The Future of Gait Analysis is Wearable - But Not Perfect | What This Meta-analysis Means for Clinicians and Runners | Find Your Stride | Edinburgh Podiatrist

Can Wearable IMUs Replace the Gait Lab?

Wearable technology has transformed endurance sport. GPS watches, foot pods and motion sensors now promise laboratory-level biomechanics almost anywhere from the clinic to the running trail. For podiatrists, this is particularly appealing. Imagine performing objective gait analysis during an outdoor run rather than relying solely on a few minutes on a treadmill.

A new systematic review and meta-analysis by Bissenov et al. (2025) tackles one of the biggest questions in sports biomechanics: Are wearable inertial measurement units (IMUs) actually accurate enough for clinical decision-making? The answer is encouraging but with several important caveats.

Are wearable inertial measurement units (IMUs) actually accurate enough to replace traditional gait analysis hardware?

What Did the Authors Do?

The authors systematically searched four major databases and identified 27 validation studies, including 441 participants, comparing wearable IMUs against laboratory gold standards:

• Optical Motion Capture (Vicon, Qualisys etc.)

• Force plates

  
  

Twenty-two studies were suitable for meta-analysis, pooling measurement errors using Root Mean Square Error (RMSE). Their primary focus was sagittal plane joint motion of the:

• Ankle

• Knee

• Hip

  
  

They also examined whether IMUs could estimate ground reaction forces (GRFs).

Methodologically, the review scores highly. The protocol was prospectively registered on PROSPERO, PRISMA guidelines were followed, and study quality was assessed using QUADAS-2. These are all hallmarks of a well-conducted systematic review.

The Good News

Ankle and Knee Measurements Are Surprisingly Accurate

The pooled results are impressive:

Joint	Root Mean Square Error (RMSE)
Ankle	4.62°
Knee	4.60°

Both fall within the commonly accepted clinical threshold of ≤5°. For clinicians, this means wearable IMUs can often provide clinically useful estimates of sagittal plane movement during walking without the need for expensive motion capture laboratories. For podiatrists involved in rehabilitation, orthotic assessment, gait analysis and exercise prescription this represents a significant opportunity.

Overground Walking Is Better Than Treadmill Walking

One particularly useful finding is that IMUs consistently performed better during overground walking. That’s important because real-world walking is precisely where clinicians increasingly want objective assessment. Ironically, many gait labs still rely heavily on treadmills.

The authors suggest treadmill motors create magnetic interference that reduces sensor accuracy, especially when magnetometers are used for orientation.

The Less Good News

The Hip Remains Difficult

Accuracy deteriorated proximally. Hip RMSE reached 5.79°, exceeding the predefined threshold. Why? Several likely reasons exist:

• increased soft tissue movement

• greater joint complexity

• sensor alignment errors

• accumulated error travelling up the kinetic chain

  
  

This mirrors previous biomechanical literature and reminds us that the closer sensors move toward the trunk, the more difficult accurate joint angle estimation becomes. For clinicians assessing pelvic control or proximal running mechanics, IMUs should therefore be interpreted cautiously.

Ground Reaction Forces Still Need Work

Perhaps the biggest disappointment concerns ground reaction forces. Only four studies were available. Errors were relatively large:

• Vertical GRF: 8.37% body weight

• Anterior-posterior GRF: 5.50% body weight

The authors conclude that estimating kinetics remains considerably less mature than measuring joint angles. Machine learning approaches are promising, but they’re not yet ready to replace force plates. This is important because many wearable systems increasingly claim to estimate impact loading, braking forces and propulsion. The evidence simply isn’t there yet.

The Biggest Limitation Nobody Should Ignore

One statistic jumps off the page. Heterogeneity was enormous. Most pooled analyses had I² values (used in meta-analyses to quantify the proportion of total variation across different studies) between 90% and 96%. That tells us individual studies differed substantially. Those differences included:

• Sensor brand

• Sampling frequency

• Calibration methods

• Number of sensors

• Processing algorithms

• Walking protocol

  
  

Pooling highly heterogeneous studies is statistically acceptable under random-effects modelling, but it limits how confidently we can generalise the findings. In other words: Not all IMUs perform equally well. Some systems demonstrated excellent accuracy. Others performed considerably worse. Unfortunately, clinicians looking for “the best IMU” won’t find that answer here.

A Running Perspective

One surprising omission is running itself. Although running studies were searched for, there simply weren’t enough high-quality validation studies to include in the meta-analysis. That matters. Walking biomechanics are relatively slow, repetitive and stable.

Running introduces:

• Greater accelerations

• Higher impact forces

• Increased soft tissue movement

• Larger sensor drift

• More rapid joint excursions

• 
  

We therefore cannot confidently assume these walking results apply to running.

Given the growing popularity of wearable running biomechanics platforms, this represents an important research gap rather than evidence that these devices perform equally well during running.

Our Verdict

This is an excellent systematic review. The methodology is transparent, the analyses are rigorous and the conclusions are appropriately cautious. Unlike many technology papers, the authors avoid over-selling wearable sensors. Instead, they acknowledge where IMUs currently excel and where they still fall short. For podiatrists, the take-home message is reassuring:

Wearable IMUs are becoming genuinely useful tools for assessing lower-limb kinematics during walking, particularly at the ankle and knee. They have the potential to expand gait analysis beyond specialist laboratories and into everyday clinical practice. However, they remain adjuncts, not replacements, for comprehensive biomechanical assessment. Clinical reasoning still matters more than technology.

Clinical Take-Home Points

• IMUs demonstrate acceptable accuracy for ankle and knee sagittal plane motion during walking.

• Overground assessments appear more reliable than treadmill measurements.

• Hip kinematics remain considerably less accurate.

• Current wearable systems should not replace force plates for measuring ground reaction forces.

• Differences in hardware and calibration remain a major barrier to standardisation.

• Evidence for running biomechanics remains surprisingly limited despite rapid commercial growth.

• Wearable sensors should complement not replace expert clinical assessment.

Bottom Line

Wearable gait analysis is no longer science fiction. For podiatrists working in sports medicine, rehabilitation and running clinics, IMUs offer exciting opportunities to collect objective movement data outside the biomechanics laboratory. But before we allow wearable technology to drive treatment decisions, we need more high-quality validation studies particularly during running, in injured athletes, and using standardised testing protocols.

Until then, the best clinicians will continue to combine wearable data with experience, clinical examination and sound biomechanical reasoning.

Find Your Stride!

Citation

Bissenov, A., Zhubi, E., Engh, M. A., Gresits, O., Tóth, R., & Terebessy, T. (2025). Concurrent validity of wearable IMUs for sagittal plane lower-limb range of motion during walking and estimated ground reaction forces: A systematic review and meta-analysis. Journal of Orthopaedic Surgery and Research, 20, 891. https://doi.org/10.1186/s13018-025-06322-w