The development of an accelerometer-based measure of human upright static anterior- posterior postural sway under various sensory conditions: test-retest reliability, scoring and preliminary validity of the Balance Accelerometry Measure (BAM).

Background: Accelerometers are being used to assess postural control in adults, but there is little to support their reliability and validity.

Objective: To estimate the test-retest reliability of the balance accelerometry measure (BAM) and to describe the known-groups validity of the BAM composite score.

Methods: Two measures of standing postural sway were taken across six sensory (vision/stance surface) and motor stance (feet together or tandem) positions from eighteen patients with vestibular disorders and 84 healthy subjects. Test-retest reliability for postural sway was estimated across all conditions using intraclass correlation coefficient (ICC). A composite measure of sway standardized to young healthy subjects on eyes open firm surface stance was compared between groups.

Results: Test-retest reliability of postural sway was good (ICC ⩾ 0.74) under all sensory conditions except eyes closed/tandem stance, which was slight to poor. Analysis of the receiver operating characteristic curve for composite scores indicated significant accuracy at identification of subjects in the vestibular/balance disorder groups. Composite standard scores equal or greater than 21.1 identified subjects with vestibular disorders with an accuracy of 72% sensitivity and 68% specificity.

Conclusions: The BAM displays good-excellent reliability for five of six sensory-motor conditions. The composite score appears to differentiate healthy from subjects with vestibular disorders.