RESUMEN
Dual-task costs of gait (variability) parameters are frequently used to probe the grade of automaticity of walking. However, recent studies reported contradicting dual-task costs for different gait variability measures within the same cohorts. The effects of a dual-task on the gait pattern are, thus, not fully understood. The aim of the current study was to analyze the different gait variability components ('Tolerance', 'Noise', and 'Covariation') during dual-task walking compared to single-task walking. In an experimental study, 21 young and healthy adults (11 males, 10 females, age: 24 ± 3 years) were included. The participants completed three experimental conditions: (a) single-task walking, (b) dual-task walking (serial-seven subtractions), and (c) cognitive single task in sitting position. To analyze different gait variability components, we applied a method which distinguishes the three components: 'Tolerance', 'Noise', and 'Covariation' (TNC). To test for differences, we used the statistical parametric mapping method. Compared to single-task walking, the results depict lower gait variability of the result parameters during the dual-task condition at 0-15% (p = 0.010) and 94-100% (p = 0.040) of the stance phase and 0-63% (p < 0.001) during the swing phase. The decreased result parameter variability was due to less (sensorimotor) 'Noise' (stance: 2-100%, p < 0.001; swing: 2-59%, p < 0.001) during the dual-task walking condition. In further studies, the sources of the reduced unstructured (sensorimotor) noise in the dual-task condition should be analyzed to better understand the effect of a cognitive dual task on the gait pattern.