TABLE 3.

Summary of Hierarchical Logistic Regression Analyses Predicting Suicide Attempt Status

Variable	β	SE	Adjusted Odds Ratio	95% Confidence Interval	P Value
Model 1 adding parent-reported sleep data at step 2
Step 1
Antidepressant use	2.50	0.75	12.18	(2.78, 53.37)	.001
Antipsychotic use	1.14	0.75	3.13	(0.72, 13.58)	.128
Victim of bullying	0.73	0.73	2.08	(0.50, 8.73)	.315
Affective disorder	0.10	0.75	1.11	(0.26, 4.80)	.893
Step 2a
Child waking up a lot at night	−0.08	1.00	0.93	(0.13, 6.62)	.939
Step 2b
Total sleep score	0.77	0.38	2.15	(1.01, 4.57)	.046
Model 2 adding self-reported sleep data at step 2
Step 1
Antidepressant use	1.97	0.62	7.17	(2.12, 24.26)	.002
Antipsychotic use	1.07	0.66	2.91	(0.80, 10.52)	.104
Victim of bullying	1.08	0.67	2.94	(0.80, 10.85)	.105
Affective disorder	−0.24	0.63	0.78	(0.23, 2.70)	.699
Step 2c
Waking up a lot at night	1.42	0.64	4.13	(1.19, 14.35)	.026
Trouble getting back to sleep	0.37	0.65	1.44	(0.40, 5.18)	.575
Step 2d
Total sleep score	0.35	0.19	1.42	(0.97, 2.06)	.070

^aIncludes all variables in Step 1 of model + Parent-reported child wakes up a lot at night

^bIncludes all variables in Step 1 of model + Parent-reported total sleep problems score

^cIncludes all variables in Step 1 of model + Self-reported waking up a lot at night and Self-reported trouble getting back to sleep

^dIncludes all variables in Step 1 of model + Self-reported total sleep problems score