Temporal relationships between changes in depression and suicidal ideation: A mediation analysis in a randomized double‐blinded clinical trial

Abstract Objectives There is a considerable debate regarding the possible dependence between depression and suicidal ideation treatments. The present study used a novel mediation approach in a randomized comparison of pharmacotherapy and combined therapy to explore whether depressive symptoms mediate the association between treatment and suicidal ideation and whether it depends on the treatment condition. Design This study is a randomized, controlled, parallel group (1:1), clinical trial using a novel mediation approach for longitudinal data. Latent difference score modelling was utilized to investigate whether changes in depressive symptoms drive subsequent changes in suicide ideation. Method Participants were 94 depressive suicidal outpatients who were assessed regarding depressive symptoms and suicidal ideation over the course of an experiment (0–2‐7 months). Direct and indirect associations between (change in) depressive symptoms and (change in) suicidal ideation were explored using Pearson's correlations and latent difference score model. Results The results showed that depression treatment affects not only suicidal ideation directly but also its influence on suicidal ideation occurs via improvement in depressive symptoms. It was found a more significant effect of combining pharmacotherapy and PPT (in comparison with the pharmacotherapy alone) on the early and late improvements of suicidal ideation (Δ 0–2 and Δ 2–7) via the early improvement of depressive symptoms (Δ 0–2). Conclusions The findings indicate that changes in depressive symptoms preceded changes in suicidal ideation. Our results highlighted that improving depressive symptoms could be a primary target in treating patients with depression experiencing suicidal thoughts.