At one time or another, most people have experienced a painful joint, such as a sprained ankle or other source of lower extremity discomfort. In order to walk under such conditions, people will generally adapt their walking style, such as walking with a limp, to avoid generating movements and forces that may exacerbate the painful injury. Once the injury heals, normal walking patterns are usually re-established.

In the case of chronic conditions such as arthritis or low back pain, however, the adaptations may persist. Despite whatever benefit patients receive by adapting their gait, it is largely unknown what the long-term effects of gait adaptations are on other joints of the musculoskeletal system and what impact such adaptations may have on neuromuscular control of the body. Using information from the Centers for Disease Control and Prevention, it can be estimated that approximately 48 million people in the United States experience activity limitations due to arthritis, and about 63 million are limited due to chronic back conditions. Clearly this qualifies as a significant public health concern. Although researchers are busy finding solutions to the problems of therapeutic management of arthritis and other musculoskeletal problems, little attention has been paid to the study of movement adaptations to chronic lower extremity dysfunction.

From a biomechanical analysis perspective, functional adaptations due to musculoskeletal impairments may be subtle and may vary from person to person. Hence, they can be difficult to detect. As a result, we have limited understanding of the biological mechanisms that engender compensatory locomotor strategies. Furthermore, rehabilitation for lower extremity musculoskeletal dysfunction traditionally targets the impairment (pain, range of movement, muscle strength) with the assumption that function will improve. However, without a clear understanding of how impairments cause functional adaptations and what effects those adaptations have on whole body function, our current assessment practices of therapies meant to improve function and reduce disability are incomplete.

This article -- though brief and barely constituting an introductory text -- will highlight some of the current thinking in disability modeling and in human movement biomechanics research and illuminate how these two distinct fields are converging.

To put the concept of disability in its proper context, we need to think about disability in terms of the role we play as behaving humans. This encompasses a spectrum of activities from personal care to professional endeavors. How well we function in a real-world environment, in essence, establishes our role and, hence, our disability status. With this definition, a deaf individual might not be considered disabled because they can function superbly in the real world, maintaining their role despite their hearing impairment. An individual with end-stage osteoarthritis (OA), on the other hand, may be severely disabled insofar as their ability to maintain their role.

Disability is, therefore, caused by limitations in the ability to function. Functional limitations are restrictions of whole body systems or subsystems (e.g., inability to crouch, lift or walk a sufficient distance) that can directly impact upon a person's role. Functional limitations are, in turn, caused by impairments of organ systems (e.g., in OA the organ is the synovial joint and its cartilage). From a musculoskeletal point of view, impairments may include loss of muscle strength, joint pain and stiffness, or laxity of joint tissues. Finally, at the heart of the process is the pathology. Thus, we have a process that resembles Figure 1 as described in several reports in the public health literature, which shows the process from impairment to disability is not unidirectional (Jette et al., 1998; Nagi, 1991).

A compelling question with regard to preventing disability is at what level of the disablement process should rehabilitation interventions be applied? In diseases where there is no cure for the pathology (as with OA), the traditional rehabilitation approach is to apply treatments at the impairment level. Using OA as our model, relieving pain and stiffness -- and strengthening muscles -- should improve function, thereby reducing disability. Interestingly, several recent randomized, controlled treatment trials of strengthening in patients with knee OA have shown only small to moderate effects on patient function (e.g., Rogind et al., 1998). While there may be numerous reasons for these findings -- and it is clear that physical activity has far-reaching benefits to all body systems -- it is not currently known how these interventions affect function. Furthermore, it is unknown whether intervening at the functional limitations level (functional training) may confer greater, or fewer, benefits than strengthening alone. In any event, to answer these compelling questions, researchers need to measure whole body, multi-segment, human movements during activities of daily living (ADLs).

Although the word gait refers specifically to one's walking pattern, the modern gait laboratory routinely captures myriad human movements, including such common ADLs as walking, arising from a chair, and ascending and descending stairs. There are a number of reasons capturing these data are considered important, not the least of which is to better understand how humans control movements during functional activities. While the exact neurophysiology of human movement control presently eludes us, gait analysis offers an objective way to document the patterns of multi-segment movement, often with the purpose of better understanding locomotor function in populations or individuals with musculoskeletal and neurological dysfunction.

Biomechanical measurements used in contemporary human movement research take a number of forms. Spatio-temporal measurements (time and distance mea-sures) are often used to describe the gross movements of the body (and, incidentally, are the most common measures of gait function in exercise intervention assessments in older people with musculoskeletal impairments); however, they are nonspecific about the origin of functional losses or gains. Joint kinematics (the rotations, velocities and accelerations of body segments) can be used to identify range of movement restrictions of painful joints, such as the osteoarthritic knee (Stauffer et al., 1977). However, the information gleaned from these data is still incomplete because, by definition, they neglect the forces that cause, or result from, these movements. Joint forces and moments (a term used to describe the rotary action of a force), when calculated from gait data, are thought to be an index of muscle forces acting at a joint (Kaufman et al., 2001), and thereby can provide valuable information about the dysfunctional joint (e.g., the impairment); however, they are not easily assembled into a coherent picture of multi-segment movement dysfunction (e.g., the functional limitation).

Although the impairment may directly affect only a single joint, the loss of function may involve the whole body's response to the impairment. As a result, to better understand loss of function, we need to focus on biomechanical measurements that can link multi-segment movements to one another. Such measurements might allow us to identify compensatory movement strategies in response to impairments, thus enabling a clearer picture of the functional consequences of those impairments. In addition, we may gain insights into how interventions at the functional limitations level of the disability model can be implemented and monitored.

One approach that appears to have some promise for studying functional adaptations with aging and disability is mechanical power and energy analysis. Unlike force and moment, which are vector quantities having both direction and magnitude, power is a scalar quantity and thus has no direction in the usual sense (such as left-right or up-down). Rather, power has direction relative to the system; that is, either adding to the energy of the system or removing energy from the system. The rate of change of energy (power) of body segments is controlled by a combination of intersegmental transfer of energy and energy produced or absorbed by the muscles. In this regard, it is possible to map the transfer of energy from segment to segment, which reveals how the muscles work to control and regulate segmental movements (McGibbon et al., 2001). Recent studies suggest that mechanical power and energy transfer analyses can be used to identify gait adaptations that occur with aging (DeVita and Hortobagyi, 2000; Judge et al., 1996) and in disabled elders' gait (McGibbon et al., 2001; McGibbon and Krebs, in press).

Judge et al. (1996) reported that healthy elders increase concentric (muscle shortening from positive work) hip power to compensate for reduced ankle power production during gait. DeVita and Hortobagyi (2000) recently showed that a redistribution of joint moments and powers occurs with aging, resulting in higher angular impulse and positive work at the hip. They also found those same measures at the knee and ankle decreased with age. Although the exact mechanism for this age-related adaptation is unclear, it may be in part due to loss of muscle strength of the ankle plantar-flexors and knee quadriceps group (Wolfson et al., 1995) and possibly hip muscle contractures (Kerrigan et al., 2001).

There are far fewer studies of this nature that include elders with physical disabilities. Our group recently compared the mechanical energy transfers of the lower extremities and low back between a group of healthy elders (age 73.7 years±4.6 years) and a group of disabled elders (age 74.7 years±7.4 years) referred to physical therapy services for a variety of mobility problems, primarily orthopaedic (McGibbon et al., 2001). As with prior studies, a significant decrease in ankle power was found between the healthy elders and disabled elders, as well as an increase in the mechanical energy expended at the hip and low back. Interestingly, however, the increase seen at the hip was in eccentric (muscle lengthening from negative work) power.

In a separate study, we compared the same healthy elders to a more homogenous group of elders diagnosed with unilateral knee OA (McGibbon and Krebs, in press). Results were similar to the 2001 study with two important emphases: elders with knee OA had significantly reduced knee power and had an even greater increase in hip eccentric energy expenditure. The latter finding might be caused by hyperextension of the hip muscles when the hip is in extension (causing more eccentric contraction), as illustrated in Figure 2. However, it is a commonly held belief that older people develop hip muscle contractures that prevent them from extending the hip (Kerrigan et al., 2001). Hyperextension of the hip may therefore function as a passive-elastic mechanism (like a rubber band) for assisting with advancement of the leg into swing phase, but may increase the load experienced by the hip cartilage and perhaps destabilize the pelvis and upper body.

Clearly, more studies are needed on a wider range of healthy and disabled older adults to isolate the true nature of these adaptations and how to therapeutically manage them if necessary.

The studies briefly described highlight the current challenges in human movement research in aging and disability. Although more research is needed, investigators in the field appear to be converging on a promising line of inquiry, and the focus of these outcomes suggests the hip plays an especially important role in gait function. Future studies should examine in more detail the relationship between hip adaptive characteristics and aging and specific lower extremity impairments caused by arthritis and chronic back pain. Other lower extremity impairments caused by osteoporosis (loss of bone mineral density) and sarcopenia (loss of muscle mass and strength) need to be examined as well. A priority goal should be how to separate the adaptive characteristics caused by the healthy aging process (such as reduced strength and mobility) and those caused by unhealthy aging.

Finally, detailed biomechanical analyses should be included in physical activity intervention studies to explore the means by which patients attain functional gains (faster walking speed and other improvements in ADL performance measures). Moreover, comparing the biological mechanisms for functional improvements from different types of physical activity-based interventions (such as conventional impairment-based rehab versus functional training-based rehab) may prove to be very useful in establishing what therapies have the best chance to eliminate functional limitations and halt the progression toward disablement.

Dr. McGibbon is technical/assistant director of the Biomotion Laboratory at the Massachusetts General Hospital and associate professor in the graduate programs in physical therapy at MGH Institute of Health Professions in Boston.

DeVita P, Hortobagyi T (2000), Age causes a redistribution of joint torques and powers during gait. J Appl Physiol 88(5):1804-1811.

Jette AM, Assmann SF, Rooks D et al. (1998), Interrelationships among disablement concepts. J Gerontol A Biol Sci Med Sci 53(5):M395-M404.

Judge JO, Davis RB, Ounpuu S, Davis RB 3rd (1996), Step length reductions in advanced age: the role of ankle and hip kinetics. J Gerontol A Biol Sci Med Sci 51(6):M303-312.

Kaufman KR, Hughes C, Morrey BF et al. (2001), Gait characteristics of patients with knee osteoarthritis. J Biomech 34(7):907-915.

Kerrigan DC, Lee LW, Collins JJ et al. (2001), Reduced hip extension during walking: healthy elderly and fallers versus young adults. Arch Phys Med Rehabil 82(1):26-30.

McGibbon CA, Krebs DE (in press), Compensatory gait mechanics in patients with unilateral knee arthritis. J Rheumatol.

McGibbon CA, Krebs DE, Puniello MS (2001), Mechanical energy analysis identifies compensatory strategies in disabled elders' gait. J Biomech 34(4):481-490.

Nagi SZ (1991), Disability concepts revisited: implications for prevention. In: Disability in America, Pope AM, Tarlov AR, eds. Washington, D.C.: National Academy Press.

Rogind H, Bibow-Nielsen B, Jensen B et al. (1998), The effects of a physical training program on patients with osteoarthritis of the knees. Arch Phys Med Rehabil 79(11):1421-1427.

Stauffer RN, Chao EY, Gyory AN (1977), Biomechanical analysis of the diseased knee joint. Clin Orthop (126):246-255.

Wolfson L, Judge J, Whipple R, King M (1995), Strength is a major factor in balance, gait, and the occurrence of falls. J Gerontol A Biol Sci Med Sci 50(Spec No):64-67.

Biomechanics in Disability Research: Gait Dysfunction in the Elderly