Locomotive dysfunction tracked with both falls and stumbling in community-dwelling adults
A cross-sectional analysis of 531 community-dwelling adults draws a sharp line between two events that are often collapsed into a single endpoint: falls and stumbling. Only falls directly injure, yet both reflect instability in daily movement, and separating them exposed a clearer clinical signal than a combined outcome ever could.
In this cohort from the 2021 Iwaki Health Promotion Project, 83 participants reported a fall, 165 reported stumbling, and 188 reported either event. The pattern is straightforward but clinically important. A person can stumble repeatedly without ever sustaining an injury, yet those episodes may still flag declining locomotive reserve, impaired balance, or compensatory movement strategies that precede a true fall.
GLFS-25 moved with every outcome in every age stratum
The strongest and most consistent association came from the 25-question Geriatric Locomotive Function Scale (GLFS-25), a questionnaire that can be administered without equipment. Across age strata, GLFS-25 was associated with all outcomes. Each 1-point increase corresponded to an adjusted odds ratio of 1.095 (95% CI 1.055–1.137) for either event and 1.062 (1.028–1.098) for falls alone.
Those figures matter because they point to a scalable bedside and community-screening tool. Rather than requiring imaging or specialized balance platforms, the GLFS-25 captures the patient’s own lived experience of movement limitation, pain, and functional difficulty. In practical terms, it offers a way to identify adults whose musculoskeletal decline is already translating into real-world instability.
The signal persisted even when the questionnaire was stripped of its two anxiety items. Area under the curve moved only from 0.702 to 0.682, suggesting that the scale’s predictive value did not depend heavily on those questions. The core message is anatomical and functional rather than psychological: locomotor impairment itself appears to be driving much of the risk signal.
Stumbling and falls did not behave identically
The study design matters because the two outcomes were asked separately, and they did not line up perfectly. Falls and stumbling are related, but they are not interchangeable. A person may stumble frequently because of minor gait disruption, postural correction, or spinal compensation, yet never progress to an injurious fall. Another may have a single catastrophic fall without a long history of stumbling.
This separation exposes a clinical nuance that is often lost in aggregate fall research. If the same predictor tracks both events, it likely reflects a broad instability phenotype. If it tracks one more than the other, it may be pointing to a more specific biomechanical pathway.
Spinopelvic alignment signaled stumbling, not a clean fall phenotype
Radiographic measures told a narrower story. Spinopelvic sagittal alignment was significantly associated only with stumbling: a smaller sagittal vertical axis in participants younger than 65 years and a larger T1 pelvic angle in those 65 years and older. Those findings are biologically plausible. Sagittal imbalance can alter the body’s center of mass, force compensatory trunk and pelvic strategies, and make steady forward progression more precarious. A subtle mismatch in alignment may be enough to provoke near-misses before it produces an injurious fall.
Even so, multinomial regression did not distinguish the radiographic associations between stumbling and falls. In other words, alignment abnormalities appeared relevant, but the data did not support a clean separation between the two outcomes once modeled together. That pushes the radiographic findings into a hypothesis-generating category rather than a basis for immediate clinical screening or intervention.
Why the questionnaire may outperform the x-ray in routine practice
There is a real-world advantage to the GLFS-25. It is immediate, inexpensive, and portable. A radiograph, by contrast, captures structure at one point in time and may miss dynamic instability, neuromuscular compensation, or fatigue-related gait failure. A questionnaire can integrate the lived effect of pain, stiffness, fear, and reduced mobility across the entire day, which may be exactly why it correlated with both falls and stumbling so consistently.
The implication is not that imaging is unimportant. It is that structure alone may not be the best first-line screen for community risk. Sagittal alignment can inform mechanistic thinking, especially in older adults with progressive postural change, but the functional phenotype appears to be easier to detect and more directly tied to self-reported instability.
What this means for fall-risk assessment
In a community setting, the study supports a simple hierarchy: ask about falls and stumbling separately, and measure locomotive dysfunction directly. The GLFS-25 identified risk across age groups and linked to both outcomes, which makes it attractive as a practical screening instrument for orthopedic and geriatric clinics, as well as for population health surveys. Adults with higher scores may already be compensating for pain, weakness, reduced flexibility, or postural inefficiency long before a major injury occurs.
The radiographic findings deserve caution. A smaller sagittal vertical axis in younger adults and a larger T1 pelvic angle in older adults may reflect age-dependent compensatory mechanics, but the analysis does not justify treating those parameters as stand-alone predictors of falls. Their value at this stage lies in generating mechanistic questions about posture, balance, and forward stability.
The cohort was cross-sectional, so cause and effect cannot be inferred. Still, the data clarify an important point for clinicians: locomotive dysfunction is visible before the fall, not just after it. In a prevention framework, that matters more than any single x-ray measurement, because the earliest intervention opportunity may be a functional complaint that patients can already describe themselves.