Knee pain when squatting: what's likely, what to change today and how to build back to depth

Who gets it, and what usually lies behind itEmerging evidence

Lifting sports record about 1 to 4 injuries per 1,000 hours of training, from mostly low-quality studies, and the knee is one of the three commonest sites, with the spine and the shoulder. In 6 years of training records from elite US weightlifters, 85% of knee injuries were tendon pain (tendinitis), and 90.5% of all injuries cost a day or less of training. Those were elite lifters: no study has followed ordinary gym-goers in the same way.

No trial has tested what causes knee pain in the squat. In 38 published powerlifting injury reports, only 3 blamed technique, and those named heavy loads or fatigue as well. We look first for a recent jump in weight, depth, volume or frequency, or a return to an old working weight after a break (our suggestion; no study has tested it).

Sources 1, 2, 3, 4

How the common causes differ

Kneecap pain (patellofemoral pain) is a common knee problem in teenagers and young adults. It hurts around or behind the kneecap, and squatting, stairs and long sits with the knees bent bring it on. Patellar tendon pain sits at one point at the lower tip of the kneecap and rises with the demand on the thigh muscles, especially jumping and landing. A worn (degenerative) meniscus tear is a problem from about the mid-30s, and knee arthritis from about 45.

Weak thigh muscles predicted later kneecap pain in military recruits, but weak hips did not, in a meta-analysis of 18 prospective studies. Hip exercise still helps kneecap pain, so do the hip work without assuming weak hips caused it.

Sources 5, 6, 7, 8, 9

Are deep squats bad for your knees?Mixed evidence

A review of more than 164 articles concluded that worries about deep squats harming healthy knees were unfounded: stress behind the kneecap peaks around 90° of bend and falls as you go deeper, as the contact area grows. An older review preferred the parallel squat for healthy athletes, because forces on the ligaments and menisci "may increase" deeper. They disagree on the best depth, not on harm, and no trial has counted injuries by depth.

In 117 former elite athletes, 31% of the weightlifters had knee arthritis on X-ray against 3% of the shooters, with the most arthritis behind the kneecap, partly explained by the lifters' high body weight. That doesn't show that ordinary squatting causes arthritis, and the National Institute for Health and Care Excellence (NICE) recommends exercise as a core treatment for knee arthritis.

Sources 10, 11, 12, 8

Is it the squat, or how it's loaded?Emerging evidence

In lab studies of healthy people, the load on the knee rises as it bends and peaks near the deepest point, while the stress behind the kneecap peaks around 90°. Tipping the shins forward, as raised heels do, gives the knee more of the work, and leaning the trunk forward and sitting back shifts it to the hips.

In 7 weight-trained men squatting their body weight, blocking the knees from passing the toes cut the turning load at the knee by about 22% but raised it at the hip about tenfold and tipped the trunk forward. Turning the feet out about 30° roughly halved the inward (valgus) load on the knee, and a wide stance raised it by about 23%. These are lab measures, mostly in healthy people, not injury outcomes: no trial has counted injuries by depth, stance or knee travel.

Sources 11, 10, 13, 14

Knees caving inMixed evidence

We found no study linking knees drifting in during a barbell squat with injury. In landing tests, knees moving in (valgus) didn't predict a torn anterior cruciate ligament (ACL) across 9 prospective studies, and it was a poor screening test in 710 elite female footballers and handball players. In small groups of adolescent girls, more valgus on landing went with later kneecap pain. So it's worth coaching, often by lightening the bar for now, but it isn't a reason for alarm.

Sources 15, 16, 17, 18

When a scan helps, and what it can't tell youStrong evidence

Kneecap and tendon pain are diagnosed from the story and an examination: tendon changes on scans also turn up in people whose pain comes from somewhere else. After an injury, a short checklist called the Ottawa knee rule decides whether you need an X-ray; it missed almost no fractures (sensitivity 98.5%). A suspected tendon tear needs same-day assessment, and an X-ray from the side often shows a complete tear. Knee arthritis is diagnosed without an X-ray from 45.

Scans of knees without pain often show changes: cartilage defects in 11% of people under 40 and 43% from 40, meniscal tears in 4% and 19%, and bone marrow changes in 30% of young athletes in weight-bearing sports. In 991 adults aged 50 to 90, 61% of those with a meniscal tear had no knee pain, aching or stiffness in the previous month. A report has to match your symptoms before it explains them.

Sources 6, 19, 20, 8, 21, 22

Exercise reduces kneecap pain, and hip plus knee work does moreStrong evidence

In a Cochrane review of 31 trials (1,690 people), exercise reduced pain during activity by about 1.5 points out of 10 in the short term compared with no exercise, and improved function. Each trial was small and low quality, but the results were consistent. Hip and knee exercises together reduced pain during activity by a further 2.2 points compared with knee exercises alone (3 trials, 104 people).

The 2018 international consensus recommends exercise, especially hip- and knee-focused work combined. An NHS leaflet uses 3 sets of 8 to 12 and mini squats of less than 30° of bend to start.

Sources 5, 23, 24

A staged programme helps patellar tendon painEmerging evidence

In the largest trial (76 athletes), a staged programme of holds, then slow heavy strength work, then springy loading and sport, keeping pain at 3 out of 10 or less, improved a tendon score by 28 points against 18 with painful eccentric squats at 24 weeks. More returned to sport (43% against 27%), though that difference could be chance, and only 40 to 49% stuck to their programme.

In 39 men, heavy slow resistance kept its gains at 6 months, while a steroid injection's faded. A review warns that tendon rehab can be slow and sometimes frustrating.

Sources 25, 26, 6

Some pain during exercise is acceptableEmerging evidence

In a meta-analysis of 7 trials (385 people) with long-lasting muscle and joint pain, exercise that allowed some pain did slightly better in the short term than pain-free exercise, and no worse later. People with Achilles tendon pain who kept running under a pain-monitoring model did as well as those who rested for 6 weeks. In a kneecap pain trial that used pain monitoring, 85% of 40 young women were back in sport without pain at 12 months, though there was no untreated group.

Sources 27, 28, 29

Changing the squat: what lab studies showEmerging evidence

No trial has tested squat changes for knee pain in lifters. In a lab model of 10 healthy people, stress behind the kneecap was lowest in squats from about 45° of bend to straight, and in leg extensions from 90° to 45°; in the last 30° of straightening, the leg extension loaded the kneecap more than the squat. That is why the plan starts with a high box and part-range leg extensions.

Turning the toes out and avoiding extremes of stance lower the load on the knee in lab studies: in 42 squatters, very wide stances with the toes forward and narrow stances with the toes turned far out gave the largest knee and hip loads. Bouncing out of the bottom raised the shearing load at the knee in studies a powerlifting review cites.

Sources 30, 14, 31, 4

What to change in the squat today (written steps)

Our 3D figure can't show these changes. No trial has tested them for knee pain in lifters: they come from lab studies in mostly healthy people, the condition trials and expert advice, so keep what passes the pain rule.

• 1. Load: drop to a weight you can lift for 8 to 12 reps with 2 or 3 to spare, inside the pain rule. Around half your usual working weight is a common starting point (our suggestion).
• 2. Depth: squat to a box or bench at the deepest height that keeps pain within the rule. With kneecap pain, start at about 45° of knee bend (a high box) and lower it a notch at a time. With pain above the kneecap, keep away from the deepest bend early.
• 3. Stance and toes: feet about shoulder width or a little wider, toes turned out about 15 to 30°, knees following the toes. Avoid the extremes: very wide with the toes pointing forward, or narrow with the toes turned far out.
• 4. Trunk and shins: early on, sit back a little more to the box, with the shins nearer upright. Later, go back to your normal squat.
• 5. Heels: leave heel wedges and lifting shoes out while the kneecap or the tendon above it is sore, and bring them back last if they help your depth.
• 6. Tempo: pause on the box rather than bouncing out of the bottom, and lower over about 3 seconds (the 3 seconds is our suggestion).
• 7. Volume: fewer squat sets for 2 to 4 weeks, keeping the rest of your training going (our suggestion). The tendon trial cut painful loading for at least 4 weeks.

Sources 30, 14, 31, 4, 25, 24

Back to the barbell: the load steps

No trial has tested how fast to bring the bar back after knee pain. Our suggestion borrows the step sizes the American College of Sports Medicine gives healthy lifters: add about 2 to 10% when you can do 1 or 2 more reps than planned. The squat technique workout suits this last stage; it includes the heel-elevated deep squat and the deep squat hold.

• 1. Go back to a barbell squat to a box or bench at the deepest height that passed the pain rule in stage 2, at about half your old working weight.
• 2. Lower the box a notch at a time over the weeks until you reach your usual depth, keeping the weight the same.
• 3. Then add load: about 2.5 to 5 kg at a time when every set finishes with 2 or 3 reps to spare and the knee is back to normal the next morning.
• 4. Change one thing at a time: depth, then load, then volume, then heel wedges or lifting shoes if you use them.
• 5. With tendon pain, keep the holds or slow heavy work twice a week, as the largest tendon trial did after the return to sport.

Sources 33, 25

Sleeves, wraps, injections and surgery

Knee sleeves and wraps: the studies we found measured performance, not pain. In 15 trained men, sleeve density didn't change squat force or power, and we found no trial of sleeves for knee pain in lifters; the 2018 consensus was uncertain about kneecap taping and bracing. They're fine for comfort, but they aren't a treatment (editorial).

Injections: for patellar tendon pain, a steroid injection's early gains had faded by 6 months, unlike slow heavy strength work, and the American Academy of Orthopaedic Surgeons (AAOS) links steroid use with weaker tendons. For knee arthritis, NICE says a steroid injection into the joint gives short-term relief of 2 to 10 weeks, and can be used to support exercise.

Surgery: for a worn meniscus or knee arthritis, a guideline panel strongly recommends against keyhole surgery, except for a knee that is truly locked, and in 140 adults with a worn meniscus tear, 12 weeks of supervised exercise did as well as keyhole surgery at 2 years. A complete tear of the quadriceps or patellar tendon is different: it almost always needs repair.

Sources 34, 23, 26, 20, 8, 35, 7, 36