Knee pain on the treadmill: the kneecap, incline and speed, the outside of the knee and what helps

What a treadmill asks of the knee, and what no study has countedEmerging evidence

Walking or running on a belt is walking or running indoors. In 18 healthy runners at their own pace, the load behind the kneecap didn't differ between the treadmill and running outside, though the Achilles tendon took about 12.5% more force on the belt. Across 33 studies of 494 adults, treadmill running was largely comparable to running outside, with a slightly flatter foot at landing, on low certainty evidence.

Walking is closer still: across 22 studies the measures were largely comparable, and across 55 studies of 1,005 adults people took shorter, quicker steps on the belt, which the authors read as a more cautious walk. Adults aged 65 to 81 needed time to get used to the belt, and in 10 adults aged 50 to 73 treadmill walking cost about 23% more energy than walking outside, so the same speed can feel harder.

No study or survey counts knee pain caused by treadmills. US emergency departments saw about 70,000 injuries from home exercise machines in 5 years, two-thirds of them linked to treadmills, with sprains and strains common in adults and cuts and soft tissue injuries in young children, but no figures for knee pain. So the map on this page is built from lab studies of slope, speed and stride and from what each condition is known to do; that a jump in time, speed, slope or decline is the usual story is our reading.

Sources 1, 2, 3, 4, 5, 6, 7, 8

Incline: easier on the main joint, not always on the kneecapEmerging evidence

Running uphill didn't raise the load behind the kneecap in the one study that measured it: in a small lab study of 20 healthy runners on a treadmill (we could read only its summary), a 6 degree slope (about 10.5%) gave the same kneecap stress as level running, while running downhill raised it.

Walking is different. Walking slowly up a 6 degree slope at about 2.7 kilometres an hour, against brisk flat walking at about 5.4 at a similar effort, cut the load through the main knee joint by 23% in adults with obesity and 35% in adults without, and at about 4.8 kilometres an hour a marker of inner-knee load fell on slopes of 10% or more in 15 young men and in 12 older men. Neither of those studies measured the kneecap.

Uphill, the hips do most of the extra work, but the knee bends more at landing and the front of the thigh works much harder as the slope rises: in 6 healthy adults, its activity rose by up to 154% on steep grades. The authors' suggestion that a grade just over 12% might be kindest to the kneecap was a suggestion, not a finding. Our reading: a slower walk up a gentle slope is a fair swap for brisk flat walking with an arthritic or inner-knee ache, but pain around the kneecap may not ease, and a steep, long 12-3-30 session is a big jump for a sore knee.

Sources 9, 10, 11, 12, 13, 14, 15

Speed, decline settings and strideEmerging evidence

Running faster raised the load behind the kneecap and on the patellar tendon with each step in 12 adults. Running slower cut the load per step but added steps, so the total load on the knee over a kilometre was 80% higher at a slow jog than at a fast run in 16 runners: slower is gentler only if you don't add distance.

Some treadmills have a decline setting, and downhill is the hard part for the front of the knee. Running down an 8.3% slope nearly doubled the braking work of the thigh muscles in 7 runners, impact peaks rose by 54% at 9 degrees in 10 runners, and in 15 female distance runners the knee absorbed more power and landed straighter on steeper declines. Walking down a 19% ramp roughly doubled the estimated force through the main knee joint, from 3.9 to 8 times body weight. These downhill studies used steeper grades than most home treadmills offer, so the size of the effect on a gentle decline is likely smaller. A straighter knee at landing is also when the IT band rubs: runners with IT band syndrome landed with the knee bent about 21 degrees, near the friction zone, and downhill running reduces that angle.

Stride matters when running. Reaching the foot well ahead of the body raised kneecap stress by 31% a step in 23 women runners, and steps 10% shorter cut it by 22%. Taking 5 to 10% more steps a minute at the same speed lowered the load absorbed at the knee in 45 runners and cut the peak force behind the kneecap by 14% in 30 adults. These are lab measures in healthy runners. When walking, people already take shorter, quicker steps on a belt, so overstriding matters mainly when running (our reading).

Sources 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27

Who gets it, and how the causes are told apartStrong evidence

Kneecap pain is common: about 23% of adults and 29% of adolescents had it in a year, pooled across studies. In runners, the knee is the commonest injured area, kneecap pain the commonest single diagnosis and IT band syndrome next, affecting an estimated 5% to 14% of runners. An orthopaedic page puts kneecap pain down to repeated stress and to a sudden change in how often, how long or how hard you exercise. So knee pain on the treadmill is most likely after a jump in time, speed, slope or decline, in a new routine such as 12-3-30 or a walking pad at a desk, in new runners, after a switch from outdoors, and from mid-life with arthritis (our reading).

Kneecap pain is defined by pain around or behind the kneecap brought on by loading a bent knee: squatting, stairs, running or hopping. A tendon is one point you can press, at the lower tip of the kneecap or its top edge, and patellar tendon pain builds with how hard the thigh muscles work. The outside of the knee at the same point in each run is usually the IT band. A stiff ache from mid-life, with morning stiffness of 30 minutes or less, is often arthritis, which the National Institute for Health and Care Excellence (NICE) diagnoses without an X-ray. Catching, locking or swelling at the joint line points to the meniscus, though 61% of people aged 50 to 90 with a meniscus tear on a scan had no knee pain in the month before. A tender bump below the kneecap in a young teenager is Osgood-Schlatter disease, which affected about 1 in 10 adolescents aged 12 to 15 in one population.

Sources 28, 29, 30, 31, 32, 33, 34, 35, 36, 37

When a scan or test helpsStrong evidence

Not usually: kneecap, IT band and tendon pain are diagnosed from the story and an examination, and scans of knees that don't hurt often show changes, with signs of arthritis on magnetic resonance imaging (MRI) in 4% to 14% of adults under 40 and 19% to 43% from 40. A report has to match your symptoms before it explains them.

After a fall from the belt, the Ottawa knee rule decides who needs an X-ray: it picked up all 63 fractures in 1,096 adults with knee injuries, and 98.5% across 6 studies of 4,249 adults. A possible stress fracture, a knee that locks or won't straighten, or catching at the joint line that hasn't settled after about 3 months of non-surgical care are when a clinician may arrange an MRI scan.

Sources 38, 36, 35, 39, 40

How it usually goesEmerging evidence

No study follows knee pain from the treadmill. NHS inform says many new knee problems begin to settle within 6 weeks, and an NHS leaflet says kneecap pain can take 3 to 6 months to improve, though it is hard to predict. In a kneecap pain trial in 40 young women, 85% were doing sport without pain at 12 months. But in a 5 to 8 year follow-up in which 60 of 310 people replied, 57% reported an unfavourable recovery, and the consensus calls kneecap pain a condition that can persist for years. In IT band syndrome, 22 of 24 runners were back running pain-free after 6 weeks of hip strengthening, with no comparison group.

No one can promise a date, which is why the plan keeps strength work going after your sessions come back.

Sources 41, 42, 43, 44, 32, 45

Changing the treadmill: lab readings, never tested against painEmerging evidence

No trial has tested a change to treadmill settings for knee pain. The settings on this page are our reading of lab studies in healthy people: flat or gentle slopes rather than declines (decline running raised kneecap stress and the thigh's braking work), a slower speed without adding distance (slower running cut the load per step but not per kilometre), and shorter, quicker steps when running (shorter steps cut kneecap stress by 22% a step, and 10% more steps a minute cut the force behind the kneecap by 14%). A slow walk up a gentle slope eased the main knee joint against brisk flat walking, but the one study of incline running and the kneecap was small and read only as a summary.

Sources 9, 11, 16, 17, 24, 27, 18

A quicker step: worth trying, not a cureMixed evidence

A 5 to 10% higher step rate lowered knee loads in lab studies of healthy runners, and in 12 runners with kneecap pain one session of running to a metronome at a 10% higher step rate was followed by less pain and more running at 3 months, with no comparison group.

But across 37 studies, the evidence on pain and injury was too thin to be sure, and a consensus panel was uncertain about gait retraining as a treatment. A quicker step can feel awkward at first. It is a load tweak to try, not a cure.

Sources 26, 27, 46, 47, 48

Strength work is the treatment for kneecap painStrong evidence

In a Cochrane review of 31 trials (1,690 people), exercise reduced kneecap pain during activity by about 1.5 points out of 10 against no exercise in the short term (5 trials, 375 people), on very low quality evidence. Two international consensus statements recommend exercise, with hip and knee exercises together in preference to knee exercises alone; foot orthoses can ease pain in the short term, and the panels were uncertain about taping and bracing.

An NHS leaflet for kneecap pain gives 3 sets of 8 to 12 of exercises including bridging, clams and mini squats to less than 30 degrees, and names the cross trainer, swimming and the stationary bike as gentler options than high-impact exercise. None of these trials was on a treadmill: the evidence is borrowed from kneecap pain in general.

Sources 49, 50, 48, 51

Hip strength for IT band painEmerging evidence

Across 13 small studies of 201 runners, mostly hip strengthening, pain fell by 27% to 100% over 2 to 8 weeks; the studies were too different to pool. In 24 distance runners with IT band syndrome, the sore side's hip muscles were weaker, and a return to running went with regaining that strength over 6 weeks, without a comparison group. An earlier review called the research on IT band treatment poor and conflicting, so this is a sensible start rather than a proven fix.

Sources 52, 45, 30

Walking is good for arthritic kneesEmerging evidence

In 1,212 people aged 50 or over with knee arthritis, those who walked for exercise were less likely to develop new frequent knee pain over 4 years (odds ratio 0.6), an association that can't show cause. NICE recommends exercise, including local muscle strengthening and general aerobic fitness, for everyone with osteoarthritis, and says joint pain may rise when you start. After a single run, the changes in knee cartilage were small and appeared short-lived in a review. In 30 people with knee arthritis, uphill treadmill walking added to physiotherapy improved stride and walking speed; the summary reports no difference in pain.

For a worn meniscus tear in mid-life, 12 weeks of supervised exercise did as well as keyhole surgery at 2 years in 140 adults.

Backward walking: across 13 trials, adding it to rehabilitation reduced pain and disability in knee arthritis, though the authors say its effect on its own, over the long term and the certainty of the evidence still need study, and in one trial of 68 adults it did no better than walking forwards. Our treadmill guide leaves walking backwards on a moving belt out.

Sources 53, 35, 54, 55, 56, 57, 58

Which machine is gentlest inside the kneeEmerging evidence

In 3 people with knee replacements fitted with force sensors, a stationary bike gave low knee forces, jogging high ones, and the elliptical lower forces than jogging but not lower than treadmill walking. So if the belt keeps hurting, a bike is the gentlest swap; the elliptical takes away impact but isn't gentler inside the knee than walking. An NHS leaflet suggests the cross trainer, swimming and the stationary bike for kneecap pain.

Sources 59, 51

What to change on the treadmill today (written steps)

Our 3D figure can't show the speed, the slope, a decline setting or your stride. These steps come from lab studies in healthy people, our treadmill, incline walking and running knee guides, and NHS pages; none has been tested against knee pain. Keep what passes the pain rule.

• 1. Flatten the belt and turn off any decline: walk at 0 to 1%, as our treadmill guide suggests. Decline running raised kneecap stress and nearly doubled the thigh's braking work, though those studies used steeper slopes than most home treadmills.
• 2. Slow down, and don't add distance to make up for it: a talk-but-not-sing pace for walking; for running, slower cut the load per step but not per kilometre, so run for less time while it settles (editorial).
• 3. Shorter, quicker steps when running: try 5 to 10% more steps a minute at the same speed. Count your steps for 30 seconds and double it, then use a metronome app set about 5% higher, as our running knee guide suggests. Land with your foot under your body rather than reaching ahead (editorial).
• 4. Don't switch to landing on your forefoot to fix the knee: our running knee guide explains that it moves load to the calf and Achilles.
• 5. A slope only if it suits your knee: with an achy, arthritic or inner knee, a slower walk up 3 to 5% is a fair swap for brisk flat walking at the same effort; with pain around the kneecap, stay flat if a slope hurts, because the knee bends more uphill.
• 6. Split long sessions and change one thing at a time: time first, then speed, then slope, as our treadmill guide teaches. The NHS says do a similar run at least 3 or 4 times before you increase your pace or distance.
• 7. Hands light on the rails: a fingertip touch for balance is fine, as our treadmill guide says. Our treadmill guide also says leave walking backwards on a moving belt out.
• 8. A different machine for now: a recumbent or upright bike gave low knee forces in people with sensor knees; the elliptical is a fair low-impact swap, though it wasn't gentler inside the knee than treadmill walking. An NHS leaflet also suggests swimming.
• 9. Keep moving: an NHS leaflet says avoiding activity weakens the leg muscles, which can add to the pain.
• 10. Safety: clip on the key, start from the side rails and let the belt stop before you step off (our treadmill guide). Treadmills were linked to two-thirds of the home exercise machine injuries seen in US emergency departments.

Sources 9, 18, 16, 17, 24, 27, 25, 11, 14, 63, 59, 51, 8

Stairs, sitting, running outside and walking pads

While the kneecap is sore, lead with your good leg going upstairs and your sore leg going down, as our elliptical knee guide suggests, and stretch the leg out or get up every half hour on long sits (editorial). Arthritis UK says rest the knee but avoid long spells of not moving at all. Running outside comes back by our running knee guide, flat routes first and downhills last; at a walking-pad desk, short spells at a slow pace on the flat (editorial).

Source 64

Medicines, braces and surgery

Our running knee guide covers painkillers and advises against taking them to run through pain. A consensus panel was uncertain about kneecap braces and taping, so treat them as comfort, not a treatment. Surgery isn't a treatment for kneecap pain or IT band syndrome in the guidance we read, and for a worn meniscus tear in mid-life, exercise did as well as keyhole surgery.

Sources 48, 56