Intra-subject sample size effects in plantar pressure analyses.

Area of Science:

• Biomechanics
• Gait Analysis
• Human Movement Science

Background:

• Plantar pressure analysis often uses small sample sizes (e.g., 10 steps).
• Large datasets reveal significant step-to-step variation in peak plantar pressure.
• The impact of sample size on plantar pressure data accuracy is under-explored.

Purpose of the Study:

• To investigate the sensitivity of plantar pressure metrics to intra-subject sample size.
• To determine the minimum number of steps required for reliable plantar pressure analysis.
• To quantify the potential for error with small sample sizes in gait studies.

Main Methods:

• Random subsampling analysis of large plantar pressure datasets (>500 steps).
• Evaluation of metric sensitivity at sample sizes ranging from 5 to 400 steps.
• Analysis of whole-record pressure, single-pixel values, and pixel-level variability (MSE).

Main Results:

• Whole-record mean/maximum pressure showed minor sensitivity above 200 steps.
• Sample sizes <50 steps significantly increased the range of pressure values, risking qualitative error.
• Pixel-level variability (MSE) was highly sensitive to sample size, exponentially increasing in smaller subsamples.
• Low sample sizes (<50 steps) captured significantly less variance than large datasets (>500 steps).

Conclusions:

• Small sample sizes (n < 20) in plantar pressure studies likely capture a low proportion of true variance.
• Results suggest that current small sample sizes may not accurately reflect population means.
• Recommendations for adequate sample sizes are needed to ensure reliable gait and plantar pressure research.