The association of short-term variability in suicidal ideation with subsequent suicidal thoughts and behaviours: Longitudinal cohort study

Behav Res Ther. 2026 Apr 20;202:105049. doi: 10.1016/j.brat.2026.105049. Online ahead of print.

ABSTRACT

BACKGROUND: There is emerging evidence that within-person variation (“intraindividual variability”) in suicidal ideation may have strong predictive power for suicide attempt. However, evidence is limited to underpowered studies and retrospective, cross-sectional designs. This study aimed to characterise trajectories of suicidal ideation frequency and intraindividual variability in suicidal ideation, identify health and sociodemographic factors associated with variability, and test whether variability longitudinally predicted suicide attempt, self-harm and severity of suicidal ideation over 12 months.

METHODS: Adults with SI (n = 831, 66% women) were assessed in the LifeTrack study, a 3-year prospective longitudinal cohort. Quadratic growth mixture modelling identified distinct trajectories of suicidal ideation, based on daily assessments of ideation frequency over 7 days. Regression models identified factors associated with these trajectories and intraindividual variability, and assessed whether variability predicted suicide attempt, self-harm and suicidal ideation over 12 months.

RESULTS: Four short-term trajectories of suicidal ideation were identified: “low stable” (55%), “low increasing” (7%), “high stable” (27%) and “high decreasing” (11%). Cox proportional hazards models showed no significant association between different trajectories and suicide attempts over 12 months (p = 0.51). Intraindividual variability was modestly associated with subsequent suicide attempts (p = 0.03), explaining minimal variance beyond initial ideation severity. Variability also had no consistent association with self-harm or suicidal ideation at 12 months. Factors associated with greater variability included higher perceived burdensomeness and less pain interference, and participants with higher distress or depression symptoms also had greater odds of poorer trajectories.

CONCLUSION: Suicidal ideation variability had limited predictive value as a risk factor for subsequent suicidal ideation, self-harm or suicide attempt in a large population-based sample. Consequently, frequent and repeated assessment of suicidal ideation may have little clinical utility. Predictors of better short-term suicidal ideation trajectories indicate that supporting mental health, physical health, and social connection is critical for people experiencing suicidal thoughts and behaviours.

PMID:42019098 | DOI:10.1016/j.brat.2026.105049