The first time someone puts them in your arms, you are told a single sentence: "mind the head". Then they leave you there, with three and a half kilos of human being who holds nothing up. And for weeks, every time you lift them, there is half a second of fear.
Here is the substance straight away. Supporting the head is not a technique to learn: it is something you gradually stop doing when they start doing it on their own, usually between three and four months. What does need learning concerns two things nobody talks about: the side you alternate and your back.
Why the head needs supporting (and for how long)
The reason is mechanical. At birth the head is large compared with the rest of the body and the paraspinal muscles of the neck are still weak: that combination produces head lag, the head staying behind the trunk when the baby is pulled up to sitting. In the first weeks, that is the rule.
Postural control then arrives, and it arrives from the top down: first the neck, then the trunk. The clinical reference is that head lag should be absent between 3 and 4 months. From that moment on, when you pick them up, they stiffen their neck and trunk and take part in the lift: it is why, at some point, you notice that the feeling of dead weight from the first month is no longer there. Nobody tells you, and it happens on an ordinary week.
One clarification that will take a weight off you: this is maturation, not training. You do not make them stronger by holding them in a certain way. What you really influence is how symmetrical the way you handle them is ,and on that there is even a randomised study.
The holds, one by one
There is no single right hold: there are holds that serve different purposes, and the craft lies in changing them often.
The cradle hold. The most natural: the head in the crook of your elbow, the forearm along the back, your other hand under the bottom. This is the one in which you look at each other's faces. It has one feature that matters, however: the crook of the elbow keeps the head slightly turned to one side, always the same one until you switch arms. It is the hold in which alternation matters most.
The shoulder hold. Their chest against yours, the head poking up over your shoulder, one hand behind neck and head, the other under the bottom. It is the position for burping, for crying and for the skin-to-skin contact of the first days. It is also the easiest to keep for a long time, because its weight rests on your trunk rather than your arms.
Tummy down along the forearm. The body along your forearm, the head near the elbow or in your hand, the legs astride. It serves for short trips around the house and gives a few seconds of prone position to a baby who would refuse it on the floor. It is sold as an "anti-colic" hold: no study shows it cures colic, but if it calms your baby, use it because it calms them.
Sideways on the hip. The legs astride your hip, one hand around the back. It arrives when the head is stable and the trunk holds: before 4 to 5 months it makes no sense, after that it becomes the hold for the next two years.
Facing the world. For short periods, from around 4 to 6 months: without your body in front to support them, it is a hold for minutes at a time.
How you lift them and how you lay them down
To pick them up from the cot: first turn them onto their side, one hand under head and neck and the other under the bottom, bring them close to your body and only then lift. Pulling them up by the back with straight arms is the manoeuvre that triggers the Moro reflex ,and the protest.
To lay them down when they have fallen asleep there is a measured detail that is worth its weight in gold. In a study in Current Biology the researchers filmed that moment second by second: the moment that most consistently wakes them is the start of the detachment from the mother's body ,not the end, which instead calms them once the support works; and the likelihood that sleep holds depends on how much they had already slept before. Put simply: do not lay them down as soon as they fall asleep. Keep them a few minutes longer, then detach slowly, legs first and head last.
Month by month: what really changes
| Age | What they do | What you do |
|---|---|---|
| 0 to 6 weeks | Marked head lag, no head control | One hand always behind head and neck. Cradle, shoulder, skin to skin |
| 6 weeks to 3 months | Lifts and holds the head for a few seconds, still wobbly | You keep supporting, but allow more freedom. This is the period when alternating sides matters most |
| 3 to 4 months | Head lag disappears, takes part when you pick them up | The hand behind the head becomes "on standby". Shoulder hold a little higher |
| 4 to 6 months | Steady head, trunk beginning to hold | The hip hold and a few minutes facing the world arrive. You free up a hand |
| 6+ months | Grabs, leans, twists to look | The problem is no longer their neck: it is your kilos to move |
The column nobody writes is the third: what changes is not only them, you change too. And your load, unlike their head control, does not improve by itself.
The side you alternate: the part that matters most
In the most recent meta-analysis of risk factors for plagiocephaly ,19 studies, 14,808 babies ,a preference for head position was among the strongest modifiable factors: odds ratio 4.75 (95% CI: 3.36 to 6.73), alongside little tummy time (3.51), bottle-only feeding (4.65) and sleep position (3.12).
But the data you really need comes from a Dutch prospective cohort of 380 healthy newborns. Among the factors associated with flattening of the skull at 7 weeks there were not only characteristics of the baby: there were care habits. Keeping the head always on the same side on the changing mat. Positioning them always on the same side during the bottle feed. Sleep positional preference. Tummy time less than three times a day. Gestures you make thirty times a day without thinking, and that add up.
In practice:
- switch arms at every feed and every long hold, not "when you get tired" ,because you always get tired on the same side;
- on the changing mat alternate the direction you lay them in: once with the head to the right, once to the left;
- alternate the side you pick them up from and the side from which you talk to them;
- if they always look at the same thing (the window, the door, you), move them, not the room.
Here the limit needs stating. No study has randomised "switch arms" against "always use the same arm": that comparison does not exist. What has been tested is the complete package, and it worked. In a Finnish randomised trial, the babies whose families received ,besides standard safe sleep instructions ,detailed instructions on environment, positioning and how to handle the baby, had at 3 months a prevalence of plagiocephaly of 11% against 31% (15% against 33% in 3D), with milder asymmetries and better neck mobility. Alternating sides is one piece of that package, not an isolated remedy.
And there is a threshold beyond which it is no longer a question of habits: if they turn their head always and only to one side and you cannot get them to turn it to the other, or if they keep it tilted towards one shoulder even when held, you are not looking at a preference. The reference guidelines on congenital torticollis are explicit already in the 2018 edition: these babies should be referred to the physiotherapist as soon as postural asymmetries are identified, and the 2024 update keeps referral among its recommendations. We discuss it in Congenital torticollis in newborns: why timing matters more than anything. If instead what you see is the head flattening, the real data, helmet included, are in Flat head in newborns: what the studies really say.
When they cry: does holding them actually work?
Yes, but not in the way you were told.
In the historic 1986 trial, on 99 mother-baby pairs, half were asked to carry the baby also outside feeds and crying: at the peak of crying, at 6 weeks, those babies cried 43% less (1.23 against 2.16 hours a day) and 51% less in the evening.
Now the reverse. The same researchers tried again with 66 babies with colic: the mothers in the intervention group ended up carrying their babies 2.2 hours a day more than the others, and the result was no difference ,3 minutes less crying a day, with an interval running from 37 minutes less to 32 more. Arms regulate physiological crying, but they are not the therapy for colic: if yours cries a lot and arms are not enough, you are not holding them wrong.
There is also a detail that changes the gesture. In a study in Current Biology, babies under 6 months carried by a mother who is walking stopped moving and crying straight away, with a rapid drop in heart rate, compared with being held by a seated mother. It is not the arms: it is the movement. A follow-up study by the same group confirmed that crying eases with carrying or with a moving cradle but not with holding still, and proposed a pattern ,5 minutes walking, then 5 to 8 minutes seated with the baby on you, and only then lay them down ,that the authors themselves say needs confirming. It costs nothing: you can try it tonight.
And your back, which is the bit nobody tells you about
In the first months you will lift your baby dozens of times a day, and nobody teaches you how. A biomechanics study from 2025 measured 11 mothers lifting their own 15-week-old baby in three ways:
- from the floor: the most demanding gesture, with the greatest ranges of knee, trunk and hip movement and the greatest power generation. It requires a genuine deep squat;
- from the car seat placed on the floor: about 4 kg more and more muscle activity, but less knee and trunk movement, because the handle is higher;
- from the changing mat: the least demanding gesture for the legs ,and that is precisely why the mothers did it leaning and rotating the trunk sideways, asymmetrically.
The point to take home is this: the gesture that catches you out is not the heavy one, it is the one that feels light. The changing mat does not make you bend your legs, so you end up twisting. Thirty times a day. With a limit to state: these are 11 women and no clinical outcome was measured, so the study describes how people move, not how much harm it does.
What to change, concretely:
- face them when holding them in the cradle or at the changing mat, and turn your feet rather than your trunk;
- bring them close to your body before lifting, never with arms straight;
- from the floor, go down into a squat: feet wide, bottom back, long spine;
- carry the car seat as little as possible: four extra kilos, on one arm only and always the same one;
- alternate the side too on which you rest them, once you start carrying them on the hip.
On how common it is: in a Canadian cohort, between 15% and 21% of women reported low back pain at 1, 3 and 6 months after birth, and the only risk factor identified was having had back pain during pregnancy. The other half of the figure needs saying too: between 76% and 80% of participants had no pain. The study is small (46 to 64 women per survey), so the percentages should be taken broadly.
Then there is the wrist, the great forgotten one. De Quervain's tenosynovitis ,pain on the thumb side when you lift under the armpits ,in a South Korean population study of 1,601,501 pregnancies affected about 2.1% of women between pregnancy and the postpartum period. But the useful figure is another one: in an analysis of 63 women with postpartum De Quervain compared with 630 without the condition, the associated factors were a pregnancy beyond 40 weeks (OR 5.81), being a first-time mother (OR 2.23) and a BMI above 25 (OR 2.08), while the baby's birth weight and twin births were not associated. Put simply: it is not that your baby is too heavy. It is the repetition. If your wrist already hurts, keep it straight and lift with the forearm rather than hooking with the thumb: the practical side, together with neck and shoulders, is in Back and neck pain while breastfeeding: why it happens.
Slings and carriers, in three honest lines
A systematic review from 2026 of 62 studies on slings and baby carriers says two things together: there are documented benefits for the baby and for the carrier, and there are rare accidental deaths ,mainly from asphyxia ,and falls, which occurred with carriers that were poorly fitted or poorly positioned; the same carriers that fit badly were associated with pain for the adult wearing them.
So it is not "yes or no": it is how. Face visible, chin off the chest, the baby high enough to be able to kiss the top of their head, and the thighs supported up to the knee. On this last point there is a reason: hips in flexion and abduction favour the development of the joint, while forced immobilisation in extension and adduction is a recognised risk factor for hip dysplasia. It is the same principle behind holding them with your hand under the bottom and thighs.
When to have the baby seen
There are two levels, and it is worth keeping them separate.
Things to get checked early, without anxiety, because the earlier you look the less work is needed later:
- after 4 months the head keeps wobbling backwards when you pull them up to sitting by the hands;
- they keep their head always tilted towards one shoulder or always turned to the same side, and you cannot get them to turn it the other way;
- on their front they push with one arm only or always fall to the same side.
Things for which you call the paediatrician now, without waiting for the next health check:
- the baby feels very stiff or very floppy in your hands, as if they were slipping away;
- they have lost something they could do before: they no longer do it as before, they simply do not do it at all.
For these two groups I am not telling you to book physiotherapy: I am telling you to call the paediatrician, who must see them first.
On head lag after 4 months precision is needed, because it is what frightens most. Persistence of head lag beyond 4 months has been associated with poorer developmental outcomes, and in an exploratory study of siblings of children with autism, head lag at 6 months was associated with a diagnosis at 36 months. But that study is preliminary, on 40 babies already at high familial risk: it is not a test and it is not a screening. It says only one thing: if at 4 to 5 months the head still wobbles, get it checked rather than waiting for the next health check.
What a physiotherapist can do (and what they cannot)
A therapy is rarely needed. What is needed is for someone to watch how you pick them up, which side you lay them on at the changing mat, how they turn their head when they are in your arms and how you are holding yourself while you do it ,and then tell you two or three things to change this week. It is work that is done well by video call, because there is nothing to do to the baby: there are thirty gestures to correct in your own home. With Ri-Hub Junior you start with a free 20-minute introductory consultation with a physiotherapist on the Italian professional register who works with children and teenagers: you show us how you pick them up and how you hold them, and we tell you what to change. For children up to 10 years old the sessions last 50 minutes and cost 59,90 €.
If there is a complex neurological or neuromuscular condition behind it, the right place is not us: those pathways require in-person assessments and specialist centres, and we tell you that straight away rather than wasting your time.
If you were told the collarbone broke during birth, it heals sooner than you think, and there is a way of picking them up in the meantime: Collarbone fracture at birth.
Stopping supporting their head is not a date, it is something that happens on its own: Head control.
Arms are only half of the day: the pushchair, the bag and the car seat are the other half: Pushing the pushchair and carrying the bag without hurting yourself.
Sources
- Infant Head Lag, StatPearls, 2023
- Assessing the Evidence for Nonobstetric Risk Factors for Deformational Plagiocephaly: Systematic Review and Meta-Analysis, Interactive Journal of Medical Research, 2024
- Risk factors for deformational plagiocephaly at birth and at 7 weeks of age: a prospective cohort study, Pediatrics, 2007
- Preventing deformational plagiocephaly through parent guidance: a randomized, controlled trial, European Journal of Pediatrics, 2015
- Physical Therapy Management of Congenital Muscular Torticollis: A 2024 Evidence-Based Clinical Practice Guideline, Pediatric Physical Therapy, 2024
- Physical Therapy Management of Congenital Muscular Torticollis: A 2018 Evidence-Based Clinical Practice Guideline, Pediatric Physical Therapy, 2018
- Increased carrying reduces infant crying: a randomized controlled trial, Pediatrics, 1986
- Carrying as colic "therapy": a randomized controlled trial, Pediatrics, 1991
- Infant calming responses during maternal carrying in humans and mice, Current Biology, 2013
- A method to soothe and promote sleep in crying infants utilizing the transport response, Current Biology, 2022
- Immediate or early skin-to-skin contact for mothers and their healthy newborn infants, Cochrane Database of Systematic Reviews, 2025
- Characterizing kinematics, kinetics, and muscle activity of postpartum mothers lifting their own infants during three everyday tasks, Journal of Biomechanics, 2025
- Prevalence of Low Back Pain, Pelvic Girdle Pain, and Combination Pain in a Postpartum Ontario Population, Journal of Obstetrics and Gynaecology Canada, 2020
- Incidence and Risk Factors for Pregnancy-Related de Quervain's Tenosynovitis in South Korea, Clinics in Orthopedic Surgery, 2023
- Risk Factors Associated With de Quervain Tenosynovitis in Postpartum Women, Hand, 2024
- Safety and benefits of adult-worn slings and baby carriers: a narrative systematic review, BMJ Paediatrics Open, 2026
- Developmental Dysplasia of Hip and Post-natal Positioning: Role of Swaddling and Baby-Wearing, Indian Journal of Orthopaedics, 2021
- Head lag in infants at risk for autism: a preliminary study, American Journal of Occupational Therapy, 2012
Last source check: September 2026.
