CORTICAL EVENT-RELATED POTENTIALS TO SPEECH SOUNDS IN 3- MONTH OLD BREAST AND FORMULA-FED INFANTS

Authors: LI, JUAN - ACNC/UAMS; PIVIK, R - ACNC/UALR; JING, HONGKUI - ACNC/UAMS; Gilchrist, Janet

Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 1/17/2006
Publication Date: 3/21/2006
Citation: Li, J., Pivik, R.T., Jing, H., Gilchrist, J.M. 2006. Cortical event-related potentials to speech sounds in 3-month old breast and formula-fed infants. Online Methods in Children's Language Processing, March 21-22, 2006, CUNY Graduate Center, New York, New York. Available: http://qcpages.qc.cuny.edu/~efernand/childlang/posters.htm.

Interpretive Summary: Studies of babies’ responses to speech sounds have not looked at how early diet may change these reactions. We studied brain responses to speech sounds in healthy 3-month-old babies who were breast or formula (milk or soy) fed. Brain responses were larger in girls. Breast-fed babies reacted faster to these sounds, and these babies spent more time processing speech sounds than those fed soy formula. Formula diets had similar effects on how babies react to speech sounds at this time. These are early results, and more data are needed to confirm and interpret these findings.

Technical Abstract: Studies showing that infants can differentially process and discriminate speech stimuli have not considered the influence of diet as an experimental variable. To investigate this relationship, cortical auditory evoked potentials (ERPs) to syllables were recorded (128 electrodes) from 3 month old infants exclusively breast fed (BF: n=59 , males=31), or fed milk (MF: n=64 , males=37) or soy (SF: n= 49, males = 23) formula since at least 2 months of age. Two syllables were presented in an oddball paradigm (/pa/, 80% and /ba/, 20%, duration: 300 ms, 72 dB, ISI: 2550 ms). This report focused on responses to the frequent syllable. As previously described, two positive peaks were observed at ~200 ms (P1) and ~350 ms (P2) in response to our consonant-vowel syllable (/pa/). These peaks have been associated, respectively, with processing of acoustic and phonemic stimulus features. Peak amplitudes and latencies were analyzed using ANOVAs with group, gender, sites (frontal, central, and anterior temporal), and hemisphere as factors. Across groups, P1 occurred earlier at central relative to temporal sites. Furthermore, peak amplitudes were greater at central relative to temporal sites (ps ' .001). This pattern is consistent with maturational processes occurring earlier in primary auditory cortex, which generates the response recorded near midline over the fronto-central region, and later in the secondary auditory cortex. For both peaks, larger amplitudes were observed for left compared to right hemisphere centrally (ps < .01).The reverse pattern was present at temporal sites (ps < .01). Significantly larger amplitudes were observed in female compared to male infants for both peaks, indicating the presence of gender differences in responses to auditory stimuli at this early age. Since responses to non-speech sounds were not studied, we cannot determine whether this difference reflects a female advantage in speech sound perception, and / or the presence of brain morphological differences noted in females, e.g., greater density of neurons in posterior temporal cortex, or larger extent of neuronal processes. In terms of diet-related effects, group differences were present for both peak latencies. BF infants had a shorter P1 latency than the SF group (p < .01). For P2, a group by gender interaction reflected longer latencies in BF compared to SF males (p < .05), and in BF relative to MF females (p < .05). Based upon the presumption that breast milk is the best diet for infant development, the longer interval between the two peaks may support more extensive phoneme processing in the BF infants. The response similarities of the MF and SF groups suggest a general equivalence of these formula types on speech sound processing at this age.