Added value of a triaxial accelerometer assessing gait parameters to predict falls and mortality among nursing home residents: A two-year prospective study

Article type: Research Article

Authors: Buckinx, F.^{a; b; *} | Beaudart, C.^{a; b} | Slomian, J.^{a; b} | Maquet, D.^c | Demonceau, M.^c | Gillain, S.^d | Petermans, J.^d | Reginster, J.Y.^{a; b} | Bruyère, O.^{a; b; c}

Affiliations: [a] Department of Public Health, Epidemiology and Health Economics, University of Liège, Liège, Belgium | [b] Support Unit in Epidemiology and Biostatistics, University of Liège, Liège, Belgium | [c] Rehabilitation and Movement Sciences, University of Liège, Liège, Belgium | [d] Geriatric Department, CHU of Liège, Liège, Belgium

Correspondence: [*] Corresponding author: Fanny Buckinx, University of Liège, Belgium. Tel.: +32 4 366 49 33; E-mail: fanny.buckinx@ulg.ac.be.

Abstract: Background:Gait impairment seems to be a risk factor for falls and mortality. Because gait change cannot be determined easily with classical clinical tests, some authors have suggested that it might be useful to use a gait-analysis system among elderly community-dwelling people. Objective:The main objective of the present study was to determine the predictive value of a quantitative evaluation of the gait characteristics in nursing home residents for the occurrence of falls and death performed using a tri-axial accelerometer (Locométrix®). Material and methods:One hundred elderly nursing home residents (80 women and 20 men, mean age 86.4 ± 6.04 years) were included in this study with the aim to follow them for 2 years. Deaths and falls were systematically recorded. A quantitative evaluation of a 10-second walk was performed with a tri-axial accelerometer (Locometrix®). Demographic data (i.e age, sex, body mass index) and clinical data (i.e. fall risk evaluated by the Tinetti test) were also recorded. Results:During the two years of follow-up, 27 patients died. After adjustment on all potential confounding variables, only body mass index was significantly associated with the risk of mortality with an odds ratio of 0.86 (95% CI: 0.77–0.96, p = 0.04). At the end of the study period, 440 falls had occurred (mean: 4.44 ± 6.79 falls per patient) but no single factors were independently associated with fall incidence. Conclusion:Our results show that a quantitative gait analysis performed using a tri-axial accelerometer is not predictive of long-term falls and mortality among nursing home residents.

Keywords: Mortality, falls, risk factors, nursing home

DOI: 10.3233/THC-140883

Journal: Technology and Health Care, vol. 23, no. 2, pp. 195-203, 2015