#453 - [Journal Club] - 📌 Can We Rewire a Preterm Baby's Brain for Language?

In this Journal Club, Daphna takes the reins with the MIND randomized controlled trial from Nathalie Maitre and colleagues in The Journal of Pediatrics. Can a multisensory bundle, combining infant-directed voice, a parent's scent, holding, and gentle containment, do more for a preterm baby's developing brain than recorded voice alone? Using event-related potentials to track how infants tell speech sounds apart, the team followed language outcomes all the way to age two. Daphna and Ben unpack the design, the Bayley and PLS-5 findings, and a takeaway every clinician can act on tomorrow. Talk to the baby, every single time.

----

The MIND Randomized Controlled Trial: An Intervention to Improve Neural Speech Processing and 2-Year Language Outcomes of Infants Born Preterm. Maitre NL, Kjeldsen CP, Jeanvoine A, Lukemire J, Slaughter JL, Key AP.J Pediatr. 2026 Jun 5:115187. doi: 10.1016/j.jpeds.2026.115187. Online ahead of print.PMID: 42250747

Transcript

Ben Courchia, MD (00:00.888)
Hello everybody, welcome back to the Incubator Podcast Journal Club. Daphna, good morning. How's everything?

Daphna Yasova Barbeau (00:05.414)
Morning. I'm going well. I'm going well. I'm doing well. I'm going well. I wonder if people had to listen to your last Journal Club in two parts.

Ben Courchia, MD (00:08.863)
Mm-hmm. You're going well.

Ben Courchia, MD (00:16.514)
You know what you should do? Listen to it once and then go read the freaking paper. That's all I can hope for, that it breaks it down a little bit for you. And it's important for people like Dr. Mitra and Brett Manley to leverage fancy, innovative statistical approaches and study design to actually answer questions that have been bugging us. And for us readers,

Daphna Yasova Barbeau (00:20.668)
That's right. That's the goal. That's the goal.

Ben Courchia, MD (00:41.71)
it's a bit difficult to understand, because we're not familiar with it. If you're not doing a Student's t-test for me, I have to dive a bit deeper into what a win ratio is. So it's going to take a bit of work. That's okay, but it's valuable work. Go do it.

Daphna Yasova Barbeau (00:44.349)
Yeah.

Daphna Yasova Barbeau (00:48.136)
Figure it out.

Yeah.

Daphna Yasova Barbeau (00:55.368)
For sure, for sure. All right. Well, you've had enough time to talk about your papers. It's my channel. I have this article entitled the MIND, M-I-N-D, Randomized Controlled Trial, an Intervention to Improve Neural Speech Processing and Two-Year Language Outcomes of Infants Born Preterm. It's in The Journal of Pediatrics. The lead author is, say it for me. Maitre. Did I get it?

Ben Courchia, MD (01:00.163)
 Mm-hmm.

Ben Courchia, MD (01:19.583)
 Nathalie Maitre. Maitre. Yeah.

Daphna Yasova Barbeau (01:24.646)
 I'm not sure. The Duolingo is not doing as much for me as I thought in my French practice. Maître. Did I say "ray"? Yes. Okay. Senior author Alexandra Key. And this is a really interesting paper,

Ben Courchia, MD (01:29.262)
 Ha ha, Nathalie.

What? Yeah.

Daphna Yasova Barbeau (01:43.168)
 because the first line of the background says, "Despite advances in neonatal care, infants born prematurely still have high rates of developmental delays and impairments." And I thought that was interesting, because we recently released episode number 448, Are NICU Outcomes Actually Getting Better Over Time, with Dr. Camp. And the answer is no.

"No, but" is the answer, I guess. And we got a lot of feedback on social media about this, because people said, wow, we've made so many advances. Look at what we're doing. We're saving these little babies. And we had to respond that that was the point, right? Mortality has decreased significantly. We had a great talk in the first Delphi, out of Japan, that mortality has decreased significantly, but

the neurodevelopmental outcomes are not changing. We're really not moving the needle on developmental outcomes. And that matters to families. So, anyway, I thought that was an interesting highlight. What they tried to do in this study, they recognized that maternal voice is very valuable. Maternal voice while holding is the gold standard of developmental care. But

for example, here in the United States, it's really difficult for us to have parents at the bedside. So could we create a proxy for parents at the bedside? That gets to the rest of the background. And I quote, "singing during NICU kangaroo care increased exposure to coached parental singing and was associated with better outcomes."

"Later in admission, parental singing and reading positively influence linguistic outcomes of former preterm infants." And so this underlies how important exposure to infant-directed voice is. Infant-directed voice means the type of language, the type of talking we use with infants, which is a little different from the intonation, the flow that we use with adults. So that specific infant-directed voice is really important

Daphna Yasova Barbeau (04:05.732)
 in shaping auditory processing of language. And of course, systemic and structural factors in the NICU limit parental presence, and that makes live infant-directed voice by caregivers, familial caregivers, really difficult. So the MIND intervention, multisensory support to improve neural processing and development, MIND,

was a way to try to expose babies to their parents' infant-directed voice and holding, even if the parent wasn't there. I'll explain exactly what the MIND intervention looked like, but they compared that to their quote-unquote enhanced standard care, where infants heard just recordings of maternal voice, but without

these other components. Okay, fine. Let me just tell you what the other components are for the MIND. So I'll tell you about the caregiver voice recordings. Mothers were trained by a board-certified music therapist to do infant-directed speech and singing.

And then they recorded these, and they made recordings that they were able to play that included lullabies, nursery rhymes, and children's books. They were able to play this from speakers. So this passive caregiver voice recording is standard of care. All the babies are being exposed to these voice recordings of their mothers, and it's not just voice recordings, specifically infant-directed

voice recordings. So the control group was hearing these caregiver voice recordings via a customizable speaker placed near the baby's head. And the nurses were instructed to do this one to two times a day. Specifically, they were instructed to activate the speaker when the infant was in a quiet alert state. Now, the MIND intervention includes this pre-recording of mommy's

Daphna Yasova Barbeau (06:25.564)
 voice that's inventoried, but it additionally includes smell. So moms wore cotton t-shirts for an hour in the NICU, and then the cotton t-shirt was cut up into pieces. Part of the MIND intervention included exposure to a piece of this cotton t-shirt. So during sessions, babies were exposed to their mother's scent

and the therapists held the infant against their chest. They were wrapped in a Zaky, which is like a kangaroo care binder, and the infant's head was on their mommy's scent cloth. We have, yes, we have. Perfect. Yeah, that's a great interview. So basically, one group just got the recordings, the other group

Ben Courchia, MD (07:07.63)
 We interviewed the creator of the Zaky, if I'm not mistaken. I'll find the episode number, but keep going.

Daphna Yasova Barbeau (07:23.432)
 got the recordings while therapists were holding them in the Zaky binder, in a quiet alert state, and wearing a piece of their mother's scent cloth. So basically, they're trying to convince the baby that it's being exposed to its mother, even when the mother can't be present, versus not being held,

not being contained in the binder, and just being exposed to their mom's voice through the speaker. Does that make sense? Did I explain that right? Do you think so?

You're muted, but I see you saying yes. Great. Okay. So they always assessed infant readiness. These sessions were attempted before or after routine care. Sessions were discontinued if the infant did not engage. Part of the engagement was using this PAL, the, what's it called, pacifier-activated music player. So by the baby sucking on the pacifier,

Ben Courchia, MD (08:06.743)
 Yeah, I was muted, but I said very well. Yes, you did that very well.

Daphna Yasova Barbeau (08:36.166)
 that's what activated the speaker to do its thing in both groups. So either the baby was being engaged just with auditory stim, or engaging on the pacifier with auditory and the additional sensory smell, touch, and proprioception. Okay. And it was also discontinued, obviously, if there was any autonomic instability, which they defined, and they made sure that consecutive sessions were at least

three hours apart. The goal dose was two to three times daily over two to three weeks, with a goal of achieving 20 sessions. And they considered them compliant with the study protocol if they were able to get in at least 12 sessions. I'll argue that it will be hard for a therapist to do this three times a day, just as potentially difficult as getting parents at the bedside, but it's an interesting concept.

And then, what were they measuring? They were looking at auditory event-related potentials. These are ERPs, using EEG, collected within 72 hours of the first and the last study session. So they took a baseline of ERPs for the babies, and then at the end of the two to three weeks of sessions, they got an additional measurement. And so they were looking at these auditory event-

related potentials. So what is that? What is an ERP? They use the scalp electrodes from the EEG, and they're looking at a measurable brain response. And then, specifically, what they're looking at, and I'll get into this, is something about consonant contrast. So is the baby able to distinguish that

the B sound is different from the D sound, which is different from the P sound? So those were the things they were comparing. And in terms of the leads they were using, based on previous studies, using the 10-20 EEG system, they used a left frontal lobe lead and a right frontal lobe lead, F3 and F4. Yes, we're following so far.

Daphna Yasova Barbeau (11:01.096)
 And

Ben Courchia, MD (11:01.611)
 Yes, I'm following. I would have a question about the significance of that, in terms of, why is it, the fact that the baby can differentiate, you said the B sound and the other sound, or something. Yeah.

Daphna Yasova Barbeau (11:10.656)
 Yeah, perfect. So that speech sound differentiation, consonant differentiation, is a fundamental building block for language development. So they're trying to see, at baseline, do the babies have this differentiation? Does it get better with the intervention? And, hopefully, is there a difference between the MIND intervention and their standard control of just the auditory sound?

Now, there are some units where babies aren't getting exposed to language at all. So the bare minimum is getting babies exposed to language. But how does that change with maternal infant-directed voice, and then the MIND intervention, maternal infant-directed voice, maternal smell, and being contained by being held on the chest? Phew. So there, I guess.

What they're trying to prove is, if we can improve these building blocks of speech, how does it improve speech outcomes at two years? So again, they looked for these ERPs, they looked at the differences between the pre and post, between different babies. And, I gave you the wrong consonants earlier, the consonants are B, G, D,

G and B and D. So they were looking at the different consonant contrasts. And then they looked at the two-year behavioral measures. They looked at 22 to 26 months corrected gestational age, the Bayley Scales of Infant and Toddler Development, Third Edition, and they used the PLS-5, the Preschool Language Scale, Fifth Edition, to look at outcome convergence. So did any of the markers in the infancy period

predict outcomes at two years? Okay. I spent a lot of time working through that. I think we're good here. In total, 120 infants were randomized to the intervention MIND group. 108 completed the intervention. 100 had EEG data available at both assessment points, and about an 85% follow-up rate at two years, which again is pretty remarkable.

Daphna Yasova Barbeau (13:32.275)
 They had 122 infants in the control group. 116 completed the control intervention, 98 with EEG data, and again, an 87% follow-up rate at two years for the control group. Importantly, no infants failed their hearing screens during those first two years. They did have some pretty impressive COVID-19-related attrition, but they were able to mitigate that.

The major baseline characteristic differences showed just that the mean educational duration for mothers in the control group was one year greater than in the intervention group. This was statistically significant. And the control group had a slightly shorter delta EEG. So the time between the pre- and post-EEG time points was slightly shorter, with a p-value of 0.049. Okay, so what did they find?

They found that there was a significant difference at post-EEG assessment between groups. They saw improvement in both groups, but, as I'll get into, more improvement in the MIND group. They found differences in the locations, so the different frontal leads, F3 and F4, and in the time window. So I didn't mention this, but they also looked at different time points

past the intervention, in milliseconds. They used two different time points from the exposure to the auditory stimulus. So at the subject level, the ANCOVA revealed significant differences between groups across the temporal lobe locations over time. And, same thing, the contrast-level ANCOVA, looking at the different consonant contrasts, also showed

differences between the groups. What did this look like at two years? There were direct positive effects of the MIND intervention on the Bayley receptive and expressive scores, even when controlling for gestational age and maternal education. All of the scores are here, but I'm not sure that everybody's that familiar with them. You should just know that the scores improved.

Daphna Yasova Barbeau (15:52.627)
 And there was also improvement on the Preschool Language Scale five auditory comprehension and expressive communication scores. And the scores represent small to medium-sized effects, about 0.5 standard deviations on the scaled score measures for both the Bayley and the PLS-5, which across published literature is a clinically meaningful effect size.

So at the participant level, the total effects modeled showed the intervention effects on speech sound differentiation. So improvements in speech sound differentiation on the pre and post also predicted better scores on the two-year language outcomes. The MIND effects on differentiation of especially the B and G consonant contrast, measured in the frontal locations,

were significant for every single language score. So if babies were able to have that differentiation, especially for B and G, they had improved scores across the board at two years. So their overall conclusion is that the MIND intervention, language plus scent plus touch and proprioception,

had a much larger effect on speech sound differentiation than enhanced standard care. And this improvement in neural effect mediated better performance on the two standardized measures of receptive and expressive language at age two. So I think this is really exciting. Maybe people are like, yeah, no duh, that makes sense. But I think it underscores that, one, exposure to language is really important. Two,

infant-directed speech is really important, and every single one of us can do that every time we engage with the baby. There's no reason you can't tell the baby, good morning, hello, you're so cute, you're so strong, in baby voice and baby speak every time we enter the isolette. So every single one of us can do that tomorrow when we go into the NICU. And that the combination of kangaroo care,

Daphna Yasova Barbeau (18:08.693)
 and being held and being talked to, and smell, and vibration, all at the same time, probably has a much bigger impact on development than just exposure to any one of these things in isolation. And of course, being near their familial caregivers is optimal. It looks like we can do a proxy with therapists, but we're pretty, we're pretty

aggressive, I would say, with developmental care in our unit. And there's no way that we would be able to have therapists do this three times a day. So

Ben Courchia, MD (18:43.827)
 Yeah. But to correct you on one thing, the meaning was not what was tested. You could say whatever you want. You could even read out the news. You can just open your mouth. Yeah.

Daphna Yasova Barbeau (18:47.273)
 Please.

Daphna Yasova Barbeau (18:52.381)
 I can say whatever I want. That's true. That's true. That's not what was tested.

Daphna Yasova Barbeau (19:00.189)
 That's right. That's right. Yeah, you're absolutely right. What you're saying is, it doesn't have to be lullabies. It doesn't have to be a baby book. Yeah.

Ben Courchia, MD (19:09.85)
 For sure. Sometimes I've seen it, I've done it myself. You're roaming around with students and you say, let's get away from the bedside, let's not bother the baby. And it's like, no, speak around the baby. It's good. Yeah. All right, buddy. Thank you very much. We'll see you tomorrow.

Daphna Yasova Barbeau (19:19.687)
 Yeah, for sure. I love that. I love that. All right, buddy.

Daphna Yasova Barbeau (19:28.233)
 Sounds good?