Development of running is not related to time since onset of independent walking, a longitudinal case study

Abstract

Introduction: Children start to run after they master walking. How running develops, however, is largely unknown.

Methods: We assessed the maturity of running pattern in two very young, typically developing children in a longitudinal design spanning about three years. Leg and trunk 3D kinematics and electromyography collected in six recording sessions, with more than a hundred strides each, entered our analysis. We recorded walking during the first session (the session of the first independent steps of the two toddlers at the age of 11.9 and 10.6 months) and fast walking or running for the subsequent sessions. More than 100 kinematic and neuromuscular parameters were determined for each session and stride. The equivalent data of five young adults served to define mature running. After dimensionality reduction using principal component analysis, hierarchical cluster analysis based on the average pairwise correlation distance to the adult running cluster served as a measure for maturity of the running pattern.

Results: Both children developed running. Yet, in one of them the running pattern did not reach maturity whereas in the other it did. As expected, mature running appeared in later sessions (>13 months after the onset of independent walking). Interestingly, mature running alternated with episodes of immature running within sessions. Our clustering approach separated them.

Discussion: An additional analysis of the accompanying muscle synergies revealed that the participant who did not reach mature running had more differences in muscle contraction when compared to adults than the other. One may speculate that this difference in muscle activity may have caused the difference in running pattern.

Keywords: children; clustering; development; kinematics; muscle synergies; neuromuscular control; running.

FIGURE 1

Overview of age and walking age (time since onset of walking) for each session/child. A total of seven sessions were recorded for each child. At the end, six sessions were analyzed for each child, and these were matched based on walking age (bottom plot) in months. A few weeks separate the different sessions between P1 and P2. +2 is hatched as not analysed due to poor data quality. FS: First steps, +2, +6, +9, +13, +19, +32 refers to the number of months since first steps, i.e., time since onset of independent walking.

FIGURE 2

Temporal gait parameters. (A) Flight phase and double support phase as a percentage of the gait cycle (mean ± std). Double support phases are on the top with flight phases being the negative percentages below. (B) The normalized speed expressed as the Froude number (v²/g ⋅ l) for each session and participant (mean ± std). ^†Denotes a significant difference between current session and all other sessions for that participant as well as adults (p < 0.007). Horizontal lines denote significant differences between the Froude numbers for the sessions. A: adults, FS: first steps, %GC: percentage gait cycle, +6, +9, +13, +19, +32 refers to the number of months since onset of independent walking.

FIGURE 3

Principal component analysis (PCA). In the two left panels, each dot represents one stride. (A) PCA results in PC1-PC2 space. (B) The outcome of the PCA in PC2-PC3 space. (C) The weightings of all sessions ordered based on walking age, so not ordered per participant. PC1 distinguishes the FS sessions from the other sessions with PC2 distinguishing “mature” from “immature” running. A: adults, PC: principal component, FS: first steps, +6, +9, +13, +19, +32 refers to the number of months since onset of independent walking.

FIGURE 4

Clustering output. (A) Output of clustering ordered based on walking age (time since onset of walking in months) with the youngest session on the right and increasing in walking age in anticlockwise direction. The size of the clusters depends on the number of strides they contain (the larger the node the more strides they contain), similarly are the lines from each node to a cluster a representation of the number of strides (thicker lines equals more strides) from that session that belongs to each cluster larger than 10%. (B) Average pairwise correlation distance from each session to those of the adults as a function of walking age (months) for P1 and P2, respectively. Sizing of dots follow the sizing of lines in panel. A: adults, FS: first steps, +6, +9, +13, +19, +32 refers to the number of months since onset of independent walking.

FIGURE 5

Muscle synergies for P1, P2, and adults. The top graphs are the grand average temporal activation patterns for each session as a function of the gait cycle and the three synergies. Amplitude is in arbitrary units. The lower bar graphs are the weighting coefficients for the muscle synergies. The naming of the muscles can be seen below. TA: tibialis anterior, GM: gastrocnemius medialis, GL: gastrocnemius lateralis, SOL: soleus, RF: rectus femoris, VM: vastus medialis, VL: vastus lateralis, BF: biceps femoris, TFL: tensor fascia latae, GLM: gluteus maximus, ES: erector spinae, +6, +9, +13, +19, +32 refers to the number of months since onset of independent walking.

FIGURE 6

Full-width half-maximum and center-of-activity of muscle synergies. Color notation as to the right. (A) The full-width half-maximum (FWHM) was calculated for the main peak of each stride for each muscle synergy. It is represented as a percentage of the gait cycle. (B) The center-of-activity (CoA) were was expressed as the percentage of the gait cycle. The horizontal lines represent a significant difference, p < 0.007 within a participant and the adults. The dagger (†) represents a significant difference between session and all other sessions. A hash (#) represents a significant difference between session and all later sessions and adults. GC: gait cycle, FWHM: full-width at half-maximum, A: adults, FS: first steps, +6, +9, +13, +19, +32 refers to the number of months since onset of independent walking.