Evaluation of toddlers different strategies during the first 6 months of independent walking: A longitudinal study

Introduction and aim: Many studies have been observing infants at the onset of walking in order to evaluate the development of different strategies and coordination [1,2]. These studies regard most of the times small groups (<10) and only few studies observed longitudinally the evolution of independent walking (on 2 or 5 subjects) [2–4]). Given the small numbers of subjects, different strategies were only described qualitatively [5] and the quantifications made for describing the development of gait were representative of the whole group analysed at a specific developmental stage. The aim of the present study is to observe longitudinally a large group of infants (>20) using inertial sensors over a 6-months period after onset of independent walking. These data will allow evaluating the changes in gait temporal parameters and coordination at the beginning of independent walking. Moreover they will permit to evaluate quantitatively differences in strategies at the very beginning of walking [5] and to eventually correlate these differences with children characteristics. Patients/materials and methods: Twenty healthy infants (77 ± 3 cm, 10 ± 2 kg, 13 ± 2 months) were included in the study. All of the infants were full-term at birth and had no known developmental delays. Tests on the infants were scheduled once a month after the onset of independent walking for 3 months and one after 6 months. When possible, a test during the very first week of independent walking was performed. Three tri-axial wireless inertial sensors (OPALS, Apdm, USA) were mounted respectively on the lower back and on the two legs. The participants were asked to freely walk in a corridor. Ten consecutive strides were analyzed. Right heel strike (HS) and toe off (TO) instants were estimated from the angular velocity of the lower limb [6]. Stride (strT), swing (swT), stance (stanceT) times, cadence (Cad) and normalized cadence (nCad) were calculated. Trunk accelerations were used to estimate step and stride regularity (stepR and strR) [7], peak to peak range and variability of trunk accelerations. Median, 25° and 75° percentiles of estimated parameters were calculated at each developmental stage. Results: StrT decreased from month 0 to month 2 and then increased with months of experience. The opposite trend was shown by Cad and nCad. StT followed the same trend of StrT, while SwT increased constantly with months of experience. StepR and strR did not show a general trend for all the children. Medio-lateral trunk accelerations showed a constant decrease in peak to peak range and variability in all the children. Discussion and conclusions: The trend of Cad and nCad found is in contrast to what found by Looper (2012) [2] in longitudinal study on 8 children but is in agreement with Sutherland [8] who analysed a large number of subjects of different ages (children to adults). An analysis of different children characteristics (e.g. age at the first test) could be interesting for explaining these different results. The low swT at the very beginning of walking could be an evidence of children fear of falling. StepR and strR did not show a general trend among the whole group: future works will focus on evaluating if there are two or more typical trends in the whole group or if regularity is not a significant descriptive index for toddlers. Variability and peak to peak range of trunk medio-lateral accelerations decreased with experience indicating a more and more regular and less and less oscillating gait. Future works will evaluate the possibility of identifying the strategies descripted by McCollum [5] by sensor data and to describe quantitatively how these strategies develop towards the pendulum mechanism with experience. Acknowledgments: This research was funded by the project “Fall risk estimation and prevention in the elderly using a quantitative multifactorial approach” (project ID number 2010R277FT) awarded by MIUR.