Letter to the Editor of Molecular Psychiatry (submitted on 24 of April, 2015; rejected for publication on 27 of April, 2015):
José Carlos Bouso, PhD.
Telling true from false: cannabis researchers show increased susceptibility to biased conclusions.
To the Editor,
In a recent study Riba et al. (1) compared 16 recreational cannabis users with 16 cannabis naïve controls, finding that the former “had a higher susceptibility to memory illusions” according to both behavioral and fMRI results using the Deese Roediger McDermott (DRM) paradigm, a episodic memory test. As will be argued below, these researchers did not discuss crucial information regarding the characteristics of the cannabis group, nor did they discuss some important limitations of their study that raises serious doubts (and concerns) about their reported conclusions.
I was involved in both participant recruitment and data collection for this study (as noted in the Acknowledgements section). Additionally, I am the anonymous clinical psychologist who conducted interviews to collect psychiatric, neurological and history of drug use information about the subjects (see the Participant section). I would like to share important information that is missing from the final version of this paper, some of which may cast doubt on the results as reported:
1) Seven of the 16 cannabis users were identified as having current or past medical conditions that could interfere with optimal neuropsychological performance: cancer, childhood asthma, migraines, depression (2 subjects), thyroid problems (treated with thyroxine at the time of the study), and anxiety.
2) Almost all the cannabis users (15 of 16) were found to have current or past records of concomitant use of other psychoactive drugs including MDMA, LSD, psilocybin mushrooms, ketamine, opiates, and cocaine, among others. Six subjects had long-term experience with potentially neurotoxic amphetamine derivatives (2); and/or wide experience with several hallucinogenic drugs, which may produce cortical changes in the default mode network (3), a brain network involved in episodic memory retrieval (4). These unreported variables raise significant doubt as to whether the study’s results are uniquely due to cannabis use.
3) Participants in the study of Riba et al. (1) also filled out questionnaires assessing psychopathology, personality and other psychological variables, and performed some other different memory and neuropsychological tests. Since the authors did not report any data gathered from these tools, it is unknown if the two groups were also matched according to other psychological conditions (as they were for IQ). These data would allow a more adequate interpretation of the results.
4) Researchers also collected hair samples from all subjects. These results were also unmentioned in the paper, which is unfortunate in that they could have provided more precise information concerning the potential relationships between past and/or current drug use and the results reported.
In sum, taking all this missing information into account, and considering that there is a lack of information regarding potential premorbid conditions of the cannabis subject group, it is simply not possible to establish causal relationships between cannabis use and alterations in the DRM paradigm. Although the authors reported correlations between lifetime cannabis use and activity in the medial temporal lobe, these results may be also biased by these confounding variables left out of the paper.
Beyond the issues outlined above, there are other significant limitations to the study not mentioned by its authors. First, the apparently limited sample size. The article does not stipulate the principal variable being studied, and the authors do not show their analysis regarding calculation of the sample size, so it is difficult to know the degree of extrapolation of their results. Second, although the DMR paradigm has more than 15 years of use, the reasons why individual differences in the DRM paradigm are related to eventual false memories are not well established yet (5). On the other side, so far it is not clear if these two phenomena are cognitively identical, and if they can be allocated to the same underlying neural processes (5). Recent studies show that errors in the DRM paradigm do not necessary imply errors in autobiographical memories (6), and that both type of errors do not seem to correlate (7). Therefore, the ecological validity of the DRM paradigm still needs to be better established. For example, it seems that using words in the DRM paradigm, as Riba et al. (1) do, shows less ecological validity that using, for instance, images (8).
A final issue regarding this article concerns the specific series of words they used. The authors provide no information relating to the reliability and validity of the stimulus they selected. Since normative data for the general population or for different subpopulations seems not to exist, it is difficult to know if the behavioral results observed in the cannabis group even correspond to a dysfunctional performance. The original protocol of the study published by the authors was designed to incorporate a third group of ayahuasca users, a brew composed, among other alkaloids, by the hallucinogenic drug dimethyltriptamine (N,N-DMT) (9). Since the authors do not report data from that third group, it is impossible to establish whether the results obtained for the cannabis group imply a particular effect derived from the use of cannabis alone.
In sum, the research reported by Riba et al. (1) is riddled with biases which should inspire caution before assuming any definitive conclusions regarding the impact of cannabis use on possible detrimental effects on reality monitoring, as the authors conclude. The results obtained by them may well be nothing more than laboratory artifacts devoid of any relationship with the actual cognitive capacities of the subjects and without any real clinical or practical implications.
(1). Riba J, Valle M, Sampedro F, Rodríguez-Pujadas A, Martínez-Horta S, Kulisevsky J, Rodríguez-Fornells A. Telling true from false: cannabis users show increased susceptibility to false memories. Mol Psychiatry 2015; 20: 772–777
(2). Yamamoto BK, Moszczynska A, Gudelsky GA. Amphetamine toxicities: classical and emerging mechanisms. Ann N Y Acad Sci 2010; 1187: 101-21.
(3). Bouso JC, Palhano-Fontes F, Rodríguez-Fornells A, Ribeiro S, Sanches R, Crippa JA, Hallak JE, de Araujo DB, Riba J. Long-term use of psychedelic drugs is associated with differences in brain structure and personality in humans. Eur Neuropsychopharmacol 2015; 25(4):483-92.
(4). Sestieri C, Corbetta M, Romani GL, Shulman GL. Episodic memory retrieval, parietal cortex, and the Default Mode Network: functional and topographic analyses. J Neurosci 2011; 31(12): 4407-20.
(5). Gallo DA. False memories and fantastic beliefs: 15 years of the DRM illusion. Memory & Cognition 2010; 38(7): 833-848.
(6). Ost J, Blank H, Davies J, Jones G, Lambert K, Salmon K. False Memory ≠ False Memory: DRM Errors Are Unrelated to the Misinformation Effect. PLOS ONE 2013; 8(4): e57939.
(7) Zhu B, Chen C, Loftus EF, Lin C, Dong Q. The relationship between DRM and misinformation false memories. Mem Cognit 2013; 41(6): 832-8.
(8). Baioui A, Ambach W, Walter B, Vaitl D. Psychophysiology of False Memories in a Deese-Roediger-McDermott Paradigm with Visual Scenes. PLOS ONE 2012; 7(1): e30416.
(9). McKenna DJ. Clinical investigations of the therapeutic potential of ayahuasca: rationale and regulatory challenges. Pharmacol Ther 2004; 102(2): 111-29.