Remote virtual reality assessment elucidates self-blame-related action tendencies in depression

Abstract

Darwin stated that humans have a strong and involuntary tendency to perform certain actions when a specific state of mind is induced. Such "action tendencies" are key to understanding the maladaptive impact of self-blame-related feelings in depression. For example, feeling like "hiding" and "creating a distance from oneself" in a text-based task were previously associated with recurrence risk in remitted depression. Despite their functional importance, action tendencies have not been systematically investigated in current depression, which was the aim of this pre-registered study. To this end, we developed the first virtual reality (VR) assessment of blame-related action tendencies and compared current depression (n=98) with control participants (n=40). The immersive VR-task, pre-programmed on devices sent to participants' homes, used hypothetical social interactions, in which either participants (self-agency) or their friend (other-agency) were described to have acted inappropriately. Good concurrent validity was demonstrated for VR- versus the text-based task. As predicted, compared with controls, people with depression showed a maladaptive profile: Particularly in the other-agency condition, rather than feeling like verbally attacking their friend, they were prone to feeling like hiding, and punishing themselves. The depression group showed a more pronounced slowing of response times in the other- versus the self-agency condition, which may reflect a stronger ambivalence about externalising blame in depression. Interestingly, feeling like punishing oneself was associated with a history of self-harm but not suicide attempts. Current depression and self-harm history were thus linked with distinctive motivational signatures, paving the way for remote VR-based stratification and treatment.

Virtual reality-assessment of social interactions and prognosis in depression

“…In line with this, Tangney et al (2007) postulated that adaptive self-blame-related action tendencies are more associated with behavioural activation and involve proactive pursuit such as reparative actions, whereas maladaptive self-blame-related action tendencies motivate social withdrawal and interpersonal separation (Duan, Lawrence, et al, 2022; Tangney et al, 2007). In keeping with this hypothesis, using a text-based task, we recently showed that withdrawal-related maladaptive action tendencies including hiding and creating a distance from oneself were more pronounced in patients with remitted depression compared with control participants (Duan, Lawrence, et al, 2022) and were associated with subsequent recurrence risk (Duan, Lawrence, et al, 2022; Duan, Valmaggia, et al, 2022).…”

“…Exclusion criteria were: a previously diagnosed or likely bipolar disorder [Hypomanic Checklist-16 (Forty et al, 2010) score > 8, with symptoms lasting ≥ 2 days, and endorsing two of the first three screening questions of the Composite International Diagnostic Interview bipolar screening scale (Kessler et al, 2006)], a personal history of schizophreniform symptoms [three clinical screening questions to exclude schizophreniform disorders (Lythe et al, 2015)], drug or alcohol abuse over the last 6 months [Primary Care Evaluation of Mental Disorders (Spitzer et al, 1994), modified to screen for drug abuse], a suspected central neurological condition, a planned or current pregnancy, or currently being treated by a mental health specialist in secondary care. More information about the inclusion/exclusion reasons can be found in the Supplementary Materials of Duan, Valmaggia, et al (2022).…”

“…We also used the Maudsley Modified Patient Health Questionnaire-9 (Harrison et al, 2021) as a secondary outcome. More details on assessments of clinical characteristics can be found in Duan, Valmaggia, et al (2022).…”

“…For the present prospective study, we used a novel VR task developed and validated in our separately reported cross-sectional study comparing people with and without depression at baseline to examine our pre-registered hypothesis 1 (NCT04593537). The cross-sectional study revealed that people with depression showed a distinctive tendency to attack/punish themselves, which was specifically associated with a history of self-harm but not suicide attempts (Duan, Valmaggia, et al, 2022). Here, we prospectively probed the prognostic role of maladaptive self-blame-related action tendencies for depressive symptoms over four-months of follow-up whilst being treated as usual in primary care.…”

“…Here, we prospectively probed the prognostic role of maladaptive self-blame-related action tendencies for depressive symptoms over four-months of follow-up whilst being treated as usual in primary care. Previous measures of action tendencies used abstract descriptions of social scenarios which limited engagement in the scenarios and ecological validity (Duan, Valmaggia, et al, 2022; Mu & Berenbaum, 2019). Our immersive VR task made it possible to act out the action tendencies as well as to increase the engagement of participants.…”