Table 4.

Multilevel models examining the associations of relevant treatment outcomes and physiological reactions entered as time-varying covariates after controlling for medications

Predictors	Dependent Variables
	Nights/Wk w/NM		NM Severity		Sleep Quality (PSQI)		Hours of Sleep		Fearful of Sleep		Health Symptoms (PILL)		CAPS past Month		Depression (TSI)
	Est.	SE	Est.	SE	Est.	SE	Est.	SE	Est.	SE	Est.	SE	Est.	SE	Est.	SE
Intercept	1.08	0.31	1.79	0.17	5.89	0.85	7.20	0.22	0.41	0.17	61.89	7.13	24.34	4.89	6.60	1.22
Corr EMG	1.10	1.32	-0.72	0.83	1.68	2.65	-1.08	0.86	0.32	0.62	28.75	12.22	19.34	15.34	5.40	3.03
LF EMG	1.03	4.13	5.03	2.55	1.65	8.72	2.07	2.72	1.04	1.98	36.77	40.57	30.77	49.34	10.86	9.92
HR	0.07	0.04	0.02	0.02	0.20	0.08	-0.08	0.02	0.00	0.02	0.07	0.35	0.52	0.44	-0.02	0.09
SC	0.37	0.22	0.35	0.14	1.67	0.46	-0.18	0.15	0.41	0.11	4.13	2.21	9.79	2.66	0.68	0.53
Δχ2 =	19.91		20.87		51.41		35.58		33.08		24.57		38.73		21.04
η2 =	0.16		0.17		0.37		0.19		0.23		0.22		0.30		0.11

Positive associations with PSQI sleep quality indicate that reductions in physiological responses are associated with improvements in sleep quality (higher scores on the PSQI indicate poorer sleep quality). All models with a Δχ² > 15.51 were statistically significant (i.e., were significant improvements over an intercept only model). η², eta squared effect size (proportion of variance explained in dependent variable). Corr, corrugator muscle. LF, lateralis frontalis muscle. HR, heart rate. SC, skin conductance. EMG, electromyogram. Est., unstandardized estimate of fixed effect (relationship between predictor and criterion). SE, standard error for the parameter estimate. Estimates in bolded text are significant at p < 0.05.