Recently, an important and landmark paper was published in PLoS ONE (hooray open access!) titled, "Investigating the Mechanisms of Hallucinogen-Induced Visions Using 3,4-Methylenedioxyamphetamine (MDA): A Randomized Controlled Trial in Humans". It sounds daunting, but trust me, it's a very cool, approachable study.
Now, in the spirit of full-disclosure, the lead author Dr. Matthew Baggott (hereafter referred to as "Matt"), is a friend of mine from grad school and he's been kind enough to grant me a very thorough interview for this post. The interview is quite long, so I'll give a brief overview of the research and some of Matt's comments, but I've posted the entire interview at the bottom of this post. And please, take some time to read the entire interview. Matt gives some fascinating historical context about drug research, mathematical models of visual hallucinations, the origin of the phrase "set and setting", and some personal insight into the difficulty of getting a study such as this approved by the DEA, FDA, and IRB!
At the bottom of this post you will also find an hour-long YouTube interview with Matt about his research as part of "Dr. Kiki's Science Hour".
That said, my real-life association with Matt is not biasing my opening statement; his paper is truly landmark for many reasons. This study is the first experiment in which humans were given MDA in over thirty years, and one of only a handful of peer-reviewed publications involving the administration of hallucinogens in as long. Regardless of my views (or yours) on the politics of drugs, the broad class of hallucinogens have incredibly interesting perceptual and cognitive effects (quite the understatement!) that merit neuroscientific research. Why? Because there are subtle and not-so-subtle cognitive and behavioral differences that can be fairly consistent within and across these drugs across people. We neuroscientists can leverage these differences to help us examine questions of perception, cognition, and behavior in a way that is otherwise impossible. As Matt says:
[T]here is a staggering diversity of hallucinogens out there: hundreds of compounds, each affecting different combinations of binding sites in the brain. Research in rats and anecdotal reports from humans tell us that these drugs have really different phenomenology. This is really obvious in the perceptual domain. For example, some hallucinogens cause profound visual distortions, making it almost impossible to see real objects during peak effects. Others have almost no visual effects, just the non-perceptual hallucinogen effects. The range of reported phenomenology is so broad, it starts to seem like we're discussing multiple pharmacological classes, which I think we are.
The range of drugs really is quite amazing, and many hallucinogenic compounds were first synthesized by a fellow Berkeleyan, Alexander Shulgin. Matt continues:
There is even one drug (diisopropyltryptamine) that is said by people to mainly just distort auditory perception, like some kind of flanger or nonlinear pitch shifter, with few other effects. The relationship between phenomenology of these drugs and their pharmacological profiles in the brain is really mysterious. And we're never going to figure it out by just studying one drug. Studying MDA is part of a long-term, audacious, and probably doomed-to-fail project of trying to relate phenomenology and cognitive neuroscience, on the one hand, to molecular-level pharmacology, on the other.
Matt used three tasks (such as the one illustrated below) to assess perception in his subjects during a placebo condition as compared to the drug-challenge. Overall, they found that:
Participants reported that MDA administration was followed by a number of prototypical hallucinogen effects… [that] support our approach of using MDA to understand hallucinogen-induced visions despite the caveat that MDA has complicated pharmacology and is relatively unstudied.
MDA had significant perceptual effects… Magnitude of closed-eye visuals after MDA was associated with lower performance on the two measures of perceptual organization. In both cases, there was also a significant interaction with dosing condition, suggesting that individuals who saw more intense closed-eye visuals both had poorer overall performance on these tasks and also had greater MDA-induced changes in perceptual performance.
As Matt's study demonstrated, these perceptual effects can tell us a bit about visual processing while simultaneously demonstrating, in a controlled, scientific setting, that these effects can be studied without any adverse medical effects. In fact, in this study, the participants were in a very controlled, scientific (read: not very stimulating) setting:
I'm quite certain that these studies are a lot less fun and a lot more tedious than taking hallucinogens with their friends or loved ones. Clinical research is often pretty boring, especially when the researchers are asking you to do strange computerized tasks that have no clear relevance to the psychedelic trip. We usually try to explain, as much as we can, study tasks so that they do not seem arbitrary and we try to make them as interesting as we can. But they're never as fun as real computer games. On the other hand, this was a fairly short study – just one long weekend. So it probably didn't get too boring. And, for many hallucinogen users, there is something interesting about getting a known dose of a pharmacologically pure drug.
Now realize, I don't make my statements about drug research lightly. I understand full-well the negative aspects of drug abuse and addiction. It is important to note that there are clinical differences between drug "use", "abuse", "dependence", and "addiction", as compared to casual/recreational use. In my past neuroscientific life, I worked as a PET scanner operator at UCLA for two years with methamphetamine abusers and published several publications on brain changes associated with methamphetamine poly-drug interactions and recovery from chronic methamphetamine abuse. These studies are what initially got me interested in the role of neuroplasticity in cognition.
Overall, this is a great study and really an important step for future human neuroscientific research with cognitively-active compounds. While there are certainly limitations to this study (about which the authors are quite forthcoming), they don't detract from the overall importance of the paper.
The Griffiths study gained a lot of attention and really helped reignite the scientific study of hallucinogens. It strikes me that the Griffiths et al. experiments are purely behavioral, whereas yours add a cognitive/perceptual component. How do you see your study as expanding on Roland's work and why do you consider this an important next step in human drug research?
I think you hit the nail on the head. We take an approach that uses cognitive neuroscience and psychophysics to investigate brain mechanisms. Another obvious, but I think important, difference is that we are looking at a different drug. Almost all legitimate human hallucinogen research uses psilocybin. Yet there is a staggering diversity of hallucinogens out there: Hundreds of compounds, each affecting different combinations of binding sites in the brain. Research in rats and anecdotal reports from humans tell us that these drugs have really different phenomenology. This is really obvious in the perceptual domain. For example, some hallucinogens cause profound visual distortions, making it almost impossible to see real objects during peak effects. Others have almost no visual effects, just the non-perceptual hallucinogen effects.
The range of reported phenomenology is so broad, it starts to seem like we're discussing multiple pharmacological classes, which I think we are. There is even one drug (diisopropyltryptamine) that is said by people to mainly just distort auditory perception, like some kind of flanger or nonlinear pitch shifter, with few other effects. The relationship between phenomenology of these drugs and their pharmacological profiles in the brain is really mysterious. And we're never going to figure it out by just studying one drug. Studying MDA is part of a long-term, audacious, and probably doomed-to-fail project of trying to relate phenomenology and cognitive neuroscience, on the one hand, to molecular-level pharmacology, on the other. MDA seems to be a drug with similarities to classical serotonergic hallucinogens (like psilocybin and LSD). But it also has many similarities to its chemical cousin, MDMA ("ecstasy"), which many people argue is not really a hallucinogen at all.
Unlike classical hallucinogens, MDMA has very consistent emotional effects, which many people describe as feelings of empathy and emotional closeness to others. In our study, MDA seemed to share some of these consistent emotional qualities. Dave Nichols has suggested the term "entactogen" (which basically means "touchy-feely", or "generating touching within" if you want to be precise about it) for drugs with these unusual emotional effects. We and others have evidence that these so-called entactogen effects are related to serotonin release. I take a state-space or dimensional approach to understanding these drugs. Instead of seeing them in discrete categories, I think it is helpful to consider them as having different amounts of underlying characteristics, although it is not yet clear how many of these characteristics or dimensions there are.
Most classical hallucinogens share a tendency to cause altered sense of self — what the psychiatrists call "depersonalization" and mystics call "ego loss" — and the feeling of changed reality. And, as I said earlier, there is a perceptual effects dimension along which different hallucinogens very. How many more dimensions of effects we'll find is anyone's guess. Personally, I suspect that much of the remaining variance will be captured in a small number of dimensions, many of them having parallels in the literature on affect and personality. For example, the MDMA-like dimension may end up relating to extraversion and decreased harm sensitivity.
So one way that we are building on Roland's work is by characterizing another hallucinogen and seeing that it is also shares some potential for making mystical-type experiences more likely. In a way this isn't surprising. After all, religions like the Native American Church and the Santo Daime regard certain hallucinogen-containing substances as sacraments. Yet, I'm not certain we as a society have sufficiently grappled with the implications of this. Some of Roland's colleagues have suggested that this makes it possible to conduct rigorous empirical research on the psychology of religious experience. Instead of just looking at statistical correlations, one can prospectively assign people to pharmacological conditions in which mystical-type experiences are likely. The possibility that your study might cause someone to find religion or change their life direction is a sobering thought. I know that this is an area that they have thought through very carefully over at Hopkins.
One of the reasons that my studies only include experienced hallucinogen users is that any big hallucinogen-related insights or life changes have probably already occurred. Thus, being in the study is not likely to cause a participant to change their name to Suzy Creamcheese, go on tour with a jam band, invent PCR, or write "One Flew over the Cuckoo's Nest".
Other than Roland Griffiths, what other researchers are currently and actively studying the behavioral effects of hallucinogens?
To my knowledge, most of the human studies with hallucinogens focuses on whether hallucinogens can help specific patient groups. For example, there are two groups in New York looking at whether psilocybin (the psychoactive chemical in psychedelic shrooms) can help people with cancer feel less distress. Someone in Switzerland is doing a small trial of LSD psychotherapy, sponsored by MAPS.
There are fewer examples of groups doing what we do, that is, using hallucinogens as scientific tools rather than potential medicines. Franz Vollenweider's group in Switzerland, with funding from the Heffter Research Institute, has been very productive in this area for many years. Right now Robin Carhart-Harris is lead researcher on an interesting psilocybin fMRI study in England, sponsored by the Beckley Foundation. (In this answer, I am not considering MDMA as a hallucinogen. If we included it, there would be a couple more groups in both categories, including some NIH-funded projects.)
As you can see, aside from some research with MDMA, human studies with hallucinogens are mostly privately funded by groups like the Beckley Foundation, MAPS, and the Heffter Research Institute. These are, in turn, supported by small foundations and, as they say, "viewers like you." Much of the Renaissance in human hallucinogen research in the late 90s and early 2000s can be traced back to a small number of people, such as the now deceased Bob Wallace, an early Microsoft employee. The dominance of private funding in this field is interesting. It probably allows greater innovation, but it also may subtly influence scientists to study accessible, easily publicized topics and run smaller (cheaper, but often statistically underpowered) studies. By the way, as I note in the paper, I received some Beckley Foundation funding that was very important for helping me to develop study measures.
You mention it briefly in your paper, but can you give a lay-introduction to the Ermentrout and Cowan model of hallucinations?
Bard Ermentrout worked with Jack Cowan in the late 1970s to develop a formal model of the simple pattern visuals that often occur shortly after hallucinogenic drug administration. Their research elegantly connected some observations first made by German neurologist Heinrich Klüver to a field of research pioneered by British genius Alan Turing. Klüver is now famous for co-discovering Klüver-Bucy syndrome in monkeys, but he came upon that syndrome while trying to understand the effects of the hallucinogen mescaline. As part of his research, he took mescaline many times himself, enough times to start to notice commonalities in how the visuals progressed over a trip. He organized the visuals into categories he called "form constants" and wrote a lovely little book about it.
Over the years, this work was more or less forgotten (when Ronald Siegel at UCLA started categorizing the perceptual effects of hallucinogens, he initially didn't know about Klüver and independently replicated many of Klüver's categories). Ermentrout and Cowan realized these patterns could be related to work on something called diffusion-reaction systems, which had been pioneered by Alan Turing. After he made important breakthroughs in computer science and helped to win World War II (and shortly before he was persecuted to death for being homosexual), Turing got interested in explaining the complex mathematical patterns that we see in nature.
He figured out that you can get spontaneous formation of elaborate patterns under certain conditions of diffusing excitatory and inhibitory processes. For example, the patterns on many different animals' coats can be explained if you hypothesize that there are two chemicals diffusing through the animal as it develops: an excitatory chemical that makes the skin or fur darken and an inhibitory chemical that prevents this. If the excitatory chemical causes cells to make more of both chemicals and the inhibitory chemical causes cells to make less of both chemicals, then — with some model parameters — you can get elaborate patterns like zebra stripes and leopard spots that seem to appear out of nowhere.
A really cool thing about this is that the pattern complexity in these models depends on how much space the pattern has to form in. And you can really see this by looking at animals. It is only the medium large animals that have neat patterns. Really small or really large animals don't have patterned surfaces at all. You don't see tiger-striped mice or whales. Kind of small animals, like hamsters, at best have pathetic wannabes patterns, like large irregular blotches. And if you look at an animal that has nice patterns on its body, the patterns often break down in the limbs where there is less space for pattern formation.
Ermentrout and Cowan realized that a similar process in the brain could explain hallucinogenic visuals. Only instead of excitatory and inhibitory chemicals diffusing, you have excitatory and inhibitory neurons firing and making neighboring neurons more or less likely fire. They proved that, if you take a simple idealized sheet of cortex, you can get spontaneous formation of stripes of alternating firing and silent neurons. Then they asked what these stripes would look like if the pattern formed in one's visual cortex. The answer depends on the angle of the stripes on the cortex. This is because the large "primary visual cortex," where visual information first enters the cortex, gets a lot more detail from the center of your vision than it does from the edges, which means you want to use more of the cortex to process the center of the vision. The result is a funny (roughly log-polar) transformation between images hitting the retina and how they get represented in primary visual cortex.
Ermentrout and Cowan reversed this transformation to calculate how stripes would appear if they were "projected" into visual space. It turns out that, depending on the stripe angle, you see spirals and tunnels and other things that are pleasingly similar to form constants described by Klüver! This type of modeling is now a substantial area of research. In fact, Ermentrout just co-authored a nice textbook with David Terman on the mathematical foundations of neuroscience.
These models lead to a spontaneous pattern formation theory of hallucinogen-induced visuals. One implicit aspect of this theory that I haven't seen clearly discussed in the literature is that the approach may be able to explain the increasing complexity of visuals that occurs over the course of a single hallucinogen experience. Just as tiger stripes form most readily on the large body of an animal and not on its limbs, you would expect stripes of neural excitement to also form most readily in the large primary visual cortex and not in the smaller visual areas. Thus, the first spontaneous patterns to form during the hallucinogen trip would be in primary visual cortex and would represent the sorts of computations done in that brain region: detailed line segments. This would explain why Klüver's form constants happen relatively early in a trip. Once you get stable patterns there, the perceived patterns may be passed on to hierarchically higher and physically smaller visual areas that make more advanced interpretations of the visual scene. That might be when you start to see objects embedded in the abstract visuals. Ultimately, the pattern might spread to areas representing scenes, at which point the individual would hypothetically experience detailed immersive virtual reality experiences, as some people report after high doses of hallucinogens.
It is a beautiful theory. The thing is, there is almost no empirical evidence supporting it. A critic might say that just about anything projected from cortical to visual space will look trippy. One of the only pieces of evidence I can think of is that this pattern formation theory can also explain the existence of flicker-induced visuals and we know that visual flicker potentiates the visual effects of low dose hallucinogens, suggesting a possible common mechanism.
So one thing that we tried to do in this MDA study was to make a very modest step in the direction of testing this theory. We took a visual illusion that is thought to involve primary visual cortex and tried to see if the magnitude of the illusion was altered in people when they were on MDA. Unfortunately, we didn't get conclusive evidence one way or the other. However, I'm really proud of the attempt to find links between experience and formal theory, even if we failed to prove or disprove anything this time.
We’ve talked about it informally, but can you say anything "on record" about the difficulty in getting all the required approvals to run this experiment?
Studies where you give controlled substances to humans are closely regulated by a whole bunch of federal, state, and local organizations. I'm not against this: We definitely should be cautious about research that can change people's lives. There is a lot of potential for things to go wrong. One unfortunate side effect of having so many agencies involved, however, is that it is difficult to estimate how long it will take to navigate the regulatory maze. You need to get permission from the FDA, DEA, IRB, state of California, and probably your department chair. A one-month delay from one agency can domino into a six-month delay in starting the project and if anyone says no, the project is finished. This makes initiating hallucinogen-related projects high risk for young scientists who need to publish. You might spend a year killing trees and never have anything flower.
Tell me a little about the set and setting the subjects were in.
The physical setting was nothing particularly fancy. It was the General Clinical Research Center at San Francisco General Hospital. I sometimes wish that we had the budget to deck out a chill space with a sweet sound system and artwork like they did at Johns Hopkins, but I suspect the people are more important than the decor anyway. When I first started conducting studies with MDMA, I really worked to make the experimental rooms as nice as I could. I eventually learned that people on MDMA will be pretty content in any room.
Interestingly, I don't think anyone has ever actually shown that mental set and physical/social setting are more important for predicting the effects of hallucinogens than they are for other drugs. In general, correlations between pre-dose psychological measures and peaked effects have been pretty unimpressive. Of course, this might just be because most studies do a pretty good job of being nice to their participants.
Or maybe "set and setting" are not that important. The phrase has an interesting history. There was a great deal of interest in early psychopharmacology in explaining the variability in drug effects, both hallucinogenic and non-hallucinogenic. Psychopharmacologists like Hans Eysenck and Max Rinkel looked at "non-specific influences" on drug response and many researchers used the individual words, set and setting. But the pairing of those two words as a phrase was popularized by Timothy Leary much later, at a time when opinion about hallucinogens was very polarized.
In a way, it seems to me that the phrase partly served to shift blame for "bad trips" or negative clinical trials from the drugs and onto the drug users and researchers. It was sometimes used as a way of saying that, if you didn't achieve mystical enlightenment, you weren't "doing it right." But the truth is, I don't think anyone can reliably predict pleasant versus unpleasant hallucinogen experiences. That being said, I think it is useful to consciously remind ourselves at the beginning of experimental sessions that the purpose of the session is scientific: that our goal is to bring back knowledge that we hope will be useful to society. I don't know if this "establishing intent" actually has any effects on the participant or improves the quality of data we collect, but I suspect it does.
These were all experienced drug users. How was the experience for them in such a controlled, three-day, medical, scientific setting?
I'm really grateful to our heroic participants, who loaned their consciousness to science for the duration of the study. I'm quite certain that these studies are a lot less fun and a lot more tedious than taking hallucinogens with their friends or loved ones. Clinical research is often pretty boring, especially when the researchers are asking you to do strange computerized tasks that have no clear relevance to the psychedelic trip. We usually try to explain, as much as we can, study tasks so that they do not seem arbitrary and we try to make them as interesting as we can. But they're never as fun as real computer games. On the other hand, this was a fairly short study – just one long weekend. So it probably didn't get too boring. And, for many hallucinogen users, there is something interesting about getting a known dose of a pharmacologically pure drug.
We are now working on analyzing the blood samples we collected and comparing our MDA results to our past studies with MDMA. We are also reaching a good publishing point in a project of analyzing the language people use to describe hallucinogen experiences. More generally, I remain very interested in how MDMA alters emotion and what this can tell us about the function of serotonin.
Anything important you’d like to add?
While studies in a controlled clinical setting with carefully screened and prepared volunteers are low risk, MDMA and related drugs have real potential for toxicity. People sometimes have idiosyncratic reactions for reasons no one understands. Women seem to be particularly susceptible to developing hyponatremia (low salt in the blood), which can be fatal, after MDMA. We are currently studying this issue. Right now, no one knows if this only happens to people who drink a whole lot of water. One possibility is that MDMA may cause release of antidiuretic hormone and make it harder for your body maintain the right water balance even when you drink normal amounts of water. Some are suggesting that sports drinks might be the safest thing to drink on MDMA. This is an important area that needs more research.
Like much of science, hallucinogen research is an area where you can only succeed by collaboration. The research participants, the visionaries funding this area, the nurses and technicians, and very many formal and informal scientific advisors were all crucial to the study.
Baggott, M., Siegrist, J., Galloway, G., Robertson, L., Coyle, J., & Mendelson, J. (2010). Investigating the Mechanisms of Hallucinogen-Induced Visions Using 3,4-Methylenedioxyamphetamine (MDA): A Randomized Controlled Trial in Humans PLoS ONE, 5 (12) DOI: 10.1371/journal.pone.0014074
Griffiths RR, Richards WA, McCann U, & Jesse R (2006). Psilocybin can occasion mystical-type experiences having substantial and sustained personal meaning and spiritual significance. Psychopharmacology, 187 (3) PMID: 16826400
Voytek B, Berman SM, Hassid BD, Simon SL, Mandelkern MA, Brody AL, Monterosso J, Ling W, & London ED (2005). Differences in regional brain metabolism associated with marijuana abuse in methamphetamine abusers. Synapse (New York, N.Y.), 57 (2), 113-5 PMID: 15906384
Berman SM, Voytek B, Mandelkern MA, Hassid BD, Isaacson A, Monterosso J, Miotto K, Ling W, & London ED (2008). Changes in cerebral glucose metabolism during early abstinence from chronic methamphetamine abuse. Molecular psychiatry, 13 (9), 897-908 PMID: 17938635