Establishing the Validity of Low Sampling Rates for Assessing Force/Torque-Time Variables During Isometric Strength...

Skeletal-muscle function is commonly assessed during isometric strength tests for both athletic and clinical populations. To assess neuromuscular performance, these are often analyzed by calculating peak force/torque and rate of force/torque development (RFD/RTD) variables using dynamometers, load cells, or force plates. These devices record a specific number of data points per second, known as the sampling rate, measured in samples per second (in hertz). It is unclear what the validity of sampling rates lower than the recommended 1000 Hz is for practical neuromuscular strength assessments.

This study examined the validity of sampling rates below previous recommendations of 1000 Hz for force/torque-time variable analysis. These variables were calculated at low sampling rates for maximal isometric strength tests of knee extension, knee flexion, and midthigh pull. The difference between each variable at 1000-Hz data and down-sampled data at 500, 250, 100, and 50 Hz was evaluated (ie, mean bias).

We obtained 1547 trials of individual repetitions from 4 previously published studies and 1 unpublished study (80 participants in total), with 1451 of these trials eligible for force/torque-time analysis. Mean bias of down-sampled data did not exceed -0.1% for peak force/torque, -2.1% for peak RFD/RTD, and -15.3% for fixed time epoch RFD/RTD variables.

Practitioners may adopt lower sampling rates into practice for force/torque-time analysis, particularly for variables of peak force/torque or peak RFD/RTD calculations.