The influence of divided attention on walking turns: Effects on gait control in young adults with and without a history of low back pain

Highlights

• We investigated the effects of divided attention during walking turns.

• We compared healthy adults to asymptomatic adults with a history of low back pain.

• Turn performance consistency improved with divided attention in healthy adults only.

• Changes in cognitive control of gait persist between painful episodes.

Abstract

The cognitive control of gait is altered in individuals with low back pain, but it is unclear if this alteration persists between painful episodes. Locomotor perturbations such as walking turns may provide a sensitive measure of gait adaptation during divided attention in young adults. The purpose of this study was to investigate changes in gait during turns performed with divided attention, and to compare healthy young adults with asymptomatic individuals who have a history of recurrent low back pain (rLBP). Twenty-eight participants performed 90° ipsilateral walking turns at a controlled speed of 1.5 m/s. During the divided attention condition they concurrently performed a verbal 2-back task. Step length and width, trunk-pelvis and hip excursion, inter-segmental coordination and stride-to-stride variability were quantified using motion capture. Mixed-model ANOVA were used to examine the effect of divided attention and group, and interaction effects on the selected variables. Step length variability decreased significantly with divided attention in the healthy group but not in the rLBP group (post-hoc p = 0.024). Inter-segmental coordination variability was significantly decreased during divided attention (main effect of condition p < 0.000). There were small but significant reductions in hip axial and sagittal motion across groups (main effect of condition p = 0.044 and p = 0.040 respectively), and a trend toward increased frontal motion in the rLBP group only (post-hoc p = 0.048). These findings suggest that the ability to switch attentional resources during gait is altered in young adults with a history of rLBP, even between symptomatic episodes.

Introduction

Functional, goal-oriented gait requires attention to navigate environmental obstacles. Even during steady-state walking, some attentional resources are utilized for gait [1]. This is demonstrated experimentally by cognitive-motor interference when gait is performed concurrently with another attention-demanding task. Cognitive-motor interference may be increased by a reduction in the automaticity of gait, an increase in the executive control resources utilized for steady-state gait [2], or as a result of impaired attentional capacity and processing [1]. Increased cognitive-motor interference during gait is evident in older adults and individuals with multiple clinical conditions including persistent pain. This is associated with impaired functional gait performance and increased risk of falls [3], [4].

The effect of cognitive-motor interference can be quantified at multiple levels of gait control. Spatiotemporal characteristics such as gait speed and step length provide information on control of overall task performance. Kinematic characteristics, such as coordination between segments (inter-segmental coordination) and individual joint excursions demonstrate how task performance is accomplished. In healthy adults, gait performance deteriorates during divided attention [5], [6], [7], [8], [9]. Existing evidence has demonstrated unchanged inter-segmental coordination [10], reduced [5] and unchanged lower-limb joint excursion [7] and reduced trunk excursion [11] during divided attention. However, as previous studies examined divided attention paradigms that involved additional mechanical demands, or did not control gait speed, the extent to which cognitive load alone accounts for these kinematic adaptations is unclear [5], [7], [11].

Adaptable locomotor behavior is facilitated by stride-to-stride variability [12]. Healthy individuals demonstrate increased step length variability [5], [8], reduced step width variability [13], decreased variability in trunk-pelvis coordination [10] and decreased variability in trunk motion when attention is divided during steady-state gait [14]. Changes in variability during attention-demanding gait perturbations such as walking turns may provide further insight into cognitive-motor interference. Ipsilateral walking turns are changes in direction that occur toward the side of the stance limb. The reorientation into the new line of progression during ipsilateral turns may be accomplished within the stance phase of a single step (ipsilateral pivot strategy) and is achieved through rapid, axial rotation in the trunk, pelvis and hip [15], [16]. As there is an immediate return to steady-state walking after the reorientation phase of the turn [15] the successful, consistent performance of the ipsilateral walking turn can be characterized by the length and width of the step immediately following the turn, and the variability of those parameters.

In people with low back pain (LBP), altered executive control of gait may result in an exaggerated response to divided attention [17]. Adults with chronic LBP demonstrate greater increases in stride length variability [17] and decreases in axial plane trunk-pelvic coordination variability [10] during divided attention compared with healthy controls. Existing studies have investigated middle-aged, symptomatic patients. In these individuals, executive function may be persistently impaired [18], or pain and fear of pain may demand additional attentional resources [1], [2], [19]. Many young individuals with persistent LBP have recurrent rather than chronic symptoms (rLBP) [20]. To better understand the mechanisms underlying cognitive-motor interference and LBP it is important to establish how individuals with rLBP respond to divided attention when they are asymptomatic. In particular, measures of trunk-pelvic inter-segmental coordination and joint excursion may provide valuable indices of cognitive-motor interference specific to the painful body region in this population.

The purpose of this study, therefore, was to compare the influence of divided attention on step length and width, inter-segmental coordination and joint excursion, and stride-to-stride variability during ipsilateral walking turns between asymptomatic young adults with a history of rLBP and healthy individuals. We hypothesized that in response to divided attention, all participants would demonstrate reduced step length and increased step width, increased step length/width variability, reduced trunk-pelvic and hip excursion and reduced trunk-pelvic coordination variability compared with baseline. We also hypothesized that these changes would be greater in the individuals with a history of rLBP.