Can people with Parkinson’s disease improve dual tasking when walking? Sandra G. Brauer a, *, Meg E. Morris b a Division of Physiotherapy, School of Health and Rehabilitation Sciences, The University of Queensland, Building 84A, Brisbane, Qld 4072, Australia b School of Physiotherapy, The University of Melbourne, Carlton, Australia 1. Introduction When asked to perform a concurrent task when walking, people with Parkinson’s disease (PD) exhibit reduced step length [1–4], gait velocity [1–5], and increased gait variability [6]. Previously, people with PD have been recommended to avoid dual tasks [7,8]. Galletly and Brauer [4] however, demonstrated that people with PD could use visual cues to improve stride length while concurrently performing an added task, suggesting that training dual tasking may be possible. Further support is evident from Canning et al. [9] who reported that a pilot program of multi-task walking training in five people with PD led to an improvement in multi-task walking speed and low levels of fatigue, difficulty and anxiety. One important issue to consider in the training of dual tasking during gait is the balance between specificity and variability of the skill practice and context. This influences ability to transfer skills [10]. It is not feasible for dual task training programs to include all tasks likely to be performed when walking in daily life. Thus it is important to determine if training dual tasking with one type of task known to interfere with gait results in improvements in dual tasking when performing other types of tasks. Previous research indicates that older adults can improve dual tasking with training, and the skills they learn are able to be retained, and transferred to a novel dual task situation [11,12]. Research on skill acquisition during dual task learning situations indicates that success in transferring skills may be dependent on several factors [12–14]. Specificity of practice is important in the acquisition of skills under dual task situations [15], with the type of tasks practiced able to influence dual task learning and transfer of training in older adults [12]. In people with PD, the complexity [16] and type of secondary task [4] performed when walking has been shown to influence dual task interference. A recent systematic review has found that people with PD demonstrate a reduced ability to perform visuospatial tasks [17]. Furthermore, the domain of the information being processed (e.g. spatial items vs. letters) can also influence performance in people with PD [18]. Thus, there is potential that the type of task or domain of information being processed in the second task may influence the transferability of dual task training during gait. There were two aims of this study. The first was to determine whether practice enables people with PD to walk with large steps Gait & Posture 31 (2010) 229–233 ARTICLE INFO Article history: Received 17 December 2008 Received in revised form 20 October 2009 Accepted 25 October 2009 Keywords: Parkinson’s disease Gait disorders Transfer Dual task Attention ABSTRACT Background: Gait disorders in people with Parkinson’s disease (PD) are accentuated when they perform another task simultaneously. This study examines whether practice enables people with PD to walk with large steps while performing added tasks, and to determine if training people with PD to walk with added working memory tasks leads to improvements in gait when walking and performing other tasks simultaneously. Methods: Walking patterns were recorded pre and post a 20 min dual task training session in 20 people with PD. Participants performed a series of 10 m walking trials under seven conditions: gait only, and with six different added tasks varying by task type (e.g. motor, cognitive), domain (e.g. postural, manual manipulation, language, calculation, auditory, visuospatial), and difficulty level. Dual task training aimed to improve step length while simultaneously undertaking a variety of language and counting working memory tasks that were different to those used in assessment. Results: Following training, step length increased when performing five of the six added tasks, indicating transfer of dual task training when walking occurred across task types and domains. Improvements in gait speed occurred in three of the six added tasks. When other gait variables were examined, such as step length variability, few improvements with training were found. Conclusions: Training can lead to larger steps when walking under dual task conditions in people with PD. The gait variable emphasised during dual task training appears to be an important factor in enabling the transfer of training improvements across tasks. ß 2009 Elsevier B.V. All rights reserved. * Corresponding author. Tel.: +61 7 3365 2317; fax: +61 7 3365 1622. E-mail address: s.brauer@uq.edu.au (S.G. Brauer). Contents lists available at ScienceDirect Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost 0966-6362/$ – see front matter ß 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.gaitpost.2009.10.011