Investigating the Mechanisms of Hallucinogen-Induced Visions Using 3,4-Methylenedioxyamphetamine (MDA): A Randomized Controlled Trial in Humans

This double-blind, placebo-controlled study (n=12) found large individual variability in MDA-induced visions and suggests that they could have similar mechanisms as other hallucinatory syndromes.

Abstract

“Background: The mechanisms of drug-induced visions are poorly understood. Very few serotonergic hallucinogens have been studied in humans in decades, despite widespread use of these drugs and potential relevance of their mechanisms to hallucinations occurring in psychiatric and neurological disorders. Methodology/Principal Findings: We investigated the mechanisms of hallucinogen-induced visions by measuring the visual and perceptual effects of the hallucinogenic serotonin 5-HT2AR receptor agonist and monoamine releaser, 3,4-methylenedioxyamphetamine (MDA), in a double-blind placebo-controlled study. We found that MDA increased self-report measures of mystical-type experience and other hallucinogen-like effects, including reported visual alterations. MDA produced a significant increase in closed-eye visions (CEVs), with considerable individual variation. Magnitude of CEVs after MDA was associated with lower performance on measures of contour integration and object recognition. Conclusions/Significance: Drug-induced visions may have greater intensity in people with poor sensory or perceptual processing, suggesting common mechanisms with other hallucinatory syndromes. MDA is a potential tool to investigate mystical experiences and visual perception.”

Authors: Matthew J. Baggott, Jennifer D. Siegrist, Gantt P. Galloway, Lynn C. Robertson, Jeremy R. Coyle & John E. Mendelson

Summary

The mechanisms of drug-induced visions are poorly understood, despite widespread use of these drugs.

Introduction

We administered 3,4-methylenedioxyamphetamine (MDA) to healthy drug-experienced volunteers to study drug-induced visual hallucinations. The results show that MDA causes altered form and depth perception, prolonged afterimages, motion-processing impairments, vivid pseudo-hallucinations, and only very rarely actual hallucinations.

Some of the ”Ecstasy” sold in Australia contains MDA instead of MDMA. Animal studies show that MDA has both typical hallucinogenic (LSD-like) effects as well as unusual effects similar to those of 3,4-methylenedioxymethamphetamine (MDMA).

Loss of sensory or perceptual ability may contribute to drug-induced hallucinations. This includes poor vision and perceptual difficulties, such as those seen in people with schizophrenia and ketamine users.

Hallucinations may sometimes arise from abnormally increased neural activity, possibly involving abnormal interactions between brain areas. Serotonergic hallucinogens likely alter the balance of excitation and inhibition in the cortex.

A third factor that may contribute to drug-induced hallucinations is impairment in perceptual inference, which can be tested by providing top-down cues for recognizing images.

General study design for MDA experiment

A double-blind, placebo-controlled within-subjects crossover study was carried out at the UCSF Clinical Research Center at San Francisco General Hospital.

Twelve healthy individuals with self-report experience with MDMA were screened for safety. They were asked to practice effective contraception and had comprehensive safety screening procedures performed before drug administration.

After a 2-hour fast, MDA was administered in a dose of 98 mg/70 kg body weight.

Self-report measures

We used visual analog items to measure the time course of rapidly changing general drug effects and visual changes. Maximum post dose changes (Emax) were used as the primary outcome measures, and summary measures of peak general drug effects and peak visual changes were made by averaging Emax for the four questions in each category.

We measured MDA effects with two self-report questionnaires, the Altered States of Consciousness (ASC) and the Hood Mysticism questionnaires. The ASC includes three scales: the Oceanic Boundlessness (OB) scale, the Dread of Ego Dissolution (DED) scale, and the Visionary Changes (VC) scale.

We selected the self-report Mysticism scale developed by Hood [82] from the work of Stace [83] to assess possible mystical-type experiences. Three empirically derived factors were measured: interpretation, introvertive mysticism, and extrovertive mysticism.

Tilt illusion task

The tilt illusion (TI) is an orientation-based visual illusion caused by lateral inhibition between orientation-selective cortical neurons in the occipital cortex. Hallucinogens may cause hallucinations by altering the balance of excitation and inhibition in the cortex.

We used a two-dimensional sine wave and a grating to measure the strength of the tilt illusion. We then used two one-up-one-down staircases to estimate the point of subjective verticality.

Participants were tested binocularly, seated approximately 0.110 m away from a 19 in. Dell monitor in a dimly lit testing room, and were presented with a 1.2u diameter target and a 5.2u diameter surround.

We determined the mean points-of-subjective-verticality (PSV) for each surrounding orientation, and corrected for head angle variation, then estimated the tilt illusion.

Contour integration task

We used a contour detection task to assess potential hallucinogen-induced changes in perceptual ability. We predicted that individuals with lower baseline performance on contour-detection tasks would have increased susceptibility to drug-induced hallucinations.

We used a contour integration task to measure perceptual organization. Participants made unspeeded two-alternative forced-choice (2AFC) about the orientation of an egglike shape that was obscured by a background of evenly-spaced randomly-oriented Gabor gratings.

Participants were tested binocularly with a 19″ Dell monitor at a distance of 0.11 m, and with increasing orientation jitter varying between 7u and 24u across five difficulty levels.

We used a generalized linear model to analyze contour integration results, and then used threshold as the dependent measure.

Object recognition task

A task was designed to assess the ability to use top-down information to facilitate perception. If drug-induced hallucinations involve increased efficacy of top-down influences of perception, then widening differences in accuracy should occur.

Stimuli were black-and-white drawings of objects placed on a random noise background. The noise background was chosen to approximate the spectral characteristics of natural scenes.

We used the Random Image Structure Evolution technique to progressively distort the images, averaging the amplitudes of all images together and recombining the phase information. This produced images with the same frequency spectrum and differing only in phase.

Pilot recognition experiments were conducted in 10 healthy volunteers to identify images that were particularly easy or difficult to identify. The experiments included 60 images at each administration.

Participants were tested binocularly, seated approximately 0.11 m away from a 19 in. Dell monitor, and were presented with stimuli for 2 seconds.

Participants pressed a key to begin a trial, and were shown an image, which they had to identify. The image was shown with 87.5% distortion, and decreased by 5% until the participant believed they could identify the image.

We used a linear model with a sequence-condition interaction term and a fixed-effect for drug condition and participant. We made pairwise comparisons using Tukey’s post hoc tests.

Self-report measures

MDA had significant effects on participants’ maximum ratings of all four general drug effects VAS measures.

Approximately half of the participants reported significant closed-eye visions on the four visual changes VAS measures. There was no significant correlation between mean Emax for visual questions and mean Emax for general drug effects.

We examined whether time of maximum change varied between patterns and other categories of visuals during hallucinogen inebriation, but found no significant differences.

Tilt illusion task

One participant had unusable data on their first session because they misunderstood the direction of the judgments being made, preventing the staircase from converging on a threshold. The magnitude of the tilt illusion was not significantly affected by dosing condition, but did increase with dosing sequence.

Contour integration task

The study found that peak CEVs predicted threshold orientation jitter, but did not predict object recognition accuracy. Participants on MDA were more impaired by image degradation, and the study found no significant effects of dosing sequence or days elapsed since last exposure to an MDMA-like drug. Examining only the accurate trials, we found significant effects of dosing condition and cue type on degradation level of correct identification.

We found that self-report visual change was significantly associated with performance for both MDA and placebo. The relationship was larger for MDA than placebo, and was specific to visual effects rather than general drug effects.

Discussion

In a controlled study, participants reported experiencing prototypical hallucinogen effects after MDA administration. This is the second published controlled experiment to assess a hallucinogen using a well-validated measure of mystical-type experience.

MDA had significant perceptual effects, such as visual percepts of smoke rings. However, subsequent researchers have emphasized the unusually consistent social and emotional effects of MDA rather than the visual changes. After MDA, individuals who experienced more intense CEVs had lower performance on the two measures of perceptual organization and had greater MDA-induced changes in perceptual performance. This finding is consistent with evidence from other populations linking hallucinations to decreased sensory fidelity and impaired perceptual organization.

Although never previously studied, increased occipital cortex excitation was predicted by Ermentrout and Cowan’s model of geometric hallucinations. We detected some evidence of increased occipital cortex excitation using the tilt illusion, but the effect was not significant.

MDMA users have reported persisting visual percepts and increased cortical excitability, which may explain the geometric pattern hallucinations sometimes reported by such individuals.

MDA-induced hallucinations may be caused by increased influence of top-down factors on perception. However, we were not able to confirm a statistical interaction between cue type and dosing condition in the object recognition task.

The object recognition task only manipulated one kind of top-down influence, but it likely minimized the type of top-down attentional changes that are normally measured, for example, in the Posner cuing task.

The overall study design limited the generalizability of the results because participants were experienced users of MDMA and hallucinogens, and dosing sessions were on consecutive days to maximize measurement of the pharmacokinetics of a long-half-life drug while minimizing blood loss in participants.

Use of noise masking and signal detection theory to examine mechanisms of hallucinogen-induced visual changes could help us understand the relationships between neuropharmacology and hallucinogen-induced phenomenology.

We conducted the first study in over thirty years to confirm that the hallucinogen MDA induces mystical-type experiences and CEVs. CEVs were associated with lower performance on measures of contour integration and object recognition.