Fall risk assessment has been incorporated into the national standard: When the TUG test does not reveal any breakdown in gait or balance, how can we determine why the person fell?
Release time:
2026-10-20 15:16
Source:
Effective September 1, 2026, the health industry standard WS/T 887—2026, “Technical Standard for Fall Prevention and Control Among Community-Dwelling Older Adults,” will come into force. Assessments of “gait and balance function” become mandatory components in evaluating fall risk among older adults living in the community, with risk categorized into low, medium, and high levels; individuals at high risk must undergo follow-up reassessment within one to three months. However, the most commonly used tools in institutional settings remain scales such as the TUG (Timed Up and Go) test—providing only a time measurement or score without explaining “why they fell.” This article clarifies three key points by reviewing nine recent studies published between 2025 and 2026: Why fall risk assessment has been incorporated into national standards, what is missing in the TUG test’s timing metrics, and what does the plan look like to advance from “classification” to “decomposition”? 。
01 Why is fall risk assessment included in the national standard?
WHY GAIT AND BALANCE ENTERED THE STANDARD
First, consider the extent of the burden. In Germany, approximately 24% of people aged 65 and older experience at least one fall per year, a figure that rises to about 33% among those aged 80 and older. [1] ; The annual economic burden of falls in the United States exceeds $280 billion, with direct medical costs exceeding $50 billion. [2] . Falls are not isolated accidents; rather, they represent one of the endpoints in a measurable chain of functional decline.
Why specifically “gait and balance”? Because this is where interventions can be targeted. A prospective study of elderly patients discharged after a fall found that the association between depressive symptoms and fall risk (OR = 1.46 in a meta-analysis of 20 studies) was primarily mediated by impairments in gait and dynamic balance. [1] — The intermediate links through which risk propagates are precisely measurable, trainable motor functions. Gait speed has long been referred to as the “sixth vital sign,” and wearable inertial sensors can achieve an AUC of 0.75–0.88 in predicting fall risk among individuals with vestibular and neurological impairments. [3] 。

Figure 1: Digitized gait biomarkers output by wearable devices: step count, distance, speed, cadence, stride length, and variability metrics (adapted from reference [3]).
The quantitative feasibility in a community setting has also been validated: the path length of the center of pressure (COP) measured using a low-cost balance board was significantly correlated with the fall index from professional posturography (r = 0.349–0.561), with an AUC of 0.783–0.865 for discriminating risk status. [4] The national standard lists gait and balance as mandatory assessment items, predicated on the premise that such objective measurement methods have matured sufficiently to be implemented at the grassroots level.
02 Challenges Encountered in TUG Assessment
WHAT A STOPWATCH CANNOT TELL
The TUG test excels in grading and follow-up. A one-year community-based follow-up study of older adults showed that a decline of ≥2 points in TUG-assessed gait stability over the year was associated with falls. [5] However, it follows a “total‑sum” logic: a single stopwatch cannot pinpoint whether the slowdown originates in a particular leg, joint, or control strategy.
The first layer of information that was erased was the left–right difference. Analysis of local dynamic stability in 131 older adults revealed significant left–right interaction differences in long-term stability in the medial–lateral direction between fallers and non-fallers. [6] — Asymmetry is embedded in the division of labor between the two legs; it doesn’t appear in TUG times.
More sensitive signals are found in angular velocity rather than time. When elderly individuals with knee osteoarthritis and age-matched controls perform balance board, single-leg stance, TUG, and walking tasks, the absolute angular velocity of most body segments is generally higher. [7] — Compensatory postural control is exacerbated, but the TUG test may not reveal any abnormalities.

Figure 2: Balance and gait assessment based on inertial measurement units: Sensors are placed on the head, chest, waist, both knees, both ankles, and the plantar surface of the feet (adapted from reference [7]).
Scales can answer “whether to intervene,” but not “where to intervene”—and this is precisely the gap between stratification and diagnosis.
The necessity of multidimensional input has been quantified: in a study involving 276 women aged 60–80, the correlation coefficients between scores on 12 balance scales and gait parameters ranged from r=0.35 to 0.67, and an ensemble machine learning model achieved an AUC of 0.91 for classifying gait stability. [8] Only with a sufficiently large number of input dimensions can the prediction be reliable—single TUGs provide far too few dimensions.
03 Solutions for Identifying the Root Causes of Falls
FROM SCORING TO DIAGNOSING
Break the assessment down into three tiers: the national standard scale is responsible for Grading , responsible for motion capture decomposition , inverse dynamics is responsible for Attribution The decomposition layer outputs joint angle curves, left–right symmetry indices, and trunk control metrics; the attribution layer calculates “which muscle bears what load.” Multidimensional assessment studies have established a standard paradigm: a 12-meter standardized walkway equipped with dual force platforms and an optical motion-capture camera array can simultaneously acquire comprehensive kinematic and kinetic data (Figure 3). [8] 。

Figure 3: Experimental setup for standardized gait analysis: a 12-meter walkway, dual force platforms, and an array of optical motion-capture cameras (reproduced from reference [8]).
At the collection end, SMART-DX EVO Distributed Intelligent Motion Capture System Arranged in a distributed configuration across multiple positions and sharing the same time base as force plates and inertial sensors, it can be rapidly deployed in non-laboratory settings such as community rehabilitation rooms and day-care centers—compressing the laboratory‑grade setup shown above into a community‑scale solution.
On the analysis side, BOB Human Movement Biomechanics Analysis Software Inverse dynamics are performed on a musculoskeletal model containing over 600 motor units, generating joint angle and torque curves and providing symmetry metrics separately for the left and right sides. For high-risk elderly patients undergoing reassessment after one to three months, the process shifts from “recalculating the entire score” to “comparing the two curves side by side.”
The value of digital biomarkers is most evident in intervention follow-up: a vestibular rehabilitation study used a head-mounted sensor to quantify head motion stabilization during gait; after 18 treatment sessions, 30 out of 32 participants showed improvements in balance and gait, and the number of individuals at “high risk of falling” decreased from 19 to 4. [2] . The risk level is changing, and more importantly, each change itself can be quantified and recorded.

Figure 4: Head-mounted sensor device used to assess head motion stabilization during gait (adapted from reference [2])
Taking one step further leads to proactive intervention: research on the deployment of multimodal deep learning combined with wearable sensing shows that, within three months, the elderly group receiving real-time alerts experienced 62% fewer falls than the control group. [9] What’s always lacking isn’t algorithms, but rather turning every step a senior takes once they walk through the door into comparable data.
The national standard provides the mandatory questions for assessment, but “decomposition” and “attribution” are what truly answer the question of “why the fall occurred.”
References
[1] Association of depressive symptoms, gait and dynamic balance with 12-month fall risk in older adults after emergency presentation. BMC Geriatrics, 2026. DOI:10.1186/s12877-026-07670-w
[2] Quantifying Changes in Balance, Gait, Fall Risk and Sensory Reweighting With a Digital Biomarker. Medical Devices: Evidence and Research, 2026. DOI:10.2147/mder.s590344
[3] Gait as a vital sign: integrating wearables and digital biomarkers. Frontiers in Neurology, 2026. DOI:10.3389/fneur.2026.1736898
[4] Validity of Center of Pressure Path Length Measured Using a Wii Balance Board for Fall Risk Screening. Healthcare, 2026. DOI:10.3390/healthcare14121685
[5] Predictive Ability of Systems of Postural Control for 1-Year Risk of Falls and Frailty in Community-Dwelling Older Adults. Journal of Ageing and Longevity, 2025. DOI:10.3390/jal5040045
[6] Functional differences between lower limbs during gait in relation to fall risk. Frontiers in Physiology, 2026. DOI:10.3389/fphys.2026.1743503
[7] Assessment of Gait and Balance in Elderly Individuals with Knee Osteoarthritis Using Wearable Inertial Sensors. Sensors, 2025. DOI:10.3390/s25206288
[8] Multidimensional balance assessment and machine learning gait stability prediction in elderly women. Scientific Reports, 2026. DOI:10.1038/s41598-026-52352-7
[9] Intelligent fall risk prediction and real-time warning system for elderly care based on multimodal deep learning and wearable sensor fusion. Scientific Reports, 2026. DOI:10.1038/s41598-026-56750-9
Fall Risk Assessment,TUG,Gait and Balance,Motion capture,Biomechanical Analysis of Movement