Table 1 A description of the applied regression models for Aim A
Aim A objective | Method | Fixed dependent effect | Fixed independent effect | Random intercept | Reported as |
|---|---|---|---|---|---|
Changes in PPT over time | Linear mixed regression | PPT | Model 1: Time Model 2: Interaction between time and test site | Subject | Mean change value with 95% CI, both dependent and independent of the test site |
(I) Distance from test site to SMT site | Linear mixed regression | PPT change | Distance between the tested vertebrae and the closest SMT site (e.g., PPT test at C7 and SMT at C1 = distance of 6) | Subject + PPT at baseline + region of pain | A Beta coefficient with a 95% CI. A scatter plot with a best fitting regression line |
(II) Rapid responder status | Linear regression | PPT change | Responder status (e.g., much better) | – | Between group mean differences for each rapid responder status change values with 95% CI |
(III) The baseline PPT value | Linear regression | PPT change | The baseline PPT value (e.g., 5.5 kg) | – | A Beta coefficient with a 95% CI |
(IV) Number of SMTs performed | Linear regression | PPT change | Number of SMTs performed (e.g., 3 SMTs) | – | A Beta coefficient with a 95% CI |
(V) Region of pain compared to the adjacent or distant region* | Linear mixed regression | PPT change | Pain region (i.e., change in PPT in the pain region (e.g., cervical) compared to change in PPT in the i) adjacent region (e.g., thoracic) or ii) distant region (e.g., lumbar)) | Subject | Between group mean differences for the pain region and adjacent/distant region with 95% CI |
(VI) Other non-SMT treatment provided | Linear regression | Non SMT-treatment provided. Categorized as myofascial (compression technique or dryneedling), muscle energy technique, massage, other (Laser or free text field), and none | – | Between group mean differences of the different non-SMT options with 95% CI |