All 26 patients completed this prospective cross-sectional cohort study. Patients were median 72 years old (44 – 86 years, 62% [n = 16] male). Median disease duration was 2.25 years (0.5 – 10 years). Except for 4-Aminopyridine and other CNS-active-agents, co-medication was unchanged during the study period (n = 11). On the day of the examination, the patients were not taking any CNS active medication. Controls were median 52 years old (33 – 72 years, 69.2% [n = 9] female).

A linear mixed model of SPV with angle (-30°, 0°, 30°) in patients only found a non-significant effect of angles (p = 0.108) (Fig. 1). Mean SPV values were 2.58 deg/s at -30° (CI 1.85–3.31), 2.38 deg/s at 0° (CI 1.65–3.12) and 2.65 deg/s at 30° (CI 1.92–3.39).

A linear mixed model of SPV (deg/sec) with tilt angle (-30°, 0°, 30°) in patients with DBN. Least-squares means and 95% confidence intervals are shown as black and gray lines

A linear mixed model of the mean SVV with DBN (patients versus controls), angle (-30°, 0°, 30°) and its pair-wise interaction showed no statistically significant interaction (p = 0.243) (Fig. 2). Observed differences in mean SVV between patients and controls were also medically not relevant. Hence, we estimated a main effects model including DBN and angle only. We found a statistical significant difference between angles (p = 0.021), but not for DBN (p = 0.695) (Supplement Table 1).

A linear mixed model of the mean SVV with tilt angle, DBN and its pair-wise interaction showed no statistically significant interaction (p = 0.243). Hence, least-squares means from a main effects model with 95%-CI are shown. (C = controls, P = patients)

Interestingly, mean SVV ratings showed a greater variance in patients than in controls at all head tilt angles, which was most striking at 30° (Fig. 2).