Can We Prevent Future Language Delays in the NICU?

March 27, 2018
Written by

Researchers investigate the use of event related potentials to measure the effects of mother’s voice exposure on speech sound differentiation.

Preterm infants are at high risk for neurosensory impairments and developmental delays, including hearing loss, which may have lasting consequences. Compared to babies born at term, preterm infants are twice as likely to have a severe language delay.

These delays are likely to be at least partly the result of a difference in sound quality that infants in the NICU experience compared to their healthier full-term peers. In particular, sounds in the Neonatal Intensive Care Unit (NICU) can be louder, have broader frequency ranges and originate from machines and medical personnel instead of parents, all this in the earliest days and weeks of life. Speech sound exposure lays the foundation for the neural networks that are necessary for later language processing and understanding. However, preterm infants are exposed to less infant-directed speech sounds during their NICU stay, a critical problem in the development of future communication skills.

Nathalie Maitre, MD, PhD, neonatologist and principal investigator in the Center for Perinatal Research at Nationwide Children’s, is one of the neuroscience researchers looking to change these outcomes. In a recent study published in Developmental Neuropsychology, her team showed that event-related potentials (ERP) provide valid and predictive biomarkers of infant speech-sound differentiation in hospitalized infants and that they are a useful tool for measuring the impacts of interventions in this vulnerable population.

ERP methodology – also called time-locked electroencephalogram (EEG) – measures the electrical potential generated by neurons and groups of neurons in the brain, in response to specific stimuli. The stimulus is repeated many times and time-locked, to confirm that the electrical brain activity observed is associated with the stimulus and not a simply a non-specific or background response. Dr. Maitre’s team measures ERP with high density EEG caps, which have a soft flexible nets covered in electrodes with microsponges, to monitor nearly the entire sphere of electrical activity in the brain. This approach is gentler for the fragile infants and enables the researchers to see much more precisely than they would with a traditional EEG using 10-20 electrodes.

In this study, researchers used the ERP to measure the infants’ responses to /ba/-/ga/ and /du/-/gu/ contrasts after a period of exposure to either passive or active intervention – that is, the mother’s voice delivered through some new technological devices developed at NCH.

“These vowel sounds are some of the earliest to be differentiated by babies – even those born very preterm,” says Dr. Maitre. “The ‘b,’ ‘d’ and ‘g’ sounds are some of the earliest consonants.”

The infants were randomized into two groups. The passive group received standardized environmental mother’s voice exposure; the active group used experimental active, suck-contingent mother’s voice exposure. All of the infants were preterm and hospitalized in the NICU; some even had brain injury.

“It’s not that passive listening doesn’t have an effect,” explains Dr. Maitre. “It would be a misconception to think that passive exposure to mother’s voice does not help infants develop speech sound differentiation to some degree. However, we know that active learning is more powerful than passive learning, and we hypothesized that the active group would have a stronger response to the intervention.”

Music therapists worked with the mothers to create the infant-directed song and speech recordings used in the intervention. Dr. Maitre notes that the most important thing about the recordings is infant-directed speech, and mothers are encouraged to speak to their babies and think about them as they record.

“While the simplicity of the pattern and musical envelope and limited frequency and amplitude range are important – because that’s all the infant’s ear and brain can process at this early age – the emotional context that comes with infant-directed communication is critical,” says Dr. Maitre.

In addition to confirming the efficacy of ERP as a biomarker for speech-sound processing, the study indicated that the intervention was beneficial for the infant.

“This study is an important early step in developing interventions that could dramatically change outcomes for these patients and families,” says Dr. Maitre. “Developing and validating the right tools enables us to move the research forward in larger studies.”



Chorna OD, Hamm EL, Shrivastava H, Maitre NL. Feasibility of event-related potential (ERP) biomarker use to study effects of mother’s voice exposure on speech sound differentiation of preterm infants. Developmental Neuropsychology. 2018;43(2):123-134.