Predictive Accuracy of the 3-m Backward Walk Test for Fall Risk in Older Adults

Falls are a leading cause of injury in older adults, and accurate tools to predict fall risk are essential. This study evaluates the predictive accuracy of the 3-m Backward Walk Test (3 MBWT) for fall risk, comparing it with the Short Physical Performance Battery (SPPB), 10-m walk test (10 MWT), and fall-related questionnaires.

Cross-sectional study.

The study was conducted in 2 nursing homes in Vienna, Austria. A total of 217 participants (65.9% female, median age, 80 years) were included.

Fall history was recorded retrospectively over 12 months and prospectively over 4 months using interviews and a Cogvis 3D sensor system. Predictive accuracy was assessed with receiver operating characteristic curves and area under the curve (AUC) analyses, whereas Poisson regression was used to analyze fall incidence.

Among participants, 86 (39.6%) reported a fall within the previous 12 months. For retrospective falls, the SPPB (AUC, 0.76), 3 MBWT (AUC, 0.75), and 10 MWT (AUC, 0.73) showed similar predictive accuracy. For prospective falls, the SPPB (AUC, 0.75) and 3 MBWT (AUC, 0.74) remained similarly accurate, whereas the 10 MWT had lower predictive power (AUC, 0.67). For the 3 MBWT, cutoff points were ≥9.31 seconds for retrospective falls (32% higher incidence; 95% CI, 1.05-1.95) and ≥8.15 seconds for prospective falls (27% higher incidence; 95% CI, 1.02-1.66). Each additional second in 3 MBWT performance increased fall incidence (retrospective, 7%; prospective, 5%). SPPB cutoffs of <7.02 and <8.00 were associated with 49% (95% CI, 1.06-2.21) and 30% (95% CI, 1.02-1.74) higher fall incidence, respectively.

The 3 MBWT shows comparable accuracy with the SPPB and moderately better performance than the 10 MWT for prospective falls. Its efficiency, predictive value, and time savings make it a practical and viable tool for clinical fall risk assessment in older adults.