Psychometric Properties of the Trail Walking Test for People With Stroke

Shamay S.M. Ng (Corresponding Author), Tai Wa Liu, Joshua Tsoh, Peiming Chen, Tony S. Cheng, Marco C.H. Cheung, Anthony H.H. Leung, Liam L.Y. Ng, Ken Y.K. So, Mimi M.Y. Tse

Research output: Journal article publicationJournal articleAcademic researchpeer-review

5 Citations (Scopus)


Objective: To investigate (i) the inter-rater and test–retest reliability of the trail walking test (TWT) and the minimum detectable change in the TWT completion time; (ii) the correlations between the TWT completion time and stroke-specific impairments; and (iii) the cutoff TWT completion time to distinguish between people with stroke and healthy older adults according to dual-tasking ambulation ability. Design: Cross-sectional study. Setting: University-based rehabilitation center. Participants: In total, 104 people with stroke and 53 healthy older adults. Main Outcome Measures: The TWT, the Fugl–Meyer Assessment of Lower Extremity (FMA-LE), the ankle muscle strength test, the limit of stability (LOS) test, the Berg Balance Scale (BBS), the Timed Up and Go test (TUG), and the Community Integration Measure (CIM). Results: The mean TWT completion time in subjects with stroke was 124.906 s. The TWT demonstrated excellent inter-rater reliability [intraclass correlation (ICC) = 0.999] and good test–retest reliability (ICC = 0.876) in people with stroke. The TWT performance demonstrated significant negative correlations with the FMA-LE scores (r = −0.409), LOS movement velocity (affected and unaffected sides; r = −0.320 and −0.388, respectively), and LOS endpoint excursion (affected and unaffected sides; r = −0.357 and −0.394, respectively); a significant positive correlation with the LOS reaction time (affected side; r = 0.256); a moderate negative correlation with the BBS scores (r = −0.72); and an excellent positive correlation with the TUG completion time (r = 0.944). The receiver operating characteristic curve analysis revealed that an optimal cutoff of 69.61 s for the TWT completion time had an outstanding diagnostic power to distinguish between people with stroke and healthy older adults (area under the curve = 0.919) with high sensitivity (88.5%) and specificity (83.0%). Conclusion: Results of our preliminary study demonstrated that the TWT is a reliable, valid, sensitive, and specific clinical test for evaluating dual-tasking ambulation ability in people with stroke aged 45 years or above and without cognitive impairments. It can differentiate the dual-tasking ambulation ability between people with stroke and healthy older adults.

Original languageEnglish
Article number821670
JournalFrontiers in Neurology
Publication statusPublished - 3 Mar 2022


  • dual tasking
  • lower limb
  • outcomes
  • stroke
  • trail walking test

ASJC Scopus subject areas

  • Neurology
  • Clinical Neurology


Dive into the research topics of 'Psychometric Properties of the Trail Walking Test for People With Stroke'. Together they form a unique fingerprint.

Cite this