks in the brain. Furthermore, this type of research demonstrates how the detailed study of psychedelics, with their unique effects on conscious experience, can aid us even beyond the therapeutic sense, and help explain some of the deepest mysteries in science, such as understanding the very nature of how our consciousness is generated by the brain.
