Performance variance on walking while talking tasks: Theory, findings, and clinical implications

Roee Holtzer, Cuiling Wang, Joe Verghese

Research output: Contribution to journalArticlepeer-review

113 Scopus citations


Dual tasks that involve walking and cognitive interference tests are commonly used in mobility assessments and interventions. However, factors that explain variance in dual-task performance costs are poorly understood. We, therefore, examined the moderating effects of two putative constructs, postural reserve and hazard estimate, on performance on a walking while talking paradigm. Participants were 285 non-demented older adults (mean age=76.9 years; %female=54.4). Postural reserve was operationalized as the presence or absence of clinical gait abnormalities. An empirical factor, based on measures of executive functions, served as a marker for hazard estimate. The moderation effects of postural reserve and hazard estimate on dual-task costs were examined via two-way interactions in a joint linear mixed effect model. Significant dual-task costs were observed for gait speed (95% CI=30.814 to 39.121) and cognitive accuracy (95% CI=6.568 to 13.607). High hazard estimate had a protective effect against decline in gait speed (95% CI=-8.372 to -0.151) and cognitive accuracy (95% CI=-8.372 to -0.680). Poor postural reserve was associated with reduced decline in gait speed (95% CI=-9.611 to -0.702) but did not moderate the decline in cognitive accuracy (95% CI=-3.016 to 4.559). Assessing postural reserve and hazard estimate can help improve mobility risk assessment procedures and interventions for individuals with cognitive and movement disorders.

Original languageEnglish (US)
Pages (from-to)373-381
Number of pages9
Issue number1
StatePublished - Feb 2014


  • Dual task
  • Executive functions
  • Gait
  • Postural reserve

ASJC Scopus subject areas

  • Aging
  • Geriatrics and Gerontology


Dive into the research topics of 'Performance variance on walking while talking tasks: Theory, findings, and clinical implications'. Together they form a unique fingerprint.

Cite this