Skip to main content
. Author manuscript; available in PMC: 2014 Jan 9.
Published in final edited form as: JAMA Psychiatry. 2013 Mar;70(3):10.1001/2013.jamapsychiatry.55. doi: 10.1001/2013.jamapsychiatry.55

Table 4.

Bivariate and multivariate associations (odds-ratios) of temporally primary DSM-IV/CIDI disorders with first onset of lifetime suicide attempt in the total sample (n = 6,483)1

Bivariate model2
Multivariate model2
OR (95% CI) OR (95% CI)
I. Fear/anger disorders
 Specific phobia 2.5* (1.7–3.8) 1.2 (0.8–1.8)
 Panic disorder and/or agoraphobia 1.9 (0.9–4.3) 1.1 (0.4–3.1)
 Social phobia 2.5* (1.4–4.3) 0.8 (0.4–2.0)
 Intermittent explosive disorder 3.5* (1.7–7.5) 1.7 (0.8–3.6)
II. Distress disorders
 Separation anxiety disorder 0.8 (0.3–1.7) 0.3* (0.1–0.7)
 Post-traumatic stress disorder 7.3* (3.7–14.4) 3.3* (2.0–5.5)
 Major depressive disorder or dysthymia 12.3* (8.0–19.0) 6.2* (3.8–10.0)
 Generalized anxiety disorder 4.1* (1.4–11.8) 0.9 (0.2–3.4)
III. Disruptive behavior disorders
 Attention-deficit hyperactivity disorder 4.8* (2.9–8.0) 1.9 (1.0–3.7)
 Oppositional-defiant disorder 7.2* (4.0–12.8) 2.1* (1.2–3.6)
 Conduct disorder 4.5* (2.6–7.9) 1.0 (0.5–2.2)
 Any eating Disorder 7.4* (4.1–13.6) 3.2* (1.5–7.0)
IV. Substance abuse3
 Alcohol abuse 2.8* (1.2–6.3) 1.1 (0.5–2.4)
 Illicit drug abuse 4.8* (2.5–9.1) 1.3 (0.5–3.3)
V. Other disorders
 Bipolar I or II 8.8* (3.8–20.1) 2.9* (1.0–7.9)
*

Significant at the .05 level, two-sided test.

1

Results are based on multivariate discrete-time survival models with person-year the unit of analysis, a logistic link function, and person-year defined as a separate dummy predictor variables for each year of life. Both models control for the socio-demographic variables in Table 2. Lifetime predictor disorders are coded as present only if their AOO is less than or equal to the age when the respondent made his or her first suicide attempt.

2

The bivariate model includes only one mental disorder in each equation, whereas the multivariate model includes all 15 mental disorders in the same equation. The addition of the 15 disorders to the predictor set in the multivariate model improves on the socio-demographic model in Table 2 (Akaike Information Criterion [AIC] = 2862.4 for the socio-demographic model and 2354.8 for the model that added the 15 mental disorders; the model with the lower AIC being the preferred model).

3

With or without a history of dependence