Virtual reality-assessment of social interactions and prognosis in depression

Abstract

AbstractImportanceStratification of depression for personalised treatment is urgently needed to improve poor outcomes. Excessive self-blame-related motivations such as self-punishing tendencies have been proposed to play a key role in the onset and maintenance of depression. Their prognostic role, however, remains elusive.ObjectiveUse Virtual Reality (VR) to determine whether maladaptive self-blame-related action tendencies are associated with a poor prognosis for depression when treated as usual in primary care (pre-registered: NCT04593537).DesignRemote prospective cohort study (6/2020-6/2021) with four months follow-up.SettingsOnline recruitment from primary care and self-report.Participantsn=879 pre-screened, n=164 eligible, n=101 completed baseline (age:18-66 years, mean=32.05±12.32, n=89 female), n=98 the VR-task, and n=93 the follow-up. Main inclusion criteria: at least one antidepressant medication trial and Patient Health Questionnaire-9≥15 at screening; main exclusion criteria: screening above threshold on validated self-report instruments for bipolar or alcohol/substance use disorders.Exposure(s)All participants completed a VR assessment via headsets sent to their homes, as well as online questionnaires to measure their clinical characteristics.Main outcomes and MeasuresPrimary: Quick Inventory of Depressive Symptomatology self-reported-16 score after four months. Hypotheses in the study were formulated before the data collection and pre-registered.ResultsContrary to our specific prediction, neither feeling like hiding nor creating a distance from oneself was associated with prognosis of depression during the follow-up period in the pre-registered regression model. Using a data-driven principal components analysis of all pre-registered continuous measures, a factor most strongly loading on punishing oneself for other people’s wrongdoings (β=.23, p=.01), a baseline symptom factor (β=.30, p=.006) and Maudsley Staging Method treatment-resistance scores (β=.28, p=.009) at baseline predicted higher depressive symptoms after four months. This relationship was confirmed by a significant interaction between feeling like punishing oneself for others’ wrongdoings and time of monthly follow-up which was driven by higher depressive symptoms at last follow-up [F(1,84)=6.45, p=.01, partial Eta Squared=0.07] in the subgroup who had reported feeling like punishing themselves at baseline. Our pre-registered statistical learning model prospectively predicted a cross-validated 19% of variance in depressive symptoms.Conclusions and RelevanceFeeling like punishing oneself is a relevant prognostic factor and should therefore be assessed and tackled in personalised care pathways for difficult-to-treat depression.Key pointsQuestionCan remote virtual-reality (VR)-based measures of social interactions be used to identify risk factors of poor prognosis in depression?FindingsIn this pre-registered remote prospective cohort study in 101 participants with depression and follow-up over four months, the VR-assessed feeling like punishing oneself for other people’s wrongdoing was the sole prognostic risk factor apart from known clinical variables such as symptom severity and previous treatment-resistance.MeaningFeeling like punishing oneself is a relevant prognostic factor and should therefore be assessed and targeted in difficult-to-treat depression.