Dear Javier,

The different variables comprised traditional linear and non- linear methods as proposed by diverse authors [....]. Duarte & Freitas (2010) subdivided the parameterization methods roughly into two groups: First, the traditional parameters which refer to estimations of the overall size of COP excursions, second, structural posturographic and more sophisticated parameters which describe temporal pattern of the time series in a nonlinear manner [...]. The former ones are used as indicators of postural stability outlining COP displacements as random fluctuations. According to this theory, larger COP deflections are associated with less stable balance and in a next step with aging and disease. Consequently, the assumed random properties are treated as unwanted noise and averaged out. Temporal structures of the displacements are disregarded. In contrast, nonlinear methods determine the time de pendent structure of the time series. Against this back- ground, both analysis techniques create complementary information [...]. In my vision, these two parameters (average and SD of COP) are complementary, they are classified in the first group (traditional paraemters), however I prefer to use the SD of COP because the results are most complet. You will can to show the results of these two variables, but depending of the case it will be unnecessary. 

Look some papers which I sent to you. I believe they can clarify your ideas.

Best regards.

As COP has been used in research to determine "normal" gait by proclaiming a "normal" COP with deviations theorized as "abnormal" or "unhealthy" theoretically reflecting pathology or needing "repair, there has been a dirth of applications of clinical valueand little high level evidence has surfaced.

The articles suggested here seem to verify my hypothesis that using COP as a clinical tool to diagnose and treat biomechanical and clinical pathology and injury probably has participated in the fact that little accurate, high level, valuable evidence has surfaced in the fields of biomechanics and biomedical engineering of the lower extremity in the past decade.

There are 100's of articles using COP of the foot in stance and gait as diagnostic and treatment tool most of which lack large cohorts, clinical merit and an error rate that is meaningful.

Furthermore, conclusions such as "We demonstrate that COP evaluation irrespective of sampling duration necessitates a set of diverse parameters to gain more variance of the data. Parameter sets are—despite of a few exceptions—uniformly invariant towards sampling durations and display a consistent factor loading pattern", seem to state that none of the parameters are effective.

Average of CoP is not valid parameter because of in different conditions may affected by position changes. It is better you used mean velocity or non linear parameter such as entropy and linear parameter such as TTB.

Saeed, are you stating that entropy and TTB are valid or that they are less invalid than average COP or other parameters?

I agree with Marcelo but I recommend that before talking of parameters and diagnostics you should carefully check the metrological accuracy of the force plate and of its software. Just to explain the topic: why otherwise the majority of force platforms would have a specific "Normal parameters set", especially if they are given in mm, sec, Hz , degrees?

Then you can select either classical parameters (Sway path, Sway Area, Area of the Confidence ellipse/circle, Antero-Post and Latero-Lat oscillations std.dev or max values, COP Mean velocity, Romberg quotients ....) and/or Frequency Spectral Analysis, and/or other parameters provided by non-linear analysis (Sway Density, Attractors, Fractal, Random Walk ....).

My experience is however suggesting that one single parameter can hardly provide a useful diagnostic indication. The tendency is for a combination of parameters that is specific for every subject and that expresses the capability of solving the problem of keeping the upright unperturbed stance.

The famous Framingham Study produced many additions to the literature.

This includes an article related to CPOE.

It concluded such things as:

1. women may have flatter feet than men

2. women who have worn high heels overpronate more than men or women who have not worn high heels

3. CPOE is not affected by changes in BMI as we age

4. CPOE is not affected by changes in activity level as we age.

In addition it seems that (and I may be wrong here) the CPOE tested 14-19 for all  cohorts with an error rate of 9 (50%) or more. If I am correct, this takes the value and applicability of this article in practice off the table.

I re-suggest that COP has produced little of clinical value in the literature as to reliability and treatment value and ask why we continue to use it?.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3655877

Hi Javier,

I also agree with Marcelo and Michele.  On many occasions, we have reported mean COP and a COP deviation if we use center of pressure outcomes.  We have also reported COP velocity along with these outcomes.  While they all come from the same data, we have always believed that no single postural control measure can capture the function of the sensorimotor system and therefore like to include multiple outcomes.  We have also used the TTB outcomes that others on this thread have mentioned because TTB does not correlate with COP outcomes, suggesting that they are capturing different aspect of postural control. 

All of these outcomes can be influenced by a number of factors which makes sense given that motor control is derived from task, environmental, and individual constraints.  While this makes study to study comparison and study to clinic comparisons sometimes difficult, these outcomes allow us to gain insight into how the sensorimotor system maintains postural control in a variety of tasks and settings.

For static stance, COP outcomes are generally reliability across a range of ages and pathological populations.  Depending on your population, these outcomes demonstrate good construct validity, and are often sensitive to a number of therapeutic interventions.  In my area of research, COP outcomes have a great deal of clinical value as they are predictive of increased risk of ankle injury and are obviously a strong indicator of restoring appropriate balance control following injury. 

We do not generally use COP outcomes during gait studies but a number of my colleagues do.  Gait is a much more complex task and the validity of COP outcomes in gait cannot be compared to the validity of COP outcomes in static stance.  If you have an interest in COP outcomes during gait, the system you use to capture those data are very important (e.g. force plate, pressure mat, pressure insole).

I hope this helps

Erik 

Hi Javier,

an additional thought you might want to consider in your data analysis and study design: we recently did a study and also a review on the specificity of balance training effects, and the results indicate that the concept of a general balance ability does not seem to be very appropriate and that it is more helpful to conceptualize balance as a set of skills. I.e. someone who is good at slack-lining does not necessarily perform better than average at one-leg stance on a force plate, and if someone trains "balance" on a tilt board, his performance in other balance tasks (e.g., one-leg stance or balancing on a different device) probably does not improve much.

Best regards,

Andreas

P.S.: if you're interested in these publications: 

Article Task-specificity of balance training

Article Specificity of Balance Training in Healthy Individuals: A Sy...

First at all, i´m sorry cause i wasnt able to answer til now, and

thank you very much, to all of you ... you really help me to make a better analysis and consider some topics i didn´t even thought about before ...

We want to make an assesment of the balance with pressure insoles ... the device and software give us some parameters related to the COP displacement (COP velocity, COP trace length, average COP, and Standar Deviation of COP). We want to show each of these parameters as a different variable in the results of some future research, but we are not sure about if the average COP and the Standar Deviation of the COP are the same parameter or not, because the software show it as 2 separate values, even if - i guess - they came from the same data.