Validation of a simple mechanical accelerometer (pedometer) for the estimation of walking activity

Area of Science:

• Biomedical Engineering
• Wearable Technology
• Physical Activity Monitoring

Background:

• Accurate measurement of physical activity is crucial for health and wellness.
• Existing activity monitors vary in accuracy and reliability.
• Development of a simple, robust, and cost-effective activity tracker is needed.

Purpose of the Study:

• To evaluate the accuracy and reproducibility of a novel mechanical accelerometer for quantifying physical activity.
• To compare the accelerometer's performance in controlled laboratory settings and real-world field studies.
• To establish the device's utility in estimating daily step counts across different age demographics.

Main Methods:

• Controlled bench tests with repetitive vertical oscillations to assess linearity, reproducibility, and sensitivity.
• Field studies involving attachment of the accelerometer to the hip during walking.
• Comparison of accelerometer-derived scores with independent footstep counts from a heel-mounted resistance pad.

Main Results:

• The accelerometer demonstrated a linear (r = 0.996) and reproducible (+/- 1.5% coefficient of variation) response to controlled oscillations.
• Device sensitivity ranged from 6.2 to 7.4 units/10,000 oscillations, with response plateauing at 4 m s-2.
• Field studies showed no significant difference between accelerometer step counts and independent measures in younger (mean difference 0.29 x 10^3 steps) and older (mean difference 0.46 x 10^3 steps) groups.

Conclusions:

• The mechanical accelerometer is a valid and reliable tool for measuring physical activity, specifically step count.
• The device provides accurate and reproducible data in both laboratory and field conditions.
• Estimated daily and weekly step counts provide a useful metric for assessing activity levels in different populations.