Suicidal ideation in multiple sclerosis.

OBJECTIVE

To examine risk factors for suicidal ideation among people with multiple sclerosis (MS).

DESIGN

Cohort study linking computerized medical records with a mailed self-report survey.

SETTING

Veteran's Health Administration (VHA) region covering the northwestern United States.

PARTICIPANTS

VHA patients with MS (N=445) who returned mailed surveys.

INTERVENTIONS

Not applicable.

MAIN OUTCOME MEASURE

Suicidal ideation is assessed by the Patient Health Questionnaire (PHQ) suicide item with suicidal ideation more than half the days considered persistent.

RESULTS

One hundred thirty-one (29.4%) of 445 respondents (95% confidence interval [CI], 25.4%-33.9%) endorsed suicidal ideation, and 35 (7.9%; 95% CI, 5.7%-10.8%) endorsed persistent suicidal ideation over the last 2 weeks. In bivariate analyses, suicidal ideation was associated with younger age, earlier disease course, progressive disease subtype, lower income, not being married, lower social support, not driving, higher levels of physical disability (mobility, bowel, bladder), and depression. Analyses on persistent suicidal ideation yielded similar results. In fully adjusted multivariate logistic regression, only depression severity and bowel disability were independently associated with suicidal ideation. Only depression severity was independently associated with persistent suicidal ideation. By using the 2-question depression screen (U.S. Preventive Services Task Force) consisting of the depression and anhedonia items from the PHQ-9, sensitivity and specificity were marginal for suicidal ideation (65.6% and 79.9%) but acceptable for persistent suicidal ideation (88.6% and 71.2%).

CONCLUSIONS

Suicidal ideation is common among VHA patients with MS, and depression severity is the best risk marker. Brief screening for depression in MS should include the assessment of suicidal ideation.