Validity and reliability of inertial measurement units on gait, static balance and functional mobility performance among community-dwelling older adults: a systematic review and meta-analysis

Lulu Yin, Xin Xu, Wang Rui-yan, Li Fei-Fei, Yushan Wang, Lin WangEFORT 2025 · DOI

TL;DR

Inertial Measurement Units (IMUs) are proving to be a reliable and valid tool for assessing the gait of community-dwelling older adults, offering a promising alternative to traditional methods. While IMUs show potential for evaluating balance and functional mobility, further standardization and research are needed to fully unlock their capabilities in these areas.

What Problem Does This Paper Address?

As the global population ages, understanding and monitoring the mobility of older adults is crucial for preventing falls, maintaining independence, and improving quality of life. Traditional methods for assessing gait, balance, and functional mobility, such as observational techniques or laboratory-based equipment, have limitations in terms of subjectivity, cost, complexity, and portability. Wearable technologies like IMUs offer a potential solution, providing accurate and portable data. However, the scientific community needs a clear understanding of how valid and reliable IMU-based measurements are for this specific demographic across various mobility tasks. This paper addresses this gap by systematically reviewing and meta-analyzing existing research on IMU performance in community-dwelling older adults.

Methods

This study employed a systematic review and meta-analysis approach. Researchers searched multiple electronic databases for relevant studies. They then screened articles based on specific inclusion and exclusion criteria, focusing on studies that evaluated the validity or reliability of IMUs for measuring gait, static balance, and functional mobility in older adults, comparing IMU data to gold-standard reference systems. For quantitative synthesis, they performed meta-analyses on outcomes with sufficient data, using intraclass correlation coefficients (ICCs) for validity and Pearson correlation coefficients (r) for reliability. They also conducted qualitative analyses to interpret results where quantitative pooling was not possible.

Datasets

Not discussed in the paper. The study is a systematic review and meta-analysis of existing research, not a study that collected new data or established new datasets.

Key Insights and Ideas

  • Gait is a Strong Suit for IMUs: IMUs are demonstrating excellent validity and reliability for measuring fundamental gait parameters like walking speed, cadence, and step/stride length. This suggests they are a robust tool for tracking basic walking performance in older adults.
  • Variability and Symmetry Need Caution: While IMUs are good for measuring average gait metrics, their use for assessing gait variability and symmetry requires more careful consideration. The data suggests these metrics might be less consistently reliable or valid with current IMU technology.
  • Balance and Functional Mobility: Promising but Not Yet Perfect: For static balance and functional mobility tasks (like sit-to-stand and timed up-and-go tests), IMUs show moderate-to-good validity and reliability. This indicates potential, but the results are not as consistently strong as for gait, suggesting a need for further refinement and standardization.
  • Quality Matters: The review highlighted that many included studies were of moderate or low quality, often lacking detailed reporting on sensor calibration, sample size calculations, and statistical analyses. This inconsistency can impact the overall confidence in the findings.
  • Standardization is Key: The paper emphasizes the need for standardized protocols, IMU placement, and calibration procedures to improve the comparability and reliability of IMU data across different studies and devices.

Key Results

  • Gait Spatiotemporal Outcomes: IMU-derived metrics for walking speed, cadence, step/stride time, step time variability, stance time, and step/stride length showed excellent (ICCs ≥ 0.954) to good-to-excellent (r ≥ 0.869) agreement with gold standards for validity. For reliability, walking speed, cadence, step/stride time, stance/swing time, and step/stride length demonstrated excellent test-retest consistency (summary ICCs ≥ 0.941).
  • Functional Mobility Tests:
    • Sit-to-Stand (STS) Test: Individual STS duration showed good-to-excellent agreement (r = 0.989) for validity and excellent test-retest consistency (summary ICC = 0.935) for reliability.
  • Timed Up and Go (TUGT) Test: TUGT duration exhibited excellent agreement (r = 0.959) for validity and excellent test-retest consistency (summary ICC = 0.935) for reliability.
  • Static Balance: While no quantitative pooling was possible due to inconsistencies, moderate-to-strong validity was generally observed for IMUs in static balance measurements.
  • Study Quality: Out of 56 included studies, 36 were rated as moderate quality (MQ) and 20 as low quality (LQ). No studies were rated as high quality (HQ) or very low quality (VLQ).

Limitations and Future Work

  • Study Quality and Reporting: A significant limitation is the moderate-to-low quality of many included studies, with a lack of detailed reporting on crucial aspects like sensor calibration, sample size justification, and statistical methods. This hinders definitive conclusions.
  • Heterogeneity: The studies exhibited considerable heterogeneity in test conditions, IMU brands, sensor placement, and participant characteristics, making direct comparisons challenging.
  • Focus on Healthy Adults: The review primarily included data from healthy older adults, limiting generalizability to those with chronic conditions or in clinical settings.
  • Laboratory Settings: Most studies were conducted in controlled laboratory environments, which may not fully represent real-world IMU performance.
  • Algorithm Specificity: The review did not delve into the specifics of IMU algorithms, which can significantly influence data accuracy. Future research should explore the impact of different algorithms.
  • Future Directions: The authors recommend future research to prioritize large, well-powered studies, standardize IMU placement and calibration, and conduct further validation in real-world settings. Investigating the impact of different sensor algorithms and focusing on improving the accuracy of variability and symmetry metrics are also suggested.

Summary updated

Abstract

Due to consistently high levels of validity and reliability, the present study supported the use of IMUs for measuring gait spatiotemporal outcomes. However, caution was advised when applying spatiotemporal variability and symmetry metrics. In addition, characterized by moderate-to-good validity and reliability, current review provides evidence of a neutral nature regarding the utilization of IMUs for static balance and functional mobility performance.
BibTeX
@inproceedings{yin2025validity,
  author = {Lulu Yin and Xin Xu and Wang Rui-yan and Li Fei-Fei and Yushan Wang and Lin Wang},
  title = {{Validity and reliability of inertial measurement units on gait, static balance and functional mobility performance among community-dwelling older adults: a systematic review and meta-analysis}},
  booktitle = {EFORT},
  year = {2025},
  doi = {10.1530/eor-2024-0088},
}
Figure