Effects of ketamine on brain function during metacognition of episodic memory

This double-blind placebo-controlled fMRI study (n=53) on ketamine (r-ketamine, continuous iv) and cognition found that ketamine increased metacognitive bias, negatively impacted metacognitive sensitivity, and increased activation of posterior brain areas.

Abstract

“Only little research has been conducted on the pharmacological underpinnings of metacognition. Here, we tested the modulatory effects of a single intravenous dose (100 ng/ml) of the N-methyl-D-aspartate-glutamate-receptor antagonist ketamine, a compound known to induce altered states of consciousness, on metacognition and its neural correlates. Fifty-three young, healthy adults completed two study phases of an episodic memory task involving both encoding and retrieval in a double-blind, placebo-controlled fMRI study. Trial-by-trial confidence ratings were collected during retrieval. Effects on the subjective state of consciousness were assessed using the 5D-ASC questionnaire. Confirming that the drug elicited a psychedelic state, there were effects of ketamine on all 5D-ASC scales. Acute ketamine administration during retrieval had deleterious effects on metacognitive sensitivity (meta-d′) and led to larger metacognitive bias, with retrieval performance (d′) and reaction times remaining unaffected. However, there was no ketamine effect on metacognitive efficiency (meta-d′/d′). Measures of the BOLD signal revealed that ketamine compared to placebo elicited higher activation of posterior cortical brain areas, including superior and inferior parietal lobe, calcarine gyrus, and lingual gyrus, albeit not specific to metacognitive confidence ratings. Ketamine administered during encoding did not significantly affect performance or brain activation. Overall, our findings suggest that ketamine impacts metacognition, leading to significantly larger metacognitive bias and deterioration of metacognitive sensitivity as well as unspecific activation increases in posterior hot zone areas of the neural correlates of consciousness.”

Authors: Mirko Lehmann, Claudia Neumann, Sven Wasserthal, Johannes Schultz, Achilles Delis, Peter Trautner, René Hurlemann & Ulrich Ettinger

Summary

A single intravenous dose of the N-methyl-D-aspartate-glutamate-receptor antagonist ketamine was administered to 53 young, healthy adults to examine the effects of the drug on metacognition and its neural correlates. The results suggest that ketamine impacts metacognition, leading to larger metacognitive bias and deterioration of metacognitive sensitivity.

Introduction

In everyday life, we often think about other thoughts. These meta-thoughts are postulated to be a distinct feature of consciousness, and the ability to reflect on our own thoughts and knowledge is a key to being conscious.

Metacognitive sensitivity can be measured by assessing how well participants introspectively assess or monitor their own cognitive processes. Metacognitive efficiency can be measured as the amount of signal strength available for the metacognitive process, expressed as a fraction of the amount of signal strength available for the primary task.

Neuroimaging and lesion studies suggest that higher-order conscious functions such as metacognition may also engage a frontoparietal network. However, little is known about the pharmacology of metacognition.

One neurotransmitter likely to mediate aspects of consciousness is the glutamatergic system. Ketamine, a noncompetitive NMDA-receptor antagonist, induces psychedelic states and is a well-established research tool with an excellent safety record in both clinical and experimental applications.

Ketamine affects cognition in several ways, including a selective degradation of episodic memory. This degradation may depend on the depth of semantic processing of the encoded items, as well as on whether ketamine is administered during encoding or during retrieval.

In this double-blind, placebo-controlled fMRI study, we investigated the role of the glutamate system in metacognition and its underlying neural activity by applying a psychotomimetic dose of ketamine. We also investigated the subjective, phenomenological effects of ketamine by including a self-report measure of altered states of consciousness.

Materials and Methods

Participants

The study was approved by the Research Ethics Committee at the Department of Psychology, University of Bonn. Behavioral data and fMRI data are available upon request.

Screening procedure

Individuals who responded to study advertisements were interviewed online, and a urine drug screen and pregnancy test were performed. A medical questionnaire was used to rule out any current or past medical conditions, or diagnoses of psychotic disorders among first-degree relatives.

Study design

Twenty-four participants received a subanesthetic dose of racemic ketamine, and 29 participants received a saline solution. Ketamine was administered as a 2 mg/ml solution with a constant target plasma level of 100 ng/ml by a bolus and continuous infusion using a computerized infusion pump.

General procedure

Participants were instructed to refrain from solid food and clear fluids for 6 h before the MRI infusion, and to stay abstinent from alcohol for 24 h before the MRI.

Task design

Study Phase I

Participants were presented with 120 word items and were instructed to make one of two types of judgments about these items. One set of items was encoded in a deep manner, whereas the other set of items was encoded in a shallow manner.

The fMRI retrieval task consisted of 180 word items, 120 of which had been encoded in the previous task and 60 of which were new. Participants responded to items they considered to be old or new by predefined button presses.

There were two types of second-order ratings: participants rated their subjective confidence regarding the judgment on a 6-point Likert scale, and then followed a control condition by pressing left or right thumb.

The participants had to navigate the cursor towards a predefined number on a scale, highlighted in blue, while avoiding extreme responding bias. The duration of the decision window was 3.5 s, followed by a 0.5 s screen where the participant’s response was highlighted.

Study Phase II

Participants performed a second encoding task in which they rated the subjective pleasantness of novel words and reported the number of syllables. The tasks alternated blockwise, with 10 blocks each comprising 10 items.

Participants filled in a 5D-ASC questionnaire after the infusion, which included three primary scales and two secondary scales. The scale scores were formed following guidelines by Dittrich et al. (2006).

After an hour, participants were tested again on retrieval of the second encoding task, without infusion of ketamine, and with two exceptions: ISI was constant and there was no Follow condition.

fMRI data acquisition and analysis

After pre-processing, the onset of each stimulus was defined, and the duration was set to be the reaction time from stimulus presentation to button press. The realignment parameters were added to the model as covariates of no interest.

Overall, there were four Type 1 regressors: “Deep” (mean number of trials across participants: 49), “Shallow” (mean number of trials across participants: 27.19), “New” (mean number of trials across participants: 47.02, SD 13.5), and “Incorrect” (mean number of trials across participants: 41.85, SD 13.9).

Study Phase II used a simpler model with conditions “Deep” and “Shallow”, and did not use retrieval performance.

A second level analysis was conducted using between-subjects factors “Drug” (ketamine/placebo) and within-subjects factors “Word Type” (deep/ shallow/new) and “Rating Type” (report/follow) for second-order contrasts. Significant clusters were inferred using the anatomy toolbox.

Four participants had to be excluded from fMRI analysis because normalization failed, so 49 participants were analyzed for Study Phase I and 50 participants for Study Phase II.

Behavioral data analysis

Type 1 (retrieval) and Type 2 (metacognitive) performance was assessed in an SDT framework. Meta-d0 analysis was applied to quantify metacognitive sensitivity, and Type 2 efficiency relative to Type 1 performance was calculated.

We conducted various exploratory analyses to facilitate mechanistic understanding of the outcomes, including a comparison of Type 1 and Type 2 performance measures and metacognitive bias, and an extension of the HMeta-d toolbox to estimate group-level parameters over log(meta-d0 /d0 ) while taking into account uncertainty in model fits at the single-subject level.

All other behavioral data analyses were conducted using SPSS 22 (IBM Corp., Armonk, USA). Independent samples t-tests, paired t-tests and mixed-design ANOVAs were used to test for drug effects on 5D-ASC scales, Type 1 and Type 2 reaction times and metacognitive bias.

Results

5D-ASC

Ketamine had a significant effect on the 5D-ASC global measure of altered consciousness and on all scales. Participants scored higher on the three primary dimensions “Oceanic Boundlessness”, “Dread of Ego Dissolution” and “Visionary Restructuralization”.

Type 1 behavioral analyses

The LoP manipulation was successful: participants showed significantly enhanced retrieval performance for deeply compared to shallowly encoded items, but no main effect of “Drug”. Ketamine did not significantly alter retrieval performance.

Type 2 behavioral analyses

Participants showed enhanced metacognitive sensitivity for deeply encoded items compared to shallowly encoded items, and this sensitivity deteriorated under ketamine. However, there were no significant main effects of either “Drug” or “Encoding Level” on metacognitive efficiency.

Second-order fMRI analyses

Report vs. follow effects were found in the right calcarine and lingual gyrus, left and right cuneus, bilateral superior occipital gyrus, and posterior medial frontal cortex. The reverse effect revealed 11 clusters.

Ketamine increased BOLD in five clusters, including the superior parietal lobule, supramarginal gyrus, inferior parietal lobule, and angular gyrus. There were no significant effects for the reverse contrast and no significant interactions.

Exploratory analyses revealed that ketamine caused a higher BOLD response than placebo in the lingual, fusiform, and calcarine gyrus and right SPL.

Type 1 behavioral analyses

Items that had been encoded deeply were recognized more often than shallowly encoded items, and participants made quicker button presses when deeply encoded items were presented.

Type 2 behavioral analyses

There were significant main effects of “Encoding Level” on metacognitive sensitivity and metacognitive efficiency, but no significant effects of “Drug” on either meta-d0 or metacognitive efficiency.

Task effects

Two clusters of brain regions were significantly more active during Report than Follow; the first cluster included right calcarine gyrus, bilateral cuneus, and right lingual gyrus, and the second cluster included left pMFC, which has been linked to metacognitive judgments.

Study Phase II

Drug effects

Ketamine caused psychotomimetic effects on all scales of the questionnaire, which validates the rationale for using this pharmacological challenge.

Ketamine administration during retrieval resulted in deterioration of metacognitive sensitivity (meta-d0) and overconfidence (larger metacognitive bias), suggesting a role of various conscious processes giving rise to this ketamine effect on metacognitive bias. Ketamine did not affect retrieval (Type 1) performance.

Controlling for Type 1 performance (d0 ) by calculating metacognitive efficiency (meta-d0 /d0 ), there was no significant group difference. Metacognitive efficiency measures should be used when comparing different groups, although the theoretical assumption of the relationship between Type 1 and Type 2 performance measures is frequently violated.

Ketamine-associated deterioration of Type 2 sensitivity might be explained by non-significant group-heterogeneity in Type 1 performance, but the group-level estimation in a hierarchical Bayesian framework offers several methodological advantages over previous estimation methods for metacognitive efficiency.

Ketamine increased activity in posterior brain regions during second-order ratings, including the right-hemispheric superior-posterior cortex, left calcarine gyrus, bilateral lingual gyrus, and left IPL. These effects were observed for both second-order rating types (Report/Follow), and are therefore not specific to genuine metacognitive processes.

Ketamine may affect brain function during second-order ratings by means of an up-regulation of posterior visuospatial cortical brain areas. This may account for both the deterioration in metacognitive sensitivity as well as the increased activation in visuospatial areas during second-order ratings.

The anatomical location of our results is of interest with regards to the “hot zone” for conscious functions proposed by Koch et al. (2016). This hot zone is mainly associated with phenomenal qualities of conscious experiences, which self-reported 5D-ASC measures confirmed to be altered by ketamine.

Ketamine increases bilateral temporoparietal functional connectivity and causes a significant alpha current reduction in posterior cortical areas, which may reflect efforts to maintain ego integrity. This may explain why the ketamine-induced psychedelic state reduces the functionality of metacognitive processes.

The 5D-ASC index of altered consciousness and ketamine effects on metacognitive sensitivity cannot be directly connected, although they may take place on the same conscious level. Different causes might account for a deterioration of metacognitive sensitivity. Exploratory analysis of metacognitive bias revealed significantly higher bias for the ketamine group, which may be due to changes in conscious access as well as altered, hallucinatory-like experiences.

Drug effects

Ketamine had no effect on Type 1 sensitivity or Type 2 sensitivity and efficiency of items encoded during maintained drug infusion.

We could not reproduce the increased activation for deeply encoded items in left PFC with ketamine reported by Honey et al. (2005a), but ketamine increased metacognitive bias (overconfidence) in the ketamine group, as was the case in Study Phase I.

Limitations

The infusion protocol served to keep plasma levels of ketamine constant, but participants might have become accustomed to the ketamine-induced state of consciousness and developed mechanisms to stabilize higher-order cognitive functions over the course of the infusion.

A limitation of this study is that only trials with correctly retrieved items could be included in the fMRI analyses, and the study may have lacked sufficient power to detect a statistically significant difference between groups.

Ketamine effects on metacognition require further research to understand the underlying mechanisms. Advanced modeling could help to clarify the effects.

Conclusions

Ketamine affected metacognition, including metacognitive bias and sensitivity, and up-regulation of activity in posterior brain areas during second-order ratings compared to placebo. Ketamine did not affect metacognitive efficiency, however.