March 2017

Knee OA in amputees: Biomechanical and technological considerations 482801691

The risk of knee osteoarthritis in the intact limb of longtime unilateral amputees is much higher than in non­amputees, and the range of potential contri­buting factors is even more complex. Optimizing prosthetic fit and function, in addition to more conventional OA interventions, can help address gait issues that contri­bute to knee joint degeneration.

By Emily Delzell

Unilateral amputees are at high risk for knee osteoarthritis (OA) in their intact limbs. Most experts think the increased risk comes from asymmetric force distribution caused by gait asymmetry and problematic compensations, poor prosthetic fit or alignment, and other factors, though not all researchers believe current evidence provides a definitive etiological answer.

Whatever its causes, knee OA is a significant concern for all unilateral amputees, particularly for the approximately 20% of individuals who lose a lower limb at a relatively young age to trauma, cancer, or congenital disease1 but remain active, potentially subjecting the intact knee to increased loading for 40, 50, or more years.

“Not only are musculoskeletal conditions like knee OA prevalent among amputees, they can compound disability. I think that only recently have they begun to receive the study they deserve, and I’m not sure if, across the board on the clinical side, it’s been recognized as much as it should be,” said David Morgenroth, MD, assistant professor in the Department of Rehabilitation Medicine at the University of Washington and a staff physician and associate program director for the Amputation Rehabilitation Fellowship at the VA Puget Sound Health Care System (VAPSHCS), both in Seattle.

Among amputees, knee OA and other musculoskeletal conditions may be taking a back seat to more traditional concerns such as residual and phantom limb pain, said Morgenroth, who is also a research investigator in the VA Rehabilitation Research & Development Center of Excellence for Limb Loss Prevention and Prosthetic Engineering in Seattle.

“Those are two of the biggies, and are both quite important, but I think if clinicians are not asking their amputee patients about intact limb knee pain then they’re potentially doing limited service; I think we need to be really thinking hard about prevention and treatment strategies,” he said.


Studies have reported knee OA rates from 10 to 17 times higher in unilateral lower extremity amputees who are longtime prosthesis users than in age-matched nonamputees.1-8 Knee OA is more prevalent among those with higher amputations.1,2,4,7 Hungerford and Cockin were the first authors to note this higher incidence, reporting in 1975 that 41% of veterans with transtibial amputation (TTA) and 63% of those with a transfemoral amputation (TFA) had significant patellofemoral osteoarthritis compared with 22% of nonamputee veterans.2 Later studies, mostly done in veterans, found similar incidence.1,3-8

Knee pain, often the first symptom of knee OA, is also common. Data show intact limb knee pain is twice as common in unilateral lower extremity amputees as in able-bodied individuals5 and is the primary complaint among longtime prosthesis users (average prosthesis wear time of 24 years).6 Pain is also more common among those with higher amputations; one study by Mussman et al found 46% of transtibial amputees had knee pain versus 75% of people with transfemoral amputation.6

Some have speculated higher body weight resulting from a more sedentary lifestyle after amputation could explain some of the increased risk of knee OA in this population,7 but researchers controlling for body mass have also found higher rates of knee pain and OA among unilateral amputees than controls.

Norvell et al in 2005 investigated self-reported OA symptoms in 63 prosthesis-wearing male veterans with traumatic unilateral amputation, controlling for body weight at ages 18 and 30 years.7 After the authors adjusted for age and body weight, those with TTA were three times and those with TFA were five times more likely to have knee pain on the intact side than nonamputee controls. Investigators concluded the loads experienced by the intact limbs of people with unilateral amputation are a potential cause of the increased risk of knee OA, a conclusion supported by Mussman et al, who found that 64% of longtime prosthesis-wearing lower extremity amputees said they depended more on their intact limb than their prosthetic side during activities.6


“The unaffected leg of people with a unilateral leg amputation typically experiences higher loading than the legs of nonamputees,” said Alena Grabowski, PhD, an assistant professor in the Department of Integrative Physiology at the University of Colorado Boulder. “These higher forces, combined with the moment arm about the knee, increase torque on the unaffected knee, which likely contributes to the greater incidence of osteoarthritis in the unaffected knee of persons with unilateral leg amputation compared with nonamputees.”

Some literature supports this hypothesis, Morgenroth said.

“Research looking at ground reaction forces in the intact limb compared with the prosthetic limb shows amputees do tend to load the intact limb more,”2 he said. “Ground reaction forces are nonspecific, though, and can be distributed throughout different joints. A few studies have looked at knee adduction moment, which is more specific to knee OA, and those bear out an asymmetry of loading between the intact limb and the prosthetic limb.”9-11

These findings, said Morgenroth, coincide with the literature showing that unilateral amputees have a higher risk of knee OA in the intact limb compared with matched controls and that they develop knee OA less commonly in the prosthetic side than the intact side. In their investigation of symptomatic knee OA and knee pain, Norvell et al, for example, found unilateral amputees are five times less likely to develop knee pain in the amputated limb than in the sound limb.7

“If you have an asymmetric gait, not only do you expose yourself to excessive wear and tear on your sound side, you also have a propensity to end up with knee and low back pain,” said Mike Corcoran, CPO, cofounder of the Medical Center Orthotics and Prosthetics (MCOP) in Silver Spring, MD. “We see a lot of above knee amputees develop Trendelenburg gait, for example, where they lean over the affected side because their muscles aren’t strong enough to support them as they advance through on their sound side, changing their gait symmetry so they land on the sound side with greater force.”

Morgenroth noted these associations make sense, but pointed to a missing piece of the puzzle.

“If increased knee loading in the intact limb during walking is really responsible for this increased prevalence, then in these studies we should see an increase in knee loading of the intact limb compared to the control population, not just compared to the prosthetic side,” he said. “This is where the literature hits a bit of a rough patch.”

A 2014 study comparing young, active transtibial amputees with nonamputees found that the intact limb in individuals with TTA wasn’t associated with greater peak external knee adduction moment (EKAM) or EKAM impulses across a range of walking speeds, though those with TTA had increased peak external knee flexor moment, peak vertical force, and loading rate as speed increased.12 Investigators concluded the amputees’ sound limb displayed few biomechanical risk factors for knee OA during walking compared with able-bodied individuals.12

A 2015 study from the same group found that young, active transfemoral amputees don’t have greater sound-limb peak EKAM while walking at a self-selected speed than able-bodied controls, although they did find higher maximal loading rates in the amputees.13 Surprisingly, the investigators also found that peak EKAM and impulses relative to body mass were lower among those with TFA than controls.

“I think the theory is still a very good theory, and I think there is some evidence pointing in that direction, but I also think we need larger, methodologically sound studies that provide more definitive evidence,” Morgenroth said.

So, what else could be contributing?

Hopping, which is associated with high forces at the knee, may be a factor.

“We used to teach people how to get up and hop around the bedroom in the morning and hop into the shower,” said Robert S. Gailey, PhD, PT, professor of physical therapy at the University of Miami Miller School of Medicine in Florida. “Now we’re saying, ‘Hey, put on your prosthesis, and don’t hop unless you absolutely have to.”

Clinically, Morgenroth has observed that transfemoral amputees tend to hop more frequently than transtibial amputees.

“Getting a transfemoral prosthesis on takes quite a bit more work typically than getting a transtibial prosthesis on. So maybe they wake up to urinate, and they just hop into the bathroom,” he said. “As you can imagine, the forces on the intact limb during hopping may be even higher than those during walking. This hasn’t really been studied in the literature, so at this stage it’s more of a clinical hypothesis.”

Tasks such as rising from a chair may be hard on the knees on unilateral amputees, Gailey said.

“There’s not much hard evidence, currently, though people are working on it, but when a person gets in and out of the chair they primarily rely on the sound limb to do that, so that puts greater wear on the knee,” he said.

In a 2016 study, Gailey and colleagues calculated the symmetry of weight distribution between lower limbs as 12 transtibial amputees and 12 age-matched nonamputees did a sit-to-stand activity.14 Amputees showed an asymmetrical weight distribution pattern and a tendency to transfer weight to the intact limb during the activity, findings not seen in the nonamputees.

Morgenroth thinks sit-to-stand tasks could potentially contribute to OA risk, particularly in younger, more active amputees.

“I think rising to a standing position from a seated position, especially if you’re in a low chair or on a low bed or couch, increases the patellofemoral forces rather than the tibiofemoral forces,” he said. “This brings up another good point: When we talk about knee osteoarthritis there’s three compartments, the medial tibiofemoral compartment, the lateral tibiofemoral compartment, and the patellofemoral compartment. At least in the general population it’s known the medial tibiofemoral compartment is affected much more commonly than the lateral, but the patellofemoral is also affected commonly.15 So, I think another important aspect of future research is trying to correlate forces in specific places that are associated with degeneration in those specific places. So, if you’re thinking about sit to stand as a potential reason, it might be important to look at the related patellofemoral forces.”

Ascending and descending stairs, and even simple standing can also be problematic for amputees, Gailey said.

“When they go up and down stairs they will often skip a step, so all the work is done on the sound limb. When they come down, they put as little weight on the prosthetic limb as possible, and so three to four times body weight comes down on the sound limb. While standing, they tend to shift the weight to the sound limb,” he said. “All these things accumulate year in and year out. You’re sitting, standing, climbing stairs, and walking with increased effort, and the biomechanics of that limb are altered. The result is the knee breaks down a heck of a lot sooner than it should.”

Device fit and physical therapy

Prosthetic fit is crucial for gait symmetry and to lower the risk of knee joint degeneration and OA, Corcoran said.

“If that socket is uncomfortable, you’re going to spend more time on the sound side,” he said. “As a prosthetist, you want to make the person comfortable, as all the technology in the world doesn’t mitigate a poor-fitting socket.”

All the experts interviewed for this article agreed.

“The first thing I advise people is to find a prosthetist who they like, who they believe is listening to their needs,” Gailey noted. “Because, first, you need a well-fitting socket that allows you to use the muscles in the hip—or, for a below-knee amputee, about the knee—and to use them rapidly enough to bounce over the prosthesis. If you don’t have that, you’re never going to trust the limb, you’re never going to be able to balance on it, and you’re going to have additional forces.”

Optimizing fit will help mitigate asymmetrical loading, said Morgenroth, who noted it also has a significant effect on functional mobility, comfort, and satisfaction.

“It’s really tempting to look at high-tech solutions and think they will solve all problems, but if you have a poor fit, there’s no prosthetic foot that will help you,” he said.

Corcoran said his group is using total surface-bearing sockets with a silicone liner or silicone or gel interface.

“With these sockets, every part of the limb bears weight, and that increases comfort,” he said. “Sockets also need to be aligned correctly to allow a normalized gait, and then there’s the importance of physical therapy that focuses on gait symmetry to mitigate the risk of OA. You’ve got to use the prosthetic properly and keep in the back of your mind not just how much you’re walking, but how well you’re walking.”

Physical therapy can help amputees walk, rise from chairs, and do other tasks with greater symmetry, Gailey said.

“I spend a lot of time working on posture, teaching them how to shift their weight when they’re standing—sometimes on the sound limb, sometimes on the prosthetic limb—as an able-bodied person would do,” he said. “I also advise them to spend the time to use their prosthesis properly because it will pay off over time; those that do can routinely incorporate those motor patterns into their everyday life and preserve the limb.”

Morgenroth would like to see physicians who work with amputees gain more experience with the population and the current literature.

“We have a lot of physicians treating amputees who don’t have extensive training in it. That’s very different than in the prosthetics community, where clinical prosthetists do have a lot more training, although from a different perspective,” he said. “One of the key things is for clinicians to read the research with a critical eye and incorporate it into a thoughtful approach in the clinical setting that includes not only their reading of the literature, but their clinical experience as well.”

Prostheses and push-off

Newer research on the effects of different prosthetic device types has changed Morgenroth’s preventive strategies and treatments, he said.

“Typical prosthetic feet are associated with reduced push-off on the trailing limb [compared with intact foot push-off], which is associated with higher loading in the intact knee,” he said.

Morgenroth and colleagues hypothesized greater prosthetic push-off would reduce loading of the entire intact limb and thereby reduce knee loading, and in 2011 reported on the effects of an experimental foot.16 They studied seven unilateral transtibial amputees who walked with different three prosthetic feet with varying amounts of push-off: a controlled energy and storage return (CESR) prototype that increases push-off, a conventional foot, and the participants’ currently prescribed foot.

The CESR foot on average produced 68% and 137% more push-off, respectively, than the prescribed foot and the conventional foot. The prosthetic foot condition with the least push-off demonstrated the largest EKAM, which was reduced by 26% with the prosthetic foot producing the most push-off. Trailing prosthetic limb push-off impulse was also negatively correlated with loading impulse in the intact leading limb.

However, research with commercially available feet that increase push-off on the prosthetic side have had mixed results.

Alena Grabowski and her coinvestigator in 2013 analyzed ground reaction forces and knee joint kinetics in seven transtibial amputees and seven age-, weight- and height-matched nonamputees who walked at five speeds over level ground under two conditions: with a commercial powered ankle-foot, which provided net positive mechanical work and powered ankle flexion during late stance, and with the patient’s own passive-elastic prosthesis.17

Compared with the passive device, using a powered prosthesis reduced the force and knee adduction moment in the unaffected leg during level walking, significantly decreased unaffected leg peak resultant forces by 2% to 11% at speeds of .75 m/s to 1.5 m/s, and significantly decreased first peak EKAM moments by 21% and 12% at 1.5 m/s and 1.75 m/s, respectively.

“In addition to decreasing the risk of osteoarthritis in the unaffected leg, we have also found that use of a powered prosthesis normalizes walking biomechanics, metabolic costs, and preferred speeds.18 Further, use of a powered compared to a passive prosthesis may improve the regulation of balance via the regulation of angular momentum in people with unilateral transtibial amputation,”19 Grabowski said.

Loading rates in Grabowski’s 2013 study were not significantly different between prosthetic feet.

“Loading rate may not be a risk factor for osteoarthritis,” she said. “Further, the loading rates had a large range of variability across subjects, suggesting there are many different loading rate strategies utilized by people without and with an amputation.”

In a 2014 Clinical Biomechanics study, Esposito and Wilken recruited 10 transtibial amputees and 10 able-bodied controls to compare limb loading between passive and powered prostheses, loading in the sound and amputated limbs, and loading in amputees and controls. Participants underwent gait analysis at three speeds.12

The powered prosthesis wasn’t associated with a decrease in the sound limb’s peak EKAM or impulse, but was associated with lower external knee flexor moment, peak vertical force, and loading rate as speed increased compared with the passive foot, and a slower loading rate compared with controls.

“The nice thing about the idea of increasing prosthetic push-off is there are other potential benefits, including improved gait symmetry and gait efficiency,” Morgenroth said. “I think it’s a concept that patients like and that feels good to them.”

Traditional interventions

Practitioners are also using more traditional knee OA interventions, such as braces and insoles for the intact limb.

“I always look at the sound limb and the foot and often recommend off-the-shelf insoles,” Gailey said. “Even if someone has good arches, there’s a good chance they’ll lose the arch; I base that anecdotally on the number of people that I’ve seen come through my clinic. Those who have been amputees for ten or more years have lost the arch or complain of knee pain, plantar fasciitis, or they have other kinds of ankle pain or a pronated foot.”

If sound-side knee pain develops, Gailey is likely to prescribe an unloader brace for use during high physical activity.

Morgenroth noted that patient tolerance of bulky unloader braces can be a problem, and said he is cautious in his use of wedged insoles.

“It’s really only helpful in potentially treating medial tibiofemoral OA and depends on a patient’s foot and ankle mobility,” he said. “If you have a patient with a lot of hindfoot motion, for instance, if you crank on their calcaneus relative to their tibia you see quite a bit of motion. You can imagine that might sort of soak up the wedging of the insole and won’t necessarily be translated up to the knee. That said, when you have a really rigid foot in the frontal plane, even though those forces may be translated up to the knee, it may be quite uncomfortable depending on the wedge angle. So, again, it has to be patient-specific.”

Corcoran emphasizes to his patients the need to maintain a healthy weight to keep load off the knee joint, as well as paying attention to changes in prosthetic fit.

“I stress the importance of sock management as the residual limb shrinks throughout the day—or if they gain weight and they’re growing out of their socket,” he said.

He also thinks consulting with a physical therapist every few years is a good idea.

“We see veterans that got hurt fifteen years ago coming back for prosthetic care who haven’t had PT for ten or twelve years, and we definitely see a real deterioration in their gait,” Corcoran said. “You need the complete package to reduce the risk of degenerative changes—a well-fitting, well-aligned prosthesis, an emphasis on gait quality, and an involved, educated patient.”

  1. Gailey R, Allen K, Castles J, et al. Review of secondary physical conditions associated with lower-limb amputation and long-term prosthesis use. J Rehabil Res Dev2008;45(1):15-29.
  2. Hungerford DS, Cockin J. Fate of the retained lower limb joints in Second World War amputees. Proceedings and Reports of Universities, Colleges, Councils, and Associations. J Bone Joint Surg 1975;57(B1):111.
  3. Burke MJ, Roman V, Wright V. Bone and joint changes in lower limb amputees. Ann Rheum Dis 1978;37(3):252-254.
  4. Kulkarni J, Adams J, Thomas E, Silman A. Association between amputation, arthritis and osteopenia in British male war veterans with major lower limb amputations. Clin Rehabil 1998;12(4):348-353.
  5. Struyf PA, van Heugten CM, Hitters M, Smeets RJ. The prevalence of osteoarthritis of the intact hip and knee among traumatic leg amputees. Arch Phys Med Rehabil 2009;90(3):440-446.
  6. Mussman M, Altwerger W, Eisenstein J, et al. Contralateral lower extremity evaluation with a lower limb prosthesis. J Am Podiatr Assoc 1983;73(7):344-346.
  7. Norvell DC, Czerniecki JM, Reiber GE, et al. The prevalence of knee pain and symptomatic knee osteoarthritis among veteran traumatic amputees and nonamputees. Arch Phys Med Rehabil 2005;86(3):487-493.
  8. Lemaire ED, Fisher RF. Osteoarthritis and elderly amputee gait. Arch Phys Med Rehabil 1994;75(10):1094-1099.
  9. Russell Esposito E, Wilken JM. Biomechanical risk factors for knee osteoarthritis when using passive and powered ankle-foot prostheses. Clin Biomech 2014;29(10):1186-1192.
  10. Russell Esposito E, Aldridge Whitehead JM, Wilken JM. Sound limb loading in individuals with unilateral transfemoral amputation across a range of walking velocities. Clin Biomech 2015;30(10):1049-1055.
  11. Agrawal V, O’Toole C, Gaunaurd IA, Gailey RS. Analysis of weight distribution strategies in unilateral transtibial amputees during the stand-to-sit activity. Ergonomics 2016;59(1):121-129.
  12. Kim YM, Joo YB. Patellofemoral osteoarthritis. Knee Surg Relat Res 2012;24(4):193-200.
  13. Morgenroth DC, Segal AD, Zelik KE, et al. The effect of prosthetic foot push-off on mechanical loading associated with knee osteoarthritis in lower extremity amputees. Gait Posture 2011;34(4):502-507.
  14. Grabowski AM, D’Andrea S. Effects of a powered ankle-foot prosthesis on kinetic loading of the unaffected leg during level-ground walking. J NeuroEng Rehabil 2013;10:49.
  15. Herr HM, Grabowski AM. Bionic ankle-foot prosthesis normalizes walking gait for persons with leg amputation. Proc Biol Sci 2012;279(1728):457-464.
  16. D’Andrea S, Wilhelm N, Silverman AK, Grabowski AM. Does use of a powered ankle-foot prosthesis restore whole-body angular momentum during walking at different speeds? Clin Orthop Relat Res 2014;472(10):3044-3054.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.