Table 2.

Primary and secondary efficacy outcomes^a.

Efficacy outcomes	Ubiquinol	Placebo	Least squares mean differences (ubiquinol−placebo) (95% confidence interval)	P value
Number of patients	61	68
Primary efficacy outcomebc
UMSARS part 2
Mean (±SD) baseline score	23.2 ± 8.7	23.3 ± 7.9
Change from baseline at 48 weeks (LS mean ± SE)	5.4 ± 0.7	7.1 ± 0.6	−1.7 (−3.2, −0.2)	0.023
Secondary efficacy outcomesbd
UMSARS part 1
Mean (±SD) baseline score	20.5 ± 7.0	20.5 ± 5.9
Change from baseline at 48 weeks (LS mean ± SE)	4.7 ± 0.8	4.9 ± 0.7	−0.2 (−2.1, 1.7)	0.864
Barthel index
Mean (±SD) baseline score	79.4 ± 22.6	76.5 ± 20.3
Change from baseline at 48 weeks (LS mean ± SE)	−16.5 ± 2.9	−21.8 ± 2.7	5.3 (−2.0, 12.6)	0.150
SARA
Mean (±SD) baseline score	18.5 ± 6.8	18.0 ± 4.9
Change from baseline at 48 weeks (LS mean ± SE)	4.0 ± 0.6	5.3 ± 0.6	−1.4 (−2.8, 0.1)	0.067
Time required to walk 10 m (sec)
Mean (±SD) baseline (sec)	35.7 ± 16.1	27.3 ± 42.8
Change from baseline at 48 weeks (LS mean ± SE)	31.9 ± 16.1	55.9 ± 15.0	−24.1 (−67.7, 21.7)	0.273

^aThe full analysis set population included all the patients who had undergone randomisation and had completed efficacy assessment at least once after the baseline.

^bThe least squares men changes from baseline between the ubiquinol and the placebo groups at week 48 were estimated using a mixed model for repeated measures. Treatment group, visit, and treatment-by-visit interaction were included as fixed effects and with adjustment for the baseline value of each endpoint, presence of the V393A variants in COQ2, and disease subtype (MSA-C or MSA-P). There was no imputation of missing data for the primary efficacy outcome or secondary efficacy outcome. The results for secondary efficacy outcomes were not corrected for multiple comparison.

^cParameters were converged by the covariance structure AR (1) with random effects.

^dParameters were converged by the covariance structure UN with no random effects.