Validity, test-retest reliability, sensitivity to change and feasibility of motor-cognitive dual task assessments in patients with dementia

Objective: To investigate validity, test-retest reliability, sensitivity to change, and feasibility of dual task (DT) assessments in patients with dementia.

Design: Validation study.

Setting: Post ward-rehabilitation.

Participants: Geriatric patients (n=105) with dementia (age 82.7±5.9, MMSE score 21.9).

Main outcome measures: Psychometric quality of DT performance of different DT-tests. Analyses were performed for motor and cognitive performance, and relative DT costs (DTCs).

Results: Spearman's rank correlations (r_s) between examined DT-tests were moderate-high for motor tasks (r_s=0.29-0.90), small-high for cognitive tasks (r_s=0.12-0.55) and small-high for relative DTCs (motor DTCs r_s=0.02-0.61, cognitive DTCs r_s=-0.19 to 0.06, combined DTCs r_s=-0.11 to 0.31). Correlations with external assessment were moderate-high for motor tasks (r_s=0.25-0.84), small-moderate for cognitive tasks (r_s=-0.10 to 0.46) and small-moderate for relative DTCs (motor DTCs r_s=-0.09 to 0.17, cognitive DTCs r_s=-0.03 to 0.21, combined DTCs r_s=-0.07 to 0.26). Test-retest reliability was excellent for motor tasks (ICC=0.75-0.96), fair-excellent for cognitive tasks (ICC=0.51-0.88) and poor-good for relative DTCs (motor DTCs ICC=0.10-0.74, cognitive DTCs ICC=0.05-0.65, combined DTCs ICC=0.15-0.71). Sensitivity to change was acceptable-excellent for trained DT-tests (p≤0.01). Effect sizes were small-large for gait parameters (SRM=0.30-1.12), large for cognitive tasks (SRM=0.82-0.95) and small-large for relative DTCs (motor DTCs SRM=0.15-0.77, cognitive DTCs SRM=0.56-0.98, combined DTCs SRM=0.40-1.10). Completion time ranged from 13.1 to 16.9min.

Conclusions: All DT-tests showed acceptable-excellent psychometric properties in patients with dementia with highest quality for the gait-based tests 'Walking & Counting' and 'Walking & reciting ABC'.