Current Issues in the Diagnosis of Developmental Language Disorders

advertisement
4/4/2016
Current Issues in the
Diagnosis of
Developmental Language
Disorders
Sean Redmond, PhD, CCC-SLP
sean.redmond@health.utah.edu
 Shifting Terms and Taxonomies
 Diagnosis of Language Impairment
Agenda
 Advances in Differential Diagnosis
 Break
 Comorbidities
 Advances in Assessment of LI in Bilingual Students
 Working Conclusions/Questions
?
What should
this profile be
Research Participant (Blinded Examiner)
called?
5;9 female
Nonverbal Standard Score (NNAT): 92
Phonological Screening: Pass
CELF-4 Standard Score: 56
Reading Standard Score (TOWRE-2): 100
TEGI elicited grammar composite [52%]: Standard Score: 25
CBCL ADHD: T Score 54 (within normal limits)
1
4/4/2016
Internet
Resources
 https://www.youtube.com/user/RALLIcampaign
 http://www.asha.org/public/speech/disorders/LBLD/
 https://www.nidcd.nih.gov/health/specific-language-impairment
Shifting Terms
and Taxonomies
 Initial recommendations for the new DSM 5 Communication
Disorders section and ASHA’s response
 Special issue in International Journal of Language and Communication
Disorders, July-August 2014
 “Criteria and Terminology Applied to Language Impairments:
Synthesizing the Evidence”: CATALISE consortium project
DSM 5
Proposal
 “Communication Disorders - Restructured to now include Social
Communication Disorder plus two diagnostic categories:
Language Disorders and Speech Disorders. These categories each
contain appropriate subtypes to cover all seven of the disorders
previously proposed for this diagnostic category (Late Language
Emergence; Specific Language Impairment; Social
Communication Disorder; Voice Disorder; Speech-Sound Disorder;
Motor Speech Disorder; Child Onset Fluency Disorder)”.
 Proposed Revisions for DSM-5, June 15, 2012
2
4/4/2016
….However, we question the inclusion of SLI as a specifier. SLI is a controversial diagnosis that is not available in the vast majority of clinical settings. It is widely used in research,
but consensus among language scientists on the robustness and validity of the category has not been reached.
Nonverbal IQ is required to make the diagnosis of SLI, and this information is not available in many or most clinical settings. It is not best practice to rely on formal testing to
make a diagnosis. The problems with formal tests are well-known and pervasive. For example, culturally and linguistic appropriate measures are not available for many children
who are L2 English learners and speakers of nonstandard dialects. Therefore the diagnosis will not apply to many groups of children.
In addition, a relatively small number of papers representing a relatively narrow view have been cited to support the proposed categories. The science is not sufficiently
advanced and the controversies surrounding this label have not been laid to rest. Issues include the fact that nonverbal IQ declines with age (L. Leonard, 1998, Specific
Language Impairment, Cambridge, MA: MIT Press), thus rendering the concept of the relation between IQ and language more difficult to understand, and the concern with the
use of 85 as a cut-off IQ for normal functioning, which is higher than that used for determination of intellectual impairment.
Recommendation
ASHA recommends omitting Specific Language Impairment as a specifier of a language disorder.
ASHA’s Recommended Revisions to the DSM-5, June 2012 (p. 15)
 BUT: S(P)CD is really SS(P)CD or “Specific Social (Pragmatic)
Communication Disorder” because…..
 “..cases of intellectual disability and global developmental delay are
excluded…. unless the communication deficits are clearly in excess of the
intellectual limitations” (p. 49).
 BUT: This would require assessment of individual’s nonverbal IQ
 BUT: Notice also, the same IQ discrepancy criteria are still involved with
DSM 5’s Language Disorder (p. 43).
 S(P)CD also excludes social/pragmatic problems which are the result of
semantic or syntactic deficits (p. 48) – i.e. SLI.
How common
are language
impairments?
Tomblin, Smith, & Zhang (1997).
Epidemiology of specific language
impairment: Prenatal and perinatal risk
factors. Journal of Communication
Disorders, 30, 325-344.
Nonverbal IQ
+
Language
_
+
_
Typical & Above
“Spared Language”
Low Cognition w/ TL
75%
11.9%
“Specific Language Impairment”
“Concomitant LI”
Low Cognition w/ LI
8.1%
5.0%
3
4/4/2016
Speech
delays?
How common
are language
impairments?
Tomblin, Smith, & Zhang (1997).
Epidemiology of specific language
impairment: Prenatal and perinatal risk
factors. Journal of Communication Disorders,
30, 325-344.
Shriberg,Tomblin, & McSweeny(1999).
Prevalence of speech delay in 6-year-old
children and comorbidity with language
impairment. Journal of speech, Language, and
Hearing Research, 42, 1451-1481.
SLI: 8%
CLI: 5%
+
Language
_
Nonverbal IQ
+
_
Typical & Above
“Spared Language”
Low Cognition w/ TL
75%
11.9%
“Specific Language Impairment”
“Concomitant LI”
Low Cognition w/ LI
8.1%
5.0%
Classroom A
Classroom D
Classroom B
Classroom E
2.3%
1.25%
Classroom C
SLI-CLI: 13.8%
CLI-SLI: 10.5%
Tomblin & Nippold (2014). Understanding
Individual Differences in Language
Development Across the School Years. pg.
124).
Classroom A
Classroom D
Classroom B
Classroom E
Classroom C
4
4/4/2016
How big is the
gap between
SLI and CLI?
Who receives
SLP services?
Zhang & Tomblin (2000). The association
of intervention receipt with speechlanguage profiles and social-demographic
variables. American Journal of SpeechLanguage Pathology, 11 (9), 345-357.
17% of LI cases were
receiving services in
kindergarten
Who receives
SLP services?
Zhang & Tomblin (2000).The association of
intervention receipt with speech-language
profiles and social-demographic variables.
American Journal of Speech-Language
Pathology, 11 (9), 345-357.
Morgan et al. (2016). Who receives
speech/language services by 5 years of age in
the United States? American Journal of
Speech-Language Pathology, 25 (2), 1-17.
Morgan et al. (2015). Minorities are
disproportionately underrepresented in
special education: Longitudinal evidence
across five disability conditions. Educational
Researcher, 44 (5), 278-292.
SLI > (CLI+ Autism + Down
Syndrome + Stuttering + Traumatic
Brain Injury)
 Why haven’t we heard more about SLI?
 Bishop, D.V.M. (2010). Which neurodevelopmental disorders get
researched and why? PLoS ONE, 5 (11), 1-9.
Classroom A
Classroom D
Classroom B
Classroom E
Classroom C
Classroom A
Classroom D
Classroom B
Classroom E
Classroom C
LI + Speech
White
Have health insurance
> Middle SES
5
4/4/2016
 So if most children with SLI and CLI are not receiving SLP services then who are?
 One possibility is that within the clinical context, services are reserved for the
most severe cases of LI…..
 ….. Or maybe not.
 Schmitt, et al. (2014) examined the correspondence between students’ symptoms of
language disorder as measured by standardized assessment (grammar, vocabulary,
listening comp, literacy) against goals on their IEPs.
Schmitt, M.B., Justice, L.M., Logan,
J.A., Schatschneider, C., & Bartlett
C.W. (2014). Do the symptoms of
language disorder align with
treatment goals? An exploratory
study of primary-grade students’
IEPs. Journal of Communication
Disorders, 52, 99-110.
 There was no correspondence between students’ goals and their test
profiles.
 For example, students with significant vocabulary deficits were as likely to
have vocabulary goals as students whose vocabulary skills were within
normal limits.
 Adjusting for different severity levels had no impact on the observed lack of
correspondence.
 Another possibility is that our caseloads are overly burdened with students who
do not have speech or language impairments but who have other problems that
trigger referrals (e.g. Reading, Behavior Problems).
 Data from the Utah Verbal Screening Study
 181 participants from the grades 1, 2, and 3 groups (mean age 8;1)
 LI defined as “CELF-4 core language < 85” during blinded assessment
 Combined parent report of services received with school records
Ash, A., Pfaff, T., & Redmond, S.
(2015). When language impairment
isn’t enough: Factors that influence
service status for school-age
children. Poster presentation at the
Symposium on Research in Child
Language Disorders, University of
Wisconsin, Madison
 SLP services +/ Other Conditions +/-
 “Typical Language” (n = 136)
 typically developing and “spared language” with other conditions
 “Language Impairment” (n = 45)
 SLI and concomitant LI with other conditions
Mother’s
NNAT
Table 1. ParticipantEducation
information: TL groups
TL:SLP(n = 102)
56.9% male
86.3% Caucasian,
non-Hispanic
TL:SLP+
(n = 34)
70.6% male
85.3% Caucasian,
non-Hispanic
t-test (2, 134)
CELF4
TEGI
AWMA
Reading
Composite
CCC-2
Pragmatic
Composite
CBCL
ADHD
3.77
(.90)
2-5
110.20
(12.55)
82-145
104.79
(9.94)
87-123
98.32
(2.41)
89-100
101.61
(10.86)
78-124
106.30
(11.40)
80-132
42.60
(8.08)
11-57
53.22
(5.57)
50-75
3.62
(.82)
2-5
112.00
(16.60)
81-161
99.71
(8.89)
87-117
93.85
(16.32)
10-100
99.30
(11.27)
83-124
94.15
(13.77)
64-119
35.35
(9.99)
12-55
59.76
(8.64)
50-77
t = .900,
p = .370
t = -.667
p = .506
t = 2.651
p = .009
t = 1.59
p = .121
t = 1.062
p = .290
t = 5.103
p = <.001
t = 4.260
p <.001
t = -4.142
p <.001
Mother’s Ed: 5-point scale of 1 = some high school; 3 = some college; 5 = advanced degree
NNAT, CELF4 Core Language Score, AWMA, Reading Composite standard scores (M = 100; SD = 15)
TEGI Early Grammatical Composite based upon percentage correct finiteness production
CCC-2 Pragmatic Composite sum of pragmatic subscales (initiation, nonverbal communication, social relations, interests)
CBCL ADHD T score (M = 50, SD = 10)
6
4/4/2016
Mother’s
Education
NNAT
CELF4
TEGI
AWMA
Reading
Composite
LI:SLP(n = 14)
3.07
(.73)
2-4
96.64
(13.25)
83-125
75.93
(6.94)
62-85
90.71
(7.93)
71-100
88.75
(8.91)
74-102
90.46
(11.24)
71-108
LI:SLP+
(n = 31)
3.06
(.93)
1-5
92.19
(13.79)
73-125
67.19
(14.55)
42-85
77.16
(24.76)
4-100
85.22
(9.57)
71-105
79.21
(13.64)
54-102
42.9% male
71.4% Caucasian, nonHispanic
61.3% male
71% Caucasian, nonHispanic
t-test (2, 43)
t = .025,
p = .981
t = 1.01,
p = .316
t = 2.76,
p = .009
t = 2.75,
p = .009
t = 1.17,
p = .248
t = 2.698,
p = .010
CCC-2
Pragmatic
Composite
CBCL
ADHD
38.29
(10.14)
16-49
59.64
(10.40)
50-75
35.94
(9.44)
17-54
57.68
(7.65)
50-75
t = .756,
p = .454
t = .712,
p = .481
Mother’s Ed: 5-point scale of 1 = some high school; 3 = some college; 5 = advanced degree
NNAT, CELF4 Core Language Score, AWMA, Reading Composite standard scores (M = 100; SD = 15)
TEGI Early Grammatical Composite based upon percentage correct finiteness production
CCC-2 Pragmatic Composite sum of pragmatic subscales (initiation, nonverbal communication, social relations, interests)
CBCL ADHD T score (M = 50, SD = 10)
Demographic
factors
 Chi Square Analysis: SLP+ in TL
 Sex: ns., Race and Ethnicity: n.s.
 Hx of Other Conditions (13% vs. 66.7%) [χ2 = 34.05, p <.001]
 Chi Square Analysis: SLP+ in LI
 Sex: n.s., Race and Ethnicity: n.s., Hx of Other Conditions, n.s.
 Backwards Stepwise Regression: SLP+ in TL
 Initial Analysis: Hx for Other Conditions, CELF-4, Reading, CCC-2 Pragmatics
and CBCL ADHD
Developmental
factors
 Final Solution: HX for Other Conditions and Reading accounted for 39.2% of
the variance [χ2 = 41.807, df = 2, N = 136, p <.001]
 Backward Stepwise Regression: SLP+ in LI
 Initial Analysis: CELF-4, TEGI, Reading
 Final Solution: Reading accounted for 20.6% of the variance [χ2 = 7.117, df =
1, N = 45, p = .008]
 In other words, on their own, the severity of children’s primary language
symptoms did not affect their likelihood of receiving SLP services. Rather,
the presence of reading difficulties and other clinical conditions predicted
receipt of sLP services.
7
4/4/2016
LI
TD
Number of children
How should
we diagnose
Language
Disorders?
Performance level
Number of children
Clinical marker
distribution
LI
TD
Performance level
Brown’s 14 Morphemes
Finite Verbal Forms
Present progressive –ing
Finite = TNS, AGR
“tense marking”
In
Also includes: modal
On
auxiliaries, auxiliary DO
Plural “s”
and HAVE
Irregular past tense
Possessive “s”
Uncontractible copula
Articles
We have known for at least 20
years that children with
language impairments have
difficulty with only a subset of
Brown’s grammatical
morphemes..
Regular past tense
3rd person regular present
3rd person irregular
Uncontractible auxiliary
Contractible copula
Kamhi, A.G. (2014). Improving clinical
practices for children with language and
learning disorders. Language, Speech, and
Hearing Services in Schools, 45, 92-103
Contractible auxiliary
8
4/4/2016
Ash, A.C., & Redmond, S.M.(2014).
Using finiteness as a clinical marker
to identify language impairment.
Perspectives on Language Learning
and Education, 21, 148-158.
Pawlowska, M. (2014). Evaluation of
three proposed markers for
language impairment in English: A
meta-analysis of diagnostic
accuracy studies. Journal of Speech,
Language, and Hearing Research, 57,
2261–2273.
Adapting
clinical
markers into
screeners for LI
Redmond, S.M. (2005). Differentiating SLI
from ADHD using children’s sentence
recall and past tense morphology. Clinical
Linguistics and Phonetics, 19, 109-127.
 Proposed clinical markers of language impairment:
 Grammar (tense-marking)
 Verbal Memory (nonword repetition, sentence recall)
 Ongoing Limitations: more unselected samples needed,
independent blinded assessments needed
 Utah Verbal Screening Study (n = 1060)
 Ethnicity: 88.5% non-Hispanic, 10.7% Hispanic
 Race: 88.6% White; 4.1% Asian; 3.9% Black; Native American 1.4%; 0.8% Pacific
Islander
 Gender: 52% Male; 48% Female
 Screening Measures
 Sentence Recall (Redmond, 2005)
 Past Tense Marking (PT probe from Test of Early Grammatical Impairment)
TEGI Protocol: https://cldp.ku.edu/ricewexler-tegi
9
4/4/2016
Not Screened
124
Not Returned
1399
Flyers Sent
2907
Parental
Permission
1184
41%
Absenteeism,
Moved out of district,
Student declined = 9
89.5%
Screened
1060
Declined
313
Unmarked
11
Excluded
207
Bilingual, gifted/ELP,
siblings, missing service
information
80%
Screened
1060
75.8%
Passed
647
Included
853
Regular Ed. = 844
LD, CD, EBD, RD, Resource = 77
Pass/fail criteria for each age
based on regular ed means (SDs)
Failed
206
Cutoffs: “10%ile on one screener or
15%ile on both [PT, SR]”
Supplemental Cases:
Jordan District: 48
U of U Clinic: 3
Referred: 7
Passed
647
22%
Failed
206
25%
Confirmatory
Testing (SLC)
144
PASSED:
Confirmatory
Testing
165
Confirmatory
Testing (SLC)
52
FAILED:
Confirmatory
Testing
89
10
4/4/2016
 I. Inter-rater Consistency (n = 50):
#agreement/(#agreements/#disagreements) * 100
 Sentence Recall: 98.8% [r = .990, p <.001]
Reliability of
Screeners
 Past Tense: 93.5% [r = .921, p <.001]
 II. Test-retest Reliability (n =37): < 4 weeks [M = 1.03, SD = 1.0]
 Sentence Recall: r = .946, p < .001
 Past Tense: r = .905, p <.001
 III. Screening-to-Confirmatory Testing Stability (n = 96): < 20
weeks [M = 14.72, SD = 4.31]
 Sentence Recall: r = .906, p < .001
 Past Tense: r = .883, p <.001
Zero-order correlations among screeners and confirmatory measures (N = 254)
Validity of
Screeners
SR
PT
CELF-4
TEGI
SR
--
PT
.606***
--
CELF-4
.804***
.614***
--
TEGI
.695***
.905***
.694***
--
*p <.05, **p <.01, ***p < .001
Zero-order correlations among screeners and confirmatory measures (N = 254)
Validity of
Screeners
SR
PT
CELF-4
TEGI
SR
--
PT
.606***
--
CELF-4
.804***
.614***
--
TEGI
.695***
.905***
.694***
--
*p <.05, **p <.01, ***p < .001
11
4/4/2016
Zero-order correlations among screeners and demographic variables from those who participated in
confirmatory testing (N = 254)
SR
SR
PT
--
Gender
PT
Gender
Ethnicity
Race
--
.016
-.070
.091
--
.082
.606***
-.006
--
Ethnicity
Race
Age
-.052
-.032
.065
.122
Age
.274***
.220***
--
-.029
.108
-.007
--
*p <.05, **p <.01, ***p < .001
PASSED:
Confirmatory
Testing
165
FAILED:
Confirmatory
Testing
89
PASSED:
Confirmatory
Testing
165
FAILED:
Confirmatory
Testing
89
CORRECT
REJECTION:
156
*LI/NON-LI status determined by CELF-4
core language score <85*
MISS:
9
HIT:
59
FALSE
ALARM:
30
CORRECT
REJECTION:
156
Sensitivity = 87.87%
Specificity = 83.51%
PPV = 65.17%
NPV = 95.15%
*LI/NON-LI status determined by CELF-4
core language score <85*
MISS:
9
HIT:
59
FALSE
ALARM:
30
Atypical (LD,
ADHD, EBD,
Autism, head
trauma/seizur
es)
40%
Typical
Developing
60%
12
4/4/2016
 A. CELF 4 < 85
Screening for
Different
definitions of
LI
 (TL = 188, LI = 66)
 Sensitivity = 87.87%
 Specificity = 83.51%
 PPV = 65.17%, NPV = 95.15%
 C. “Receiving SLP services”
 (TL = 165, LI = 89)
 Sensitivity = 60.67%
 Specificity = 78.78%
 PPV = 60.67%, NPV= 78.78 %
 B. CELF 4 < 80
 (TL = 203, LI = 51)
 Sensitivity = 92.15%
 Specificity = 79.31%
 PPV = 52.18%, NPV = 97.58%
 D. “2 of the following: CELF4 <
85, TEGI below criteria, CCC-2 <
85, SLP services”
 (TL = 185, LI = 69)
 Sensitivity = 82.61%
 Specificity = 82.70%
 PPV = 60.04%, NPV = 92.72%
 Clinical markers of S(P)CD have yet to be proposed.
How should
we diagnose
S(P)CD?
 Rather there is a collection of pragmatic symptoms that are
homologous to symptoms historically associated with autism. The
S(P)CD designation is reserved for individuals who have these
symptoms but do not meet the full criteria for autism.
 Suggested protocol (after ADOS and ADI-R ruled out autism):




Children’s Communication Checklist-2 (use SDI)
Test of Narrative Language (esp. story grammar items)
Figurative language subtests
Observational protocol (e.g. Prutting & Kirchner, 1983)
 Is it possible to have a “social skills” deficit that isn’t language
based (e.g. impulsivity from ADHD, social anxiety/withdrawal)?
What language measures would differentiate these from S(P)CD?
 “The identification of a specific disorder when several diagnoses are
possible because of shared symptoms and signs” (McCauley, 2001).
Differential
Diagnosis
 Needed to address the threats of missed diagnosis, misdiagnosis,
diagnostic overshadowing as well as offset potential educational and
health disparities
 Needed for the clinical management of true comorbidity
 Needs clearly articulated diagnostic constructs
 Needs clinical markers and protocols that are capable of differentiating
pediatric language disorders from other conditions
13
4/4/2016
SLP Textbooks [14]: Angell (2009); Bernstein & Tigerman-Farber (2009);
Farber & Radziekicz (2008); Hedge (2006); Kaderavek (2011), Lahey
(1988); McCauley (2001); Nelson (2010); Owens (2014); Paul & Norbury
(2012); Pence & Justice (2008); Tomblin et al (2002); Yoder (1988)
8
A Neglected Topic in
SLP pre-professional
training…
7
Differential diagnosis is
not covered on ASHA’s
online portal/practice
issues
6
5
4
3
2
1
0
Not Mentioned
Less than a Paragraph
1-2 Paragraphs
> 2 Paragraphs
ADHD
Differentiating
Language
Disorder from
ADHD
 “Developmentally
inappropriate levels of
inattention, hyperactivity,
and impulsivity”
 DSM5: symptoms must be
present in non-academic
contexts.
ADHD
Parent-rated behavior,
Visuospatial Memory,
Speaking rate,
Vocal abuse,
Mazes
SLI
 “Diminished language
proficiencies in the absence
of significant limitations in
hearing acuity, cognitive
development, or social
development”
SLI
Reading,
Academics,
Teacher-rated behavior,
Peer Difficulties,
Pragmatics,
CPT,
EF
Tense-Marking,
Nonword Rep.,
Sentence Recall,
Verbal Memory,
Narratives
14
4/4/2016
References*
 Diagnostically Neutral Symptoms and Signs: Barkley (2006); Brock
& Knapp, (1996); Camarata & Gibson (1999); Cardy et al., (2010); Catts
et al., (2002); Charach et al., (2009); Conti-Ramsden & Botting (2004);
Finneran, Francis, & Leonard (2009); Henry, Messer, & Nash (2012);
Humphrey, Storch, & Gefken, 2007; Johnson et al., (2002); Knox &
Conti-Ramsden (2007); Milberger et al., (1995); Nigg et al., (2005);
Parriger (2012); Rabiner & Cole, 2000; Redmond (2011); Redmond &
Rice (1998); Redmond & Rice (2001); Riccio & Jemison (1998); Riccio,
Reynolds, & Lowe (2001); Reilly et al (1999); Spaulding, Plante, &
Farinella (2006); Wiener & Mak, 2009; Willcutt et al., (2005);
Willoughby & Blair (2011); Zelnik et al. (2012)
 Clinical Markers SLI: Archibald & Joanisse,(2009); Bishop, North, &
Donlan, (1996); Conti-Ramsden, (2003); Conti-Ramsden, Botting, &
Faragher (2001); Dollaghan & Campbell, (1998); Leonard (2014); Oram
et al., (1999); Palowska (2014); Parriger (2012); Rice, Wexler, & Cleave,
(1995); Rice, Wexler, & Hershberger (1998); Redmond, Thompson, &
Goldstein (2011);SLI Consortium (2002); Spaulding et al., 2006; TagerFlusberg & Cooper (1999)
 Clinical Markers ADHD: Barkley (2006); Biederman, Keenan, &
Farone (1990); Garcia-Real et al., (2013); Hamdan et al., (2009);
Martinussen et al., (2005); Merwood et al., (2013); Nigg (2006);
Redmond (2004); Redmond & Ash (2014); Rooderyns (2006)
*Redmond, S.M. (2016). “Language impairment in the Attention-Deficit/Hyperactivity Disorder context.” Journal of Speech
Language Hearing Research, 59, 133-142.
X (Other Conditions)
Disassociation
Table
Rice, M.L. (2016). Specific language
impairment, nonverbal IQ, attentiondeficit/hyperactivity disorder, autism
spectrum disorder, cochlear implants,
bilingualism and dialectical variants:
Defining the boundaries, clarifying
conditions, and sorting out causes.
Journal of Speech, Language, and Hearing
Research, 59 (1), 122-132
+
+
_
Typical & Above
“Spared Language”
TL w/ X
“Specific Language Impairment”
“Concomitant X+LI”
Language
_
 Sometimes referred to in the literature as “non-specific language impairment (NLI)”
Language
Impairment +
Low
Nonverbal IQ
 Standardized tests, narratives: NLI lower than SLI (Wetherell, Botting, & Conti-Ramsden,
2007; Tomblin & Nippold, 2014)
 Tense marking: NLI lower and slower than SLI (Rice, Tomblin, Hoffman, Richman, &
Marquis, 2004).
 Expository samples: NLI = SLI (Nippold, Mansfield, Billow, & Tomblin, 2008).
 NLI increased risk for socioemotional behavioral problems (e.g. ADHD): (Elbro et al.,
2009; Snowling et al, 2006)
 NLI increased risk for reading problems: (Bishop & Adams, 1990; Catts, Fey, Tomblin, &
Zhang, 2002).
 Children with NLI less responsive to reading interventions: (Bowyer-Crane et al., 2011)
 Children with NLI and SLI may respond differently to language intervention
strategies… (Goorhuis-Brouwer & Knijff, 2002)
 ….or maybe not (Bowyer-Crane et al., 2011; Cole, Dale, & Mills, 1990; Fey, Long, & Cleave,
1994). BUT: small study samples and moderate treatment effects limit synthesis.
 NLI = SLI + “spared language” with low cognition
 i.e. low nonverbal IQ and LI are additive comorbid disorders
15
4/4/2016
Language
Impairment +
Reading
Disability
 Tense marking, nonword repetition, standardized language
tests: SLI+RD = SLI (Bishop et al., 2009; Catts et al., 2005).
 Standardized reading tests: SLI+ RD = RD (Bishop et al., 2009; Catts
et al., 2005).
 SLI+ RD increased risk for teacher (not parent) reported behavior
problems (Tomblin, Zhang, Buckwalter, & Catts, 2000).
 SLI+RD = SLI.
 i.e. LI and RD are non-additive, non-interactive comorbid
disorders
Language
Impairment +
ADHD
 Standardized tests, tense marking, nonword repetition, sentence recall:
ADHD+LI = SLI (Redmond, Ash, & Hogan, 2015).
 Modest (but significant) positive correlations (were found between ADHD
symptoms and some of the language measures. ADHD is protective??
 Narratives: ADHD+LI = SLI (standardized scores on TNL)…..but children with
ADHD+LI may have fewer story grammar elements than SLI (Hamilton, Ruetschle,
Mong, Timler, & Redmond, 2015).
 Peer victimization: Elevated ADHD symptoms in SLI study samples associated with
higher rates of peer victimization (i.e. being bullied) (Redmond, 2011)
 Impact on Intervention: ADHD+LI more likely to receive SLP services than SLI
(Sciberras et al., 2014)
Paradis, Emmerzael, & Sorensen.
(2010). Assessment of English
Language Learners: Using Parent
Report on First Language
Development. Journal of
Communication Disorders, 43, 474497.
 Alberta Language and Development Questionnaire (ALDeQ)
 Parents compare their child to other ELL children they know
 Early language milestones and family history items in particular
differentiated Bi-SLI from Bi-TD
 http://www.linguistics.ualberta.ca/en/CHESL_Centre/~/media/ling
uistics/Media/CHESL/Documents/Paradis_ALDeQ_15Apr11.pdf
16
4/4/2016
 The common reality of mixed dominance in bilinguals linguistic
abilities challenges the traditional logic of “testing in the student’s
dominant language”.
BESA
 The Bilingual English-Spanish Assessment (BESA) protocol
accommodates for mixed dominance.
 Age range: 4:0 to 6;11
 Phonology, Morphosyntax (cloze, sentence recall) and Semantics
(receptive, expressive) Subtests in both English and Spanish
 Morphosyntax subtests are designed around each language’s common
LI symptoms
 Subtests standard scores based on child’s best performance (English or
Spanish)
 Parent/Teacher questionnaires of language input/output used to
determine if testing should be in Spanish, English, or both
 Administered by a bilingual examiner over 2-days (1 day for each
language – Spanish first)
 Accuracy against reference standard [language samples measures +
parent and teacher report + clinical observation] (Se = 92-96%, Sp =
85-92%).
European
COST Action
IS0804
 Measures of bilingual-SLI need to be able to accommodate a wide
variety of language combinations
 Comprehension: Exhaustive Wh-Questions: “who (all) is holding a
soccer ball; “who is giving what to whom”
 Elicitation: subject-verb agreement, case marking, object clitics
 Sentence repetition: negation, biclausal sentences, wh-questions,
relative clauses
 Nonword repetition: a common “quasi-universal” list of words (2-5
syllables in length) built from CV syllables from a limited range of
common phonemes (e.g. zibalita). Native informants select 16 from
the set that are not a real word in their language.
Carrizo &
Redmond
(under review)
 Could performance on the COST Action quasi-universal NWR
predict children’s performance on the BESA subtests as well as
their English, Spanish or Best NWR?
 What are the trade-offs involved in using percent whole word
correct (PWWC) rather than percent phonemes correct (PCC)?
 26 Spanish-English speaking bilingual children (5:0-6:11) (3 were
receiving SLP services).
17
4/4/2016
 Results:
 All of the NWR tasks were significantly and moderately-highly
correlated with BESA subtests and Language Index Score
Carrizo &
Redmond
(under review)
cont’d
 PPC r range: 0.66 to 0.87 (higher values with Morphosyntax)
 PWWC r range: 0.5 to .76 (higher values with morphosyntax)
 Linear regression PPC: Quasi-universal NWR accounted for 85%
of the variance on BESA Language Index Score
 Linear regression PWWC: Quasi-universal NWR and Best NWR
accounted for 83% of the variance on BESA Language Index Score
 i.e. PPC with one NWR task was as good as PWWC with three
(BEST NWR requires both Spanish and English to determine). This
could be used by monolingual SPLS to screen for Bi-LI
 Our evidence base on the diagnosis of LI remains modest relative
to what’s available for other neurodevelopmental disorders.
Working
Conclusions
 For now, grammatical and verbal memory measures represent our
strongest choices for the diagnosis of LI and its differential
diagnosis (other clinical conditions, bilingualism). More work is
needed to complete the picture across the life-span
 There are ongoing disconnects between cases of LI identified by
standardized, research-based protocols and cases identified in the
field by SLPs.
 S(P)CD is currently a “safety net” designation for cases which
don’t meet enough autism criteria but would benefit from
services. Evidence that S(P)CD can be reliably differentiated from
SLI, ADHD, Selective Mutism, neglect, PTSD, and other clinical
conditions is needed.
 Concomitant impairments impact LI in different ways:
Working
Conclusions
cont’d
 Low Nonverbal IQ and LI are additive disorders: lower language
levels and slower growth than SLI; may be less-responsive to some
kinds of therapy (mixed evidence)
 Reading Disorder and LI are non-additive, non-interactive
disorders: same levels and growth as SLI; increased risk for teacher
reported behavioral problems
 ADHD and LI are non-additive, non-interactive disorders: same
levels as SLI; impact on growth unknown; may benefit from
preferential access to SLP services; may be at increased risk for peer
difficulties
18
4/4/2016
Working
Conclusions
cont’d
 The largest obstacle to advancing the diagnostic integrity of LI maybe
ourselves. We (ASHA) need to do a better job of raising the public
visibility of “language impairments”.
 Terminological disputes (SLI vs. LI vs. LLD vs. LBLD vs. CD) compromise
the message and confuse parents.
 Educational/health disparities in receipt of LI services exist and will likely
persist.
 Individuals with LI and their families need to have a stronger voice in the
policies that affect them.
 We (ASHA) need to install mechanisms to keep track of rates of LI over
time and across different regions. Aggregates of students “receiving SL/CD services” (cf. ASHA and USOE) can’t be used to advance these
agendas.
 Updated, multi-site, epidemiological studies of LI are long overdue.
19
Download