- EMA captures real-time suicidal urge fluctuations and significantly improves prediction of next-session suicidal ideation beyond retrospective session measures.
- Both mean and maximum EMA-assessed suicidal urges independently predicted next-session ideation (p < .001).
- Predictive utility remains nearly unchanged from twice daily to once every other day; further reductions substantially reduce utility, suggesting moderate sampling balances burden and accuracy.
J Affect Disord. 2026 Jun 17:122133. doi: 10.1016/j.jad.2026.122133. Online ahead of print.
ABSTRACT
OBJECTIVES: Suicidal risk fluctuates rapidly, yet suicide monitoring primarily relies on retrospective assessments. Ecological momentary assessment (EMA) offers a promising approach for capturing real-time fluctuations in suicidal urges and identifying elevated suicidal risks beyond clinical settings, but poses substantial participant burden. This study aimed to examine the incremental predictive value of EMA and its optimal sampling frequency.
METHODS: Military service members and veterans (N = 62) with recent suicidal ideation or attempts received outpatient psychotherapy weekly. Participants completed measures of suicidal ideation at each session and four daily EMA surveys of suicidal urges over 28 days. Mixed-effects models examined whether EMA-assessed suicidal urges during the inter-session window predicted suicidal ideation at the subsequent therapy session beyond retrospective measures. We simulated reduced EMA frequencies by probabilistically thinning EMA observations and re-estimated model performance with different EMA frequencies.
RESULTS: Both the mean and maximum levels of EMA-assessed suicidal urges significantly improved the prediction of next-session suicidal ideation beyond prior session-based measures (p < .001). Simulation analyses revealed a nonlinear relationship between EMA frequency and its predictive utility. The predictive utility of EMA remained nearly unchanged as sampling frequency decreased from highly frequent (twice per day) to moderately frequent (once every other day) schedules but declined more sharply with further reductions in EMA frequency.
CONCLUSIONS: EMA enhanced the prediction of near-term suicidal risks beyond session-based assessments. Moderate EMA sampling frequencies (once every other day) may preserve most predictive utility while reducing participant burden, offering a practical approach for scalable suicide monitoring in clinical populations.
PMID:42309190 | DOI:10.1016/j.jad.2026.122133
Share Evidence Blueprint
Save to Google Notes
Search Google Scholar
Save as PDF
⭐ My Revision List
