Visual Impairment and Mobility Performance in Older Adults
Visual Impairment and Mobility Performance in Older Adults
We found that VI participants in the SEE Study had slower speeds than their NVI counterparts at every study visit; however, there is no evidence that the decline in speeds over time differed between these 2 groups. These results suggest that the difference in walking speeds between the VI and the NVI remained over the study period and did not increase over time. We also found that the VI were more likely than the NVI to be classified as having mobility disability at each study visit. Similar to our analyses of speeds, the change over time in the odds of being classified as having mobility disability was similar between the VI and the NVI.
The results of this study were contrary to our a priori hypothesis that the VI would have greater speed declines and steeper mobility disability trajectories than the NVI over the SEE Study period. We examined the following 2 possible explanations of why we did not observe a difference in mobility trajectories between the VI and the NVI: 1) differential loss to follow-up of the VI with the slowest performance speeds, and 2) sensitivity to the cutpoints of visual impairment and disability.
We posited that speed trajectories in the VI might have been attenuated (i.e., the slope of this trajectory would have been brought closer to the slope of the NVI) if there were a differential loss of these individuals. However, our cross-sectional models determining the odds of being lost to follow-up compared with the odds of remaining in the study at each study visit indicate that the interaction terms assessing differential losses to follow-up of VI participants with slowest speeds were not significant. This suggests that our observed results are likely not due to differential loss to follow-up of the VI participants with slowest speeds over the SEE Study period.
An advantage of the SEE Study is that it provides performance-based mobility measurements that allowed us to assess change in speed over time. However, there are no clinical standards to classify individuals as disabled on the basis of mobility performance. Previous studies have suggested that, for walking on flat surfaces, speeds of 0.6 m/second or slower indicate poor health and functioning. In our analyses, we defined walking disability as 1 standard deviation below the baseline population means, which corresponded to 0.6 m/second for the 4-m task. Therefore, our cutpoint likely identified performance values on the stair tests that were abnormal. Our data did not support the hypothesis that the odds of being classified as disabled on the basis of performance speeds would increase at a greater rate over time in the VI compared with the NVI.
We examined the effect of shifting our definition of disability to speeds slower than 0.5 standard deviations below the population mean. These analyses resulted in the same inference for all of the covariates included in our primary models. Similarly, when we changed our definition of visual impairment to distance visual acuity worse than 20/60, we again observed the same inference as in our primary models. These observations indicate that our results are robust to the cutpoint of disability and visual impairment used.
We can offer only potential explanations for why the VI had slower speeds than the NVI at each study visit, but the trajectory of these speeds was similar in these 2 groups. Cesari et al. have shown that comorbidity is associated with worse physical functioning, and speeds were slower as the number of comorbidities increased. In the SEE Study, 41% of the VI and 31% of the NVI had 3 or more comorbid conditions; therefore, it is possible that the mobility trajectories diverged prior to study enrollment and prior to the accumulation of multiple health conditions. Further research to determine this would require a longitudinal study of individuals without comorbid conditions who develop incident visual impairment.
This study found that the largest difference in walking speeds between the VI and the NVI was observed for the stair descent task (−0.11 steps/second) (Table 2). The differences in speeds were similar for both walking up stairs and walking 4 m (−0.08 steps/second and −0.08 m/second, respectively) (Table 2). This may suggest that walking down stairs is the most difficult of the 3 tasks for the VI. Prior research has suggested that slower walking speeds in the VI may be partially driven by the inability to recognize changes in terrain, such as a step or a ramp. Additionally, the riser of steps has better contrast than the top surface of steps, meaning that it is easier to see a step's riser than the top of a step. When an individual walks down stairs, the step riser is not visible, and this may explain why the biggest difference in speed between the VI and the NVI was observed for this task.
The results from this study imply that walking speeds in the VI remain significantly slower than in the NVI as people age, and that those with visual impairment are more likely than the NVI to be classified as having mobility disability. It is possible that the slowing of walking speed is an instantaneous adaptation at the onset of visual impairment, and that the VI walk slowly in an effort to maintain or improve their perception of mobility safety. This hypothesis makes sense, because the VI have greater fear of falling than do the NVI. It is possible that efforts to improve mobility speeds alone may not be effective at improving the perception of mobility disability. Instead, this could suggest that, if the goal is to reduce mobility disability in the VI, rehabilitation efforts should include interventions aimed at improving both mobility speed and mobility safety, such as with the use of mobility aids.
Discussion
We found that VI participants in the SEE Study had slower speeds than their NVI counterparts at every study visit; however, there is no evidence that the decline in speeds over time differed between these 2 groups. These results suggest that the difference in walking speeds between the VI and the NVI remained over the study period and did not increase over time. We also found that the VI were more likely than the NVI to be classified as having mobility disability at each study visit. Similar to our analyses of speeds, the change over time in the odds of being classified as having mobility disability was similar between the VI and the NVI.
The results of this study were contrary to our a priori hypothesis that the VI would have greater speed declines and steeper mobility disability trajectories than the NVI over the SEE Study period. We examined the following 2 possible explanations of why we did not observe a difference in mobility trajectories between the VI and the NVI: 1) differential loss to follow-up of the VI with the slowest performance speeds, and 2) sensitivity to the cutpoints of visual impairment and disability.
We posited that speed trajectories in the VI might have been attenuated (i.e., the slope of this trajectory would have been brought closer to the slope of the NVI) if there were a differential loss of these individuals. However, our cross-sectional models determining the odds of being lost to follow-up compared with the odds of remaining in the study at each study visit indicate that the interaction terms assessing differential losses to follow-up of VI participants with slowest speeds were not significant. This suggests that our observed results are likely not due to differential loss to follow-up of the VI participants with slowest speeds over the SEE Study period.
An advantage of the SEE Study is that it provides performance-based mobility measurements that allowed us to assess change in speed over time. However, there are no clinical standards to classify individuals as disabled on the basis of mobility performance. Previous studies have suggested that, for walking on flat surfaces, speeds of 0.6 m/second or slower indicate poor health and functioning. In our analyses, we defined walking disability as 1 standard deviation below the baseline population means, which corresponded to 0.6 m/second for the 4-m task. Therefore, our cutpoint likely identified performance values on the stair tests that were abnormal. Our data did not support the hypothesis that the odds of being classified as disabled on the basis of performance speeds would increase at a greater rate over time in the VI compared with the NVI.
We examined the effect of shifting our definition of disability to speeds slower than 0.5 standard deviations below the population mean. These analyses resulted in the same inference for all of the covariates included in our primary models. Similarly, when we changed our definition of visual impairment to distance visual acuity worse than 20/60, we again observed the same inference as in our primary models. These observations indicate that our results are robust to the cutpoint of disability and visual impairment used.
We can offer only potential explanations for why the VI had slower speeds than the NVI at each study visit, but the trajectory of these speeds was similar in these 2 groups. Cesari et al. have shown that comorbidity is associated with worse physical functioning, and speeds were slower as the number of comorbidities increased. In the SEE Study, 41% of the VI and 31% of the NVI had 3 or more comorbid conditions; therefore, it is possible that the mobility trajectories diverged prior to study enrollment and prior to the accumulation of multiple health conditions. Further research to determine this would require a longitudinal study of individuals without comorbid conditions who develop incident visual impairment.
This study found that the largest difference in walking speeds between the VI and the NVI was observed for the stair descent task (−0.11 steps/second) (Table 2). The differences in speeds were similar for both walking up stairs and walking 4 m (−0.08 steps/second and −0.08 m/second, respectively) (Table 2). This may suggest that walking down stairs is the most difficult of the 3 tasks for the VI. Prior research has suggested that slower walking speeds in the VI may be partially driven by the inability to recognize changes in terrain, such as a step or a ramp. Additionally, the riser of steps has better contrast than the top surface of steps, meaning that it is easier to see a step's riser than the top of a step. When an individual walks down stairs, the step riser is not visible, and this may explain why the biggest difference in speed between the VI and the NVI was observed for this task.
The results from this study imply that walking speeds in the VI remain significantly slower than in the NVI as people age, and that those with visual impairment are more likely than the NVI to be classified as having mobility disability. It is possible that the slowing of walking speed is an instantaneous adaptation at the onset of visual impairment, and that the VI walk slowly in an effort to maintain or improve their perception of mobility safety. This hypothesis makes sense, because the VI have greater fear of falling than do the NVI. It is possible that efforts to improve mobility speeds alone may not be effective at improving the perception of mobility disability. Instead, this could suggest that, if the goal is to reduce mobility disability in the VI, rehabilitation efforts should include interventions aimed at improving both mobility speed and mobility safety, such as with the use of mobility aids.
Source...