Table 3.
Estimated means and between-arm differences on primary and secondary outcomes at each follow-up time (N=413). Estimated using linear mixed-effects regression, except for outcomes PA and NA immediately postintervention that were estimated using linear regression. All models were adjusted by gender, K10, and whether participants had searched for Web-based mental health services in the last 2 weeks. Estimates using model with no adjustment for Web-based health services search were similar (not shown).
| Outcome | Link, n=205 | Control, n=208 | Difference | 95% CI | P value | |||
| Mean | 95% CI | Mean | 95% CI | |||||
| Positive affect | ||||||||
|
|
Baselinea | 23.0 | 22.2 to 23.8 | 23.0 | 22.2 to 23.8 | —b | — | — |
|
|
Immediate | 24.4 | 22.1 to 26.7 | 24.1 | 21.8 to 26.3 | 0.3 | −1.1 to 1.8 | .65 |
|
|
1 month | 30.2 | 27.9 to 32.4 | 29.5 | 27.3 to 31.6 | 0.7 | −1.0 to 2.4 | .44 |
|
|
3 months | 31.6 | 29.4 to 33.8 | 30.6 | 28.4 to 32.8 | 1.0 | −0.7 to 2.8 | .24 |
| Negative affect | ||||||||
|
|
Baselinea | 20.4 | 19.6 to 21.3 | 20.4 | 19.6 to 21.3 | — | — | — |
|
|
Immediate | 16.7 | 14.8 to 18.6 | 18.1 | 16.2 to 19.9 | −1.4 | −2.5 to −0.2 | .02 |
|
|
1 month | 13.5 | 11.5 to 15.4 | 16.0 | 14.0 to 18.1 | −2.6 | −4.1 to −1.1 | .001 |
|
|
3 months | 15.6 | 13.6 to 17.5 | 16.1 | 14.1 to 18.0 | −0.5 | −2.0 to 1.0 | .49 |
| Psychological distress (K10) | ||||||||
|
|
Baselinea | 26.2 | 24.1 to 28.3 | 26.2 | 24.1 to 28.3 | — | — | — |
|
|
1 month | 24.2 | 22.0 to 26.5 | 25.2 | 22.9 to 27.4 | −0.9 | −2.4 to 0.6 | .23 |
|
|
3 months | 23.4 | 21.1 to 25.6 | 24.4 | 22.1 to 26.6 | −1.0 | −2.5 to 0.6 | .21 |
| Barriers to seeking help | ||||||||
|
|
Baselinea | 34.9 | 32.7 to 37.2 | 34.9 | 32.7 to 37.2 | — | — | — |
|
|
1 month | 35.4 | 33.0 to 37.9 | 35.0 | 32.6 to 37.5 | 0.4 | −1.3 to 2.0 | .65 |
|
|
3 months | 33.8 | 31.4 to 36.3 | 34.6 | 32.1 to 37.0 | −0.8 | −2.5 to 1.0 | .41 |
aEstimated mean (95% CI) of baseline outcome for the two study arms are the same because they were constrained to be equal in the mixed-effects model.
bNot applicable.