Severity of Suicidal Ideation Matters: Reexamining Correlates of Suicidal Ideation Using Quantile Regression

Objective Numerous risk factors have been identified for suicidal ideation, including perceived burdensomeness, thwarted belongingness, agitation, insomnia, nightmares, cognitive anxiety sensitivity, and rumination. However, the complexity of these associations has not been well studied; the magnitude of these effects may vary at differing levels of suicidal ideation. The present study reexamined established risk factors for suicidal ideation using quantile regression, a statistical technique that calculates the effect at numerous quantiles of suicidal ideation, as opposed to the average effect across all quantiles. Method A sample of 354 psychiatric outpatients (61.3% female, mean age = 27.01 years, standard deviation = 10.40) completed self‐report measures of their suicidal ideation and related risk factors prior to their initial intake appointments. Results The relationship between each suicide risk factor and suicidal ideation was strongest at higher (.9 quantile), as opposed to nonexistent (.5 quantile) and low‐moderate (.7 quantile), levels of suicidal ideation. The interaction proposed by the interpersonal theory of suicide was significant at nonexistent and low‐moderate, but not high, levels of suicidal ideation. Conclusion Our findings indicated that predictors of suicidal ideation differed in magnitude at varying levels of suicidal ideation. Implications, limitations, and future research directions are discussed.