Saturday, October 4, 2014

Data Normalization - Body-Size-Independent Measures

Muscle strength is strongly influenced by body size and correlated with measures such as Body Mass and Heigth; therefore, utilization of strength body-size-independent measurements for the aforementioned applications is important. The importance of this is especially apparent when comparing persons of different body sizes (ie, athletes vs nonathletes, men vs women, young vs old), or protocols where Body Mass could change between data collection periods (eg, long-term treatment). Normalizing strength measurements to measures of body size has traditionally been used to remove body-size dependence. However, there is no consensus on the method by which strength measurements should be normalized.


While some studies have not normalized strength measures, others have normalized strength to Body Mass or a combination of Body Mass and Heigth (ie, Body Mass X Heigth). Ratio standards (normalizing to Body Mass) assume that muscle strength is directly proportional to Body Mass. These assumptions, however, may result in misleading findings. An alternative to ratio standards is allometric scaling, a normalization technique based on the theory of “geometric similarity”, or that the shapes of human bodies are similar, but variable in size. Allometric scaling is a normalization technique that divides strength by Body Mass raised to a power that removes body-size effects. The equation for normalized strength is Sn=S/mb, where S is the non normalized strength measure, m is Body Mass, b is the allometric b-value (or scaler), and Sn is the body-size-independent strength measurement. 


In literature exists generic theoretical reference values for allometric adjustment of various parameters, but is more appropriate to use the specific exponent generated for your sample.


The normalization procedures included plot a linear regression with a log scale of peak strength parameters of your subjects by your respectives log scale Body Mass values aand then divide its strength parameters by their values of body mass raised to the slope found in linear regression.

Here's a link to download algorithm implemented in Matlab ® to perform the produre to calculate de allometric b-value of your sample. Matlab Code
 
To use the available function,first you must create a spreadsheet in excel with the "A" column containing the body mass in kg of its subjects and the "B" column containing the respective peaks of the strength parameters of the subjects. (ie. figure below)



The output of Allometry_gbiomech.m gives you the allometry b-value, the Confidence Interval of the measure and the R Squared value in a dialog box.

References:
- Lleonart, J.; Salat, J.; Torres, G. J. Removing allometric effects of body size in morphological analysis. Journal of theoretical biology 2000;205:1, p. 85–93.
- Jaric S. Muscle strength testing: use of normalisation for body size. Sports Med 2002;32, p. 615-31.
- Jaric, S.; Mirkov, D.; Markovic, G. Normalizing physical performance tests for body size: a proposal for standardization. Journal of strength and conditioning research / National Strength & Conditioning Association 2005;19:2, p. 467–474.
- Shingleton, A. W. Allometry: The Study of Biological Scaling. Nature Education Knowledge 2010;3:10.

Big Hug!

No comments:

Post a Comment