TRANSFERABILITY AND SUSTAINABILITY OF EFFECTS OF DUAL-TASK TRAINING IN PATIENTS WITH DEMENTIA

Abstract

Introduction: Transferability of trained to untrained dual task (DT) performances has not been studied in persons with dementia (PWD). Objective(s): The study objective was to evaluate immediate and longterm transferability of DT performances in PWD. Method(s): Multi-morbid, older PWD (n = 105) were included in a 10-week randomized, controlled trial with the intervention group (IG) performing a DT-based training ('Walking and Counting') while the control group (CG) performed unspecific, low-intensity exercises. DT-performance was measured under 3 conditions: 'Walking and Counting' (trained), 'Walking and Verbal Fluency' (semi-trained) and 'Strength and Verbal Fluency' (untrained). Motor and cognitive performance was documented as single-task (ST) and DT at baseline (T1) after training (T2) and at follow up (T3). Result(s): The trained DT 'Walking and Counting' was significantly improved for motor performances (p<0.001-0.011, effect sizes as partial eta squared eta2 = 0.071-0.249) as well as cognitive performances (p<0.001-0.047, eta2 = 0.044-0.295). Significant transfer effects were found for the semi-trained DT 'Walking and Verbal Fluency' in motor (p<0.001-0.016, eta2 = 0.068-0.150) and partly cognitive performances (p = 0.026, eta2 = 0.059). No significant transfer effects were found for the completely untrained DT 'Strength and Verbal Fluency'. Effects sustained for the trained DT for most motor (p = 0.009-0.038, eta2 = 0.051-0.089) and partly cognitive outcomes (p<0.001, eta2 = 0.157). Transfer-effects could in parts be preserved for semi-trained DTs at follow up. Conclusion(s): Study results document for the first time a transfer from trained to untrained DTs which partly sustained after follow up in patients with moderate stage dementia. With increasing difference in task conditions and time lap to intervention, transfer effects were reduced.