RESUMO
OBJECTIVE: To systematically examine stepping performance as a risk factor for falls. More specifically, we examined (i) if step tests can distinguish fallers from non-fallers and (ii) the type of step test (e.g. volitional vs reactive stepping) that is required to distinguish fallers from non-fallers. DATA SOURCE: PubMed, EMBASE, CINAHL, Cochrane Database of Systematic Reviews and reference lists of included articles. STUDY SELECTION: Cross-sectional and cohort studies that assessed the association between at least one step test and falls in older people (age ≥ 60 and/or mean age of 65). RESULTS: A meta-analysis of 61 studies (n = 9536) showed stepping performance was significantly worse in fallers compared to non-fallers (Cohen'sd 0.56, 95 % CI 0.48 to 0.64, p < 0.001, I2 66 %). This was the case for both volitional and reactive step tests. Twenty-three studies (n = 3615) were included in a diagnostic meta-analysis that showed that step tests have moderate sensitivity (0.70, 95 % CI 0.62 to 0.77), specificity (0.68, 95 % CI 0.58 to 0.77) and area under the receiver operating characteristics curve (AUC) (0.75, 95 % CI 0.59 to 0.86) in discriminating fallers from non-fallers. CONCLUSIONS: This large systematic review demonstrated that both volitional and reactive stepping impairments are significant fall risk factors among older adults. Step tests can identify fallers from non-fallers with moderate accuracy.
