Premature Baby: Correcting the Age, and What Helps
Written by the Inclusive Developmental and Therapy Center therapy team — our Speech & Language Therapist, psychology and ABA staff. Clinical reviewer for this site: Dr Muhammad Suffyan, MB BS (GMC 8023727) · Updated August 2026; this page has changed since our last clinical review.
Your baby came early. Since then you have been handed two different ages for the same child, a list of things she should be doing by now, and a strong sense that everyone else's baby is ahead.
This page covers three things: how to do the arithmetic properly, how long to keep doing it — which is the question nobody answers — and what the evidence says actually protects development, which is more within your reach than you have been led to believe.
How to work out corrected age
Take 40 weeks. Subtract the number of weeks your baby was born at. That is the correction, and you subtract it from your baby's actual age.
A baby born at 32 weeks is 8 weeks early. At 6 months old on the calendar, her corrected age is 4 months. When you read a milestone list, you read the four-month row — not the six-month row.
Do this for everything: milestone charts, developmental checks, other people's babies at family gatherings, and the sentence in your own head at three in the morning.
When do you stop correcting?
You will be told two years, almost universally. There is a study that tested it, and the answer is more nuanced than that.
Researchers assessed 75 healthy preterm children (mean gestation 32.5 weeks) and 69 term-born children using the Bayley Scales at 24 and 36 months, and compared what happened when scores were calculated with corrected age versus chronological age.
- At 24 months, corrected and chronological scores differed significantly on every scale — cognitive and both language scales.
- At 36 months, the difference had gone for the cognitive scale but remained significant for the language scales.
The practical rule that follows: by three years, correcting no longer changes much for general development — but it still matters for language. If you are worrying about how many words your three-year-old has, and she was born at 31 weeks, you are entitled to the correction on that specific question for longer than anyone told you.
The same study also found something you should hear plainly rather than discover slowly: the preterm group scored lower than the term group on cognitive and language scales at both ages, whether or not the age was corrected. Correction is an honest adjustment, not a way of making a real difference disappear. Which is exactly why the rest of this page matters.
What the risk actually is — in our part of the world
Most figures you will find come from wealthy countries with large neonatal units. There is a review that looked specifically at low- and middle-income countries, which is the honest comparison for a family in Pakistan.
It pooled 47 datasets from 12 countries covering 72,974 preterm-born children, all assessed at 12 months or older with a validated method. The pooled prevalences:
- Any neurodevelopmental impairment: 16% (95% CI 11–21%)
- Cerebral palsy: 5% (95% CI 3–6%)
- Developmental delay across the various domains: 8% to 13%
- Hearing impairment or deafness, and visual impairment or blindness: 1% each
Read that the right way round. Sixteen percent means roughly five out of six preterm children in these countries had no neurodevelopmental impairment. It also means the risk is real and one child in six needs to be found early — which is what the checks are for, and why you should keep going to them even when they feel like a formality.
The thing you can actually do
Here is the finding we would most want a parent of a premature baby to have.
An individual participant data meta-analysis published in The Journal of Pediatrics harmonised seven birth cohorts across five countries, 2,560 children, and asked whether observed sensitivity in the mother–child interaction predicted later cognitive and language scores — adjusting for the child's sex, multiple birth, being small for gestational age, age at assessment, the mother's education and age, single-parent status, and the child's own neurodevelopmental impairment.
It found sensitivity associated with later cognition (0.20, 95% CI 0.14–0.26) and language (0.15, 95% CI 0.08–0.22), both p < 0.001. And the interaction that matters most: the association with later cognition was stronger in children born at younger gestational ages.
The authors called sensitive parenting a potential protective factor, and we will keep their wording — this is observational data, so it cannot prove that sensitivity causes the difference. But it is a large, well-adjusted analysis pointing at something free, available at home, and apparently more powerful the earlier your baby was born.
What "sensitive" actually means, in practice
It is not devotion, and it is not doing more. It is a specific and learnable thing: noticing your baby's signal and responding to it, at their pace rather than yours.
- Follow her lead. Whatever she is looking at is the topic. Name that, rather than redirecting her to the toy you had planned.
- Wait longer than feels natural. Premature babies often take longer to respond. Count to five before you fill the silence — the gap is where her turn goes.
- Read the stop signals. Looking away, arching, a change in colour or breathing means "enough for now", not rejection. Backing off when she signals is the sensitive response; pressing on is not.
- Serve and return. She makes a sound, you answer as though it were a word. She moves, you notice out loud. This is the whole mechanism of early language and it costs nothing.
- Do it in the ordinary moments — feeding, changing, being carried. That is where the hours are.
- Speak your own language. Urdu, Saraiki, Punjabi. A parent talking fluently and warmly in their own language does far more than a parent working carefully in a borrowed one — see raising a child in a bilingual home.
One thing follows directly from that meta-analysis and is worth naming: sensitivity is hard when you are frightened or exhausted, and a NICU stay produces both. Support for the parent is not a soft extra here — it is upstream of the mechanism the study measured. Our page on the break you are not taking covers that side honestly.
What to watch for, at corrected age
Every one of these is read at corrected age, not calendar age.
- Not smiling socially by around 2 months corrected
- Not turning towards sound, or no reaction to a loud noise, at any age
- Stiffness or floppiness, or consistently using one side of the body far more than the other
- Hands persistently fisted beyond about 4 months corrected
- Not babbling by around 9 months corrected
- No pointing or showing by around 15–18 months corrected
- No words by around 18 months corrected — see an 18-month-old who is not talking
- Loss of any skill at any age — this one is a same-week conversation
- Feeding that never became straightforward, or persistent coughing during feeds
Vision and hearing deserve their own line. They are the two impairments in the review above that come in at only 1% each — but they are also the two that are most treatable and most often missed, and an undetected hearing loss looks exactly like a language delay. If a newborn hearing screen was not done, ask for one. Our page on the two kinds of hearing loss explains what the results mean.
What not to do
- Do not compare her with a term baby of the same calendar age. That comparison is not measuring anything real.
- Do not skip the follow-up checks because she seems fine. One in six is the figure the checks exist for.
- Do not wait to be told to worry. "He will catch up" is often true and it is not a plan. If something specific concerns you at corrected age, ask.
- Do not let anyone tell you the correction is an excuse. It is arithmetic, and there is trial data on how long it holds.
Where we fit
We are not a neonatal service and we do not replace your paediatrician's follow-up.
What we do is the developmental side, once your baby is home and the question becomes whether the gap is closing. Our developmental assessment reads everything at corrected age, and if the answer is that she is doing exactly what a baby of her corrected age should do, we will tell you that and send you away. Where support is needed, speech and language therapy and occupational therapy are the usual routes, speech delay and global developmental delay explain the two labels you are most likely to hear, and our milestone pages give you the ages to read against.
A consultation is Rs 1,500 and lasts up to 50 minutes, we work in Urdu or English, and you do not need a referral.
Sources
- Ionio C, Riboni E, Confalonieri E, et al. Paths of cognitive and language development in healthy preterm infants. Infant Behavior & Development. PubMed 27450100
- Sultana Z, et al. The prevalence of long-term neurodevelopmental outcomes in preterm-born children in low- and middle-income countries: a systematic review and meta-analysis of developmental outcomes in 72,974 preterm-born children. Journal of Global Health. PubMed 40181744
- Jaekel J, et al. The association of maternal sensitivity with subsequent cognition and language among children born preterm: an individual participant data meta-analysis. The Journal of Pediatrics. PubMed 41655627
Frequently asked questions
How do I work out my baby’s corrected age?
Take 40 weeks, subtract the number of weeks your baby was born at, and subtract that difference from your baby’s actual age. A baby born at 32 weeks is 8 weeks early, so at 6 months on the calendar her corrected age is 4 months — and you read the four-month row of any milestone chart, not the six-month row.
When should I stop using corrected age?
Later than two years for language. A study comparing 75 healthy preterm and 69 term-born children on the Bayley Scales found that at 24 months corrected and chronological scores differed on every scale, while at 36 months the difference had gone for the cognitive scale but remained significant for the language scales. So by three years correction no longer changes much for general development, but it still matters for language.
Will my premature baby catch up?
Often, but the honest answer is that correction is an adjustment rather than an erasure. In the same study the preterm group scored lower than the term group on cognitive and language scales at both 24 and 36 months whether or not the age was corrected. That is precisely why early support and the follow-up checks matter.
What is the actual risk of a developmental problem?
A meta-analysis of 47 datasets from 12 low- and middle-income countries covering 72,974 preterm-born children found pooled prevalences of 16% for any neurodevelopmental impairment, 5% for cerebral palsy, 8% to 13% for developmental delays across domains, and 1% each for hearing and visual impairment. Read that the right way round: roughly five out of six preterm children had no neurodevelopmental impairment — and the one in six is who the checks exist to find.
Is there anything I can do at home that helps?
An individual participant data meta-analysis of seven birth cohorts across five countries (2,560 children) found observed sensitivity in the parent–child interaction associated with later cognition (0.20) and language (0.15), and the association with cognition was stronger in children born at younger gestational ages. The authors called it a potential protective factor rather than a proven cause, but it is free, available at home, and apparently more powerful the earlier your baby was born.
What does sensitive parenting actually mean in practice?
Noticing your baby’s signal and responding at their pace rather than yours. Follow her lead and name whatever she is looking at. Wait longer than feels natural — premature babies often take longer to respond, and the gap is where her turn goes. Read stop signals such as looking away or arching as "enough for now" and back off. Answer her sounds as though they were words. Do it in feeding, changing and carrying, and do it in your own language.
What should I watch for?
All at corrected age: no social smile by about 2 months; no turning towards sound at any age; stiffness, floppiness or consistently favouring one side; hands persistently fisted beyond about 4 months; no babble by about 9 months; no pointing or showing by 15 to 18 months; no words by about 18 months; feeding that never became straightforward. Loss of any skill at any age is a same-week conversation.
Should I still go to follow-up checks if she seems fine?
Yes. One in six is the figure those checks exist to find, and hearing and vision are the two impairments most treatable and most often missed — an undetected hearing loss looks exactly like a language delay. If a newborn hearing screen was never done, ask for one.