Validity and Reliability of an Integrated Smartphone Measurement Approach for Balance

Clinical balance assessments vary in reliability due to subjectivity in their scoring. A valid and objective accelerometer-based smartphone evaluation could benefit patients, clinicians, and researchers.

Our objective was to assess the validity and reliability of smartphone-based standing balance.

A repeated-measures study was conducted with 23 healthy young adult participants across 2 sessions ∼7 days apart.

Participants completed 30-second standing trials during tandem-stance eyes-open, tandem-stance eyes-closed, single-leg eyes-open, and single-leg eyes-closed conditions. Android and iOS smartphones were placed vertically on the lower back via a belt with 3 retroreflective markers attached and tracked by an 8-camera motion capture system. Sway path, range, and area were calculated from smartphone accelerometer and marker data. We assessed reliability using intraclass correlation coefficients (ICC[2,k]) and validity using Pearson r correlations between the marker and smartphones from visit 1.

Across eyes-open conditions, Android (ICC = .84-.96), iOS (ICC = .82-.98), and marker-based (ICC = .84-.95) assessments demonstrated good to excellent reliability. Across eyes-closed conditions, Android (ICC = .41-.87), iOS (ICC = .34-.79), and marker-based (ICC = .31-.87) assessments demonstrated poor to good reliability. Correlations between smartphones and the marker data were moderate to very high (r = .56-.97).

The smartphone-based assessment was valid and reliable, indicating that clinicians and researchers can implement this method to measure balance with the opportunity for remote administration and increased patient tracking across various recovery timepoints.