A Mother's Recorded Voice Helped Premature Babies' Language-Wiring Grow

What a Stanford study of babies who heard their moms read Paddington in the NICU (Travis and colleagues, 2025) found — and what it means for families

When premature babies in a newborn intensive care unit were played nightly recordings of their own mother reading a storybook, a language-related pathway in their brains grew more mature than it did in babies who heard no recordings. The extra listening caused no harm to the babies' breathing, weight, or recovery, and the benefit showed up on brain scans taken before the infants went home. Researchers say this is the first strong evidence that hearing more of a parent's voice can actually shape how a premature baby's brain develops — not just calm the baby in the moment [1].

Why this question matters

Babies are built to hear their mother's voice. In the last months of pregnancy, a developing baby is surrounded almost constantly by the muffled sound of speech. A baby born many weeks early loses that. Instead of a warm bath of language, life in a busy, open-room newborn intensive care unit (NICU) is full of alarms, machines, and only brief bits of talk — sometimes as little as 20 to 50 minutes of speech a day. Researchers have long known that the more talk a premature baby hears, the more that baby tends to babble and vocalize later on [2]. This matters because children born very early — before 32 weeks of pregnancy — are more likely than other children to struggle with language and, later, with reading and learning in school. Scientists have traced much of that difficulty not to a single injury but to a slower, stalled build-out of the brain's wiring, a process that stays sensitive to a baby's surroundings [3]. That last point is the hopeful part: if the environment still matters, then changing the environment might help.

How families and clinicians used to face this problem

For decades, the main tools were comfort and closeness. NICU teams dimmed lights, set quiet hours, encouraged skin-to-skin "kangaroo" holding, and asked parents to read and talk to their babies. Studies of a mother's voice found real short-term benefits — babies' heart rates steadied and their oxygen levels held better when they heard mom [4]. Pioneering "developmental care" programs even hinted, years ago, that a gentler, more supportive environment might change the way a premature baby's brain grew [5]. But there was a stubborn gap. All of this evidence was based on watching what naturally happened — comparing babies who happened to hear more talk with babies who heard less. That kind of study can never fully rule out another explanation: maybe babies with healthier brains simply draw more talk out of the adults around them. To know whether the voice itself does the work, you have to run an experiment where chance, not the baby's condition, decides who gets extra speech. That is exactly what this study did.

What the researchers did

The team, led by Dr. Katherine Travis at Stanford's Lucile Packard Children's Hospital, ran the study between 2016 and 2019. They enrolled 46 babies born between 24 and 31 weeks of pregnancy who were medically stable. Using a method like a carefully balanced coin flip, they randomly placed each baby into one of two groups. This randomness is the whole point: it means the two groups started out similar, so any difference at the end can be credited to the treatment rather than to which babies were healthier.

Before the study began, every mother recorded herself reading the first chapter of the children's book A Bear Called Paddington, in whatever language she spoke at home. For babies in the treatment group, a small locked music player in the crib softly played two 10-minute clips of their mother's reading every hour, overnight from 10 p.m. to 6 a.m. — adding up to about two and a half extra hours of their mom's voice each night. The volume was kept low and safe, and the player was placed well away from the baby's head. Babies in the comparison group had the very same equipment set up in their cribs, and their mothers also made recordings, but nothing was ever played. This clever setup meant that families and nurses generally could not tell which babies were actually hearing the recordings, which keeps the results honest. The recordings started at around 32 weeks, the age when a premature baby's hearing is developed enough to make sense of speech [6]. By then, premature babies can already tell apart and even prefer the sing-song way parents talk to infants [7].

Before each baby went home, they had a brain scan called an MRI, which uses magnets — not radiation — to take detailed pictures. A special version of the scan let researchers look closely at the brain's "white matter," the bundles of wiring that connect different brain regions. They focused on a cable-like pathway called the arcuate fasciculus, which links the parts of the brain that recognize speech sounds to the parts that produce speech. Its left-side branch is one of the brain's most important language connections [8].

What they found

The babies who heard their mother's recorded voice had a more mature left-side language pathway than the babies who did not. On the scans, the treatment group's wiring looked measurably further along in its development, and — importantly — the same result showed up on three separate scans and three different measurements, which makes a fluke much less likely. The difference was sizable: roughly comparable to the gap you would see between a full-term baby and a very premature baby [9]. Interestingly, the effect appeared mainly in the left-side language pathway and not on the right side or in other brain regions the team checked, suggesting the extra speech was specifically nourishing the brain's language wiring rather than causing a general change.

Just as reassuring is what did not happen. The recordings caused no increase in the pauses in breathing, slow heart rates, or drops in oxygen that NICU teams watch for. Babies in both groups gained weight and went home at similar times. In short, the intervention appeared both helpful and safe.

What this means for families and their baby's care

If your baby is in the NICU, the most useful takeaway is empowering: your voice matters to your baby's developing brain, and reading aloud is not just sweet — it may be genuinely good for how the language parts of the brain grow. For parents who cannot be at the bedside for long stretches because of work, other children, distance, or cost, a recording of your voice can keep reaching your baby even when you cannot be there. Researchers point out that this could also make care fairer, because families differ in how much time they can spend at the bedside, and a recorded voice can make sure every baby gets a baseline of loving speech [10].

A few honest notes of caution are important, so families read this study for what it truly shows. It was a small study done at a single hospital, and it included mostly healthier premature babies, not the sickest or most fragile. It measured brain wiring on a scan, not the child's actual talking or reading years later — so while the brain change is very promising, we cannot yet promise it turns into better language down the road. The researchers were careful not to oversell it, and neither should we. What we can say is that a simple, inexpensive idea — playing a recording of mom reading a bedtime story — was linked to healthier growth of a key language pathway, with no signs of harm. Music and other soothing sounds have shown their own benefits too, so a baby's ears are clearly a doorway to the developing brain [11].

What researchers are working on next

Scientists now want to know how far this reaches. Larger studies across many hospitals will test whether the benefit holds for babies of different ages and for sicker infants who were not included this time. They also want to learn how much speech is enough, whether live reading works as well as recordings, and — the question families care about most — whether this early boost to brain wiring shows up later as stronger language and learning. For now, the message is warm and simple: talk to your baby, read to your baby, and know that even a recording of your voice may be helping build the connections your child will use to understand and speak for the rest of their life [1].

References

  1. Travis KE, Scala M, Marchman VA, et al. Listening to mom in the neonatal intensive care unit: a randomized trial of increased maternal speech exposure on white matter connectivity in infants born preterm. Front Hum Neurosci. 2025;19:1673471. doi:10.3389/fnhum.2025.1673471 ↩
  2. Caskey M, Stephens B, Tucker R, Vohr B. Importance of parent talk on the development of preterm infant vocalizations. Pediatrics. 2011;128:910–916. doi:10.1542/peds.2011-0609 ↩
  3. Back SA, Miller SP. Brain injury in premature neonates: a primary cerebral dysmaturation disorder? Ann Neurol. 2014;75:469–486. doi:10.1002/ana.24132 ↩
  4. Filippa M, Panza C, Ferrari F, et al. Systematic review of maternal voice interventions demonstrates increased stability in preterm infants. Acta Paediatr. 2017;106:1220–1229. doi:10.1111/apa.13832 ↩
  5. Als H, Duffy FH, McAnulty GB, et al. Early experience alters brain function and structure. Pediatrics. 2004;113:846–857. doi:10.1542/peds.113.4.846 ↩
  6. Hall JW. Development of the ear and hearing. J Perinatol. 2000;20:S12–S20. doi:10.1038/sj.jp.7200439 ↩
  7. Butler SC, O'Sullivan LP, Shah BL, Berthier NE. Preference for infant-directed speech in preterm infants. Infant Behav Dev. 2014;37:505–511. doi:10.1016/j.infbeh.2014.06.007 ↩
  8. Catani M, Jones DK, Ffytche DH. Perisylvian language networks of the human brain. Ann Neurol. 2005;57:8–16. doi:10.1002/ana.20319 ↩
  9. Dibble M, Ang JZ, Mariga L, Molloy EJ, Bokde ALW. Diffusion tensor imaging in very preterm, moderate-late preterm and term-born neonates: a systematic review. J Pediatr. 2021;232:48–58.e3. doi:10.1016/j.jpeds.2021.01.008 ↩
  10. Brignoni-Pérez E, Scala M, Feldman HM, Marchman VA, Travis KE. Disparities in kangaroo care for premature infants in the neonatal intensive care unit. J Dev Behav Pediatr. 2021;43:e304–e311. doi:10.1097/DBP.0000000000001029 ↩
  11. Anderson DE, Patel AD. Infants born preterm, stress, and neurodevelopment in the neonatal intensive care unit: might music have an impact? Dev Med Child Neurol. 2018;60:256–266. doi:10.1111/dmcn.13663 ↩