Table 3.
Results of the stratified meta-analyses
| Variable | No. of trials/comparisons | SMD | 95% CI | I2 | Tau2 | P for interaction |
|---|---|---|---|---|---|---|
| All trials, physical function | 33/43 | − 0.50 | − 0.61, − 0.40 | 64% | 0.07 | – |
| Fixed-effect model | − 0.53 | − 0.57, − 0.45 | ||||
| Intervention | 0.08 | 0.42 | ||||
| Pharmacological | − 0.53 | − 0.64, − 0.42 | ||||
| Non-pharmacological | − 0.40 | − 0.74, − 0.05 | ||||
| Intervention | 0.08 | 0.44 | ||||
| Biological treatment | − 0.57 | − 0.68, − 0.46 | ||||
| Other treatment | − 0.46 | − 0.63, − 0.28 | ||||
| All trials, pain | 30/41 | − 0.48 | − 0.66, − 0.30 | 87% | 0.28 | – |
| Fixed-effect model | − 0.50 | − 0.56, − 0.44 | ||||
| Intervention | 0.19 | < 0.001 | ||||
| Pharmacological | − 0.64 | − 0.78, − 0.49 | ||||
| Non-pharmacological | 0.26 | − 0.20, 0.72 | ||||
| Intervention | 0.29 | 0.28 | ||||
| Biological treatment | − 0.64 | − 0.80, − 0-49 | ||||
| Other treatment | − 0.41 | − 0.68, − 0.14 | ||||
| All trials, spinal mobility | 30/45 | − 0.32 | − 0.48, − 0.17 | 83% | 0.21 | – |
| Fixed-effect model | − 0.32 | − 0.38, − 0.26 | ||||
| Intervention | 0.19 | 0.28 | ||||
| Pharmacological | − 0.35 | − 0.51, − 0.19 | ||||
| Non-pharmacological | − 0.02 | − 0.61, 0.56 | ||||
| Intervention | 0.33 | 0.65 | ||||
| Biological treatment | − 0.31 | − 0.47, − 0.15 | ||||
| Other treatment | − 0.28 | − 0.52, − 0.03 | ||||
| All trials, spinal stiffness | 25/34 | − 0.59 | − 0.74, − 0.44 | 75% | 0.14 | – |
| Fixed-effect model | − 0.61 | − 0.68, − 0.54 | ||||
| Intervention | 0.18 | 0.95 | ||||
| Pharmacological | − 0.60 | − 0.76, − 0.44 | ||||
| Non-pharmacological | − 0.61 | − 0.93, − 0.29 | ||||
| Intervention | 0.16 | 0.34 | ||||
| Biological treatment | − 0.77 | − 0.88, − 0.65 | ||||
| Other treatment | − 0.55 | − 0.75, − 0.35 | ||||
| All trials, PGA | 21/28 | − 0.71 | − 0.89, − 0.54 | 83% | 0.18 | – |
| Fixed-effect model | − 0.74 | − 0.81, − 0.67 | ||||
| Intervention | 0.19 | 0.15 | ||||
| Pharmacological | − 0.77 | − 0.96, − 0.59 | ||||
| Non-pharmacological | − 0.37 | − 0.81, 0.07 | ||||
| Intervention | ||||||
| Biological treatment | − 0.84 | − 1.09, − 0.60 | 0.20 | 0.23 | ||
| Other treatment | − 0.60 | − 0.87, − 0.34 | ||||
| All trials, PJ/E | 15/15 | 0.05 | − 0.11, 0.22 | 68% | 0.06 | – |
| Fixed-effect model | 0.00 | − 0.08, 0.09 | ||||
| Intervention | 0.00 | < 0.001 | ||||
| Pharmacological | − 0.06 | − 0.15, 0.03 | ||||
| Non-pharmacological | 0.99 | 0.63, 1.35 | ||||
| Intervention | 0.04 | 0.02 | ||||
| Biological treatment | − 0.10 | − 0.20, − 0.01 | ||||
| Other treatment | 0.27 | − 0.10, 0.64 | ||||
| All trials, acute phase reactants | 27/31 | − 0.51 | − 0.70, − 0.32 | 84% | 0.22 | – |
| Fixed-effect model | − 0.62 | − 0.69, − 0.55 | ||||
| Intervention | – | – | ||||
| Pharmacological | − 0.51 | − 0.70, − 0.32 | ||||
| Non-pharmacological | – | – | ||||
| Intervention | 0.12 | 0.001 | ||||
| Biological treatment | − 0.77 | − 1.02, − 0.52 | ||||
| Other treatment | − 0.22 | − 0.37, − 0.07 | −0.07 | |||
| All trials, spine radiographs | 1/1 | 0.96 | 0.22, 1.69 | – | – | – |
| Fixed-effect model | – | – | ||||
| Intervention | – | – | ||||
| Pharmacological | – | – | ||||
| Non-pharmacological | – | – | ||||
| Intervention | ||||||
| Biological treatment | ||||||
| Other treatment | 0.96 | 0.22, 1.69 | ||||
| All trials, fatigue | 3/4 | − 0.65 | − 0.82, − 0.48 | 0% | 0.00 | – |
| Fixed-effect model | − 0.65 | − 0.82, − 0.48 | ||||
| Intervention | – | – | ||||
| Pharmacological | − 0.65 | − 0.82, − 0.48 | ||||
| Non-pharmacological | – | – | ||||
| Intervention | – | – | ||||
| Biological treatment | − 0.65 | − 0.82, −0.48 | ||||
| Other treatment |
SMD standardized mean difference, CI confidence interval, I2 inconsistency (i.e., the percentage of total variation across studies due to heterogeneity), Tau2 tau squared is an estimate of the variance of the true effect sizes, PJ/E peripheral joint count/enthesitis index