Background: Walking speed is an important clinical tool in the physical therapy examination of older adults, but not always feasible to measure. Clinicians may be limited by space, a client's functional limitations, or use of telehealth. The Walking Speed Questionnaire (WSQ) can be used to estimate walking speed, but the effects of testing protocols on the accuracy and diagnostic capability of the WSQ have not yet been investigated.
Objectives: The purposes of this study were to compare estimated with recorded walking speed and examine effects of start condition and instructions in healthy older adults.
Methods: One hundred community dwelling adults 60 years or older were tested. After completing the WSQ, participants performed two trials of six walking conditions that included standing and walking starts under instructions to "walk at a usual pace," "as if crossing a street," and "fast."
Results: Participants averaged 72.4 (9.6) years of age. The WSQ average walking speed was 1.22 (0.16) m/s and was slower than each condition, p-value = 0.001, except for those performed with instructions to walk at usual pace with both standing, 1.23 (0.29) m/s and walking starts, 1.26 (0.30) m/s. Compared to the usual pace walking start condition, the WSQ sensitivity and specificity were 31% and 94% respectively with an 82% accuracy. A modified cutoff of the WSQ to 1.17 m/s improves sensitivity.
Conclusions: If walking testing cannot be performed, the WSQ is a reasonable alternative providing an accurate estimate of a client's walking speed when walking at a usual pace.
Keywords: Gait speed; Older adults; Testing protocol.