Trial number and duration effects on standing weightbearing asymmetry measures

Side-to-side weight-bearing asymmetries during quiet bilateral stance may be important in the study of postural control and low back pain. However, clear and appropriate methods to measure and categorize these asymmetries have not been determined. The goal of this investigation was to evaluate trial number and duration when using a statistical means for categorization. Eleven healthy subjects completed ten trials of 20 second quiet stance while standing naturally with feet at shoulder width on two force platforms. The average of the first 5 seconds of vertical ground reaction force was compared with that of the entire 20 seconds. Subjects repeated the procedure within ten days. Symmetry/asymmetry categorization was determined by conducting repeated measures t-tests (p > 0.01) on the percent of body weight placed on the right foot compared to the left. Intra-class correlation coefficients (ICCr) were utilized to assess trial duration and day-to-day variations. Statistical power was used to evaluate the adequacy of ten trials. Statistical categorization was identical when using 5 and 20 seconds of data, except for two subjects that were found to be asymmetric over 5 seconds and symmetric over 20 seconds on the second day (ICCr = 0.995 day 1, 0.993 day 2). Power estimates indicated that one more trial would be necessary to categorize those with p values approaching significance (p > = 0.040) as asymmetric while 20 or more trials would be needed for others (p >= 0.117). Day-to-day repeatability was high (ICCr = 0.924 for 5 seconds, 0.927 for 20 seconds) with only two subjects changing from an asymmetric categorization on the first day to symmetric on the second day (20 seconds duration only). In conclusion, short duration trials (~5 seconds) may be adequate for use when evaluating weight-bearing asymmetries. However, due to the relatively large number of trials needed for this statistical method, other options should be pursued (which might reduce the number of trials but increase the duration of the trial). Finally, weight-bearing asymmetries are highly consistent from day-to-day, indicating measures taken on a single day should be representative of typical musculoskeletal loading during quiet stance.