Protocol for a randomized comparison of integrated versus consecutive dual task practice in Parkinson’s disease: the DUALITY trial

Parkinson’s disease (PD) is a common, multisystem neurodegenerative disease which is characterized by motor and non-motor symptoms [1]. The motor symptoms of PD are manifold but include gait and balance disorders, which have a significant impact on functional mobility and quality of life [2]. Depending on the disease stage, up to 79.2 percent of patients (advanced stage) may report to have freezing of gait (FOG) [3], which is a disabling gait disorder characterized by episodes of lack of forward progression despite the intention to walk [4]. Approximately 60% of patients fall each year and about two thirds fall recurrently [5‐7].

Dual tasking is the simultaneous performance of two attention-demanding tasks with different goals, whereby one task can be denoted as the primary and the other as the secondary task [8]. Factors that affect dual task performance are the environment in which the task takes place, the nature of the secondary task, the age and disease-specific factors of each individual [9]. Gait has been found to deteriorate during dual task (DT) performance in PD [10‐14], resulting in a decrease of gait velocity, cadence and step length [9, 13‐18], an increase in gait variability [9, 13, 17, 19] and an increase in double support time [14, 20]. Also, falling and FOG are more commonly provoked in DT conditions [11, 21, 22].

Mild cognitive impairment is thought to occur in 20-57% of PD patients, even as early as 3–5 years after the diagnosis [23, 24]. Several studies have indicated that executive dysfunction is a robust determinant of DT interference at least in mid but not in early stage PD [13, 17, 25, 26]. Executive function refers to a set of abilities which flexibly guide behavior towards goals and includes switching between cognitive sets or tasks, appropriately inhibiting and generating responses and updating working memory contents [27, 28]. Wild et al. [13] showed that cognitive performance in PD worsened during DT walking and that this was correlated to global cognitive condition.

Motor learning is highly dependent on cognitive status in PD [29, 30]. Patients with freezing of gait (FOG) showed greater executive deficits than their non-freezing counterparts [31‐33] and were also found to learn a serial reaction time task less well [34]. This raises the question as to the extent of the cognitive challenge which is appropriate for achieving robust learning in PD.

Evidence for the efficacy of physiotherapy is growing in PD [35, 36]. Various modes of gait training were found effective in improving gait velocity [37, 38]. Although more controlled studies remain needed, cognitive training shows promising benefits in several cognitive domains including executive function in PD [39]. Despite the fact that evidence-based physiotherapy guidelines discourage the use of dual task exercise [40], a number of open label studies [38, 41‐46] support the feasibility and efficacy of DT training in PD [45, 46]. Hence, a phase III randomized study is currently being undertaken in which the effect of single and dual task gait training is compared in a wide variety of PD patients [47].

The current trial proposes to test the efficacy of two strategies for DT training. The first strategy entails consecutive task training (CTT), whereby each task will be trained separately. We expect that as a result of this type of training, performance of each task may become more automatic and thus free residual brain capacity for subsequent simultaneous task performance. The second strategy proposes integrated dual task practice (IDT). We expect that this training mode may increase the efficiency of shared neural resources [48] and as such improve DT performance. IDT may have additional benefits over CTT as this training mode may also improve the efficiency of specific brain areas involved in task integration [49, 50], enhance executive function and ease the transfer of learning to daily DT situations. Moreover, an advantage of IDT is that dual tasking can realistically never be avoided entirely, so it might be wise to prepare patients for such inevitable dual task events that commonly occur in daily life. Interestingly, in balance-impaired older adults, IDT and CTT led to largely similar performance increases with the exception of better retention in IDT [51, 52]. These findings call for replication, and importantly, it is currently unknown to what extent these results apply to PD patients.

Therefore, the primary aim of this study is to compare the effectiveness of consecutive versus integrated training (tested by combining walking with cognitive exercises) on DT gait performance in PD. We hypothesize that IDT practice will resort in better dual task outcomes and better consolidated learning results, particularly in patients without cognitive impairment. As secondary questions we will examine which dual task training mode is most beneficial in terms of fall risk reduction and leads to the best dual task gains in freezers versus non-freezers. Here, we describe the design of this training study, also referred to as the DUALITY trial.

Acknowledging that dual tasking forms an integral part of daily functioning, the DUALITY study proposed here aims to investigate the efficacy and safety of a rehabilitation intervention to improve dual tasking in PD patients. We will establish robust evidence on which training modality, i.e. integrated or consecutive training, leads to the best training results and how this compares with a control period without intervention. Given the reduced capacity for consolidation of motor learning in PD [49, 73] and the presence of executive deficits [28], it will be established whether dual task learning effects transfer to improvements in daily life and are sustained for 12 weeks without training.

A paucity of dual task studies have been performed in PD patients, most of which with poor methodological designs. This pilot work has shown short-term positive effects on gait performance [42‐45], although there was considerable variation in the population, training period and tasks (motor, cognitive) that were studied. This is the first power-based and randomized trial that compares two interventions in an early to mid-stage PD population that are both aimed at improving dual task performance but with contrasting treatment strategies. In analogy with a DT training study in balance-impaired older people [51, 52], we expect that both groups will show some improvement on DT outcomes However, as integrated DT training is more cognitively challenging we hypothesize that this will result in a better consolidation and retention of the training effects in line with motor learning theory [74]. In addition, we expect that integrated training will resort in better DT outcomes as the exact practice conditions mimics real life tasks more closely and this is important for transfer of learning [74]. In this study, the evaluation of motor performance will be supplemented by assessment of cognitive performance which allows monitoring of learning effects in both motor and cognitive domains. We focus on assessing various components of executive functions, an area that is specifically shown to be affected in PD patients [28]. This implies that this trial will also provide novel evidence on the impact of cognitive training on executive function in PD which may inform future power-based studies in this area.

This will be the first dual task training study that will look at differences in dual task gains between freezers versus non-freezers after stratification for these subgroups. The results of this study will establish which training strategy is optimal in each subgroup, taking their different cognitive profiles into account. Since freezers show greater executive deficits than their non-freezing counterparts [31‐33, 75] and learn less well in dual task conditions [34], we anticipate that freezers may benefit more from consecutive training in order to achieve dual task learning.

In conclusion, in this DUALITY trial we will study two different strategies aimed at improving dual tasking in PD patients. We expect that the trial will provide novel and clinically important information on the most effective and safe training strategy in different patient subgroups and in this way will contribute to developing future directions for rehabilitation targets in PD.