Table 4.

Results of linear regression analyses predicting outcomes from usage group (intended vs minimal/low users) and separate usage measures of intervention users (N=89).

Variablea		Full modelb						Predictor
		Adjusted R2		F (df)		P		B		t (df)		P		95% CI
Psychological distress
	Adherence (minimal/low vs intended users)		0.34		23.76 (2,86)		<.001		−2.87		−2.47 (86)		.02		−5.19 to −0.56
	Number of logins		0.32		20.46 (2,86)		<.001		−0.03		−1.23 (86)		.22		−0.07 to 0.02
	Mean time logged in		0.31		20.69 (2,86)		<.001		−0.04		−1.35 (86)		.18		−0.09 to 0.02
	Total time logged in		0.32		20.40 (2,86)		<.001		−0.001		−1.20 (86)		.23		−0.01 to 0.00
	Emails sent		0.32		19.62 (2,83)		<.001		−0.06		−0.60 (83)		.55		−0.27 to 0.14
	All exercises completed (yes/no)		0.34		24.48 (2,86)		<.001		−2.80		−2.51 (86)		.01		−5.02 to −0.59
Positive mental health
	Adherence (minimal/low vs intended users)		0.49		41.86 (2,86)		<.001		5.18		2.40 (86)		.02		0.88 to 9.48
	Number of logins		0.45		37.28 (2,86)		<.001		0.03		0.90 (86)		.37		−0.04 to 0.11
	Mean time logged in		0.46		36.77 (2,86)		<.001		0.03		0.50 (86)		.62		−0.08 to 0.13
	Total time logged in		0.45		36.64 (2,86)		<.001		0.00		0.33 (86)		.74		−0.01 to 0.01
	Emails sent		0.46		34.77 (2,83)		<.001		0.05		0.28 (83)		.78		−0.32 to 0.43
	All exercises completed (yes/no)		0.51		45.51 (2,86)		<.001		5.24		2.53 (86)		.01		1.12 to 9.36

^aOf the 108 users, 19 missed either the premeasure or the postmeasure of the outcomes, resulting in a sample size of 89 for these analyses. Three participants missed a score on emails sent, resulting in a sample size of 86 for these analyses.

^bThe full models were controlled for baseline levels of psychological distress or positive mental health.