Double-Dissociation Studies in Psychiatric Research: A Scoping Review

After participating in this activity, learners should be better able to:• Evaluate the double-dissociation approach to research in neuropsychology• Assess research aiming to provide evidence of double dissociation between neurobiological abnormalities and clinical presentations in psychiatry BACKGROUND: Psychiatric neuroscience research has grown exponentially, but it has not generated the desired breakthroughs in diagnosis, treatment development, or treatment selection. In many instances a given neurobiological abnormality is found in multiple clinical syndromes, and conversely, a clinical syndrome is associated with multiple neurobiological abnormalities. To the extent that neurobiology research is conducted to explain psychiatric manifestations, however, it should also provide insight into how certain brain abnormalities lead to one and not another specific clinical presentation-that is, "double-dissociation." We hypothesized that most psychiatric research studies are not designed to identify such double dissociations.

We selected three leading psychiatric journals (American Journal of Psychiatry, JAMA Psychiatry, and Molecular Psychiatry) that are representative of high-quality psychiatry research and that also provided a sample size that was feasible to screen. We screened all original research manuscripts published over the course of one calendar year (2017) to identify those measuring brain function or biological parameters (which, collectively, we term neurobiological measures) in psychiatric disorders. We asked whether such biological research could provide evidence for a double dissociation of any kind.

We found that only 7 of 403 articles published in three psychiatry journals, constituting approximately 2% of publications, examined the dissociation of neurobiological measures relating to two psychiatric disorders or symptom clusters. Of these 7 studies, 5 used imaging as research tool; 1 used genotype array; and 1 used polymerase chain reaction (PCR). Sample sizes of the 7 studies ranged from 100 to 2876.

We report on a striking paucity of research aiming to provide evidence of double dissociation between neurobiological abnormalities and clinical presentations in psychiatry. We conclude that this paucity represents a missed opportunity for the field. Double-dissociation approaches have been used successfully in many studies in neurology and psychiatry in the past, and more widespread and explicit adoption of this design may improve the mechanistic insights obtained in psychiatry research.