Engagement as a predictor of suicidal ideation and suicide attempts: Evidence from a 12-month ecological momentary assessment study in psychiatric outpatients.

Abstract

Ecological momentary assessment (EMA) enables real-time monitoring of suicidal thoughts, offering a window into dynamic risk processes that static assessments miss. However, it remains unclear whether patterns of EMA engagement-rather than the content of responses-carry clinical significance for suicide risk. This year-long prospective cohort study examined whether EMA engagement predicted daily suicidal ideation and clinician-assessed suicide attempts during the 12-month period among psychiatric outpatients at elevated suicide risk. Participants (N = 585; M age = 44.1 years, SD = 14.4; 64.4% female) completed daily EMAs via the MEmind mobile application. Engagement was indexed by (1) total EMA responses and (2) EMA consistency, defined as the cumulative proportion of days with at least one EMA completed. Mixed-effects models estimated associations between engagement and (a) daily active and passive suicidal ideation and (b) suicide attempts obtained from electronic health records. Quadratic models revealed an inverted-U relationship between EMA consistency and active ideation (β2 = -185.55, p = .035). Higher EMA consistency predicted a lower likelihood of a suicide attempt (aOR = 0.72, p = .038), and this protective effect followed a nonlinear pattern (β2 = -10.75, p = .031). Consistent EMA engagement was associated with reduced suicide-attempt risk, independent of demographic factors. Engagement itself, beyond self-reported ideation, may serve as a behavioral marker of suicidality.