Suicide Attempts Among Depressed Adolescents in Primary Care

Abstract

Although depression is strongly associated with suicide attempts and suicide deaths, most depressed youth do not make an attempt, indicating the need to identify additional risk factors. We examined suicide attempts among 451 depressed primary care patients, 13 to 21 years of age. In bivariate analyses, youth classified as suicide attempters showed elevated levels of psychopathology, specifically depressive symptoms, externalizing behaviors, anxiety, substance use, mania, and posttraumatic stress disorder symptoms. Externalizing behaviors and depression severity uniquely contributed to the prediction of suicide attempts in multivariate analyses. High levels of environmental stress as well as a few key stressful events were associated with suicide attempts; a recent romantic breakup or being assaulted added to suicide attempt risk, beyond the effects of psychopathology. Implications of results for primary care preventive services and suicide attempt prevention are discussed.

This work was supported by Grant HS09908 to Joan Asarnow from the Agency for Health Care Research and Quality. Additional support for data management and analysis was provided by P30 MH068639 to Kenneth B. Wells, MD, MPH, from the National Institute of Mental Health.

Notes

Note: AA = African American; OR (CI) = odds ratio (confidence interval); CIDI–12, 2.1 = 12-month Composite International Diagnostic Interview Version 2.1.

^a N = 451. Sample includes intervention (n = 418) and pilot study (n = 33) participants.

^b n = 56.

^c n = 395.

^d Ethnicity variables dummy coded with Caucasian as comparison group in logistic regression analyses.

Note: CIDI–12, 2.1 = 12-month Composite International Diagnostic Interview Version 2.1; YSR/YASR = Youth/Young Adult Self-Report; BSI = Brief Symptom Inventory; PC-PTSD = Primary Care PTSD Screen; POSIT = Problem-Oriented Screening Instrument for Teenagers; YRBS = Youth Risk Behavior Survey.

∗p < .05.

∗∗p < .01.

Note: Odd-ratios for psychopathology variables reflect a 1 SD increase and are adjusted for demographic variables. OR (CI) = odds ratio (confidence interval); CIDI–12, 2.1 = 12-month Composite International Diagnostic Interview Version 2.1; YSR/YASR = Youth/Young Adult Self-Report; BSI = Brief Symptom Inventory; PC–PTSD = Primary Care PTSD Screen; POSIT = Problem-Oriented Screening Instrument for Teenagers; YRBS = Youth Risk Behavior Survey.

^a N = 451.

^b n = 56.

^c n = 395.

^d Total N = 450; no attempt n = 394.

Note: Odd ratios for psychopathology variables reflect a 1 SD increase and are adjusted for demographic variables. OR (CI) = odds ratio (confidence interval); YSR/YASR = Youth/ Young Adult Self-Report; CIDI–12, 2.1 = 12-month Composite International Diagnostic Interview Version 2.1.

Note: OR (CI) = odds ratio (confidence interval).

^a N = 451.

^b n = 56.

^c n = 395.

^d n = 450.

Note: OR (CI) = odds ratio (confidence interval); CIDI–12, 2.1 = 12-month Composite International Diagnostic Interview Version 2.1; YSR/YASR = Youth/ Young Adult Self-Report.

¹To further examine the effect of the interaction between depression and romantic breakup, the nonsignificant interaction between externalizing behaviors and romantic breakup interaction terms was dropped from the model.