You noticed it from behind. The child runs toward the slide and the tips of their feet turn toward each other, as if trying to meet halfway. Or it is the opposite: they walk like a duck, with their feet pointing outward. And now you cannot watch them walk without watching that.
Here is the conclusion up front: in the vast majority of cases it is not a deformity, it is a normal variant of growth, and time is what fixes it. Not insoles, not corrective shoes, not exercises.
But "wait and see" is unbearable when said like that. So here is what you are waiting for, how long it lasts, and the few things that do deserve a proper visit.
The right question is not "how long", it is "from where"
A foot can point inward for three reasons, at three different heights of the leg:
- the foot itself, when the forefoot is turned inward (metatarsus adductus);
- the tibia, when the shin bone is rotated inward relative to the knee (internal tibial torsion);
- the femur, when the thigh bone is rotated forward at the neck (femoral anteversion).
Understanding which of the three is involved changes everything, because they have different ages and different times to resolve. And as a rule an X-ray is not needed: imaging serves to rule out other things, not to orient you among the variants. It is enough to watch the child walk and run barefoot, and to see where the kneecaps point. If the knees point straight ahead and only the feet point inward, the rotation is below the knee; if the kneecaps also face each other, it is above.
The general picture was laid out by a classic study of 1,000 normal lower limbs in children and adults, which set the reference values of the rotational profile, later updated by more recent case series. Its conclusion, written in 1985, is the sentence you need: outward-turned feet in infants, internal tibial torsion in the child who is starting to walk, and internal femoral torsion in the young child are the extremes of a normal developmental pattern, and almost always they require no treatment at all.
The foot: metatarsus adductus
It is the form that is seen first of all, often already at the maternity ward: the outer edge of the foot, instead of being straight, is curved like a comma, and the forefoot is turned inward while the heel is in place.
It comes from the position in the womb and it is the site with the most reassuring numbers. A prospective study followed 130 feet in 83 children for an average of seven years, without treating them: at the last check, 86% were normal or only slightly deviated, and all were fully mobile; another 10% were still moderately deviated but without any symptoms; only 4% remained deviated and stiff. These resistant cases, the authors say, could not be identified before the age of three.
And the stretching exercises that are sometimes recommended to parents? They have been tested. In a randomised study of 94 newborn feet, assigned by lottery either to a stretching programme carried out by the parents or to observation alone, there was no significant difference in the success rate (p = 0.19). In fact, in moderate to severe feet the treated group tended to do worse, although the difference was not significant. The authors conclude to observe until 4-6 months and reserve corrective casting for feet that resist.
A useful detail: in a prospective study lasting 11 years, among 25 newborns with metatarsus adductus the risk of hip dysplasia or instability on ultrasound was about 1 in 25. A small sample, a figure to take with caution, but it is the reason the paediatrician looks at the hips anyway.
The tibia: internal tibial torsion
It is the most common cause in the child who is starting to walk.
The explanation is mechanical and surprisingly simple: the fetal position shapes the bones. In the womb the femur is rotated outward and the tibia inward; these effects resolve spontaneously during childhood, and only afterwards do the individual differences determined by genetics emerge. How much does this site weigh? In a historical study, about 80% of children under 18 months who walk with their toes turned in have internal tibial torsion: it is by far the most frequent explanation at that age, and it is also the one that resolves first, spontaneously, in the early years of life. In the same study, however, already after two years the most frequent cause of toes turned in became something else: femoral anteversion, in about 70% of cases.
If you are still wondering whether milestones arrive at the right times, the real windows, which are very wide, are in Not crawling and not walking: the real windows for motor milestones.
The femur: femoral anteversion
This is the story of ages 4 to 6, and it is the one that usually leads parents to search online.
The picture is recognisable: the kneecaps face each other, the child runs in a somewhat jumbled way with the feet swinging outward, sits happily in a W position, and rotates the hip inward a great deal when laid face down.
The numbers come from studies that followed with CT scans children referred for a visit precisely because of toes turned in, without treating them. In 200 limbs of 100 children between 4 and 10 years old, the angle of anteversion was highest precisely in the 4-6 age group, and in that group it decreased significantly at the follow-up check. In a second study of 126 limbs of 63 children followed for about three and a half years, the average angle went from 41.4° to 33.3°: a clear reduction, more marked in the younger children and in those who started from higher values.
Here is the uncomfortable part, which almost nobody writes: in that study most of the children, although improving, remained in a band of moderate to severe anteversion. The angle decreases, almost always enough that nobody notices any more, but not everyone returns to the average. And in 501 Danish children between 6 and 17 years old the decrease with age is there and statistically significant, but not clinically relevant: after age seven the values remain essentially stable.
In plain terms: most of the work is done by growth before school age. If at nine the rotation is still clearly visible, waiting another five years is no longer a strategy: it is the moment to have it looked at, not to operate, but to know where it stands.
And if instead the child walks with toes turned out (the "pigeon toes")?
"Pigeon toes" is how this is almost always called at home; the term you will find on the report is external rotation or externally rotated foot. They mean the same thing: the tip of the foot points outward instead of straight ahead.
The same logic as for toes turned in applies, three sites, different ages. In the first months the most frequent cause is not in the bone but in the soft tissues: a contracture of the hip in external rotation, a leftover of the fetal position. It lessens with growth, to the point that it was present in fewer than 5% of children over 18 months. And for as long as it lasts, that contracture hides any femoral anteversion that would only show up later, which is why it happens that a child has toes turned out at one year and toes turned in at three. It has not got worse: it has been discovered.
In older children, outward-turned feet come from an external twist of the tibia, much more rarely from a backward twist of the femur (none clinically evident was found in the study cited), or from a flexible flat foot that rotates the point of contact with the ground outward: on this last point the discussion is broader and you will find it in Flat foot in children: when something really needs to be done.
One exception to keep in mind in the older child: if the outward-rotated foot appears between ages 8 and 15 and comes with limping and pain in the hip, groin, thigh or knee, it is not a growth variant and it should not be waited out. It is the case you will find further on, among the signs that warrant an orthopaedist.
The calendar, in a table
| Age | Most likely site | What usually happens |
|---|---|---|
| 0-12 months | foot (metatarsus adductus) or hip | the foot almost always straightens by itself; the hip contracture loosens |
| 1-2 years | tibia (inward) | this is the stage of the first steps: it decreases as the child walks and runs |
| 2-3 years | tibia, but the femur starts to prevail | the hip contracture has loosened and femoral anteversion becomes visible |
| 4-6 years | femur (inward) | this is the period when it is most noticeable, then the angle decreases |
| 7 years and up | whatever is left | spontaneous changes slow down a great deal |
Sitting in a W: what the evidence really says
Not much. And that is already information in itself.
The advice to correct the W position circulates everywhere, almost always without a source. When it was put to the test, the result was this. A prospective study of 104 children and 196 hips, all with pelvic X-rays, compared those who sat in a W position (46% reported doing so or having done so) with those who did not. Hip dysplasia was present in 9% of the first group and 10% of the second: no difference, not even when distinguishing between those who sat that way by choice, those who did it without preferring it, and those who did it only in the past. A 2021 review of myths in paediatric orthopaedics goes further: there is no scientific evidence that sitting in a W position leads to contractures, hip dislocations or functional deficits.
Two points of honesty. The first: that case series was made up of children followed in a tertiary paediatric hospital, not taken at random from the street. The second: nobody has shown that sitting in a W position causes femoral anteversion, and the most credible reading is the opposite: children with anteversion sit in a W position because for them that is the comfortable posture, as the 2017 family doctor review also writes. If that is so, stopping it does not change the bone, and it costs you ten corrections a day. If it still bothers you, offer other positions when you think of it, without making it a battle.
Insoles, braces, corrective shoes
Here the literature is old, clear and not very commercial. Already in 1976 a study measured the immediate effect of shoe modifications on the angle of the gait: wedges had no clinically significant effect, and the torsion heel, the most effective of the group, corrected only a quarter of the apparent rotation during walking.
A systematic review from 2012 looked for everything that existed on non-surgical treatment of toes turned in and found only five studies, all of level IV and of variable quality: the deviation plates, the standardised shoes and the insoles with a plate extension improved the appearance of the gait in a statistically significant way, while wedges, torsion heels and splints reduced nothing. The authors themselves conclude that the evidence is limited and should be taken with caution. The review for American family doctors, updated in 2025, says that the evidence supporting insoles or braces is minimal; the review of myths adds that physiotherapy, insoles and braces have not been shown to be effective for these variants.
There is one study that seems to say the opposite, and it is fair to mention it: in 51 children with toes turned in who fell often, a medial-wedge insole was judged effective in 80.8% of cases. But that judgement was a score reported by children and parents, the study was retrospective, and the comparison group had only seven healthy children: it is impossible to separate the effect of the insole from that of growth.
The honest summary: an insole can change the appearance of the gait while the child wears it. It does not rotate the bone, and it has never been shown to shorten the time it takes.
"But they trip and fall all the time"
This is the reason many parents ask for help, and it is legitimate: it is not cosmetic, it is function. A 2024 study of 30 children with toes turned in due to tibial torsion found that the angle of the foot goes along with a different load on the forefoot and the heel, and, more interesting, that the children compensate on their own, keeping a symmetrical gait. That ability to self-regulate, the authors write, starts to give way when the rotation becomes more marked. Be careful not to ask more of this study than it says: it looked at pressures and gait symmetry, and it did not measure falls. It is still a reason to take "they fall all the time" seriously rather than dismissing it.
Be careful, though: falling often is not a specific symptom of toes turned in. Coordination, balance, vision, strength and simply age count at least as much. The full picture is in Child who falls often: when it is normal and when it is not.
And as an adult? Arthritis, pain, sport
The underlying fear, the one nobody says out loud, is: will it wreck their hips or knees? The best answer we have comes from an osteological collection of 1,158 adult tibias and femurs, in which tibial torsion and femoral anteversion were measured directly on the bone and compared with the degree of arthritis in the hip and knee. The result: neither was an independent predictor of arthritis, and the authors explicitly write that parents of asymptomatic children can be reassured.
A more recent study adds a nuance. They recalled 58 adults never treated who had femoral anteversion of at least 30° as children, mean age 46 years, and compared them with 24 healthy controls: they reported more hip pain and more limitations in sport and quality of life, while in the knee the only difference concerned symptoms. But only one patient out of 58 had hip arthritis, and the authors conclude that the limitations were small to moderate, not enough to change the very restrictive current indications for surgery in children.
And there is a finding that flips the perspective: comparing 50 high-school sprinters with 50 peers, the sprinters had a significantly lower thigh-foot angle, more inclined inward, and more often ran with their toes turned in. That does not mean toes turned in make you fast: it means that, within the normal range, this rotation is not a defect to correct in order to do sport.
The few situations that warrant an orthopaedist
So far the message is reassuring, and it should be. But there are signs that deserve a real assessment, few and recognisable:
- Clear asymmetry between the two sides. This is the most important one. In a series of 1,319 healthy children there was no significant difference between right and left either for tibial torsion or for femoral anteversion: a normal variant is not expected to show a gap between the two sides.
- Pain. Physiological rotational variants do not hurt.
- A boy or girl between 8 and 15 years old who limps and has pain in the hip, groin, thigh or knee, perhaps walking with a foot rotated outward. Here you do not wait: at that age slipped capital femoral epiphysis must be ruled out, the most common hip disorder of adolescence and one of the diagnoses most often missed in children. It is confirmed with an X-ray of both hips. The pain is often poorly localised and sometimes is felt only in the knee: that is why it gets missed.
- Worsening over time instead of the slow improvement expected for the age.
- Always trips in the same way, or falls out of proportion compared with peers.
- A stiff foot, which does not straighten even when the fingers guide it.
- Delay or loss of motor skills, muscle tone that is too stiff or too floppy, a complicated perinatal history.
- Measurements outside the norm for age: the formal indication is to refer the child when the values exceed two standard deviations from the norm, when there is pain, or when function is compromised.
On surgery, to close the loop: it is rare, reserved for children over eight years old with severe deformities that cause real dysfunction. It is not around the corner, and using it as a scare tactic to sell an insole is wrong.
One last thing: in a study of 48 parents who came for a visit about their child's toes turned in, after the consultation with the specialist anxiety went down and understanding went up markedly, and those who remained anxious were precisely those who said they wanted to look for other treatments. Understanding what is going on is, measurably, the therapeutic part.
What a physiotherapist can do (and what they cannot)
Let us say it clearly, even though it goes against our own interest: for most of these children no therapy is needed. No exercise has ever been shown to rotate a femur or a tibia, and anyone who promises you otherwise is selling something else. What is almost always needed is someone who watches the child walk and run and tells you where the rotation comes from, whether it is within the normal range for their age and whether anything is asymmetrical: turning "who knows, let us hope it passes" into "I know what I am looking at and what to check in six months".
This works very well by video call: it is enough to see them move barefoot in the hallway at home, and to look at how they sit down and how they lie down. With Ri-Hub Junior you start with a free 20-minute first consultation with a physiotherapist on the Italian professional register who works with children and teenagers: you film them walking, we look at it together and tell you what we are seeing. Up to age 10 sessions last 50 minutes and cost 59,90 €; from age 11 upward the standard session of 29,90 € applies.
And where a paediatric orthopaedist or a neurologist is needed, we tell you straight away: complex neurological and neuromuscular conditions are not treated online, they require in-person assessments and specialised centres.
If what strikes you instead is that they walk on their toes, the picture is different, and the numbers are in Child who walks on their toes: when to really worry.
Wear on the sole is used as a do-it-yourself diagnosis: what can really be deduced from it is in Shoes worn down on one side.
Sources
- Lower-extremity rotational problems in children. Normal values to guide management: Journal of Bone and Joint Surgery (Am), 1985
- Rotational profile of the lower limb in 1319 healthy children: Gait & Posture, 2008
- Lower Extremity Abnormalities in Children: American Family Physician, 2025
- Lower Extremity Abnormalities in Children: American Family Physician, 2017
- Pediatric orthopedic mythbusters: the truth about flexible flatfeet, tibial and femoral torsion, W-sitting, and idiopathic toe-walking: Current Opinion in Pediatrics, 2021
- Slipped Capital Femoral Epiphysis: Diagnosis and Management: American Family Physician, 2017
- Minimal Out-Toeing and Good Hip Scores of Severe SCFE Patients Treated With Modified Dunn Procedure and Contralateral Prophylactic Pinning at Minimal 5-year Follow up: Journal of Pediatric Orthopaedics, 2022
- Hip Dysplasia Is Not More Common in W-Sitters: Clinical Pediatrics, 2020
- The natural history of hooked forefoot: Journal of Bone and Joint Surgery (Br), 1978
- Does the parental stretching programs improve metatarsus adductus in newborns?: Journal of Orthopaedic Surgery, 2017
- Neonatal foot deformities and their relationship to developmental dysplasia of the hip: an 11-year prospective, longitudinal observational study: Journal of Bone and Joint Surgery (Br), 2009
- External rotation contracture of the extended hip. A common phenomenon of infancy obscuring femoral neck anteversion and the most frequent cause of out-toeing gait in children: Clinical Orthopaedics and Related Research, 1975
- Change of Femoral Anteversion Angle in Children With Intoeing Gait Measured by Three-Dimensional Computed Tomography Reconstruction: One-Year Follow-Up Study: Annals of Rehabilitation Medicine, 2018
- Change of Femoral Anteversion Angle in Children With Intoeing Gait Measured by Three-Dimensional Computed Tomography Reconstruction: 3-Year Follow-Up Study: Annals of Rehabilitation Medicine, 2023
- Angular alignment, rotational profile, and joint range of motion in the lower limb of typically developing children from 7-16 years of age: Acta Orthopaedica, 2025
- The effectiveness of shoe modifications for intoeing: Orthopedic Clinics of North America, 1976
- Non-surgical management of a pediatric "intoed" gait pattern: a systematic review of the current best evidence: Journal of Multidisciplinary Healthcare, 2012
- Effectiveness of medial-wedge insoles for children with intoeing gait who fall easily: Turkish Journal of Physical Medicine and Rehabilitation, 2019
- Impact of foot progression angle on spatiotemporal and plantar loading pattern in intoeing children during gait: Scientific Reports, 2024
- Femoral Version and Tibial Torsion are Not Associated With Hip or Knee Arthritis in a Large Osteological Collection: Journal of Pediatric Orthopaedics, 2017
- Long-Term Outcome of Idiopathic Increased Femoral Anteversion in 58 Untreated Individuals at a Mean Age of 46.2 Years: Journal of Pediatric Orthopaedics, 2026
- Sprinting and intoeing: Journal of Pediatric Orthopaedics, 1996
- Parental understanding of intoeing gait: A preliminary study: The Foot, 2019
Last source check: September 2026.
