Table 2.

Two regression models for % change in BFMDRS-m that were created in a data-driven approach. A full model containing all covariates is available in the Supplementary Material (Table S3).

	Effect	95% CI	p
Model 1 (89 patients with CMCT measurements and BFMDRS outcomes)
(Intercept)	34.621	[24.126; 45.117]	<0.00001
Acquired dystonia	−29.234	[−41.794; −16.673]	0.00002
SEP abnormal	−15.231	[−27.792; −2.671]	0.01850
Model 2 (51 patients with SEP measurements and BFMDRS outcomes)
(Intercept)	35.964	[27.741; 44.187]	<0.00001
Acquired dystonia	−30.513	[−40.338; −20.689]	<0.00001
CMCT abnormal	−16.163	[−29.886; −2.440]	0.02150