Table 3.

Predicting the presence of suicidal behavior at a given follow-up assessment from CORT AUC_i during the TSST, person-standardized stress at the present follow-up assessment, and their interaction

	Outcome: current suicidal behavior
	Model 1: AUCi × Peer Stress		Model 2: AUCi × Academic Stress		Model 3: AUCi × Mother–Child Stress
	γ	SE	γ	SE	γ	SE
Fixed effects
Intercept	−4.13	2.32	−3.56	2.41	−4.22	2.55
Level 2 (between-subject predictors)
Age	−0.10	0.16	−0.16	0.16	−0.13	0.17
Oral contraceptive use	0.26	0.47	0.49	0.51	0.45	0.50
Corticosteroid use	0.53	0.63	0.49	0.65	0.74	0.67
Psychotropic use	−0.79	0.46	−0.83	0.48	−0.71	0.48
Baseline suicidal behavior (y/n)	3.08	0.53	3.25	0.62	3.30	0.65
Mean depression	0.32	0.24	0.49	0.21	0.43	0.21
Mean stressa	0.22	0.24	−0.18	0.19	0.15	0.22
TSST CORT AUCi	−0.38	0.19	−0.43	0.20	−0.50	0.22
Time of day	0.17	0.002	0.19	0.12	0.14	0.11
Level 1 (within-subject predictors)
Time	−0.65	0.11	−0.62	0.12	−0.70	0.14
Current stressa	0.30	0.17	0.33	0.19	0.12	0.21
Current SI (y/n)	4.97	0.70	5.22	0.78	5.35	0.85
Cross-level interaction
TSST CORT AUCi × current stressa	−0.33	0.13	−0.28	0.18	0.06	0.20
Covariance parameters
Intercept	1.24	0.34	0.91	0.63	0.86	0.66
Current stressa	0.083	0.17	0.11	0.46	0.88	1.01
Simple slopes of peer stress on suicidal behavior risk, at high (+ 1 SD) and low (−1 SD) CORT AUCi
	γ	SE	Odds ratio	OR 95% CI
At low TSST CORT AUCi (−1 SD)	0.59	0.23	1.80	1.14–2.84
At high TSST CORT AUCi (+1 SD)	0.01	0.009	1.00	0.63–1.61

Gamma values are analogous to unstandardized beta weights in logistic regression. Significant estimates are bold

^aEach model substitutes a different subscale of the CCSQ stress measure