Table 2.
Association Between the Energetic Cost of Walking and Gait Speed Decline Modeled Using Generalized Estimating Equations With a Working Autoregressive Correlation Structure and Stratified at a Baseline Age of 65 Years
| Dependent Variable: Usual Gait Speed | Model A | Model B | ||
|---|---|---|---|---|
| 40–65 years; n = 140 | 65 years and older; n = 317 | |||
| Independent Variables | β | p Value | β | p Value |
| Intercept | 1.234 | <.001 | 1.456 | <.001 |
| EEM (per 0.01mL/kg/m) | 0.002 | .55 | −0.008 | <.001 |
| Centered age | 0.001 | .69 | −0.012 | <.001 |
| Body composition | −0.140 | .01 | −0.191 | <.001 |
| Non-White race | −0.050 | .03 | −0.060 | .01 |
| Lower extremity arthritis pain | −0.058 | .01 | −0.020 | .15 |
| Poor balance | −0.039 | .34 | −0.055 | <.001 |
Notes. The model shows the β coefficient and p value from a population average model assessing the decline in usual gait speed (m/s) for each 1-year increase in age, adjusting for body composition (ratio of fat-to-lean mass), race, lower extremity arthritis pain, and self-reported balance difficulty. Age was centered at 40 years in Model A and at 65 years in Model B. Variables for sex, BMI, history of smoking, cardiovascular disease, cerebrovascular disease, pulmonary disease, and diabetes were not significant and were not included in the final model.
BMI = body mass index; EEM = energy expenditure per meter.