Table 3.

Baron and Kenny, Sobel and bootstrapping analyses for sleep disturbances as a mediator of the association between RA and QST outcomes.^†

QST Measure	Baron and Kenny Criteria			Sobel’s Test (P)	Boot-strapping (95% CI for the indirect effect of RA on QST measures, through sleep)
	Association between RA and QST (P)	Association between RA and sleep problems (P)	Association between sleep problems and QST, adjusted for RA (P)
Central Pain Measure
CPM	0.04	<0.0001	0.03	0.04	−0.74, −0.04
Pain Threshold
Left Wrist	0.05	<0.0001	0.49	0.49	−0.76, 1.22
Right Wrist	0.03	<0.0001	0.83	0.83	−1.13, 0.83
Left Knee	0.10	<0.0001	0.63	0.63	−0.67, 0.92
Right Knee	0.01	<0.0001	0.65	0.66	−0.72, 1.02
Pain Tolerance
Left Wrist	0.02	<0.0001	0.41	0.42	−0.58, 1.04
Right Wrist	0.02	<0.0001	0.89	0.89	−0.79, 0.77
Left Knee	0.03	<0.0001	0.98	0.98	−0.78, 0.65
Right Knee	0.001	<0.0001	0.67	0.67	−0.64, 0.88

^†Each row is a separate model. RA: rheumatoid arthritis, QST: quantitative sensory testing, CPM: conditioned pain modulation.