Table 4.

Logistic Regression ORs and CIs for Predictors

	Comorbid psychopathology vs. Low (c)		High externalizing vs. Low (c)		Comorbid psychopathology vs. High externalizing (c)
Model Predictors	Logit (OR)	95% CI	Logit (OR)	95% CI	Logit (OR)	95% CI
Model 1: Control variables
Race	−.75 (.47)*	[.24, .96]	.20 (1.22)	[.68, 2.17]	−.94 (.39)*	[.18, .87]
Sex	.65 (1.92)†	[.95, 3.87]	.54 (1.72)*	[1.04, 2.85]	.11 (1.12)	[.51, 2.43]
Model 2: Maltreatment dimensions a
Threat only	1.06 (2.88)	[.79, 10.52]	−.33 (.72)	[.21, 2.46]	1.39 (4.01)†	[.84, 19.26]
Deprivation only	1.14 (3.12)*	[1.15, 8.45]	−.25 (.78)	[.35, 1.73]	1.39 (4.00)*	[1.28, 12.48]
Threat and deprivation	.61 (1.84)	[.73, 4.62]	−.32 (.73)	[.39, 1.36]	.93 (2.53)†	[.91, 6.99]
Model 3: Emotion regulation
Lability-negativity	.42 (1.52)	[.80, 2.89]	.56 (1.76)*	[1.05, 2.93]	−.14 (.87)	[.44, 1.73]
Emotion regulation	−.08 (.92)	[.36, 2.35]	.10 (1.10)	[.61, 1.99]	−.18 (.84)	[.30, 2.36]

Note. Race is coded as 1 = Black and 0 = Other. Sex is coded as 1 = male and 2 = female. ‘C’ refers to the comparison group.

^aMaltreatment dimensions were dummy coded and entered simultaneously with the non-maltreated group as the reference category.

The predictors for each model are listed below the model number. Models 2 and 3 also contained control variables (race and sex).

^†p < .10;

^*p < .05;

^**p < .01;

^***p < .001.