Table 9.

Regression analysis prediction of a depressive symptom composite from digital biomarker variables, with covariance for age and gender.

		B	Std. Error	β	t	Sig.	R	ΔR²
Model 1	(Constant)	50.803	10.803		4.703	<.001	0.231	0.006
	Age	-0.205	1.155	0.225	1.451	0.155
	Gender	4.394	3.028		1.867	0.069
Model 2 Step 1	(Constant)	25.252	13.525		1.867	0.069	0.460	0.151
	Age	-0.113	1.068	-0.015	-0.106	0.916
	Gender	4.352	2.798	0.223	1.555	0.128
Step 2	LF	0.199	0.071	0.398	2.800	0.008
Model 3 Step 1	(Constant)	29.031	13.010		2.231	0.32	0.549	0.228
	Age	0.697	1.082	0.094	0.644	0.523
	Gender	6.909	2.908	0.354	2.376	0.023
Step 2	LF	0.167	0.069	0.346	2.515	0.016
	Sad.TF:time	-0.987	0.449	-0.343	-2.211	0.033

LF, Low- Frequency Power; Sad.TF:time, The total duration of gazing at sad facial expressions.