Table 4. Predicted pain effect values using the multiple linear regression model for significant and non-significant studies.
| Intercept | FREQUENCY (x/wk) | TIME (min/wk) | DURATION (wks) | Dur*Time | Dur*Freq | Pain Effect Sig | Pain Effect Not Sig | ||
|---|---|---|---|---|---|---|---|---|---|
| Coefficient | 0.374 | -0.36 | 0.01 | -0.03 | -0.001 | 0.04 | 0.743 | 0 | |
| Change FREQ by 1 day | |||||||||
| Less 2 day | 1 | 120 | 15 | 1800 | 15 | 0.307 | -0.436 | ||
| Less 1 day | 2 | 120 | 15 | 1800 | 30 | 0.547 | -0.196 | ||
| Average Values | 3 | 120 | 15 | 1800 | 45 | 0.787 | 0.044 | ||
| Add 1 day | 4 | 120 | 15 | 1800 | 60 | 1.027 | 0.284 | ||
| Add 2 day | 5 | 120 | 15 | 1800 | 75 | 1.267 | 0.524 | ||
| Add 3 day | 6 | 120 | 15 | 1800 | 90 | 1.507 | 0.764 | ||
| Change TIME by 30 min | |||||||||
| Less 1.5 hr | 3 | 30 | 15 | 450 | 45 | 1.237 | 0.494 | ||
| Less 1 hr | 3 | 60 | 15 | 900 | 45 | 1.087 | 0.344 | ||
| Less 0.5 hr | 3 | 90 | 15 | 1350 | 45 | 0.937 | 0.194 | ||
| Average Values | 3 | 120 | 15 | 1800 | 45 | 0.787 | 0.044 | ||
| Add 0.5 hr | 3 | 150 | 15 | 2250 | 45 | 0.637 | -0.106 | ||
| Add 1 hr | 3 | 180 | 15 | 2700 | 45 | 0.487 | -0.256 | ||
| Add 1.5 hr | 3 | 210 | 15 | 3150 | 45 | 0.337 | -0.406 | ||
| Change DURATION by 1 wk | |||||||||
| Less 6 wk | 3 | 120 | 9 | 1080 | 27 | 0.967 | 0.224 | ||
| Less 4 wk | 3 | 120 | 11 | 1320 | 33 | 0.907 | 0.164 | ||
| Less 2 wk | 3 | 120 | 13 | 1560 | 39 | 0.847 | 0.104 | ||
| Average Values | 3 | 120 | 15 | 1800 | 45 | 0.787 | 0.044 | ||
| Add 2 wk | 3 | 120 | 17 | 2040 | 51 | 0.727 | -0.016 | ||
| Add 4 wk | 3 | 120 | 19 | 2280 | 57 | 0.667 | -0.076 | ||
| Add 6 wk | 3 | 120 | 21 | 2520 | 63 | 0.607 | -0.136 |
The last two columns show predicted pain effect sizes for significant studies versus non-significant studies as a result of the exercise prescription shown in the second column. The highlighted cells indicate the variables being changed compared to the average values from the included studies.