1310-GRP Linking Gait Patterns to Acute Glucose Dynamics Using Wearable Sensors

Location

https://campus.kennesaw.edu/colleges-departments/ccse/events/computing-showcase/fa26-cday-program.php

Document Type

Event

Start Date

20-11-2026 2:00 PM

Description

Passive wearable sensing may complement continuous glucose monitoring if gait changes track short-term glucose changes within a person; most evidence, however, comes from chronic diabetes. In this pilot study, ten Hispanic/Latino adults walked about 10 m (up to 3 walks per round) at baseline and every 20 minutes for 2 hours after a glucose beverage, while glucose was measured with a Dexcom G7 and gait with Noraxon inertial sensors (200 Hz). We analyzed 10 gait variables drawn from the diabetes gait literature: cadence, stride length, stance time, walking speed, peak hip, knee and ankle angles, peak foot clearance, mean step width and stride-time variability. For each variable, within-person Spearman correlations were combined in a two-stage meta-analysis and cross-checked with a mixed model using within-person-centred glucose; findings were classified by decision rules fixed in an analysis plan. No variable was classified as solid. Stride-time variability was the only weak signal (pooled rho = -0.32; mixed-model p = 0.021; meta-analysis p = 0.063). It did not hold after removing low-quality rounds (p = 0.22 and 0.13), nor after correction for testing 10 variables (q = 0.21 and 0.63, respectively). Walking speed showed a non-significant tendency (8 of 10 participants positive). Differences between people explained 58-94% of each variable’s variance. With 10 participants only large effects can be detected, so these results are hypothesis-generating and inform the design of a larger study.

This document is currently not available here.

Share

COinS
 
Nov 20th, 2:00 PM

1310-GRP Linking Gait Patterns to Acute Glucose Dynamics Using Wearable Sensors

https://campus.kennesaw.edu/colleges-departments/ccse/events/computing-showcase/fa26-cday-program.php

Passive wearable sensing may complement continuous glucose monitoring if gait changes track short-term glucose changes within a person; most evidence, however, comes from chronic diabetes. In this pilot study, ten Hispanic/Latino adults walked about 10 m (up to 3 walks per round) at baseline and every 20 minutes for 2 hours after a glucose beverage, while glucose was measured with a Dexcom G7 and gait with Noraxon inertial sensors (200 Hz). We analyzed 10 gait variables drawn from the diabetes gait literature: cadence, stride length, stance time, walking speed, peak hip, knee and ankle angles, peak foot clearance, mean step width and stride-time variability. For each variable, within-person Spearman correlations were combined in a two-stage meta-analysis and cross-checked with a mixed model using within-person-centred glucose; findings were classified by decision rules fixed in an analysis plan. No variable was classified as solid. Stride-time variability was the only weak signal (pooled rho = -0.32; mixed-model p = 0.021; meta-analysis p = 0.063). It did not hold after removing low-quality rounds (p = 0.22 and 0.13), nor after correction for testing 10 variables (q = 0.21 and 0.63, respectively). Walking speed showed a non-significant tendency (8 of 10 participants positive). Differences between people explained 58-94% of each variable’s variance. With 10 participants only large effects can be detected, so these results are hypothesis-generating and inform the design of a larger study.