Runner’s Knee: Why the VMO Theory Is Out and Hip Strength Is In

Educational journalism, not medical advice. Every claim here is checked against its cited sources by editor Tim Bunce — a health writer, not a physician. It isn’t specific to your situation: for health decisions, talk to your own clinician. How we work →

The VMO myth

For decades, patellofemoral pain treatment focused on the vastus medialis obliquus (VMO), the medial quadriceps muscle responsible for the “tear drop” visible just above the kneecap. The theory: if the lateral quad (vastus lateralis) was relatively stronger than the VMO, the kneecap would track laterally, irritating the patellofemoral joint. The treatment: terminal-range leg extensions, “VMO activation” drills, taping.

The evidence stopped supporting this picture by the mid-2000s. EMG studies couldn’t reliably show selective VMO activation in any standard exercise; imaging studies couldn’t show consistent VMO atrophy in PFPS patients; and RCT comparisons of VMO-focused vs. general-quad programmes showed no advantage for VMO targeting Collins 2018.

Where the field landed: the hip

The biomechanical work that changed treatment was Ireland and colleagues’ 2003 paper documenting the “dynamic Q angle” pattern in PFPS. During single-leg landing, women with PFPS showed:

• Greater hip adduction (femur moving inward)
• Greater hip internal rotation
• Greater knee valgus (knee moving inward over the foot)
• Weaker isometric hip abduction on the affected side Ireland 2003

The pattern matters because the kneecap doesn’t track in space — it tracks in the trochlear groove of the femur. If the femur rotates inward under the patella during loading, the patella appears to track laterally relative to the leg, but it’s actually the femur that moved. Strengthening the hip abductors restores the femur position; the “maltracking” resolves Rabelo 2018.

“Hip-focused exercise programmes produce greater short- and medium-term reductions in patellofemoral pain than knee-focused programmes. The effect is largest in patients with measurable hip abductor weakness on the affected side.”

— Collins et al., Br J Sports Med, 2018 view source

The drills that actually work

The published rehabilitation trials use variations on a small set of hip-abductor and hip-rotator drills:

• Side-lying hip abduction. Lie on the non-affected side, lift the top leg straight up about 30-45°, keep the toe pointed forward (slight internal rotation activates gluteus medius rather than tensor fasciae latae). 3 sets of 15-20 per side.
• Clamshells. Side-lying, knees bent at 45°, heels together. Lift the top knee while keeping the heels touching. Pelvis stable. 3 sets of 15-20 per side.
• Banded lateral walks. Resistance band around ankles or above knees, slight athletic squat, walk sideways 10-15 steps each direction. 2-3 sets.
• Single-leg deadlifts. Balance on one leg, hinge at the hip, opposite leg extends back as torso lowers. Trains hip stability + posterior chain in one movement. 2-3 sets of 8-10 per leg.
• Step-downs with control. Stand on a 15-20 cm step, slowly lower the opposite heel to the floor, return. Knee must track straight over the second toe — not collapse inward. 3 sets of 10 per leg.

Two 20-minute sessions weekly for 6-8 weeks is the standard prescription. The published RCT response rates are 60-80% pain reduction at 12 weeks for adherent patients Rabelo 2018.

What else matters

• Running form: cadence. Increasing step rate by 5-10% reduces patellofemoral joint load 20-30% without changing pace. Most runners with PFPS benefit from a metronome-paced 5-10% cadence increase during recovery Heiderscheit 2011.
• Volume management. Most PFPS is overuse. Cap weekly running volume at 70-80% of normal during recovery, then progress.
• Footwear. No single shoe type has been shown to prevent or treat PFPS. Comfort matters more than category.
• Pain-free activity. Cycling and swimming usually remain pain-free during PFPS recovery. Don’t stop conditioning — substitute the modality.

When to see a clinician

• Pain that doesn’t improve after 4-6 weeks of consistent hip-focused work
• Swelling, locking, or instability (not typical of PFPS; suggests intra-articular pathology)
• Night pain or rest pain
• Pain following a specific traumatic event
• Bilateral symptoms in someone who’s only training one side hard

Practical takeaways

• The dominant driver of patellofemoral pain in active adults is hip-abductor and hip-rotator weakness, not localised VMO weakness. The VMO theory is outdated.
• The drills that work: side-lying hip abduction, clamshells, banded lateral walks, single-leg deadlifts, controlled step-downs. 6-8 weeks of consistent practice produces 60-80% pain reduction in published RCTs.
• For runners with PFPS, increase cadence 5-10% — reduces patellofemoral load 20-30% without changing pace.
• Substitute cycling or swimming during recovery. Don’t stop conditioning — the deconditioning prolongs recovery.
• See a clinician if pain doesn’t improve in 4-6 weeks, or if you have swelling, locking, night pain, or post-traumatic onset.

Don't ditch the quads: the real winner is hip AND knee

It is tempting to read "the VMO theory is out" as "stop training the quadriceps." That is not what the evidence says, and it is worth correcting before anyone abandons the front of the thigh entirely. The shift was away from the narrow idea that you can selectively fire one slice of the quad (the vastus medialis obliquus, or VMO) to re-rail the kneecap. It was not a verdict that quadriceps strength is irrelevant. The strongest synthesis to date makes this clear: a 2018 systematic review and meta-analysis pooling 14 trials and 673 people found that combined hip-and-knee strengthening outperformed knee strengthening alone for reducing pain and improving activity, and it did so without necessarily producing a measurable strength gain — suggesting the benefit comes from how the whole limb is coordinated under load rather than from any single muscle getting bigger Nascimento 2018.

An earlier 2015 meta-analysis reached a complementary conclusion: adding proximal (hip and trunk) work to a programme reduced pain in the short and medium term, and proximal-plus-quadriceps training delivered the best pain and function results at one year Lack 2015. In plain terms, the hip is the part most programmes used to neglect, so emphasising it produces the biggest marginal improvement — but the quadriceps still belong in the plan. The practical takeaway is additive, not substitutive: keep the squats, step-ups, and leg work that load the knee through a comfortable range, and add the side-lying abduction, clamshells, and banded walks described earlier. The 2015 best-practice guide, built on level-1 evidence plus expert clinical reasoning, frames exactly this as the core of care — an individually tailored, multimodal programme combining gluteal and quadriceps strengthening with education and activity modification Barton 2015.

How hard, how deep, and how much pain is okay

The earlier drills section gives a sensible starting dose. Here is what governs whether you can push harder — and the single most useful concept is that different movements stress the kneecap by wildly different amounts. A 2023 biomechanics study modelled patellofemoral joint load across 35 weight-bearing exercises and everyday activities and found it ranged from roughly 0.6 times body weight during walking to about 8.2 times body weight during a single-leg decline squat, with a consistent pattern: the deeper the knee bends under load, the higher the force on the joint Song 2023. Walking, shallow double-leg squats, and low step-ups sit at the gentle end; lunges and running sit in the middle; deep single-leg squats and decline squats sit at the punishing end.

That hierarchy is your dial. If a movement flares the knee, you do not have to stop training it — you can usually keep the same exercise and simply reduce the depth, the step height, or the load until the force drops into a tolerable band, then climb back up as symptoms allow. This is where a pain-monitoring rule helps. A widely used model, tested in a randomised trial of loaded rehabilitation, permits training into mild discomfort: pain during and after the session may reach but should not exceed roughly 5 on a 10-point scale, it should settle back toward baseline by the next morning, and it should not creep upward week to week Silbernagel 2007. If the morning-after pain is climbing or symptoms worsen across successive weeks, the load is too high and should be scaled back. That model was developed for tendon rehabilitation, so treat the exact number as a guide rather than a law — but the principle that some pain during strengthening is acceptable as long as it settles is a genuine departure from the old "if it hurts, rest it" advice, and it is why progress tends to be faster when load is adjusted rather than avoided.

A worked example makes the loading dial concrete. Suppose body-weight step-downs from a normal step provoke a flare. Rather than dropping the exercise, lower the step to a thin book's height so the knee bends less — which cuts the peak force Song 2023 — and rebuild the height over a few weeks as the morning-after pain stays quiet. The same logic governs squat depth: a quarter squat loads the joint far less than a deep single-leg squat, so chronic flare-ups are often a depth-and-load problem rather than a sign the movement itself is harmful. People sometimes ask whether to push through pain on days when symptoms are higher; the safer answer is to keep moving but reduce the dose that day rather than skipping entirely, because complete rest tends to decondition the limb and prolong the problem.

What about taping, braces, and shoe inserts?

Plenty of people arrive with a knee brace, a roll of tape, or a pair of orthotic insoles and want to know whether these are doing anything. The honest answer is that they can help in the short term but are not a cure, and none of them replaces strengthening. Foot orthoses have the best trial evidence. A four-arm randomised trial of 179 adults compared prefabricated orthoses, flat inserts, physiotherapy, and a combination. At six weeks, more people improved with orthoses than with flat inserts (about four people needing treatment for one extra success); by 52 weeks that edge had vanished and all groups had improved similarly Collins 2008. The reasonable reading is that orthoses can buy short-term relief, particularly for people whose feet roll inward noticeably, but they speed recovery rather than change the destination — and they are not better than exercise.

Patellar taping sits in a similar place. The 2015 best-practice guide includes taping as a reasonable adjunct, mainly because it can reduce pain enough in the early going to let someone exercise and move more comfortably — not because it permanently fixes how the kneecap tracks Barton 2015. The framing that matters for a reader is this: tape, brace, and insoles are short-term symptom tools that can make the active treatment tolerable. If you find yourself relying on any of them months later with no progress, the plan is incomplete, because the durable gains come from the loading and movement work, not the hardware.

How long it takes, and why "it just goes away" is a myth

The most damaging belief about this condition is that it is self-limiting and will resolve on its own with rest. The long-term data say otherwise, and being honest about this protects readers from quitting their programme too soon. A multicentre study that followed people 5 to 8 years after diagnosis found that 57 percent still reported an unfavourable recovery, and a longer baseline duration of pain plus a worse initial function score predicted the poorer outcomes Lankhorst 2016. A 2025 evidence-and-gap map of the prognosis literature reinforced that recovery is highly variable and that a meaningful subgroup stays symptomatic, while noting that the psychological and behavioural factors influencing recovery remain understudied Neal 2025. The encouraging counterpoint is that most improvement happens in the first several months of consistent rehabilitation, and the people with the best odds are those who start sooner and stick with the loading work rather than waiting for spontaneous resolution.

Adolescents deserve a specific note. Patellofemoral pain is common in active teenagers, and their prognosis is at least as stubborn as adults', with a substantial share reporting persistent symptoms years later Neal 2025. The instinct to simply pull a young athlete out of all activity is understandable but counterproductive; guided load management that keeps them moving within tolerable limits is the better path, and a clinician should steer it. A brief word on who else should loop in a clinician early: this article is general education, not a personal treatment plan, and anterior knee pain is not always patellofemoral pain. Anyone with knee pain alongside a separate diagnosed condition, anyone managing pain with regular medication, pregnant readers, children and teenagers, and older adults in whom early osteoarthritis can mimic these symptoms should have the diagnosis and the exercise plan confirmed by a physiotherapist or physician before leaning on self-directed loading. For the large majority of otherwise healthy, active people, though, progressive strengthening guided by the pain rules above is both the safest and the most effective starting point the evidence supports — expect to invest two to three months of consistent work before judging whether the approach is failing.

References