Table 2.

Bivariate and multivariate logistic regression in patients under 18 years of age with suicidal behavior.

	Not suicidal behavior (N = 251)	Yes suicidal behavior (N = 91)	p-Value	Univariate OR	CI
Females (%)	149 (59.6)	80 (87.9)	<0.001*	4.93	2.50–9.72
Age (SD)	15.4 (1.6)	15.2 (1.6)	0.402	ns
Marital status (%)
Single	239 (97.6)	84 (96.6)	0.295	ns
Married/coupled	6 (2.4)	3 (3.4)	0.388	ns
Living
Alone	3 (1.2)	0	0.568	ns
With relatives	185 (74)	84 (92.3)	0.003*	4.28	1.88–9.72
Institution	62 (24.8)	7 (7.7)	<0.001*	0.25	0.11–0.58
Previous follow-up (%)	214 (85.3)	78 (85.7)	0.535	ns
First visit period				ns
Before lockdown	35 (13.9)	11 (12.1)	0.371	ns
During first lockdown	36 (14.3)	15 (16.5)	0.212	ns
During second lockdown	180 (71.7)	65 (71.4)	0.089	ns
Diagnosis axis I (%)
Childhood onset disorder	82 (32.7)	7 (7.7)	<0.001*	0.19	0.07–0.45
Illegal drugs abuse	18 (7.2)	1 (1.1)	0.031*	ns
Psychotic disorders	9 (3.6)	1 (1.1)	0.206	ns
Depressive disorders	35 (13.9)	51 (56)	<0.001*	7.86	4.55–13.59
Bipolar disorder	1 (0.4)	0	1.000	ns
Eating disorders	26 (10.4)	6 (6.6)	0.401	ns
Impulse control deficit	22 (8.8)	3 (3.3)	0.102	ns
Tr adjustment disorders	21 (8.4)	5 (5.5)	0.491	ns
Anxiety disorders	24 (9.6)	4 (4.4)	0.179	ns
None	13 (5.2)	12 (13.2)	0.018*	2.78	1.21–6.34
Diagnosis axis II (%)
None	202 (80.8)	61 (67)	0.013*	0.493	0.28–0.84
Cluster B	45 (18)	30 (33)	0.005*	2.251	1.30–3.87
Cluster C	3 (1.2)	0	0.568	ns
Hospital admission (%)	72 (28.6)	52 (57.1)	0.001*	–	–

Forward stepwise logistic regression model
	Variable	Wald X2a	p-Value	OR corrected	CI
First step	Depressive disorders	54.43	<0.001*	7.83	4.53–13.53
Second step	Female	10.79	0.001	3.31	1.62–6.77
	Depressive disorders	40.96	<0.001*	6.26	3.57–0.97
Third step	Female	8.16	0.004	2.88	1.39–5.98
	Living with relatives	7.56	0.006	3.49	1.43–8.54
	Depressive disorders	40.11	<0.001*	6.34	3.58–11.24

^aThe Hosmer–Lemeshow test was not significant for the third step. (X² = 2395, df = 5, p = 0.792).

* p<0.05.