Understanding “the whirling ball of comorbidity”: Reading Disability and ADHD Erik Willcutt, Ph.D. University of Colorado, Boulder
Dec 18, 2015
Understanding “the whirling ball of comorbidity”:Reading Disability and ADHD
Erik Willcutt, Ph.D.University of Colorado, Boulder
Funding Colleagues in Training Collaborators at CU and other institutionsNIH Research Grants (PI) Graduate Students M. Banich E. Arnold J. Keenan
R01 MH 62120 (Willcutt) H. Barnard (DU) A. Bryan P. Asherson H. Kraemer
R01 MH 63941 (Willcutt) C. Bidwell G. CareyT.
Baneschewski J. Kuntsi
R01 HD 47264 (Willcutt) K. Brodsky N. Chhabildas J. Biederman B. Lahey
P50 HD 27802 (Olson) L. Duncan R. Corley B. Blachman J. Loney
R01 HD 38526 (Olson) R. Gaffney-Brown T. Curran B. Byrne K. McBurnett
R01 MH 63207 (Hewitt) C. Hartman J. DeFries L. Cardon R. Nicolson
R01 DC05190 (Wadsworth) K. Mackiewicz M. Ehringer C. Carlson J. Nigg
R01 MH 70037 (Banich) L. McGrath (DU) N. Friedman S. Cherny D. PaulsR01 MH 66115 (Ross) G. McHaffie J. Hewitt D. Davalos W. Pelham
L. Santerre-Lemmon (DU) C. Hopfer A. DePrince B. Pennington
NIH Training Grants J. Schacht K. Hutchison A. Doyle R. Plomin
T32 MH 16880 (Hewitt) M. Shanahan (DU) T. Ito J. Epstein S. Samuelsson
F31 MH 78514 (Bidwell) K. Krauter S. Faraone C. Schatschneider
F31 DC 008726 (McGrath) Postdoctoral Fellow D. Miklowitz J. Fletcher L. Seidman
F31 MH 77482 (Shanahan) R. Betjemann A. Miyake C. Francks J. Sergeant
Y. Munakata M. Frank S. ShirkNIH Project Officers Project Coordinators R. Olson M. Friedman S. SmalleyP. McArdle, B. Miller T. DiLeo R. O’Reilly J. Gayan S. SmithJ. Rumsey, F. Tuma S. Hitt-Laustsen S. H. Rhee J. Gruen E. Sonuga-Barke
J. Keith R. Ross C. Hartung J. Stevenson
Other Funding W. Legg A. Smolen D. Hay Jim Swanson
Ackerman Foundation M. Stallings W. Heller R. Todd
Australian Res. Counc. S. Wadsworth S. Hinshaw R. Tannock
University of Colorado S. Young P. Jensen I. Waldman
“That’s it! Just me, alone at home, trying to be a good parent for my little whirling ball of comorbidity”
Too active
Anxious
Learning
Aggressive Comorbidity
• What is comorbidity? Answer: Two or more disorders in the same individual.
• Why does comorbidity matter?
• What causes comorbidity between reading disability and ADHD?
• What are the implications of comorbidity for assessment and treatment?
• Suggestions for the future
Questions we will answer today
Why should we care about comorbidity?
Comorbidity is the rule, not the exception
0
10
20
30
40
50
60
70
80
90
100
AcademicImpairment
Retained in school Social Impairment Occupationalimpairment
Justice System
Pe
rce
nt
Comorbidity is associated with greater initial impairment and more extensive and severe negative outcomes
Control ADHD RD+ADHDRD
Comorbidity has important implications for treatment
RD ADHD
ADHD RD
Shared risk factor RD + ADHD
Our ongoing studies most relevant to comorbiditybetween RD and ADHD
• Colorado Learning Disabilities Research Center (Director: Olson)
- 8 - 18 year old twins screened through schools• RD only (N = 400), ADHD only (N = 275), RD + ADHD (N = 125)
• Comparison group without RD or ADHD (N = 1,000)
- extensive initial testing and five-year follow-up assessment
• International Twin Study of Early Reading Development (PI: Olson)
- unselected sample of approximately 2,000 twins in US, Australia, Scand.- Tested yearly from preschool until 4th grade
• Validity of ADHD in adults (PI: Willcutt, Banich)
- screening: 4,000 undergraduates- individual testing: 200 students with ADHD, 100 without ADHD
Defining RD and ADHD
• Reading Disability- unexpected difficulty learning to read.- not explained by environmental deprivation, inadequate
education, or low cognitive ability.- 1.5 SD below the estimated population mean on a test of single-
word decoding (PIAT or WJ-III).
• ADHD- inattentive and / or hyperactivity-impulsive behaviors that are
inconsistent with an individual’s developmental level.- Parent interview (DICA-IV) and teacher ratings (DSM-IV)- Similar results for DSM-IV subtypes, so pooled for this
presentation.
What causes comorbidity?
more than 25 different explanations have been proposed(at last count)
Competing Explanations for Comorbidity(e.g., Caron & Rutter, 1991; Neale & Kendler, 1995)
Clinical Sample Bias: Comorbidity is only present in clinical samples.
Comorbidity is present in our community samples
5%8%
3%
Tests of the Competing Hypotheses
Artifactual Hypothesis
Clinical Sample Bias
Results
Not supported
Competing Explanations for Comorbidity(e.g., Caron & Rutter, 1991; Neale & Kendler, 1995)
Clinical Sample Bias: Comorbidity is only present in clinical samples.
Method bias: Comorbidity occurs because both disorders are assessed with the same measure.
Tests of the Competing Hypotheses
Artifactual Hypothesis
Clinical Sample Bias
Method Bias
Results
Not supported
Not supported
Artifactual Hypotheses
Competing Explanations for Comorbidity(e.g., Caron & Rutter, 1991; Neale & Kendler, 1995)
Clinical Sample Bias: Comorbidity is only present in clinical samples.
Method bias: Comorbidity occurs because both disorders are assessed with the same measure.
Rater bias: raters are more likely to endorse symptoms of a second disorder if the child has the first disorder.
Secondary symptom (phenocopy) hypothesis: Disorder #1 causes an individual to exhibit the symptoms of disorder #2 when they do not actually have the disorder.
Attention Problems in Individuals With and Without RD(Willcutt, Chhabildas, & Pennington, 1998; Willcutt et al., under review)
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
Mother Father Teacher Self-Report
Z-s
core
RD
Control
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
ParentChildhood
ParentCurrent
SelfChildhood
Self Current
Z-s
core
RD
Control
8 - 18 year old twins Undergraduates
Tests of the Competing Hypotheses
Artifactual Hypothesis
Clinical Sample Bias
Method Bias
Rater Bias
Secondary Symptom
Results
Not supported
Not supported
Not supported
Not supported
Competing Explanations for Comorbidity(e.g., Caron & Rutter, 1991; Neale & Kendler, 1995)
Clinical Sample Bias: Comorbidity is only present in clinical samples.
Method bias: Comorbidity occurs only because both disorders are assessed with the same measure.
Rater bias: raters are more likely to endorse symptoms of a second disorder if the child has the first disorder.
Secondary symptom (phenocopy) hypothesis: Disorder #1 causes an individual to exhibit the symptoms of disorder #2 when they do not actually have the disorder.
Causal Hypothesis: Disorder #1 directly causes disorder #2.
A plausible causal model
ADHDWeak phonological
developmentRD
Poor attention to instruction
Tests of the Competing Hypotheses
Artifactual Hypothesis
Clinical Sample Bias
Method Bias
Rater Bias
Secondary Symptom
“True Comorbidity” Models
Causal Hypothesis
Result
Not supported
Not supported
Not supported
Not supported
More work is needed
Competing Explanations for Comorbidity(e.g., Caron & Rutter, 1991; Neale & Kendler, 1995)
Clinical Sample Bias: Comorbidity is only present in clinical samples.
Method bias: Comorbidity occurs only because both disorders are assessed with the same measure.
Rater bias: raters are more likely to endorse symptoms of a second disorder if the child has the first disorder.
Secondary symptom (phenocopy) hypothesis: Disorder #1 causes an individual to exhibit the symptoms of disorder #2 when they do not actually have the disorder.
Causal Hypothesis: Disorder #1 directly causes disorder #2.
Common Etiology Hypothesis: The two disorders co-occur due to shared risk factors.
Causes of RD and ADHD
• Family studies-RD and ADHD are each significantly familial (e.g., Wadsworth et al., 2000;
Willcutt et al., 2000)
-RD and ADHD run in the same families (Friedman et al., 2003)
Causes of RD and ADHD• Family studies
-RD and ADHD are each significantly familial (e.g., Wadsworth et al., 2000; Willcutt et al., 2000)
-RD and ADHD run in the same families (Friedman et al., 2003)
• Twin Studies: disentangle three causes of individual differences in reading or ADHD-Heritability: genetic influences-Shared environment: environmental factors that affect both twins in a similar way.-Nonshared environment: environmental factors that are specific to the child with the disorder.
Twin studies of individual differences in ADHD symptoms(N > 15,000 twin pairs)
Twin studies of individual differences in reading
Comorbidity of RD and ADHD is explained by shared genes(Trszenlewski et al., 2006; Willcutt et al., 2000, 2003, in press a and b)
Genetic and environmental risk factors specific to
RD
RD ADHD
Genetic and environmental risk factors specific to
ADHD
Shared Genes
Causes of RD and ADHD• Family studies
-RD and ADHD are each significantly familial (e.g., Wadsworth et al., 2000; Willcutt et al., 2000)
-RD and ADHD run in the same families (Friedman et al., 2003)
• Twin Studies-ADHD: highly heritable, minimal shared environment-RD: highly heritable + shared environment-The association between reading deficits and inattention is due primarily to common genetic
influences (e.g., Willcutt et al., 2007)
-other genetic and environmental influences are unique to each disorder (Willcutt et al., 2007a, 2007b)
Tests of the Competing Hypotheses
Artifactual Hypothesis
Clinical Sample Bias
Method Bias
Rater Bias
Secondary Symptom
“True Comorbidity” Models
Causal Hypothesis
Common Etiology
Result
Not supported
Not supported
Not supported
Not supported
More work is needed
Supported:
Shared genetic risk factors
Sounds good, but one final detail...
Where are these shared genes and what do they do?
ADHD
G1
Cog 1
The “old school” model:
One Gene, One Deficit, One Disorder
RD
G2
Cog 2
Autism
G3
Cog 3
Schiz.
G4
Cog 4
• Dopamine genes– Dopamine transporter– Dopamine D2, D3, D4, D5 receptors– Dopamine-beta-hydroxylase
• Serotonin genes– Serotonin Transporter– Serotonin receptor 1A, 2B
• Other neurotransmitters– Monoamine oxidase A (Jiang et al., 2000)
• Genes whose exact location and function is still unknown– Chromosomes 2, 3, 4, 5, 11, 15, 16, 17
A Subset of the Genes Associated with ADHD
(send me an email if you would like the full list)
Causes of RD and ADHD• Family studies
-RD and ADHD are each significantly familial (e.g., Wadsworth et al., 2000; Willcutt et al., 2000)
-RD and ADHD run in the same families (Friedman et al., 2003)
• Twin Studies-ADHD: highly heritable, minimal shared environment-RD: highly heritable + shared environment-The association between reading deficits and inattention is due primarily to common genetic influences (e.g., Willcutt et al., 2007)
-other genetic and environmental influences are unique to each disorder (Willcutt et al., 2007a, 2007b)
• Molecular genetics-RD and ADHD are each influenced by multiple genetic and environmental risk factors-Each of these risk factors accounts for a small amount of the variance in reading and/or ADHD symptoms
ADHDRD
A simplified hypothetical model of the genetic and environmental etiology of RD and ADHD
(see Shanahan et al., 2006; Willcutt et al., 2005b)
G E E
ADHD
G G EG G G G E G
RD
A large pool of genetic and environmental risk factors influence RD and ADHD
...
ADHD
G GG G
RD
Shared genetic influences increase risk for both disorders, sometimes leading to comorbidity
G E E
ADHD
G G EG G G G E G
RD
Additional genetic and environmental risk factors are specific to RD or specific to ADHD
How do these causal factors influence brain functioning?
G G E
ADHD
Cognitive Deficit 1
G G E
Cognitive Deficit 3
G G G
Cognitive Deficit 2
G E E
Cognitive Deficit 4
RD
What cognitive weaknesses are intermediate between the genes and the behavioral symptoms?
(see Shanahan et al., 2006; Willcutt et al., 2005b)
Four Key Cognitive Processes
Phonological processing: recognize and manipulate the phonemic constituents in speech.- say “plig” without the “l”
Four Key Cognitive Processes
Phonological processing: recognize and manipulate the phonemic constituents in speech.- say “plig” without the “l”
Working memory: retain information in short-term memory and manipulate that information.- repeat a series of digits in reverse order.
Four Key Cognitive Processes
Phonological processing: recognize and manipulate the phonemic constituents in speech.- say “plig” without the “l”
Working memory: retain information in short-term memory and manipulate that information.- repeat a series of digits in reverse order.
Response inhibition: suppress a response when it is not correct / appropriate.
XXXXX
XXXXX
XXXXX
XXXXX
XXXXX
XXXXX
XXXXX
RED
Four Key Cognitive Processes
Phonological processing: recognize and manipulate the phonemic constituents in speech.- say “plig” without the “l”
Working memory: retain information in short-term memory and manipulate that information.- repeat a series of digits in reverse order.
Response inhibition: suppress a response when it is not correct / appropriate.
Processing Speed: process information rapidly and correctly.- name colors or common objects- rapidly process nonverbal material
G G E
ADHD
G G GG G G G E E
RD
What did we find?(Shanahan et al., 2006; Willcutt et al., 2005a, in press)
G G E
ADHD
G G GG G G G E E
RD
Weak phoneme awareness is specific to RD
PhonemeAwareness
G G E
ADHD
G G GG G G G E E
RD
Slow processing speed and weak working memory are associated with both RD and ADHD
PhonemeAwareness
ProcessingSpeed
Working Memory
G G E
ADHD
G G GG G G G E E
RD
Weakness in Response inhibition is specific to ADHD
PhonemeAwareness
ProcessingSpeed
Working Memory
Response Inhibition
Conclusions and Future Directions of our Group• RD and ADHD are frequently comorbid
– Future: longitudinal causal models
• Comorbidity is associated with multiple negative outcomes- Future: additional adult outcomes (Sally Wadsworth)
• Common genetic influences lead to comorbidity between RD and ADHD
• RD and ADHD are influenced by multiple genes, none of which is a necessary or sufficient cause- Future: What are these genes, what is their function, and how do they interact?
• These common genes may lead to slow processing speed or working memory difficulties.- Future: brain imaging
Clinical Implications and Practical Suggestions• When assessing RD or ADHD, clinicians should also screen
carefully for the other disorder– Quick ADHD screeners: Barkley, Conners, DuPaul, SNAP– More general psychopathology: Achenbach, BASC, Conners
• When RD and ADHD co-occur, both disorders are likely to require treatment:
• Needed: studies that test if comorbidity matters for treatment!– Add ADHD screener to ongoing studies
– Re-analyses of treatment studies that also measured ADHD
RD ADHD
ADHD RD
Shared risk factor RD + ADHD