Elsevier

Gait & Posture

Volume 17, Issue 1, February 2003, Pages 81-87
Gait & Posture

Normalisation of gait data in children

https://doi.org/10.1016/S0966-6362(02)00062-0Get rights and content

Abstract

The comparative effect of semi-dimensional (SD) and non-dimensional (ND) normalisation on the results of a longitudinal study of gait in 5–12-year old children was investigated. The use of both height and leg length in the normalisation was examined. Only ND analysis could be used to identify subjects with the same accelerations. ND analysis of the children's gait indicated that there was little change in the combination of step length and cadence used to achieve a particular velocity between 5 and 12. The first peak and mid-stance trough values of the vertical component of ground reaction force did not change with age. We recommend the use of ND normalisation rather that SD to allow comparisons between individuals of differing size and mass.

Introduction

To compare gait information between two people of significantly different heights and or masses one must ‘normalise’ the information in an attempt to remove all variation due to these differences. Although normalisation of gait data is routine practice it has been applied in a number of different ways, often without mathematical justification or a firm basis in dynamics; for example Ounpuu et al. [1] used normalisation in their analysis of joint moments in children. Joint moments were divided by the subject's mass that would have given moments in units of m2 s−2:Moment=force×leverarmUnitsofNmMomentmass=force×leverarmforce/gUnitsofm2/s2where g is the acceleration due to gravity.

In this example, the normalised moment, therefore, depended on the dimensions of the subjects. This type of normalisation may be justified when all subjects are approximately the same size but the size of children varies and it is likely that there would have been a significant effect on moment results.

Acceleration is a key factor in gait as it determines inertial forces which, together with gravitational force, determine ground reaction forces and joint forces and moments. It is possible to use normalisation to allow subjects who have the same accelerations to be grouped together to enable meaningful comparisons. One technique that allows grouping by acceleration is that summarised by Hof [2] who described a set of non-dimensional (ND) variables based on the principle of similarity of acceleration. It is only possible for two individuals of different heights and masses to have the same dynamic patterns of motion, with the same joint angles and joint forces and moments if all of the ND variables are identical. Pierrynowski and Galea [3] demonstrated that the normalisation proposed by Hof [2] reduced the inter-subject variability in a group of subjects whose heights and weights were significantly different. Pierrynowski and Galea [3] investigated eight scaling schemes and found that normalisation to ND quantities was one of the most successful methods for reducing inter-subject variability which supports the use of this method in paediatric populations.

In the current study an investigation has been completed of the effects of differing normalisation techniques on the results of a 7 year longitudinal study of gait of 16 children (Stansfield et al. [4], [5]).

Section snippets

Methods

Gait analysis was performed on 16 children (eight boys and eight girls) each year for 7 consecutive years. Children walked barefoot at self-selected normal velocities. Ground reaction forces (Kistler Instruments, AG Winterthur, Switzerland) and motion analysis data (Vicon, Oxford Metrics Group, Oxford, UK) were collected for approximately three sets of data for each leg.

Weight, height (standing from the top of the head to floor) and leg length (greater trochanter to lateral knee to lateral

Results

The anthropometric measurements of the children studied are shown in Table 1. Male and female children have been included together in the analysis of these data. Results of alternate year groups are presented in Fig. 1, Fig. 2, Fig. 3, Fig. 4. All results from all children within each age band were split into bands of the relevant velocity quantity and the average value of the parameter being investigated was calculated. For example Fig. 1A illustrates the SD step length of children at certain

Discussion

Hof [2] presented a summary of a ND normalisation technique that allows comparison of subjects of similar acceleration profiles. Hof [2] has not, to our knowledge presented any examples of the influence of choice of normalisation technique on outcomes. In light of the sometimes arbitrary use of normalisation methods in the literature it is useful to provide examples using data collected from individuals in a longitudinal study to illustrate the effect of using different techniques.

In the

Conclusions

The effect of using different normalisation methods when comparing gait analysis results of children of different sizes has been demonstrated.

To examine the development of children's gait patterns with age, it is necessary to use a ND normalisation technique to ensure comparison of children with the same acceleration patterns.

The use of ND quantities in the analysis of the gait patterns of 5–12-year old children revealed that with age, at any particular velocity, there were only small

Acknowledgements

This work was funded by the generosity of the James and Grace Anderson Trust. The authors would like to thank all of the subjects who demonstrated their dedication to the project by returning for 7 consecutive years of gait analysis.

References (8)

There are more references available in the full text version of this article.

Cited by (110)

View all citing articles on Scopus
View full text