BACKGROUND: Heterogeneity in suicidal ideation over time in patients with first-episode psychosis is expected, but prototypical trajectories of this have not yet been established. We aimed to identify trajectories of suicidal ideation over a 3-year period and to examine how these trajectories relate to subsequent suicidality.
METHODS: We used longitudinal data from the prospective 10-year follow-up OPUS trial of young Danish patients with first-episode psychosis. Participants were recruited between January, 1998, and December, 2000, from all inpatient and outpatient mental health services in Copenhagen and Aarhus County, were aged between 18 and 45 years, and had a diagnosis in the schizophrenia spectrum. Randomisation was done by a computer-generated randomisation list, stratified in alternating blocks for five centres. In Aarhus, a secretary drew lots when the researcher phoned her. Patients were randomly assigned to a treatment group (modified assertive community treatment) or a control group (standard treatment at community mental health centres) and assessed at treatment initiation, and after 1, 2, 5, and 10 years. We included all patients' data on suicidal ideation from the first three assessments in a latent growth mixture modelling analysis to empirically identify trajectories of suicidal ideation. Multivariable logistic regression analyses were applied to estimate associations between trajectories and subsequent suicidality. This trial is registered with ClinicalTrials.gov, number NCT00157313.
FINDINGS: Three trajectories for suicidal ideation were identified in 521 patients. 317 of 521 members (61%) of the largest trajectory, low-decreasing, consisted of patients who initially reported suicidal ideation once-to-a-few-times in the past year followed by a decrease in suicidal ideation. 172 of 521 (33%) members of the frequent-stable trajectory reported suicidal ideation sometimes-to-frequently at the first assessment and remained stable over time. Finally, 32 of 521 (6%) members of the frequent-increasing trajectory reported frequent suicidal ideation initially followed by worsening. Compared with the low-decreasing class, the risk of persistent suicidal ideation was higher for patients in the frequent-stable and frequent-increasing classes at 5 years (Odds ratio=4·5, 95% CI 2·50-8·02 and 4·7, 1·55-14·24 respectively) and at 10 years (4·2, 2·42-7·13 and 3·2, 1·20-8·70 respectively). Also the risk of suicide attempt at 5 years was higher for patients in the frequent-stable (2·8, 1·07-7·43) and frequent-increasing classes (6·6, 1·69-25·54) compared with the low-decreasing class. There was no difference in suicide rate at 5 or 10 years between trajectories.
INTERPRETATION: Because up to 40% of patients experienced frequent suicidal ideations that persisted or increased during the first years of treatment, our findings warrant special focus on suicidal issues as an integral part of treatment along with treatment to target psychotic symptoms.
FUNDING: The Danish Council for Independent Research (Medical Sciences), Trygfonden, The Mental Health Services of the Capital Region of Denmark, The Danish Ministry of Health, The Danish Ministry of Social Affairs, the Psychiatry and Social Service Department in Central Denmark Region, and The Lundbeck Foundation.