Related Experiment Video
Updated: May 8, 2026
Exergaming in Older People Living with HIV Improves Balance, Mobility and Ameliorates Some Aspects of Frailty
Published on: October 6, 2016
Frailty and technology: a systematic review of gait analysis in those with frailty
Michael Schwenk1, Carol Howe, Ahlam Saleh
1Interdisciplinary Consortium on Advanced Motion Performance (iCAMP) and Southern Arizona Limb Salvage Alliance (SALSA), Department of Surgery, College of Medicine, University of Arizona, Tucson, Ariz., USA.
Background:
New technologies for gait assessment are emerging and have provided new avenues for accurately measuring gait characteristics in home and clinic. However, potential meaningful clinical gait parameters beyond speed have received little attention in frailty research.
Objective:
To study gait characteristics in different frailty status groups for identifying the most useful parameters and assessment protocols for frailty diagnosis.
Methods:
We searched PubMed, Embase, PsycINFO, CINAHL, Web of Science, Cochrane Library, and Age Line. Articles were selected according to the following criteria: (1) population: individuals defined as frail, prefrail, or transitioning to frail, and (2) outcome measures: quantitative gait variables as obtained by biomechanical analysis. Effect sizes (d) were calculated for the ability of parameters to discriminate between different frailty status groups.
Results:
Eleven publications met inclusion criteria. Frailty definitions, gait protocols and parameters were inconsistent, which made comparison of outcomes difficult. Effect sizes were calculated only for the three studies which compared at least two different frailty status groups. Gait speed shows the highest effect size to discriminate between frailty subgroups, in particular during habitual walking (d = 0.76-6.17). Gait variability also discriminates between different frailty status groups in particular during fast walking. Prominent parameters related to prefrailty are reduced cadence (d = 1.43) and increased step width variability (d = 0.64), whereas frailty (vs. prefrail status) is characterized by reduced step length during habitual walking (d = 1.32) and increased double support during fast walking (d = 0.78). Interestingly, one study suggested that dual-task walking speed can be used to predict prospective frailty development.
Conclusion:
Gait characteristics in people with frailty are insufficiently analyzed in the literature and represent a major area for innovation. Despite the paucity of work, current results suggest that parameters beyond speed could be helpful in identifying different categories of frailty. Increased gait variability might reflect a multisystem reduction and may be useful in identifying frailty. In addition, a demanding task such as fast walking or adding a cognitive distractor might enhance the sensitivity and specificity of frailty risk prediction and classification, and is recommended for frailty assessment using gait analysis.
