Construction and validation of the Dynamic Parkinson Gait Scale (DYPAGS).

The dynamic evaluation of Parkinson's disease (PD)-related episodic gait disturbances in routine is challenging. Therefore, the aim of our study was to assess the reliability/validity of the Dynamic Parkinson Gait Scale (DYPAGS) composed of eight relevant items for the objective quantification of PD gait features: walking forwards/backwards/with dual-task, turning to both sides, imaginary obstacle avoidance with both legs and passing through narrow spaces. The scale was validated on thirty-five patients with mild to severe parkinsonism in their habitual "on-state". A shorter 6 item-version was designed on the basis of a principal component analysis. No significant floor/ceiling effect was detected. The internal consistency was excellent. The levels of interrater agreement, precision and minimal detectable change were adequate. The criterion-related validity was demonstrated by strong correlations with the DYPAGS scores and those at the gait subscales of the Tinetti Mobility Test and MDS-UPDRS. The construct validity was assessed by moderate-strong correlations with the Freezing of Gait Questionnaire, mobility index of the PD Questionnaire (PDQ-39), disease duration and levodopa equivalent daily doses. Statistical analyses using the coefficient of determination showed that both DYPGAS versions were superior to the other instruments to identify patients with gait disturbances with poorer response to dopaminergic treatment. Full and short DYPAGS are reliable instruments for the quantification of "on" PD-related episodic gait disturbances. The full version is sensitive to detect subtle disturbances in mild parkinsonism. The shorter one is easily administered and reliably quantifies gait disturbances in moderate to severe parkinsonism. We recommend their use for research and clinical practice.