Applying different mathematical variability methods to identify older fallers and non-fallers using gait variability data

Autor(es)

Marques, Nise Ribeiro ; Hallal, Camilla Zamfolini ; Spinoso, Deborah Hebling ; Morcelli, Mary Hellen ; Crozara, Luciano Fernandes ; Goncalves, Mauro

• Tipo de documento: Artigo Científico
• Fonte: AGING CLINICAL AND EXPERIMENTAL RESEARCH; v. 29, n. 3, p. 473-481, JUN 2017.
• Citações Web of Science: 2

Resumo

Background The clinical assessment of gait variability may be a particularly powerful tool in the screening of older adults at risk of falling. Measurement of gait variability is important in the assessment of fall risk, but the variability metrics used to evaluate gait timing have not yet been adequately studied. Objectives The aims of this study were (1) to identify the best mathematical method of gait variability analysis to discriminate older fallers and non-fallers and (2) to identify the best temporal, kinematic parameter of gait to discriminate between older fallers and non-fallers. Methods Thirty-five physically active volunteers participated in this study including 16 older women fallers (69.6 +/- 8.1 years) and 19 older women non-fallers (66.1 +/- 6.2 years). Volunteers were instructed to walk for 3 min on the treadmill to record the temporal kinematic gait parameters including stance time, swing time and stride time by four footswitches sensors placed under the volunteers' feet. Data analysis used 40 consecutive gait cycles. Six statistical methods were used to determine the variability of the stance time, swing time and stride time. These included: (1) standard deviation of all the time intervals; (2) standard deviation of the means of these intervals taken every five strides; (3) mean of the standard deviations of the intervals determined every five strides; (4) root-mean-square of the differences between intervals; (5) coefficient of variation calculated as the standard deviation of the intervals divided by the mean of the intervals; and (6) a geometric method calculated based on the construction of a histogram of the intervals. Results The standard deviation of 40 consecutive gait cycles was the most sensitive (100 %) and specificity (100 %) parameter to discriminate older fallers and non-fallers. Conclusion The standard deviation of stance time is the kinematic gait variability parameter that demonstrated the best ability to discriminate older fallers from non-fallers. Protocol number of Brazilian Registry of Clinical Trials: RBR-6rytw2. (AU)

• Processo FAPESP: 11/11639-7 - Efeito de Duas Modalidades de Treinamento com Exercícios Físicos no Equilíbrio, no Torque Muscular e em Variáveis Biomecânicas e Metabólicas Durante a Marcha de Indivíduos Idosos
• Beneficiário: Nise Ribeiro Marques
• Modalidade de apoio: Bolsas no Brasil - Doutorado