Immediate effects of visual feedback on the accuracy of foot landing adjustments in older people with diabetes mellitus: A cross-sectional study

Abstract

Diabetes Mellitus in older adults reduces the accuracy of foot landing adjustments and increases the risk of falling. This study investigated whether targeted visual feedback enhances the accuracy of the foot landing in older participants with diabetes. Forty-eight volunteers in three groups of young, healthy older and older adults with diabetes participated. During treadmill walking, Nexus Vicon software streamed real-time orientation data of two markers on the first toes into MATLAB scripts through a Visual3D server. The system visualised real-meaningful step length or minimum toe clearance (MTC) in each step on a monitor in front of a treadmill at eye level. Participants responded to four subject-specific step-length (±10% mean baseline step length) or MTC targets (3.5 and 5.5 cm higher than the mean baseline MTC). One target was displayed every 10 steps and removed after 10 steps when participants walked without seeing any target, then a new or the same target appeared and stayed on for another 10 steps. Tasks were increasing or decreasing baseline step length or increasing the baseline MTC in response to virtual targets. The accuracy (absolute error) of each step adjustment in each 10-step block with a target was calculated. The mean accuracy of Step 1 and the mean accuracy of Steps 2-10 were measured and compared within and between groups using three-way ANOVA tests. Errors significantly differed between conditions (Step 1 and Steps 2-10) in all groups. All groups showed reduced errors in Step 1 during Steps 2-10. Among groups, the group with diabetes presented the greatest errors in Step 1 and Steps 2-10. These findings suggest that meaningful visual feedback about spatial gait parameters (step length, MTC) can improve foot-landing accuracy in older participants with diabetes, highlighting its potential as a training tool to prevent falls in this high-risk population.

Fig 1. Foot landing adjustments in response to short and long step length targets on a monitor.

The participant used visual feedback on the monitor to match the peaks of step-length graphs with the target (the horizontal line). The Y-axis shows the real-time length of each step and a target during four steps.

Fig 2. Effects of visual feedback on errors during step shortening.

The errors in Step 1 (1) were significantly greater than those in Steps 2-3 (2). Young adults (Group I); healthy older adults (Group II); older adults with diabetes (Group III).

Fig 3. Effects of visual feedback on errors during step lengthening.

Errors in Step 1 (1) were significantly reduced in Steps 2-3 (2). Young adults (Group I); healthy older adults (Group II); older adults with diabetes (Group III).

Fig 4. Effects of visual feedback on errors during increasing the minimum toe clearance (MTC) when high MTC targets that were 3.5 cm higher than baseline MTC were presented.

The errors in Step 1 (1) were significantly reduced in Steps 2-3 (2). Group I (young adults), Group II (healthy older adults), and Group III (older adults with diabetes).

Fig 5. Effects of visual feedback on errors during increasing the minimum toe clearance (MTC) when higher MTC targets that were 5.5 cm higher than baseline MTC were presented.

The errors in Step 1 (1) were significantly reduced in Steps 2-3 (2). Young adults (Group I); healthy older adults (Group II); older adults with diabetes (Group III).