Unraveling individual patterns of passive and active suicidal ideation: A person-specific network modeling exploration

Introduction. Prediction and prevention of suicidal ideation (SI) remain challenging, largely because traditional approaches have focused less on individual variability in its moment-to-moment fluctuations. To address this gap, we examined how group-level network models capture average dynamic associations over time, while person-specific networks reveal individual deviations in both contemporaneous and time-lagged relationships, clarifying how changes in cognitive or emotional states precede within-person changes and subsequent fluctuations in SI, particularly in relation to perceived burdensomeness, thwarted belongingness, and hopelessness.Methods. Twenty-four psychiatric inpatients (75% female, mean age = 33.6, SD = 10.7) completed ecological momentary assessments (five prompts/day) for 28 days post-discharge, reporting passive and active SI, perceived burdensomeness, thwarted belongingness, andhopelessness. Data were analyzed using Group Iterative Multiple Model Estimation (GIMME) to derive group- and individual-level networks.Results. Group-level GIMME networks showed contemporaneous and lagged connections between passive and active SI. Individual-level GIMME networks showed that for 30.4% of individuals, hopelessness mediated the link between interpersonal risk factors, passive, and active SI.Conclusion. By identifying both shared and person-specific temporal pathways, these findings highlight the value of dynamic, idiographic assessment strategies for advancing precision psychiatry in suicide prevention.