Sensor-based postural feedback is more effective than conventional feedback to improve lumbopelvic movement control in patients with chronic low back pain: a randomised controlled trial

Table 3 Results of the linear regression analysis and post-hoc tests for type of feedback

Linear regression		Post-hoc multiple comparisons for type of FB
Fixed effects	p-value	Comparison	Estimated differences between groups (95% CI)	p-value
Waiter’s bow
Initial model
Health status	0.09
Type of FB	< 0.0001
Joint	0.01
Baseline score kinematics	0.06
Health status*type of FB	0.61
Health status*Joint	0.71
Type of FB*Joint	0.94
Final model
Type of FB	< 0.0001	Mirror minus Control	0.6 (− 3.1 to 4.4)	0.91
Joint	0.04	Sensor minus Control	10.6 (6.8 to 14.3)	< 0.0001^a
Joint		Sensor minus Mirror	9.9 (6.1 to 13.7)	< 0.0001^a
Lifting task
Initial model
Health status	0.20
Type of FB	< 0.0001
Joint	0.029
Baseline score kinematics	0.003
Health status*type of FB	0.65
Health status*Joint	0.44
Type of FB*Joint	0.57
Final model
Type of FB	< 0.0001	Mirror minus Control	−0.3 (− 3.7 to 3.0)	0.97
Joint	0.02	Sensor minus Control	6.9 (3.5 to 10.2)	< 0.0001^a
Baseline score kinematics	0.002	Sensor minus Mirror	7.2 (3.8 to 10.6)	< 0.0001^a

FB Feedback, Health status healthy of CLBP, Joint lumbar spine or hip, Type of FB sensor, mirror or control
^ain favour of the sensor group

ISSN: 1743-0003