Nearly one in four adults over 40 lives with painful osteoarthritis, a condition that can make everyday movement difficult and is one of the major causes of adult disability. The disease gradually wears down the cartilage that cushions joints. Once that damage occurs, doctors currently cannot reverse it. Treatment usually focuses on pain relief, with joint replacement becoming an option when symptoms become severe.
A clinical trial from researchers at the University of Utah, New York University, and Stanford University points to a different possibility: changing the way a person walks.
A Personalized Shift in Foot Angle
In the yearlong randomized controlled trial, people with knee osteoarthritis were trained to make a small, personalized change in the angle of their foot while walking. The result was striking. Participants who received the real gait retraining reported pain relief comparable to medication, and MRI scans suggested they had less knee cartilage deterioration than people in the placebo group.
The study, published in The Lancet Rheumatology, was co led by Scott Uhlrich of the University of Utah’s John and Marcia Price College of Engineering. According to the researchers, it was the first placebo controlled study to show that a biomechanical intervention could help treat osteoarthritis symptoms and potentially slow joint damage.
“We’ve known that for people with osteoarthritis, higher loads in their knee accelerate progression, and that changing the foot angle can reduce knee load,” said Uhlrich, an assistant professor of mechanical engineering. “So the idea of a biomechanical intervention is not new, but there have not been randomized, placebo-controlled studies to show that they’re effective.”
Why the Treatment Had To Be Customized
The study focused on people with mild to moderate osteoarthritis in the medial compartment of the knee, which is the inner side of the joint. This area usually carries more weight than the outer side, making it a common site for knee osteoarthritis.
But there is a key challenge: the best walking adjustment is not the same for everyone. Some people reduce knee loading by turning their toes slightly inward. Others benefit more from pointing them outward. For some, the wrong change can fail to help or even increase stress on the painful part of the knee.
“Previous trials prescribed the same intervention to all individuals, resulting in some individuals not reducing, or even increasing, their joint loading,” Uhlrich said. “We used a personalized approach to selecting each individual’s new walking pattern, which improved how much individuals could offload their knee and likely contributed to the positive effect on pain and cartilage that we saw.”
That point has become even more important as related research continues to show that foot angle changes can affect knee forces differently depending on the person, the joint, and the walking pattern. A 2024 study in Bioengineering, for example, found that inward and outward foot rotation affected different peaks of knee loading, while not significantly increasing ankle joint moments in the group studied. Other research has also shown that foot progression angle can be measured outside the lab with wearable sensors, supporting the idea that future versions of this approach could be easier to deliver in real life.
How the Trial Worked
During their first two visits, participants received a baseline MRI and walked on a pressure sensitive treadmill while motion capture cameras measured the mechanics of their gait. The researchers used those data to determine whether each person would benefit more from turning the toes inward or outward, and whether a 5° or 10° adjustment would be best.
This screening also identified people who were unlikely to benefit because none of the tested foot angle changes reduced loading in their knees. Those participants were excluded from the trial. The researchers noted that including such people in earlier studies may have helped explain why previous results on pain were less clear.
Of the 68 participants enrolled, half were assigned to the real gait retraining group. The other half received a sham treatment designed to control for the placebo effect. In the sham group, participants were assigned foot angles that matched their natural walking pattern. In the intervention group, each participant received the foot angle change that produced the greatest reduction in knee loading.
Training the New Walking Pattern
Both groups returned to the lab for six weekly training sessions. During these sessions, participants walked on a treadmill while wearing a device on the shin that provided vibration feedback. The vibrations helped them keep their assigned foot angle while walking.
After the six week training period, participants were encouraged to practice the walking pattern for at least 20 minutes each day. The goal was for the movement to become automatic. Follow up visits showed that, on average, participants stayed within one degree of their prescribed foot angle.
After one year, participants reported their knee pain levels and underwent a second MRI so researchers could measure changes in cartilage health.
“The reported decrease in pain over the placebo group was somewhere between what you’d expect from an over-the-counter medication, like ibuprofen, and a narcotic, like oxycontin,” Uhlrich said. “With the MRIs, we also saw slower degradation of a marker of cartilage health in the intervention group, which was quite exciting.”
A Drug Free Option for a Long Treatment Gap
For some participants, one of the most appealing parts of the approach was that it did not require pills, surgery, braces, or a device worn all day. One participant said: “I don’t have to take a drug or wear a device…it’s just a part of my body now that will be with me for the rest of my days, so that I’m thrilled with.”
That long term adherence could be one of the intervention’s biggest strengths. Many people develop osteoarthritis decades before they are candidates for joint replacement. During that time, they may rely heavily on pain medications and other symptom management strategies.
“Especially for people in their 30’s, 40’s, or 50’s, osteoarthritis could mean decades of pain management before they’re recommended for a joint replacement,” Uhrlich said. “This intervention could help fill that large treatment gap.”
A 2026 conference abstract in Osteoarthritis and Cartilage also highlighted continued interest in placebo controlled trials of foot progression angle retraining, underscoring that researchers are still trying to determine which gait strategies work best and for whom. However, this area remains under active study, and the 2025 Lancet Rheumatology trial is still one of the strongest clinical demonstrations of a personalized approach.
Why Patients Should Not Try This Alone
Although the findings are promising, the researchers emphasized that this is not a simple “turn your toes in” or “turn your toes out” recommendation. The benefit depended on careful measurement and personalization. For some people, the wrong adjustment could increase stress on the knee rather than reduce it.
That is why the process still needs to be simplified before it can be used widely in clinics. The motion capture system used to prescribe each person’s walking change is expensive and time consuming. The research team envisions a future version that could be delivered through physical therapy, with retraining taking place during normal walks rather than only inside a lab.
“We and others have developed technology that could be used to both personalize and deliver this intervention in a clinical setting using mobile sensors, like smartphone video and a ‘smart shoe’,” Uhlrich said. Future studies of this approach are needed before the intervention can be made widely available to the public.
Those interested in participating in future studies can contact Uhlrich’s Movement Bioengineering Lab by filling out this web form.
The study, titled “Personalised gait retraining for medial compartment knee osteoarthritis: a randomised controlled trial,” was published in The Lancet Rheumatology. Co lead authors are Valentina Mazzoli of NYU’s Department of Radiology and Julie Kolesar of Stanford’s Human Performance Lab. Coauthors include Amy Silder, Andrea Finlay, Feliks Kogan, Garry Gold, Scott Delp and Gary Beaupre of Stanford and the VA Palo Alto Medical Center. The research was supported by federal research grants from the Department of Veterans Affairs, National Institutes of Health and National Science Foundation.
