We compared peak joint power and kinematic variables between comfortable and fast walking speeds among a group of
elderly adults with impairments in physical performance. Contrary to our hypothesis, we found that both hip and ankle power outputs increased significantly with speed, suggesting that ankle muscle power capacity was not strictly a limiting neuromuscular factor in normal walking. However, ankle power output among the low-performing elders remained diminished compared with the healthy adults, ever when walking at a faster gait speed.
A number of studies have shown that ankle plantarflexor power and work done during late stance is reduced in healthy
elderly compared with healthy young adults. For example, DeVita and Hortobagyi8 found that healthy elderly generated
significantly less work at the ankle and knee, but greater hip extensor work during early stance, when walking at the same
speed as a group of healthy young adults. These changes were interpreted as a distal to proximal shift in the locus of neuromuscular function with aging. Although the exact methods and details differ, others have also reported that healthy elders seem to use adaptations at the hip to compensate for reduced plantarflexor power output. McGibbon and Krebs were the first to show that frail elderly with functional limitations show even further declines in ankle power output during walking. Similar observations have been made of elders with a history of falls and are now also shown in our analysis of elders with low physical performance. An important observation in this study was that even when the low-performance elders increased walking speed to a magnitude that was 30% greater than the healthy elderly, ankle power output remained significantly diminished. The only other significant difference in joint powers that was evident in the low-performing elderly, relative to the healthy group, was an increased hip extensor power during early stance.
McGibbon and Krebs have also reported impairment-related changes in hip kinetics, although they suggest the primary change is toward greater eccentric hip work during midstance.
Whether the difference between studies is attributable to differences in the methodology used or characteristics of the
population is not clear. During late stance, the ankle muscles are generally believed to propel the stance limb into swing, and may also produce substantial forward and vertical acceleration of the upper body. Given the relatively close timing between the ankle plantarflexor power output during late stance and the contralateral hip extensor power generation during early support (see fig 2), it is feasible that the proximal muscle actions about the hip may compensate for reduced mechanical actions at the ankle. McGibbon has recently speculated on various mechanical means by which hip muscle actions may compensate for reduced ankle power. Future studies that can quantitatively identify such compensatory mechanisms are warranted. Ankle muscle weakness has been suggested as a possible cause of reduced ankle power output in elderly gait.Indeed, maximum ankle power capacities of healthy elderly are 20% to 40% below that of young adults,6 a difference that is likely even more marked among impaired elders. However, in this study the impaired elderly tested did indeed increase ankle power output (24%) to walk faster, which strongly suggests that as a group, ankle power output was not a strictly limiting factor. Of potential importance are the kinematic changes between the healthy and low-performance elders. Relative to the healthy elderly, the low-performance group showed larger transverse pelvic rotations, reduced peak ankle dorsiflexion, and a shift to greater hip flexion and less hip extension. These
differences, which may arise from tightness in the hip flexors and ankle plantarflexors, are similar to the differences seen
between healthy young and healthy old adults.7 For example, the healthy elderly exhibited a 3° shift in hip range of motion
(ROM) toward greater hip flexion compared with young adults. In the present study, the low-performance group had an
additional 6° shift toward hip flexion, for a total of 9° hip ROM shift compared with young adults.7 It is notable that hip extension did not increase with walking speed among the lowperforming elders, suggesting that it indeed may be a limiting
factor. Kerrigan et al14 have also found that a large reduction in hip extension was the primary kinematic factor that distinguished the gait of elderly fallers from healthy young and elderly adults. The larger transverse pelvic rotations used by the low-performing elders may be a mechanism to increase step length, and conversely gai