Titre
Gait disturbances as specific predictive markers of the first fall onset in elderly people: a two-year prospective observational study
Type
article
Institution
UNIL/CHUV/Unisanté + institutions partenaires
Périodique
Auteur(s)
Mignardot, J.B.
Auteure/Auteur
Deschamps, T.
Auteure/Auteur
Barrey, E.
Auteure/Auteur
Auvinet, B.
Auteure/Auteur
Berrut, G.
Auteure/Auteur
Cornu, C.
Auteure/Auteur
Constans, T.
Auteure/Auteur
de Decker, L.
Auteure/Auteur
Liens vers les unités
ISSN
1663-4365
Statut éditorial
Publié
Date de publication
2014
Volume
6
Première page
22
Peer-reviewed
Oui
Langue
anglais
Résumé
Falls are common in the elderly, and potentially result in injury and disability. Thus, preventing falls as soon as possible in older adults is a public health priority, yet there is no specific marker that is predictive of the first fall onset. We hypothesized that gait features should be the most relevant variables for predicting the first fall. Clinical baseline characteristics (e.g., gender, cognitive function) were assessed in 259 home-dwelling people aged 66 to 75 that had never fallen. Likewise, global kinetic behavior of gait was recorded from 22 variables in 1036 walking tests with an accelerometric gait analysis system. Afterward, monthly telephone monitoring reported the date of the first fall over 24 months. A principal components analysis was used to assess the relationship between gait variables and fall status in four groups: non-fallers, fallers from 0 to 6 months, fallers from 6 to 12 months and fallers from 12 to 24 months. The association of significant principal components (PC) with an increased risk of first fall was then evaluated using the area under the Receiver Operator Characteristic Curve (ROC). No effect of clinical confounding variables was shown as a function of groups. An eigenvalue decomposition of the correlation matrix identified a large statistical PC1 (termed "Global kinetics of gait pattern"), which accounted for 36.7% of total variance. Principal component loadings also revealed a PC2 (12.6% of total variance), related to the "Global gait regularity." Subsequent ANOVAs showed that only PC1 discriminated the fall status during the first 6 months, while PC2 discriminated the first fall onset between 6 and 12 months. After one year, any PC was associated with falls. These results were bolstered by the ROC analyses, showing good predictive models of the first fall during the first six months or from 6 to 12 months. Overall, these findings suggest that the performance of a standardized walking test at least once a year is essential for fall prevention.
PID Serval
serval:BIB_FFBB8F209A19
PMID
Open Access
Oui
Date de création
2014-04-10T08:03:19.626Z
Date de création dans IRIS
2025-05-21T05:54:24Z
Fichier(s)![Vignette d'image]()
En cours de chargement...
Nom
BIB_FFBB8F209A19.P001.pdf
Version du manuscrit
published
Taille
2.64 MB
Format
Adobe PDF
PID Serval
serval:BIB_FFBB8F209A19.P001
URN
urn:nbn:ch:serval-BIB_FFBB8F209A196
Somme de contrôle
(MD5):2203b4f534ccc3bfae50aab201daa28d