Babbling Complexity and Speech Outcomes in Children with Cochlear Implants

1
Babbling Complexity and Speech Outcomes in
Children with Cochlear Implants

• Elizabeth Walker, MA
• Sandie Bass-Ringdahl, PhD
• The University of Iowa, Iowa City, IA

2
Babbling and early speech in infants with normal
hearing

• Research indicates continuity between
  prelinguistic vocalizations and early speech
  development
• Stoel-Gammon (1989)
• Children who are poor babblers (i.e., produce
  limited babbling with few consonants) may produce
  first words later and acquire lexical items more
  slowly than children with more complex babble

3
Babbling and early speech in infants who are
deaf or hard of hearing (D/HH)

• Few studies on relationship between babbling and
  early speech and language in this population
• Wallace, Menn, Yoshinaga-Itano (1998)
• Examined relationship between babbling quality
  and hearing loss
• Measured phonetic complexity in babbling for
  infants between 5 and 13 months
• Found no significant relationship between
  babbling and demographic variables (PTA) or
  outcome variables (articulation measures)

4
Rationale and Research Questions

• Clinicians need more informal measures for
  tracking speech and language progress in children
  receiving CIs under the age of 2.
• Can measures of phonetic complexity in babbling
  be used to track progress for infants with CIs?
• Finding a predictive relationship between
  babbling complexity and speech/language outcomes
  provides us with a means for assessing progress
  in very young children with CIs.
• Does babbling complexity predict speech and
  language outcomes in this population?

5
Method

• Participants
• 19 infants with prelingual SNHL
• Mean age at first visit, 13.74 months (4-27
  months)
• Mean age at initial stim, 18.39 months (11-27
  months)
• Data collection
• Vocal recordings collected at pre- and post-CI
  visits
• Infant interacted with experimenter and caregiver
  
• Sessions followed similar protocol across
  participants

6
Method Mean Babbling Level (MBL)

• Measured phonetic complexity of babbling using
  Mean Babbling Level (MBL) analysis (Stoel-Gammon,
  1989)
• Transcribed maximum of 50 vocalizations per visit
• Vocalizations isolated based on 2-second breath
  groups
• Vocalizations with greater than 2 second pause
  separating them were considered to be two
  different vocalizations

7
Method Mean Babbling Level (MBL)

• Vocalizations assigned to levels based on
  phonetic content and syllable structure
• Level I vowel, syllabic consonant or
  consonant-vowel combination consonant is glide
  or glottal
• Level II CV combination in which the consonant
  is a true consonant place/manner do not change
• Level III at least two true consonants that
  differ by place and/or manner
• MBL (Sum of of Level I) x 1 (Sum of of
  Level II) x 2 (Sum of of Level III) x 3/Total
  utterances
• Scores range from 1.0 to 3.0
• MBL increases with chronological age and is
  negatively correlated with age at onset of
  meaningful speech

8
Method Outcome measures

• Language measures collected at approximately 48
  months chronological age (mean CA 49 months
  range 39-56 months)
• Mean length of CI use 29 months (18-36 months)
• Minnesota Communication Development Inventory
  (MCDI) expressive and receptive language quotient
  (LQ) scores
• Goldman-Fristoe Test of Articulation-II standard
  scores (GFTA-II SS)
• Peabody Picture Vocabulary Test-III standard
  scores (PPVT-III SS)

9
Method Data analyses

• MBL scores for participants with multiple data
  points within time periods averaged at 5
  intervals
• Pre-implant and initial stimulation
• 2 week, 1 and 2 months post-CI
• 3, 4, 5 months post-CI
• 6, 7, 8, 9 months post-CI
• 10, 11, 12, 13 months post-CI
• Pearson Product-Moment correlations calculated
  for average MBL at 5 time intervals and MCDI
  Expressive and Receptive LQ, GFTA-II SS, PPVT-III
  SS
• Stepwise multiple regression analyses run for
  average MBL at 6-9 and 10-13 month time intervals
  and four outcome measures

10
Results Correlational analyses

• Significant correlations between MBL at 6-9
  months post-CI and PPVT SS, GFTA SS, and MCDI
  receptive and expressive LQs
• MBL PPVT r.70 (n17, plt.01)
• MBL GFTA r.52 (n18, plt.05)
• MBL MCDI Receptive r.75 (n18, plt.01)
• MBL MCDI Expressive r.75 (n18, plt.01)

11
Results Correlational analyses
r.70, plt.01
12
Results Correlational analyses

• Significant correlations between MBL at 10-13
  months post-CI and PPVT SS and MCDI receptive and
  expressive LQs
• MBL PPVT r.59 (n16, plt.05)
• MBL GFTA r.43 (n18, n.s.)
• MBL MCDI Receptive r .63 (n17, plt.01)
• MBL MCDI Expressive r.52 (n17, plt.05)

13
Results Correlational analyses
r.59, plt.05
14
Results Regression analyses

• MBL and MCDI Expressive LQ
• 67 of the variance accounted for by MBL at 6-9
  months post-CI (n15 p.0002)
• MBL and MCDI Receptive LQ
• 66 of the variance accounted for by MBL at 6-9
  months post-CI (n15 p.0003)
• MBL and GFTA-II SS
• 31 of the variance accounted for by MBL at 6-9
  months post-CI (n15, p.03)
• MBL and PPVT-III SS
• 59 of the variance accounted for by MBL at 6-9
  months post-CI (n15 p.0009)

15
Discussion

• Is mean babbling level an effective measure for
  tracking progress in very young children with
  CIs?
• MBL is simple to perform and provides clinicians
  with a gradient measure by which to monitor vocal
  development in children with CIs.
• Is there a significant relationship between
  babbling complexity and speech and language
  measures in children with cochlear implants?
• Phonetic complexity in babbling predicts later
  speech and language outcomes in children with
  CIs, after they have had at least 6 months of
  listening experience.

16
Take-home message

• Clinical implications
• Prelinguistic vocalizations may be an important
  prognostic indicator for later speech and
  language development
• Low MBL scores after 6 months of CI use may
  indicate a need to increase clinical services or
  change CI MAP
• Theoretical implications
• Babbling is continuous with speech and language
  in children with hearing loss
• Once children are exposed to auditory input, they
  will show a progression in vocal development that
  is similar to younger, normal-hearing peers

17
Acknowledgements

• Funding support
• NIH/NIDCD Grant P50 DC00242 Grant RR00059 from
  the General Clinical Research Centers Program,
  NCRR, National Institutes of Health the Lions
  Clubs International Foundation and the Iowa
  Lions Foundation
• University of Iowa Pediatric Cochlear Implant
  team
• J. Bruce Tomblin, Brittan Barker, Linda Spencer,
  Maura Kenworthy, and Tanya Van Voorst
• University of Iowa Pediatric Audiology lab
• Michele Arata, Erin Becker, Kate Larsen, Greta
  Martin, and Katie Woodard