Posted in | News | Biosensors

Sensor-Based Nonlinear Metrics Improve Fall Risk Assessment

Advanced sensor-based non-linear center-of-pressure metrics capture subtle postural control patterns missed by traditional measures. While individual metrics show limited differences, combined machine learning analysis improves fall risk classification. These findings suggest complexity-based sensor data can enhance clinical assessment of balance and aging-related fall risk.

Study: Non-Linear Center-of-Pressure Features Associated with Fall History in Older Adults: An Exploratory Analysis. Image Credit: Digineer Station/Shutterstock

In a recent article published in the journal Sensors, researchers explored how non-linear center-of-pressure features, reflecting the temporal organization of postural sway, relate to fall history in older adults, suggesting these measures may provide complementary insights beyond conventional sway magnitude metrics.

Saving this for later? Download a PDF here.

Background

Falls among older adults present a significant public health concern due to their frequent association with injury, reduced mobility, and loss of independence. Assessing and mitigating fall risk requires reliable evaluation of postural control, commonly quantified by analyzing postural sway through center-of-pressure (CoP) data.

Recent advancements in sensor technology have enabled more detailed recording of CoP trajectories, facilitating the application of non-linear analytical methods to quantify the temporal and structural complexity of postural sway.

Drawing on the loss-of-complexity hypothesis, which posits that physiological systems tend to lose multi-scale variability and adaptability with aging and pathology, researchers have begun utilizing non-linear metrics to characterize complexity, regularity, and long-range temporal correlations in CoP signals.

These metrics, such as recurrence quantification analysis (RQA), detrended fluctuation analysis (DFA), fractal dimension (FD), multiscale entropy (MSE), stabilogram diffusion analysis (SDA), and sway density curve (SDC), provide insights into the dynamical patterns governing postural control beyond simple sway magnitude.

The Current Study

This study conducted a secondary analysis of an open-access CoP dataset originally collected from community-dwelling adults aged 18 to 85 years using a force platform (AMTI). The focus was narrowed to individuals aged 60 years and older to investigate associations with fall history.

Data acquisition involved 60-second quiet-standing trials on a firm surface under both eyes-open and eyes-closed conditions. The CoP time series in mediolateral and anteroposterior directions were preprocessed with a 20 Hz low-pass Butterworth filter to remove noise. Linear features calculated included mean CoP velocity, sway area, and velocity variability, as well as frequency-based measures.

The application of non-linear analyses entailed calculating multiple indices: SDC-based measures representing postural stabilization, fractal dimension estimating geometric complexity, DFA scaling exponents for long-range correlations, SDA metrics capturing diffusion characteristics of sway trajectories, and MSE values reflecting entropy across multiple temporal scales.

Additionally, RQA was used to quantify determinism and laminarity in CoP signals via state-space reconstruction with optimized embedding parameters. To reduce confounding, propensity score matching was used to balance covariates, including age, sex, BMI, health status, and medication use, between fallers (≥1 fall in the past year) and non-fallers.

Statistical comparisons employed univariate tests following matching, and multivariate machine learning models interpreted with SHapley Additive exPlanations (SHAP) quantified feature contributions.

Results and Discussion

After propensity score matching, 18 matched pairs of fallers and non-fallers were analyzed. Conventional linear sway metrics, including velocity and sway area, did not differ significantly between groups for either visual condition.

Similarly, none of the individual non-linear metrics showed robust univariate associations with fall history, which may reflect multiple factors including sample size limitations, residual confounding, and the complexity of underlying control strategies.

Despite these null univariate results, the multivariate model interpretation revealed that certain non-linear features, particularly those related to temporal structure and multi-scale patterns - such as short-time-scale multiscale entropy, fractal dimension, and RQA indices of regularity - contributed more substantially to separating fallers from non-fallers.

Under eyes-closed conditions, longer time-scale entropy components and DFA scaling exponents were relatively more prominent, suggesting that sensory integration deficits influencing temporal complexity may relate to fall risk when visual information is absent.

These outcomes emphasize that postural sway magnitude alone may not capture fall-related changes in postural control. The sensor-derived non-linear metrics tap into the organization and quality of sway variability, reflecting subtle alterations in adaptive control mechanisms.

This aligns with sensorimotor control theories and supports the enhanced value of integrating sensor-acquired time-series data with complexity-based analyses. However, the findings should be interpreted cautiously due to the exploratory nature, small sample size, and reliance on retrospective self-reported falls, which may be subject to bias.

Additionally, challenging standing conditions that impose greater sensorimotor demands were not assessed, which could reveal stronger differences.

Conclusion

This study highlights the potential of advanced sensor-derived non-linear center-of-pressure features to complement traditional linear sway metrics in characterizing balance and fall history among older adults.

While conventional magnitude-based measures proved insufficient to differentiate fallers and non-fallers during quiet standing on a firm surface, combining multiple non-linear temporal complexity metrics extracted from force-plate CoP recordings provided additional explanatory value.

These findings suggest that incorporating sensor-enhanced complexity analyses could refine postural assessment frameworks, potentially improving fall risk evaluation. However, the results are preliminary and intended to stimulate further research with larger, prospective cohorts and diversified postural tasks.

Emphasizing reproducible sensor data processing and a parsimonious set of well-characterized features will be essential to translating complexity-based CoP analyses into clinically meaningful fall risk assessment tools.

Journal Reference

Wakabayashi D., Okada Y. (2026). Non-Linear Center-of-Pressure Features Associated with Fall History in Older Adults: An Exploratory Analysis. Sensors 26(8):2298. DOI: 10.3390/s26082298, https://www.mdpi.com/1424-8220/26/8/2298

Dr. Noopur Jain

Written by

Dr. Noopur Jain

Dr. Noopur Jain is an accomplished Scientific Writer based in the city of New Delhi, India. With a Ph.D. in Materials Science, she brings a depth of knowledge and experience in electron microscopy, catalysis, and soft materials. Her scientific publishing record is a testament to her dedication and expertise in the field. Additionally, she has hands-on experience in the field of chemical formulations, microscopy technique development and statistical analysis.    

Citations

Please use one of the following formats to cite this article in your essay, paper or report:

  • APA

    Jain, Noopur. (2026, April 14). Sensor-Based Nonlinear Metrics Improve Fall Risk Assessment. AZoSensors. Retrieved on April 18, 2026 from https://www.azosensors.com/news.aspx?newsID=16811.

  • MLA

    Jain, Noopur. "Sensor-Based Nonlinear Metrics Improve Fall Risk Assessment". AZoSensors. 18 April 2026. <https://www.azosensors.com/news.aspx?newsID=16811>.

  • Chicago

    Jain, Noopur. "Sensor-Based Nonlinear Metrics Improve Fall Risk Assessment". AZoSensors. https://www.azosensors.com/news.aspx?newsID=16811. (accessed April 18, 2026).

  • Harvard

    Jain, Noopur. 2026. Sensor-Based Nonlinear Metrics Improve Fall Risk Assessment. AZoSensors, viewed 18 April 2026, https://www.azosensors.com/news.aspx?newsID=16811.

Tell Us What You Think

Do you have a review, update or anything you would like to add to this news story?

Leave your feedback
Your comment type
Submit

Sign in to keep reading

We're committed to providing free access to quality science. By registering and providing insight into your preferences you're joining a community of over 1m science interested individuals and help us to provide you with insightful content whilst keeping our service free.

or

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.