Construct and criterion validity of the functional gait assessment—Brazil in community-dwelling older adults

• Published in: Revista brasileira de fisioterapia (São Carlos (São Paulo, Brazil)) Brazil)), 2021-03, Vol.25 (2), p.186-193
• Main Authors: Marques, Larissa Bragança Falcão, Moreira, Bruno de Souza, Ocarino, Juliana de Melo, Sampaio, Rosana Ferreira, Bastone, Alessandra de Carvalho, Kirkwood, Renata Noce
• Format: Article
• Language: English
• Subjects: Accidental falls; Aged; Balance; Clinimetric properties; Gait; Original Research; Rehabilitation
• ISSN: 1413-3555; 1809-9246
• DOI: 10.1016/j.bjpt.2020.05.008

•FGA-Brazil showed moderate to high correlation with gait and balance tests.•FGA-Brazil differentiate between older adults with high and low concern about falls.•FGA-Brazil is able to predict future falls with high accuracy in the older population.•The optimum cutoff score of the FGA-Brazil for predicting falls is ≤22.•FGA-Brazil did not exhibit ceiling and floor effects. The functional gait assessment (FGA) is a reliable instrument to evaluate walking balance in the Brazilian older population. However, other measurement properties need investigation. To determine the construct and criterion validity of the FGA-Brazil and its ceiling and floor effects. Sociodemographic, clinical, and anthropometric data were collected from 126 older adults. Participants completed the Mini-mental state examination followed by the FGA-Brazil, Berg balance scale (BBS), gait speed, and the Falls Efficacy Scale-International. Six months later, the participants were interviewed by telephone about their fall history. Exploratory factor analysis was used to determine the structural validity. We also determined the construct validity of the FGA-Brazil, using hypothesis testing, by investigating the differences between groups using the Mann–Whitney U test. Criterion validity was determined using the Spearman correlation between the FGA-Brazil and the other balance and gait measures, and using the Receiver Operator Characteristic curve. Participants’ mean age was 69.3 ± 7.4 years, and 84 (69.4%) were female. Factor analysis resulted in two factors explaining 53.3% of the total variance. Moderate and high significant correlations were found between the FGA-Brazil and gait speed (r = 0.65) and BBS (r = 0.80). A significant difference in the FGA-Brazil median score between older adults with low and high concern about falls was observed. The cutoff score recommended for predicting falls was 22 or less. No ceiling and floor effects were observed. We recommend the FGA-Brazil to determine the risk of falls in community-dwelling older adults.