Baseline Characteristics of MTA Subjects

The ADHD Explosion Part 1:
Causes, Models,
Rising Prevalence,
and Policy Implications
Stephen P. Hinshaw
University of California, Berkeley
Help Group Summit
10/17/14
ADHD: Key Themes
Newsworthy
 Cause of ADHD is SpongeBob Square Pants
 Cause of ADHD is starting kindergarten at age 4
 Stimulants lead to heart attacks
 New York Times 2012/2013 opinion pieces:
 Sroufe, Kureishi, Friedman, Brooks: Back to the past
Too much of the news and opinion is
mythical (see subtitle of book)
4
Facts
 ADHD is a neurodevelopmental disorder
with high genetic liability
 ADHD incurs huge costs to those with high
levels of symptoms
 All too few people with ADHD have
excellent life outcomes—if it’s a gift, in the
words of Ned Hallowell, it’s hard to unwrap
Myths
 Medications are poisons, destroying developing brains
 Meds help in 80% of cases
 May actually be neuroprotective for youth with ADHD
 Medication alone is a sufficient treatment
 Need family/school intervention for skill building
 SEE PART 2 TOMORROW!
 ADHD can be assessed and diagnosed in a 10’ office visit
 Yet this, far too often, is the national standard
 Results in both overdiagnosis and underdiagnosis
Impairment
 Academic (school failure)/Vocational
 $100 billion/year (youth) indirect costs (justice, sp. ed, SUD)
 $200 billion annually (adults) indirect costs (job problems)
 Social/peer (most peer-rejected condition)
 Family (reciprocal chains of bidirectional influences)
 Accidental injury (across the age span)
 Impairment often independent of comorbidity…AND key
comorbidities don’t respond optimally to ADHD tx
 E.g., LD, delinquency, depression
DSM-5 vs. RDoC
 DSM-5 changes:
 Neurodevelopmental disorder
 Types (Inattentive, HI, Combined) now ‘presentations’
 Adult examples of most symptoms
 Age of onset of impairing symptoms: < 12 years, not < 7
 **Each successive edition of DSM has loosened criteria
somewhat, which is one reason for “ADHD explosion”
 Research Domains Criteria
 Dimensional, multiple levels (genes to culture)
 Search for underlying mechanisms
ADHD Cross Culturally
 Appears in nearly all cultures (that feature
compulsory education)
 Polanczyk et al. (2007), AJP:
 Diagnostic prevalence strikingly similar across world regions: 5%
 Disparities linked to dx practices (ICD vs. DSM; informants; etc
 Hinshaw et al. (2011)
 Within-country variation high in many nations
 However, treatments and systems of care vary radically across
regions and cultures
 MANY NATIONS ‘CATCHING UP’ WITH U.S. MEDICATION TRENDS
 But some not: politics, history, penetration of Big Pharma
Nature of ADHD: Models
 “Cognitive” models: Attention deficit, EF
 “Inhibitory” models: Barkley (1997)
 “Motivation” models: Reward undersensitivity
 E.g., Volkow et al. (2009): large medication-naïve adult
sample, PET scans of transporters and receptors
(Motivation)
(Attention)
(Motivation)
(Attention)
Transporter PET Image
(Motivation)
(Attention)
Combination Models
Sonuga-Barke et al. (2010):
 Top-down executive control
 Bottom-up delay aversion
 Time management
ADHD clearly implicates multiple brain
regions and paths for different facets of
symptomatology
Neural profiles
 Structural/anatomical:
 Overall lowered cerebral volume; caudate, cerebellum…
 Key research: Shaw et al. (2006, 2007, 2009, 2012)
 Delayed patterns of cortical thickening/thinning in ADHD vs.
comparison samples, longitudinally
 Roughly 3 year delay for ADHD groups: Immaturity come to life
 Immaturity persists; thickness correlated with symptoms
 Functional: Frontal-striatal paths
 Until recently: must ‘scan’ during active cognitive performance
 Default mode: reliable differences when S’s not ‘doing
anything’; more ‘intrusions’ into task performance in ADHD
ADHD: Causes
Heritability and Genes:
H2 of ADHD near .8
**What is heritability?
‘genetic liability,’ but not inevitability
 Too often, assumption is that ADHD is ‘fixed’
and largely immutable
PKU example
Height example
IQ example
Which genes?
 Seemed a simple question 10-15 years ago: Genes
related to dopamine systems and pathways in brain
 But any single gene variant explains only a tiny
fraction of “ADHD-ness”
 ‘Dark matter’ of genetics: missing heritability!
 Recent discoveries: genes conferring risk for
ADHD are SAME as those conferring risk for
schizophrenia, mood disorders, and autism
 MUST BE that early influences are epigenetic
Other Risk Factors
 Low birthweight
Predicts ADHD, LD, Tourette’s, CP, retardation
 Teratogenic effects
FAE: Many are nearly identical to ADHD symptoms
Smoking/nicotine: genetic mediation, too
 Early parenting: No consistent evidence as causal
Middle-class; few prospective studies from early years
 Insecure attachment?
Does NOT predict later ADHD, independent of comorbid
aggression
Risk Factors: Equifinality
 Carlson et al. (1995):
In low-income sample, early maternal insensitivity
predictive of ADHD symptoms to a greater extent than
early temperament
Need genetically informative design
 Institutional deprivation (Kreppner et al., 2001)
English and Romanian Adoptive Study Team:
Inattention/overactivity associated with length of severe
institutional deprivation in first 4 years
Specific effect: Conduct problems and internalizing
symptoms not similarly associated with deprivation
Yet, different “feel” from typical ADHD presentation
AND, EF deficits may be distinct from ‘typical’ ADHD
presentation
 Hence, equifinality apparent
Role of Parenting
 Maintaining cause, if not primary cause
 Parents tend to fight fire with fire
 Coercive discipline (too lax, too harsh)
 Cycles of dysregulated emotion
 Given heritability of ADHD, parents likely to have
ADHD symptoms themselves
 Parent management: PART 2, TOMORROW!
Important New Findings
Harold et al. (2013a, 2013b)
 Adoption study in UK
 Controls for biological relatedness
 Even in adoptive families, kids’ levels of ADHD
elicit overcontrolling parenting from parents
 AND, levels of harshness predict further ADHD
symptoms, over time
 It’s not all in the genes!
Ultimate cause?
 The “real” cause of ADHD has to be
compulsory education (same as for LD)
 Certainly, ‘attention’ or ‘impulse control’ genes
have been around for the history of our species,
but extremes not salient until we made children
sit and learn to read
 If it’s true that achievement pressure “reveals”
ADHD, is it also true that current high rates of
pressure are fueling the recent explosion?
Developmental Paths
 Infancy/temperament:
 Activity level vs. effortful control
 Preschool Manifestations (S. Campbell)
 Careful evaluations of 3 and 4 year olds
 See AAP Guidelines (2011)
 Prospective predictions to mid-late childhood:
 PPP = .5! Hence, multifinality apparent
 That is, suggestions of (a) “he’ll grow out of it” and (b)
“medicate today” are each fraught with error
 Predictors of continuation:
 (a) severity of early ADHD
 (b) negativity of early parent/child interaction, controlling for
severity of child’s ADHD
Parenting Influences on Positive Peer Status
Hinshaw, Zupan, et al. (1997)
 Aim: Predict peer acceptance from parenting
 Ideas About Parenting (Heming et al., 1989)
 3 factors = Authoritarian, Authoritative, Permissive
 Authoritative Factor: 15 items
 Warmth, Limits, Autonomy Encouragement--e.g.,
 “I
encourage my child to be independent of me”
“I expect a great deal of my child”
“I have clear, definite ideas about childrearing”
 “Raising a child is more pleasure than work”
“When I am angry with my child, I let him know”
“I reason with my child regarding misbehavior”
Results
 Mothers of ADHD boys: lower on Authoritative
 ES = .75
 Yet variance in ADHD group equivalent to comparisons
 Tested predictive power of parenting factors,
observed overt and covert behavior, and
internalizing score (CDI, observed withdrawal) via
hierarchical regressions
 Neither Authoritarian nor Permissive beliefs predicted peer
nominations, but Authoritative beliefs did so (beta = .3),
even with diagnostic group controlled
Explained Variance in Positive Nominations
0.14
0.12
0.1
0.08
0.06
0.04
0.02
0
-Overt
-Covert
-Intern
Mom A-R
Moderation and Implications
Prediction applies only to ADHD group
(beta = .30); for comparisons, beta = .00.
Key theme: “firm yet affirming” parenting
style
Sex Differences/Female Presentation
More in Part 2, tomorrow
 Another myth: ADHD effects only boys!
 Our sample (BGALS):
 Largest in existence of preadolescent girls with ADHD
(140, with 88 matched comparison girls)
 Baseline: marked impairments across symptoms,
impairments, neuropsych measures
 Impairments maintained at 5-year follow-up
 11/11 domains, with widening gap in math
 Sources: Hinshaw (2002); Hinshaw et al. (2006), Journal
of Consulting and Clinical Psychology

10-year follow-up
 95% retention rate (vs. 92% at 5 year)
 How? Facebook, relentless staff
 Despite ‘losing’ ADHD status majority of time, impairments
maintain in academics, comorbidities, social functioning.
 Yet, self-harm findings: Different adolescent path for girls??
Suicide attempts: 22% ADHD-C
NSSI: 51% ADHD-C 29% ADHD-I
8% ADHD-I
6% comparisons
19% comparisons
BGALS Follow-up: Self-harm
10-year follow-up (M age = 20)
Hinshaw et al. (2012), Journal of Consulting and Clinical Psychology
Conclusions
 ADHD not a static “entity”
 Different pathways lead to ADHD: Equifinality
 Differential outcomes from early ADHD symptoms:
Multifinality
 What predicts, moderates, mediates differential outcomes?
 Peer deficits and social skills; EF deficits; Motivation
 Developmental, contextual factors crucial
 Parenting styles, which may not be causal, are important
determinants of outcome, even for a condition with h2 = .7/.8
 Systems, health-care, legislative, cultural, stigma-related factors
related to underutilization and disparities in care
Assessment
Full coverage requires a day-long workshop
 Brief visit: false positives and false negatives
 Must get informant ratings, for kids, teens, or adults
 Brief/narrow vs. broader scales
 Ideal to get info from past as well as present teacher
 Must get full developmental history
 Must appraise rule-out and comorbid conditions
 LD, Anxiety, Depression, etc. require different interventions
Tidal Wave/ADHD Explosion
National Survey of Children’s Health (Visser et al., 2013)
Parent-reported ADHD ‘ever diagnosed’
 For all 4-17 year olds in U.S.:

2003: 7.8%

> 40% INCREASE IN 9 YEARS!
2007: 9.5%
2012: 11.0%
 Low income rates now = middle class; Black = White

Hispanic lower (but fast growing)
Medication higher, too:
 Just under 70% of those ‘currently diagnosed ‘now receive medication
 From other sources: Largest medication increases: adolescents, adults
Earlier Explosions: 1990s
 Policy shifts:
 IDEA: ADHD as OHI
 Medicaid: authorizes ADHD
 SSI: ADHD (with other impairment) can qualify
 Late 1990s: FDA changes regs on DTC ads
 2000: Concerta (first effective long-acting form)
 More and more LBW babies survive
Huge Regional Variation Now
 Rise across entire nation, but major-league
state-by-state variation, too
 2011-12:
 Arkansas now #1, Indiana #2, NC #3
NC had been #1 in 2007
 Medication trends similar to 2007, but
slightly higher overall
37
38
What does not explain variation
Demographics
Hispanic population clearly higher in California, and
traditionally the lowest rates of diagnosis
Eliminated a little of the CA-NC difference but not most
**Hispanic rates growing FAST, esp. in California
Rates of health-care providers
Explains other disorders, but not here
State “culture”
 May explain regional differences within state -- but
not state-by-state per se
**Consequential accountability
 1970s-80s: public school reforms “input focused”
 Reduce class size, pay teachers more, etc.
 Results not consistent; shift in 1990s to “output focused”
 I.e., incentivize test score improvements per se
 Consequential accountability—districts get ‘noted’ or even
cut off from funds, unless test scores go up
 30 states implement such laws <2000
 Then, becomes law of the land for all states with No Child
Left Behind (takes effect 2002-3)
Consequential accountability laws prior to NCLB (but
not psychotropic medication laws): In the South
Region
Northeast
Midwest
South
West
United States
Number of
States
9
12
17
13
51
Consequential
Accountability
before NCLB
5
5
15
5
30
Psychotropic
High School Medication
Exit Exam
Law
4
2
3
2
13
5
8
5
28
14
Sources: Investigators' Research, Dee & Jacob 2011, Dee & Jacob 2006, and Center for
41
Education Policy
Findings
From “triple difference” model
Between 2003-2007, in the 20 “NCLB states,”
poorest children showed huge increases in ADHD Dx:
In these states, 59% increase in ADHD dx for kids within
200% of FPL
vs. only 8% in middle- or upper-class kids
Nothing like that in states with previous consequential
accountability (all kids in those states went up 20% or so)
Nothing like that in private schools
This trend reverses by 2012, with Obama’s dismantling of
NCLB
Consequential accountability introduced via NCLB was associated
with higher ADHD diagnostic prevalence increases among lowincome children aged 8-13 from 2003-2007, but there was no
association from 2007-2011 (unadjusted results)
16%
NCLB Consequential
Accountability State,
Income < 200% FPL
Pre-NCLB Consequential
Accountability State,
Income <200% FPL
All Children
15%
14%
13%
12%
11%
10%
9%
8%
2003
2007
2011
Pre-NCLB Consequential
Accountability State,
Income ≥ 200% FPL
NCLB Consequential
Accountability State,
Income ≥ 200% FPL
District of Columbia is included within the 21 No Child Left Behind
consequential accountability states.
NCLB: No Child Left Behind; FPL: Federal poverty level
N=24,982 (2003), 22,467 (2007), 24,426 (2011)
Sources: 2003, 2007, and 2011 National Survey of Children’s Health
“Unintended effect”
Accountability laws encourage ADHD diagnosis for at least
two reasons:
#1: Diagnosis may lead to treatment, which may help boost achievement
test scores
 Scheffler et al. (2009), Zoega et al. (2012)
#2: In some states/districts, diagnosed youth are excluded from
the district’s average test score!
 Gaming the system, although NCLB eventually outlaws this
Why poorest kids? NCLB targets Title I schools
Psychotropic medication laws
 In 2001, Connecticut passed a law ‘pushing back’
against rising ADHD medication use in students
 By now, 14 states have passed such “psychotropic medication
laws,” of one or more of 3 types:
 Schools are prohibited from recommending meds
 Schools cannot require meds as a condition of enrollment
 Parental refusal to medicate the child cannot, in and of itself,
be considered neglect
 IN THESE STATES, NO RISE IN ADHD DIAGNOSES
FROM 2003-2012, VS. > 50% RISE IN OTHER STATES
46
Findings
In the 14 states with these laws, essentially
no change in ADHD diagnostic prevalence
between 2003 and 2011, versus a > 50%
increase in other states!
Where have we been?
 ADHD requires multi-level thinking
 Genes matter
 Families matter
 Cultural values placed on performance matter
 Educational policies matter
 Pharma matters
 ADHD is too important and too impairing to think
about it reductionistically
 When kids, learning, schools, productivity, and
medicating young minds are in play, stakes are high
Diversion (Part 2 tomorrow)
 Define: non-prescription use
 Rates extremely high (why??)
 How effective are stimulants as ‘neuroenhancers’
for general population?
 Smith & Farah (2011), Psychological Bulletin
 Ilieva et al. (2013), Neuropharmacology
 Rates of abuse/addiction: Policy implications
Thanks…
 NIMH and NIDA grants
 Robert Wood Johnson Policy Investigator Award
 Participants in many studies
 The Help Group
 You, the audience