You change him and his legs will not open. You slide his arm into the sleeve and have to push, as if the elbow were locked. You pick him up and instead of curling against you he stays like a board, with his fists closed and his feet pointed. Then you search online for the word hypertonia and in three clicks you are reading an article on cerebral palsy.
Let us start with the answer. A normal newborn is very tense, and that is right: flexor tone is the last thing built before birth, and around six weeks it starts to loosen by itself. What distinguishes normality from true hypertonia is not how much stiffness you see, but whether it changes. A healthy baby is firm when he cries and soft when he sleeps. Stiffness that is always there, the same in every situation, is another story, and in that case you do not wait: you go to the paediatrician, not to a physiotherapist and not to a forum, because it is the situation in which getting there early changes things. You will not find a diagnosis here: you will find what studies have actually measured and the precise moment to pick up the phone.
Tone is not strength
Tone is the resistance the muscles offer when it is someone else moving the baby's arm, while he is passive. Strength is how much push he can produce himself. A baby can be stiff and weak, or soft and very strong: two independent axes, which whoever assesses him looks at separately, because together they say where the problem comes from. The exact opposite of this article is in Hypotonia in newborns: when to worry (and when not to).
And inside the word “hypertonia” there are different things. An interdisciplinary consensus gathered at the National Institutes of Health in April 2001 set the definitions still in use: spasticity is hypertonia in which resistance increases with speed at which you stretch the muscle, or catches beyond a threshold of speed or angle; dystonia is not resistance but involuntary contractions that twist the body into abnormal postures; true rigidity resists even at very low speeds and does not depend on speed.
This is not pedantry: they are three different mechanisms, and none of the three can be recognised from a video shot on the changing mat.
The newborn is flexed by construction (and tone rises from below)
The newborn's flexor tone is not a contracture: it is the last achievement of pregnancy, and it is built in a precise order. In a cohort of 42 babies born under 1,300 grams, examined every week from 25 weeks of postmenstrual age until term, and none of whom later developed cerebral palsy, researchers timed the onset of tone:
| Where | When flexor tone appears |
|---|---|
| Legs (popliteal angle, heel-to-ear) | first detectable at 29 weeks |
| Lower limbs, full picture | 33-35 weeks |
| Upper limbs | 35-37 weeks |
| Shoulder | 37-38 weeks |
| Trunk (ventral suspension) | 36-40 weeks |
Maturation is caudocephalic, from the legs to the arms, and centripetal, from the periphery to the centre. The trunk arrives last, and more than half of those babies, assessed at term, still showed a lag of the head when pulled up by the hands in a sitting position.
Translated for you: a full-term newborn curled up, with elbows bent, fists closed and legs resisting, is displaying the last part of a job that has been done well. With one caveat about the design: that cohort was made of very premature babies followed up to term. It describes the sequence in which tone is built, not the norm of a nursery.
Around six weeks, it loosens
In a study of 76 neurological examinations carried out between 3 and 10 weeks of life in term, low-risk babies, the scores changed in almost all the items, with a clear shift right around six weeks. Two things happened together: flexor tone in the limbs decreased and active tone in the neck increased.
It is the opposite of what a parent fears: the baby does not “lose tone”, he shifts it. He lets go at the periphery and begins to hold the centre, that is, the head. If at two months he is less curled up than he was at birth, you have just seen his nervous system at work.
The hand follows the same logic: palmar grasp, the fist that closes on its own around whatever touches the palm, is a primitive reflex destined to fade in the first months of life. The data we have come from a small cohort of healthy preterm babies followed for a year, in which primitive reflexes disappeared later than expected even when corrected age was used: at 6 months of corrected age only 2.7% still showed palmar grasp. Up to that point clenched fists are normal repertoire, which is why later on the threshold we give for closed hands is six months. The ages for all the others are in Primitive reflexes in newborns: what they are and when they disappear.
The real boundary: not how much stiffness, but whether it changes
What distinguishes normal variability from true hypertonia is not intensity: it is variability.
A healthy baby does different things at different moments: he stiffens when he cries and loosens when he sleeps, pushes your arm away during a feed and ten minutes later goes limp against you. The repertoire is wide.
Research on spontaneous movements says the same. The pattern that specifically predicts cerebral palsy is the cramped-synchronised one: movements in which the muscles of the limbs and trunk contract and relax almost at the same time, stiff, poor, always the same. In a study of 84 preterm babies with abnormalities on cranial ultrasound, the 33 who showed it consistently or predominantly all went on to develop cerebral palsy, and the earlier it appeared, the worse the outcome. But the other half of the table must be read too: the 13 with consistently normal movements and the 30 with a poor repertoire had a normal outcome in 84% of cases, and in the remaining 16% a mild form.
The deep dive on the movement that frightens most is in Newborn arching backwards: what it means.
On asymmetry the truth is more nuanced than they tell you
Here we are honest even when it is not convenient, because “asymmetry is a warning sign” is one of the most repeated phrases online. A systematic review of 47 studies on the predictive value of spontaneous movements measured them all. Fidgety movements between 10 and 20 weeks of corrected age have the highest sensitivity, 97% (95% CI: 93-99), with specificity 89% (83-93). The cramped-synchronised pattern in the writhing period has the best specificity, 97% (95% CI: 74-100), with a sensitivity of 70% (54-82). And then the uncomfortable fact: the presence of asymmetric postures detected by standardised neurological examinations had weak associations with later cerebral palsy.
Does that mean ignoring it? No, for a different reason: a side that is stiffer than the other often points to a local problem, such as a torticollis, a hip, or a peripheral injury. It should be looked at anyway, and quickly. But it is not the cerebral palsy warning sign that the internet describes, and that distinction spares you weeks of terror.
And there is one last caution, a decisive one. Another systematic review, 19 studies selected from 351, found for general movement observation a sensitivity of 98% and a specificity of 91%, excellent figures, but stated plainly that all the available studies concerned high-risk populations, preterm and low-birthweight babies, and that accuracy in low-risk babies remains unclear. Those percentages are not the probability that applies to your full-term baby who looks tense to you today: they are the precision of a tool in a neonatal follow-up clinic.
Hypertonia that goes away on its own: it exists, and it is little discussed
A Swiss case series described eight otherwise normal babies with dystonic postures and movements appearing in the first months, progressing for a period and then gradually disappearing, with follow-up from 3 months to 5 years. The authors write that at onset the distinction from cerebral palsy can be difficult, and that this mechanism probably explains some cases labelled as “transient cerebral palsy”. It should be taken for what it is, eight babies, a case series, the lowest rung of the evidence, but it shows that the phenomenon exists.
And here is the counterweight, because “transient” does not mean “irrelevant”. In a cohort of 103 very preterm babies, transient tone abnormalities at 6 months of corrected age (at least two criteria among postural abnormalities, tone abnormalities and hyperreflexia) affected 29.1% of the babies and were associated with altered diffusion values in the white matter on term MRI and with lower motor scores at 18 months, while in the same study there were no differences in cognitive scores. That 29.1% concerns very preterm babies, not the general population.
The takeaway: high tone in the first months can regress. But it is not decided from home by waiting.
⚠️ When to see the paediatrician, without waiting
Just one of these points is enough to book an appointment, and not for the next routine health check:
- constant stiffness: it is always there, in every position and at every moment, and does not loosen even in sleep;
- persistent asymmetry: one side of the body constantly stiffer than the other, or a limb clearly used less;
- legs that cross like scissors when you lift him under the arms, or that go rigidly onto tiptoe when you put him down;
- hands still clenched in fists with the thumb trapped inside after 6 months, without him bringing them to the centre or grasping objects;
- dressing him becomes harder over time: an arm that does not go into the sleeve, legs that will not open for a nappy;
- difficulty sucking, swallowing or feeding, or inconsolable crying that goes with the stiffness;
- delay in motor milestones relative to the real windows, which are wider than you hear said;
- regression: he loses something he already did. This is never observed by waiting, at any age, and it goes to the paediatrician the same day.
A candid note on the list: it is a clinical description built on what standardised examinations look at. The Hammersmith infant neurological examination, for example, has 37 items divided into three sections, 26 of which cover cranial nerve function, posture, movements, tone and reflexes, calibrated on a cohort of 12- and 18-month-old children with no known perinatal risk factors. But it is not a validated score for this purpose, and none of these points on its own makes a diagnosis. It serves to decide one thing only: whether to pick up the phone.
Why the rush really matters
Until a few years ago a diagnosis of cerebral palsy was made between 12 and 24 months. Today a systematic review of six reviews and two guidelines has established that, combining standardised tools, it can be made before 6 months of corrected age: under 5 corrected months the most predictive are the term MRI (sensitivity 86-89%), Prechtl's general movements (98%) and the Hammersmith examination (90%).
The problem is that in practice this does not happen: diagnosis still arrives between 12 and 24 months in high-income countries, that is, after the best window for interventions that exploit the plasticity of the developing brain.
The international recommendations on early intervention from 0 to 2 years, built on 16 systematic reviews and 27 randomised trials, put at the top of the principles of good practice immediate referral as soon as a high-risk diagnosis is made. It is also worth noting what they do not say: of the 28 recommendations across nine domains, only one concerns the management of tone. Most of the work lies elsewhere.
Two numbers for proportion. An analysis of 41 regions in 27 countries estimates the birth prevalence of cerebral palsy in high-income countries at 1.6 per 1,000 live births, and falling; an earlier meta-analysis of 49 studies put it at 2.11 per 1,000, with a risk that is far from uniform: 111.80 per 1,000 among babies born before 28 weeks (95% CI: 69.5-179.8, a very wide interval: few children in that group). And among children with a diagnosis, again in high-income countries, 2 in 3 will walk, 3 in 4 will talk and 1 in 2 will have normal intelligence.
And the older child who walks on tiptoe?
High tone has a second age at which parents notice it: when the child walks on tiptoe. Here the numbers are very reassuring. In 1,436 Swedish five-and-a-half-year-olds, 2.1% still walked on their toes and 2.8% had stopped: 4.9% in all. The same authors then went back to the cohort, restricted to the 1,401 healthy children: of the 63 who had been toe-walkers, they followed up to age ten the 26 who at five and a half were still walking on tiptoe. At ten, 50 of the 63, 79%, had stopped on their own, without treatment and without Achilles tendon shortening.
The more useful finding, however, is another: among children with a neuropsychiatric diagnosis or a developmental delay the proportion was 7 out of 17, 41.2%; and among the children who were still walking on tiptoe at ten, four had a neurodevelopmental comorbidity. The authors conclude that among those who keep walking on tiptoe these comorbidities are frequent, not that they affect everyone. Walking on tiptoe on its own, in a child who is otherwise doing well, almost always resolves. In a child who has other things going on too, it is a reason to look further. The picture is in Child who walks on tiptoe: when to really worry.
What we can do, and what we cannot
Let us say it plainly.
We do not treat complex neurological and neuromuscular conditions online. If behind the stiffness there is cerebral palsy or another condition of the nervous system, that pathway requires in-person assessment, a child neuropsychiatrist and a specialised centre: it is not us, and if we realise it, we will tell you on the first day. And if any of the signs above appear, the right place is the paediatrician, straight away, not a physiotherapy visit, not even ours.
What we can do is smaller, and often it is exactly what you need: watch your baby move and tell you whether what you see is within normal range. In most of the cases where a parent writes “my son is stiff”, behind it there is physiological flexor tone, a baby who arches because extension is the only movement he has learned, or a posture being organised badly because he spends too many hours lying the same way. These are easy to see by video call, because the baby is at home, on his own mat, and is not crying because of the clinic.
With Ri-Hub Junior you start with a free twenty-minute introductory consultation with a physiotherapist on the Italian professional register who works with children and teenagers: you show us a nappy change and two minutes of play on the floor, and we tell you honestly whether the right pathway is us, the paediatrician or a specialised centre. Up to age 10 the session lasts one hour and costs 59,90 €; from 11 years up the standard session at 29,90 € applies.
In three lines
A tense, curled-up newborn with fists closed is a normal newborn: flexor tone is the last thing built before birth, and around six weeks it loosens by itself. What matters is not how much stiffness you see, but whether it changes. If instead it is always there, asymmetric, blocks everyday movements, or comes with milestones that do not arrive or that go backwards, do not watch it for another month: call the paediatrician, because the one variable you can still control is time.
If your concern is timing, that he does not sit, does not crawl, does not walk, the real windows are here: Not crawling and not walking: the real windows of motor milestones.
Sources
- Classification and definition of disorders causing hypertonia in childhood, Pediatrics, 2003
- Tone and reflex development before term, Pediatrics, 1990
- Neurological examination in healthy term infants aged 3-10 weeks, Biology of the Neonate, 2005
- Cramped synchronized general movements in preterm infants as an early marker for cerebral palsy, Archives of Pediatrics and Adolescent Medicine, 2002
- Predictive validity of spontaneous early infant movement for later cerebral palsy: a systematic review, Developmental Medicine and Child Neurology, 2018
- A systematic review of tests to predict cerebral palsy in young children, Developmental Medicine and Child Neurology, 2013
- Transient idiopathic dystonia in infancy, Neuropediatrics, 1991
- Transient tone anomalies in very preterm infants: association with term-equivalent brain MRI and neurodevelopment at 18 months, Early Human Development, 2020
- Early, Accurate Diagnosis and Early Intervention in Cerebral Palsy: Advances in Diagnosis and Treatment, JAMA Pediatrics, 2017
- Early Intervention for Children Aged 0 to 2 Years With or at High Risk of Cerebral Palsy: International Clinical Practice Guideline, JAMA Pediatrics, 2021
- Early Diagnosis and Classification of Cerebral Palsy: An Historical Perspective and Barriers to an Early Diagnosis, Journal of Clinical Medicine, 2019
- Global prevalence of cerebral palsy: A systematic analysis, Developmental Medicine and Child Neurology, 2022
- An update on the prevalence of cerebral palsy: a systematic review and meta-analysis, Developmental Medicine and Child Neurology, 2013
- Optimality score for the neurologic examination of the infant at 12 and 18 months of age, The Journal of Pediatrics, 1999
- Primitive reflex in premature healthy newborns during the first year, Arquivos de Neuro-Psiquiatria, 2005
- The prevalence and course of idiopathic toe-walking in 5-year-old children, Pediatrics, 2012
- Idiopathic Toe-Walking: Prevalence and Natural History from Birth to Ten Years of Age, The Journal of Bone and Joint Surgery, 2018
Last source check: September 2026.
