Characterization of Postural Sway in Women with Osteoporosis and a Control Group by Means of Linear and Nonlinear Methods.
Felix StiefAnna SohnLutz VogtAndrea MeurerMarietta KirchnerPublished in: Bioengineering (Basel, Switzerland) (2023)
The mechanisms underlying the altered postural control and risk of falling in patients with osteoporosis are not yet fully understood. The aim of the present study was to investigate postural sway in women with osteoporosis and a control group. The postural sway of 41 women with osteoporosis (17 fallers and 24 non-fallers) and 19 healthy controls was measured in a static standing task with a force plate. The amount of sway was characterized by traditional (linear) center-of-pressure (COP) parameters. Structural (nonlinear) COP methods include spectral analysis by means of a 12-level wavelet transform and a regularity analysis via multiscale entropy (MSE) with determination of the complexity index. Patients showed increased body sway in the medial-lateral (ML) direction (standard deviation in mm: 2.63 ± 1.00 vs. 2.00 ± 0.58, p = 0.021; range of motion in mm: 15.33 ± 5.58 vs. 10.86 ± 3.14, p = 0.002) and more irregular sway in the anterior-posterior (AP) direction (complexity index: 13.75 ± 2.19 vs. 11.18 ± 4.44, p = 0.027) relative to controls. Fallers showed higher-frequency responses than non-fallers in the AP direction. Thus, postural sway is differently affected by osteoporosis in the ML and AP directions. Clinically, effective assessment and rehabilitation of balance disorders can benefit from an extended analysis of postural control with nonlinear methods, which may also contribute to the improvement of risk profiles or a screening tool for the identification of high-risk fallers, thereby prevent fractures in women with osteoporosis.
Keyphrases
- postmenopausal women
- bone mineral density
- end stage renal disease
- newly diagnosed
- chronic kidney disease
- magnetic resonance imaging
- high resolution
- machine learning
- body composition
- ejection fraction
- minimally invasive
- magnetic resonance
- single molecule
- optical coherence tomography
- neural network
- prognostic factors
- convolutional neural network