Table 3.
Proportions per group, and odds ratios regarding clinically relevant improvements in SF-36 variables
| Variable | Intervention (n = 47) | Control (n = 54) | Change to top-scoreA | >20 point improvement | |||
|---|---|---|---|---|---|---|---|
| SF-36 score <88 at BL (n (%)) | ImprovedA(n (%)) | SF-36 score <88 at BL (n (%)) | ImprovedA(n (%)) | OR (95 % CI)B | OR (95 % CI)B | Cohen’s kappaC | |
| Physical functioning | 28 (60 %) | 10 (21 %) | 36 (67 %) | 16 (30 %) | 0.64 (0.26–1.60) | 1.17 (0.32–4.31) | 0.79 |
| Role physical | 16 (34 %) | 13 (28 %) | 14 (26 %) | 9 (17 %) | 1.91 (0.73–5.00) | 1.68 (0.67–4.30) | 0.99 |
| Bodily pain | 26 (55 %) | 11 (23.5 %) | 31 (57 %) | 12 (22 %) | 1.07 (0.42–2.71) | 1.34 (0.55–3.24) | 0.87 |
| General health | 39 (83 %) | 6 (13 %) | 41 (76 %) | 15 (28 %) | 0.38 (0.13–1.08) | 1.14 (0.42–3.11) | 0.82 |
| Vitality | 39 (83 %) | 7 (15 %) | 42 (78 %) | 8 (15 %) | 1.01 (0.34–3.02) | 4.32 (1.29–14.55)* | 0.85 |
| 5.03 (1.42–17.85)*D | |||||||
| Social functioning | 14 (30 %) | 12 (25.5 %) | 13 (24 %) | 7 (13 %) | 2.30 (0.82–6.45) | 2.98 (0.72–12.24) | 0.89 |
| Role emotional | 15 (32 %) | 14 (30 %) | 11 (20 %) | 6 (11 %) | 3.39 (1.18–9.74)* | 2.85 (1.04–7.83)* | 0.99 |
| 4.25 (1.37–13.16)*D | 3.46 (1.17–10.23)*D | ||||||
| Mental health | 26 (55 %) | 10 (21 %) | 25 (46 %) | 10 (19 %) | 1.19 (0.45–3.17) | 2.30 (0.82–6.45) | 0.86 |
| PCS | 38 (81 %) | 10 (21 %) | 34 (63 %) | 9 (17 %) | 1.35 (0.50–3.67) | 1.70 (0.61–4.79) | 0.83 |
| MCS | 29 (62 %) | 18 (38 %) | 25 (46 %) | 11 (20 %) | 2.43 (1.00–5.88)* | 5.88 (1.54–22.42)** | 0.83 |
| 2.82 (1.09–7.33)*D | 8.24 (2.03–33.49)**D | ||||||
Note: OR: Odds ratio; PCS: physical component summary; MCS: mental component summary
AChange to top-score: a baseline value <88 and follow-up value ≥88
BNonlinear logit regression with ‘change to top-score’ or ‘>20 point improvement’ as dependent variable and group allocation as independent variable
CAgreement between the ‘change to top-score’ and ‘>20 point change’ classification methods in regard to the proportions of participants displaying clinically significant improvements in the SF-36 scores
DChange in body weight and self-reported leisure time physical activity as continuous covariates (only reported when these two covariates had a significant impact on the OR)
*p < 0.05
**p < 0.01