Which muscles take the stress off your knees

Almost every article on this subject tells you that weak hips cause runner’s knee. The largest prospective study ever done on the question — 4,818 people, followed until 483 of them developed knee pain — found that hip strength did not predict it at all. Here is what the evidence actually supports.

What is really taking the load

When you run, the knee is rarely the thing doing the work. Simulation studies that model every major muscle during running give a clear division of labour:

• Early stance — the quadriceps. They are the biggest contributor to braking and to holding you up as you land. This is the shock-absorbing job.
• Late stance — the calf. Soleus and gastrocnemius become the greatest contributors to both support and propulsion as you push off. At ordinary running speeds, up to about 7 m/s, the ankle plantarflexors provide most of your vertical support.
• Swing — the hips and hamstrings. Above 7 m/s the calf can no longer shorten fast enough, and speed comes instead from stride frequency driven by the iliopsoas, gluteus maximus and hamstrings.

For a recreational runner, that means the calf is doing more of the work than almost anyone realises, and the quadriceps are the shock absorber. The knee is the joint in the middle where the consequences show up.

The load number that explains a lot

Running produces peak knee joint contact forces of 8.02 times body weight, against 2.72 for walking. Three times harder per step.

But measured per kilometre covered, it is 0.80 body weights per metre running against 0.75 walking — not significantly different (p = 0.098). Each running step is harder, but it covers far more ground. This is why running does not produce the joint destruction the peak forces suggest, and it is covered in more detail here.

The hip weakness story is probably backwards

This is the correction that matters most, and it goes against what you will read almost everywhere.

The cross-sectional evidence is real: people who currently have patellofemoral pain do have weaker hips than people who do not. Same for runners with current iliotibial band syndrome — in one meta-analysis, lower hip abductor strength was found in female runners with current ITBS only, and not in retrospective or prospective cohorts.

The prospective evidence says it does not predict anything. A meta-analysis of 18 prospective studies followed 4,818 people until 483 developed patellofemoral pain. Hip weakness was not a risk factor — pooled effects ranged from −0.09 to −0.20 and none were meaningful. In adolescents the result actually ran the other way: greater hip abduction strength was associated with more future knee pain (SMD 0.71, 95% CI 0.39 to 1.04).

An earlier review put it plainly: “reduced hip strength may be a result of PFP rather than the cause.”

Which makes sense. If your knee has hurt for six weeks, you have been walking, sitting and moving differently for six weeks. Measuring your hip strength afterwards tells you about the consequence, not the cause.

One prospective study dissents, and it is worth knowing about: 629 novice runners tracked over a year found that higher eccentric hip abduction strength was associated with lower knee-pain risk over the first 50 km. That was an unadjusted analysis, limited to the first 50 km — but it measured eccentric strength, where the null findings all measured isometric strength. That distinction is the live open question in this area.

What does predict knee pain

From the same prospective evidence: quadriceps weakness. In military recruits, weaker knee extensors predicted future patellofemoral pain, most strongly when normalised to BMI (SMD −0.69, 95% CI −1.02 to −0.35). A separate review of 7,518 military personnel, of whom 572 developed the condition, found the same thing at the same effect size.

What did not predict it: age, height, weight, BMI, body fat and Q angle.

So if you are going to train one thing for knee pain prevention on the current evidence, it is the quadriceps — and the calf, on the load-sharing argument above rather than on direct prevention evidence.

Strengthening works — but not for the reason usually given

Two things are simultaneously true here, and holding both is the honest position.

Hip and knee strengthening genuinely helps people who already have patellofemoral pain. Adding hip work to knee work reduced pain by a further 1.5 points on a 10-point scale (95% CI 0.8 to 2.3), with benefits maintained after the programme ended. A Cochrane review found exercise reduced pain during activity by 1.46 points (95% CI −2.39 to −0.54) and improved function substantially.

But it does not appear to work by making you stronger. The same meta-analysis that found those pain improvements reported that “these outcomes were achieved without a concurrent change in strength” — no significant strength gain for any intervention. Something else is doing the work: load management, coordination, the analgesic effect of exercise itself, or simply time.

And the Cochrane review is blunt about quality: “the evidence for each of seven main outcomes for all comparisons was of very low quality, generally due to serious flaws in design and small numbers of participants. This means that we are very unsure about the estimates.”

So: do the exercises if your knee hurts — they help. Do not sell yourself the story that you are preventing future pain by building hip strength, because the prospective evidence does not support it.

Where strength training is genuinely proven

Here the evidence is strong, and it is the real reason to lift.

A meta-analysis of six randomised trials covering 7,738 people aged 12 to 40, with 177 injuries, found strength training reduced injury risk to about a third: RR 0.338 (95% CI 0.238 to 0.480). The strength of evidence was rated high, no publication bias was detected, and there was a dose-response — a 10% increase in strength training volume reduced injury risk by more than four percentage points.

That looks decisive, and we used to present it that way. The problem is the population: those trials were mostly footballers, handball players and military recruits doing supervised neuromuscular programmes, not runners. When the randomised trials of exercise-based prevention in endurance runners specifically were pooled — nine trials, 1,904 people — the result was not significant. We set that out in full in what actually prevents running injuries.

So the honest version: lift because it makes you a more durable and more economical runner, and because it preserves muscle you would otherwise lose with age. Do not lift on a promise that it will keep you injury-free, because in runners that has not been shown.

One honest caveat: those trials were mostly not in runners. The running-specific prevention evidence is much thinner and largely null. We think the general finding is strong enough to act on; you should know it is an extrapolation.

One thing to be careful about

Stronger quadriceps predict less symptomatic arthritis — knee extensor weakness carried OR 1.65 (95% CI 1.23 to 2.21) for developing symptomatic knee OA across 5,707 people. But there is a genuine complication.

An 18-month study of people with existing knee arthritis found that in malaligned knees, greater baseline quadriceps strength was associated with more structural progression (predicted probability 0.41 versus 0.19, p = 0.03), with a similar signal in very lax knees. A later study using MRI found no such effect on cartilage loss and did not reproduce the harm signal — but it is on the record.

What that means practically: if you already have diagnosed knee arthritis, known malalignment or significant joint laxity, get individual assessment before starting a heavy loading programme rather than following a generic plan off a website — including this one.

Two changes to how you run

Both are biomechanically well demonstrated, and neither has been shown to prevent anything. Take them as ways to reduce load on a knee that is currently complaining, not as prevention.

• Shorten your stride by about 10% — reduces patellofemoral load by 15 to 20% per step, and 9 to 12% per kilometre.
• Increase your step rate by about 10% — reduces peak patellofemoral stress and stress per kilometre, most effectively at slower speeds.

Worth knowing if you train on a treadmill: patellofemoral load is no different from overground running, but Achilles tendon force is 12.5% higher, with a 15.6% greater loading rate. Load taken off one structure tends to land on another.

A twenty-minute routine, twice a week

Built around what the evidence actually supports: quadriceps and calf as the priorities, hips because they help when things hurt, and progressive load because that is what produced the injury reduction.

1. Split squats — 3 sets of 8 each leg. The main quadriceps movement. Add load once bodyweight is comfortable.
2. Single-leg calf raises — 3 sets to two reps short of failure, slow, full range, off a step. The most under-trained muscle in recreational running.
3. Step-downs — 3 sets of 8 each leg, controlled on the way down. This trains the braking action the quadriceps do in early stance.
4. Single-leg glute bridges — 3 sets of 10 each side, hips level.
5. Side-lying hip abduction or banded walks — 2 sets of 15 each side.
6. Hamstring curls or Nordic curls — 3 sets of 6 to 8.

Progress it. The injury-prevention finding was dose-dependent, and the trials that worked used supervised, progressive loading. Doing the same six exercises at the same difficulty for a year is not what was tested.

When it is not just a niggle

See a doctor or physiotherapist rather than working through a routine if you have: knee pain lasting more than two weeks despite reduced running, swelling, the knee giving way or locking, pain that wakes you at night, or any pain following a specific injury such as a twist or a fall. A previous ACL or meniscal injury raises the odds of later arthritis by four to six times, and getting it properly rehabilitated matters more than any exercise on this page.

References

• Neal BS, et al. Risk factors for patellofemoral pain: a systematic review and meta-analysis. Br J Sports Med, 2019;53(5):270–81. doi:10.1136/bjsports-2017-098890
• Rathleff MS, et al. Is hip strength a risk factor for patellofemoral pain? Br J Sports Med, 2014;48(14):1088. doi:10.1136/bjsports-2013-093305
• Hamner SR, Seth A, Delp SL. Muscle contributions to propulsion and support during running. J Biomech, 2010;43(14):2709–16. doi:10.1016/j.jbiomech.2010.06.025
• Dorn TW, Schache AG, Pandy MG. Muscular strategy shift in human running. J Exp Biol, 2012;215(11):1944–56. doi:10.1242/jeb.064527
• Miller RH, et al. Why don’t most runners get knee osteoarthritis? Med Sci Sports Exerc, 2014;46(3):572–9. doi:10.1249/MSS.0000000000000135
• Lauersen JB, Andersen TE, Andersen LB. Strength training as superior, dose-dependent and safe prevention of acute and overuse sports injuries. Br J Sports Med, 2018;52(24):1557–63. doi:10.1136/bjsports-2018-099078
• Nascimento LR, et al. Hip and knee strengthening is more effective than knee strengthening alone for patellofemoral pain. J Orthop Sports Phys Ther, 2018;48(1):19–31. doi:10.2519/jospt.2018.7365
• van der Heijden RA, et al. Exercise for treating patellofemoral pain syndrome. Cochrane Database Syst Rev, 2015;1:CD010387. doi:10.1002/14651858.CD010387.pub2
• Øiestad BE, et al. Knee extensor muscle weakness is a risk factor for development of knee osteoarthritis. Osteoarthritis Cartilage, 2015;23(2):171–7. doi:10.1016/j.joca.2014.10.008
• Sharma L, et al. Quadriceps strength and osteoarthritis progression in malaligned and lax knees. Ann Intern Med, 2003;138(8):613–9. doi:10.7326/0003-4819-138-8-200304150-00006
• Amin S, et al. Quadriceps strength and the risk of cartilage loss and symptom progression in knee osteoarthritis. Arthritis Rheum, 2009;60(1):189–98. doi:10.1002/art.24182
• Ramskov D, et al. Higher eccentric hip abduction strength is associated with a reduced risk of developing patellofemoral pain among novice runners. J Orthop Sports Phys Ther, 2015;45(3):153–61. doi:10.2519/jospt.2015.5091
• Foch E, Brindle RA, Pohl MB. Iliotibial band syndrome in runners: a systematic review and meta-analysis. Gait Posture, 2023;101:73–81. doi:10.1016/j.gaitpost.2023.02.001
• Willson JD, et al. Influence of step length and landing pattern on patellofemoral joint kinetics during running. Scand J Med Sci Sports, 2015;25(6):736–43.
• Willy RW, et al. Patellofemoral joint and Achilles tendon loads during overground and treadmill running. J Orthop Sports Phys Ther, 2016;46(8):664–72.