Validade de construto estrutural, convergente e divergente e de critério preditiva da Functional Gait Assessment-Brasil em idosos da comunidade

Abstract

The aging of the population is associated with increased comorbidities, disabilities and also the increase in the number of falls. Falls cause serious injuries, fractures and fear of relapse that can lead the elderly to functional dependence, and represent a major cause of death in this population. Early identification of individuals at risk of falling is essential for promoting effective interventions and minimizing consequences. The Functional Gait Assessment (FGA) assesses postural balance during gait and determines the risk of falls in adults over 60 years. The purpose of this study was to determine the structural, convergent, divergent and predictive validity of the FGA-Brazil in community-dwelling older adults. A methodological study was conducted and data were collected at the Ambulatório da Faculdade de Ciências Médicas de Minas Gerais. One hundred and twenty-one older adults aged 60 to 90 years completed the FGA-Brazil, Mini-mental state examination (MMSE), the Berg Balance Scale (BBS), the 4-meter walk test and Falls Efficacy Scale – International (FES-I- Brazil) during one session. Exploratory factor analysis was used to determine construct validity. Mann–Whitney-test and Spearman correlation coefficients were used to determine convergent and divergent validity between the BBS, gait speed, FES-I Brazil and FGA-Brazil. Six months later, the patients were interviewed by telephone and information on their fall status collected. Based on the occurrence of not of fall, the sample was divided into two groups. Following, a receiver operator curve (ROC) and the area under the curve were used to determine the predictive validity of the FGA in classifying fall risk in older adults. A ceiling effect was considered if 15% of the sample had scores > 27, and a floor effect if 15% had scores < 3. The results showed that for the construct validity 2 common factors were extracted, explaining 53.3% of the total variance. Only items 3 and 4 (gait with horizontal and vertical rotation of the head) measured different dimensions from the other items on the scale. For convergent validity, there was a high correlation between FGA-Brazil and the Berg Balance Scale (r = 0.802, p <0.001) and moderate with the gait speed test (r = 0.651, p <0.001). For the divergent construct validity, the sample was divided into two groups based on the FES-I Brazil cutoff of ≤ 20 for low concern and >20 for high concern regarding falls. The results of the Mann Whitney test showed that there is a significant difference (p <0.001) between the groups indicating that FGA-Brazil was effective in separating the groups with low concern and with high concern about falls. Data for predictive validity were obtained from 116 participants. Data from five participants could not be obtained due to contact failure or changed phone number. Of those, 27 older adults (23.3%) reported 1 or more falls during the 6 months. The ROC curve showed that the area under the curve was significant (AUC=0.923, p = 0.001) indicating that FGA-Brazil was able to predict falls in this group of elderly. The sensitivity of the test was 74.1% and specificity was 95.5%, with a cut off score of the FGA-Brazil  17. However, we recommend the use of the cut-off point  22, with greater sensitivity and able to identify 26/27 older adults who suffered falls. The interpretation of the total FGA scores for 121 elderly revealed that none of the participants scored < 3 and only 5.8% of the older adults scored >27 on the FGA-Brazil. Therefore, the FGA-Brazil had no ceiling and floor effects in the sample studied. From the results found, we can conclude that the FGA-Brazil is a valid scale that evaluates the concept of risk of falls, besides demonstrating moderate to high correlation with other measuring instruments that evaluate the same construct. Furthermore, FGA-Brazil was able to predict the risk of falls in elderly Brazilians. Therefore, our results provide sufficient evidence to support the use of FGA-Brazil in the assessment of balance and gait in the community's elderly.