The Data behind Altered States of Consciousness: Interview with Timo Torsten Schmidt
When we want to grasp the subjective experiences of others, we count on their verbal accounts. Limited to language, quantifying how something feels for an individual is very hard. It is even more arduous for altered states of consciousness where descriptions become somewhat metaphorical. On our quest to quantify “What is it like to be in an altered mind state?”, we spoke with Timo Schmidt about his Altered States Database project.
idalab: What is the Altered States Database and what does its objective of mapping subjective experience to brain functions imply?
Timo: The Altered States Database (ASDB) is a collection of data on the subjective experiences induced by different substances, such as LSD or psilocybin mushrooms, and non-pharmacological methods, like meditation or sensory deprivation. We did not generate that data ourselves, but rather extracted it from peer-reviewed scientific publications. Multiple attempts have been made in this field of research to quantify subjective experiences – mainly with questionnaires – applying various methods to alter conscious states. While several approaches induce different types of states, those experiences seem to overlap at their core. In other words, there is a restricted space of possible altered states of consciousness, assumably limited by the architecture of our brains. How altered states map to brain functions is a fascinating question in neuroscience.
Clinical relevance of altered states
idalab: What are the requirements and clinical applications of brain mapping for pathologies involving altered states of consciousness?
Timo: Altered states occurring in different pathologies are complicated to study because patients have generally suffered from their conditions for a long time. As a result, the brain undergoes many transformations, for example in depression and schizophrenia. Since every patient is different, it is hard to directly compare patients to a healthy control group. Nonetheless, altered states of consciousness caused by pathologies and those induced by drugs or non-pharmacological methods exhibit common features, such as visual hallucinations and auditory phenomena. The idea is thus to alter the consciousness of healthy subjects in an experimental setting for a short amount of time – one or two hours – allowing researchers to study normal and altered functioning in the same brain. Insights generated by mapping specific aspects of altered states of consciousness to brain function in a controlled environment ultimately benefit our understanding of pathologies.
idalab: When and why did you become interested in altered states of consciousness and how did this inquisitiveness materialise in the ASDB?
Timo: My motivation is rooted in the clinic. Knowing of the difficulty to quantify subjective experiences in patients, I became interested in the use of substances for short-term induction of altered states. For instance, when LSD was developed and first delivered to psychiatrists, the information leaflet stated that the drug could be used by the therapist himself to temporarily induce a psychotic-like state. That way, he could experience from a first-person perspective what his patients might feel and then use that knowledge to better treat them. This was coined the “model psychosis” approach.
More generally, the literature dealing with the measurement of subjective experiences is very diverse. As a reader, it is difficult to compare reports across substances, such as LSD, ketamine or DMT. When you try to figure out their commonalities, you are typically left with studies involving only a handful of participants where experiences were assessed with several questionnaires you might not know about. After reading the work of Adolf Dittrich, it became clear to me that consolidating the data scattered across scientific articles would be really meaningful. Executing this project was made possible thanks to the great work of Master’s students from the Osnabrück University.
Data extraction process
idalab: The ASDB presents highly structured data extracted from peer-reviewed publications. However, as you said, psychometric data is not reported in a standard way in the literature. How did you design the methodology used to gather and organise that unstructured text data? What were the main challenges?
Timo: The first step was conceptual and consisted in making a decision on the type of questionnaires to include. Some questionnaires are very specific to psychedelic experiences, whereas others are broader and cover effects elucidated by psychedelics, alcohol, and meditation. We were interested in this overall comparison and selected questionnaires that span the entire spectrum of phenomena. Besides, we also focused on those that contain enough data to compare across subjects, studies, and induction methods.
We then had to analyse how the data is represented in the literature in order to design the database. In general, scientists do not publish their raw data on individuals, but rather summary statistics, such as mean, standard deviation, and number of participants. For instance, participants answer 100 items that can be grouped in 5 to 10 categories (or scales) for which the group statistics are given. Because our goal is to be perfectly transparent, we provide the reference of each original article, by linking to PubMed where you can download the publication. This also accounts for possible human errors on our side, as it allows database users to signal aspects that need correction.
idalab: The manual data extraction conducted by your team requires domain knowledge as well as much time and effort. Could there be a way to automate that process?
Timo: At the moment, it is not possible to reliably automate data extraction from articles, as it is not prepared in a digital format. The information is rather lost in text or hidden in bar diagrams. To digitize that data, we used standard tools for graphical processing, such as WebPlotDigitizer, which is a little more handy than using a ruler to assess graphs on a printout. All in all, forty students participated in that effort in the context of seminars, lab rotations or internships. That process was completed at the end of last year. The data up to 2017 is freely available online and can be downloaded from my Open Science Framework (OSF). Thanks to the ASDB, researchers can now perform meta-analyses on dose-response relationship for specific substances. It is, however, still not an option for all substances and the database also highlights the demand for further studies.
Target audience & vision for the project
idalab: You created the database not only for scientists, but also for the general public. How should non-scientists approach the database and what role does the WikiProject play in that context?
Timo: This is a big challenge. We started this project as scientists, meaning that we can interpret bar diagrams and understand measures of variability. Therefore, our idea was and still is to explain mean and standard deviations on the website so that non-scientists can make sense of what they see. The best way to present the data to the general public is not evident, but we believe that a basic understanding of the methodology is required. We clearly do not want people to use the data to make predictions about their potential drug experiences. The database is not suited for this because there are too few data points. Moreover, the data mainly stems from studies where people were administered substances in a laboratory, which is a quiet, controlled environment. Experiences can be radically different if, for example, LSD is taken on a festival where you can be overwhelmed by diverse sensory stimulations.
Our WikiProject aims at improving the quality of Wikipedia articles on commonly used drugs, such as psychedelics, as we noticed that those articles are not very detailed on the type of experiences people have. When people do their research, they land on websites that feature first-hand reports, such as PsychonautWiki or Erowid. However, this sample is biased, since reviews are most likely posted by people who had extreme experiences, either positive or negative. There, you see both ends of the spectrum while, in scientific data, you rather get an average across subjects, which is a standardised and more neutral information.
After discussing that topic with philosophers and social scientists, we still believe that open science and open data is the way to go. Despite the difficulty of doing good science communication. We want to present the data as it is and not undergo any ideology-driven interpretation. It is not fully possible to prevent people from reading the data selectively or from misinterpreting it, but we do our best to provide enough information to avoid that. We are continuously looking to improve and are welcoming suggestions.
idalab: What is your long term vision for the ASDB?
Timo: As an advocate of open science, I am trying to promote the idea of making raw data backing peer-reviewed articles publically available in their appendix or in a data repository. Having individual subject data on top of the published summary statistics would greatly increase the power of meta-analyses and help generate new insights. Ideally, the community would use a standard format to publish data on the ASDB. Increasing the visibility of their own studies might incentivize scientists to come out of their ivory tower and stop hiding their raw data. This would be beneficial for research as a whole, as it would make the data across laboratories more comparable and thus more valuable.
idalab: Timo, thank you very much for introducing us to the world of altered states of consciousness! We look forward to reading your paper on the topic.
Featured Image: ASDB logo