Outcomes of the per-protocol analysis, objective sickness absence measured at 12 and 18 months of follow-up and questionnaire data measured at baseline (T0), 6 (T1) and 12 months (T2) of follow-up.
| Variable | Intervention group, ‘never started’ group excluded (n=46) | Control group (n=70) | β, RR, HR§ (95% CI) | p Value |
| Objective sickness absence 0–12* | ||||
| At least once on sick leave (N, %)‡ | 34 (73.9) | 48 (68.6) | 1.08 (0.85 to 1.36) | 0.677 |
| Duration (calendar days); mean (SD)/median | 24.29 (34.18)/11.00 | 50.83 (75.75)/15.00 | −0.74 (−1.37 to −0.11) | 0.021 |
| Frequency; mean (SD)/median | 2.18 (1.47)/2.00 | 2.35 (1.78)/2.00 | −0.08 (−0.39 to 0.23) | 0.621 |
| Time to onset of first sickness absence spell (calendar days); mean (SD) | 165.67 (143.77) | 187.37 (146.42) | 1.18 (0.76 to 1.82) | 0.473 |
| 0–18† | ||||
| At least one time on sick leave (N, %)‡ | 39 (84.8) | 56 (80.0) | 1.06 (0.89 to 1.26) | 0.625 |
| Duration (calendar days); mean (SD)/median | 33.05 (51.18)/15.00 | 62.57 (81.89)/25.50 | −0.64 (−1.18 to −0.10) | 0.021 |
| Frequency; mean (SD)/median | 2.69 (1.89)/2.00 | 2.93 (2.17)/2.00 | −0.08 (−0.37 to 0.20) | 0.569 |
| Time to onset of first sickness absence spell (calendar days); mean (SD) | 200.76 (198.97) | 231.19 (206.57) | 1.18 (0.78 to 1.77) | 0.441 |
| Participants with sickness absence spell >28 calendar days; N (%)‡ | ||||
| 0–12* | 6 (13.0) | 17 (24.3) | 0.54 (0.23 to 1.26) | 0.160 |
| 0–18† | 7 (15.2) | 22 (31.4) | 0.48 (0.23 to 1.04) | 0.078 |
| Depressive complaints according to Beck Depression Inventory-II (0–63)¶; Mean (SD) | ||||
| Baseline T0 | 18.15 (9.42) | 14.84 (8.11) | ||
| T1 | 14.00 (11.50) | 16.30 (10.08) | ||
| T2 | 12.31 (10.01) | 16.69 (11.04) | ||
| T1–T0 | –4.50 (7.59) | 0.93 (7.06) | −5.02 (−8.34 to −1.70) | 0.004 |
| T2–T0 | –4.66 (8.76) | 2.09 (9.51) | −6.05 (−10.34 to −1.75) | 0.006 |
↵* Measured from baseline until 12 months of follow-up.
↵† Measured from baseline until 18 months of follow-up.
↵‡ Dichotomous variable (no/yes) with N (%) indicating ‘yes.’
↵§ β (95% CI) in case of continuous variable and RR (95% CI) in case of dichotomous variable and HR in case of time to onset of first sickness absence spell.
↵¶ Scale range.