Agenda | Abstracts |
Speakers | What is Public
Health? | Research |
Conclusions
Epidemiologic
Research on ADHD: What We Know
and What We Need to Learn
Peter
S. Jensen, M.D., New
York State Psychiatric Institute
Attention
Deficit Hyperactivity Disorder, or ADHD, is the most common behavioral
disorder of childhood and is generally thought to account for the lion’s
share of referrals for mental health treatment (Richters et al., 1995).
Despite the continuing ongoing controversies that have surrounded the
syndrome, it is the most widely studied of all of the behavioral/emotional
disorders affecting children and adolescents, having been the focus of
literally thousands of studies since the mid 1960’s. In its most current
diagnostic iteration (DSM-IV, American Psychiatric Association, 1994), its
prevalence has been variously estimated as ranging from 3 to 18%,
depending upon the specific methods one employs in implementing the
diagnostic criteria (e.g., face-to face diagnostic interviews vs.
checklists, single vs. multiple informants, etc.). While its exact
diagnostic criteria have changed over the last 3 decades, the general
symptoms thought to characterize the syndrome have remained largely
unchanged. Studies that have examined the relations between previous
diagnostic categorizations (e.g., DSM-III, DSM-IIIR) and DSM-IV indicate a
high degree of overlap of previous and current definitions (Lahey et al.,
1994).
Despite
lingering controversies, the many research studies conducted to date have
enabled a substantial body of information to be amassed. As a result of
the continuing controversies, the NIH (via the Office of Medical
Applications of Research, NIMH, and NIDA) sponsored a consensus conference
in November, 1998. The goal of this conference was to identify those areas
where evidence was such that definitive conclusions could be drawn, as
well as those areas where more research was needed. Of note, several of
the areas of greatest need for more research are those to be addressed in
this presentation, namely:
- validity of the syndrome itself, particularly with regard to whether
it constitutes the far end of a normal spectrum versus a qualitatively
different and distinct condition;
-
etiology
-
prevention
-
issues of over- vs. under-diagnosis in "real world" settings
-
factors underpinning differences in treated prevalence; and diagnostic
and treatment practices
Validity
of the syndrome:
Current evidence indicates that ADHD can be reliably diagnosed
under research conditions (Jensen et al., 1995; Schwab-Stone et al., 1996;
Shaffer et al., 1996). In fact, ADHD as a condition is unsurpassed among
other child psychiatric disorders, and consistently demonstrates the
highest levels of test-retest reliability coefficients. Validity
coefficients are also consistently high when done under ideal conditions
(Jensen et al., in press). When applying the classic validational criteria
outlined by Robins and Guze (1970), the weight of evidence indicates that
ADHD does indeed meet criteria as a valid behavioral syndrome that has
reasonably discrete defining characteristics and can be reliably
diagnosed. As noted by the consensus conference final statement, there is
substantial evidence for the validity of the disorder. However, like
hypertension, ADHD likely has multiple causes, and may instead reflect a
final common pathway of behavioral manifestations rather than a single
disorder. As noted by the consensus conference, further validational work
on the syndrome is needed, including determining whether the syndrome is
best characterized along dimensional lines or as a discrete category. This
determination is likely to be of great importance, since if there is
indeed no natural cut-off, additional efforts to identify subthreshold
cases or prodromal presentations would be quite important. Such efforts
could yield substantial benefits, ranging from identifying persons at
risk, to identifying a "broader" phenotype that could provide
greater scientific benefits for genetic studies, to developing preventive
interventions.
Prevalence:
Given the fact that there appears indeed to be no natural cut-off or
discontinuous function that characterizes ADHD, it should not be
surprising that wide differences in prevalence have been noted. Across
cultures, prevalence rates ranging from 1 to 20% have been reported (Bird
et al., 1999). These rates vary as a function of the diagnostic criteria,
methods for establishing the presence of symptoms, nature of the
informants, and degree of impairment (Bird, 1999). Within the United
States, prevalence rates of 4-5% are generally accepted. In Europe,
application of ICD-10 criteria ("Hyperkinetic Disorder") are
substantially more demanding, and tend to yield substantially lower rates,
usually in the range of 1% (Swanson et al., 1997). Most evidence to date
suggests that children are not particularly good reporters of the full
syndrome. Inclusion of children’s reports tends to raise prevalence
rates by 25-33% (Jensen et al., in press). Similarly, counting cases
regardless of the presence of impairment, or without requiring the
presence of symptoms in multiple settings, can raise rates by as much as
50%. Regardless, these variations indicate the need not just for standard
definitions (as in DSM-IV), but also uniform measures and methods for
instantiation of the disorder.
Evidence
to date indicates a 6:1 to 9:1 ratio of males to females in clinical
settings. However, in epidemiology studies, ratios drop to 3:1. Further,
by adulthood, prevalence rates appear to be equal (Barkley, 1999;
Johnston, 1999). There is no significant or sizable body of evidence
concerning possible differences as a function of ethnicity or SES.
However, minority and lower SES families are less likely to seek out care
and obtain treatment for their children (Zito et al., 1998). Further, SES
factors may complicate the treatment of the disorder, and/or increase the
likelihood of poorer long-term outcomes (MTA Cooperative Group, in press;
Biederman et al., 1995).
Etiology:
Current evidence implicates multiple factors in the etiology of ADHD, yet
it is unclear how these factors interact at the level of the single case.
The best studied factor is familiality/heritability. Depending upon the
nature of the study, heritability has been estimated to be between 50-80%.
Yet these factors do not appear to be "all or none," instead,
specific traits of ADHD appears to be heritable in a continuous function
(Levy et al., 1997).
Most
recently, evidence indicates that ADHD symptoms have a central nervous
system basis (as do all normal and abnormal behaviors, thoughts, and
emotions). However, such brain-behavior correlations do not constitute
proof that ADHD reflects a disordered biologic state. Nonetheless,
there now exist a range of neuropsychologic, electrophysiologic, and
neuroimaging studies that have shown fairly consistent differences in
prefrontal cortical, parietal, and basal ganglia functioning associated
with ADHD symptomatology. In addition, there now have been a number of
candidate gene studies exploring the dopamine transporter (DAT1) or
dopamine receptor (DRD4) genes, each of which has now had several
independent replications (Swanson & Castellanos, 1999). While these
studies generally indicate an association with ADHD and specific
polymorphisms, effects are quite small. It is likely that a host of other
genes are involved, and that specific components or combinations of
environmental forces are necessary to switch on these genes and lead to
the expression of the full phenotype.
Other
factors implicated by correlational evidence in the onset, maintenance,
and severity of ADHD include pre- and perinatal factors (hypoxia, maternal
smoking, hyperbilirubinemia), lead, child abuse, specific genetic mutation
associated with thyroid disease, acquired forms of ADHD due to traumatic
brain injury, and early attachment difficulties. Remarkably, however,
despite intermittent leads in these areas, none have been subjected to
systematic programmatic research (Swanson & Castellanos, 1999).
Despite
these few hints, in truth we know little concerning etiology, a conclusion
also reached by the consensus conference. As a result, we know also
essentially nothing about primary or secondary prevention. More however is
known about tertiary prevention, in that effective treatments have been
demonstrated to improve longer-term outcomes several years post treatment
(Gilberg et al., 1997; MTA Cooperative Group, in press; Loney et al., in
press; Biederman et al., 1999). The lack of research in the prevention
area is striking, even thought there is ample evidence that subthreshold
and/or pre-morbid states can be reasonably identified. In effect, the
research field appears to have unwittingly excluded this area of research
from consciousness, perhaps because of the taken-for-granted assumptions
that ADHD symptoms are fully biologic or inborn (immutable). Such
assumptions, cannot be correct, however, given the number of children who
over the course of development show significant remission (20-40%) (e.g.,
Hechtman, 1992). How and why do such significant improvements in symptoms
take place? What developmental processes are at work? Could similar
factors come into play and be used to clinical advantage early in the
course of the disorder, perhaps before full expression of symptoms, via
some targeted prevention strategies?
Avenues
of exploration that appear to be understudied include the study of a)
specific perinatal/prenatal influences in interaction with susceptibility
genes, b) the impact of home and environmental factors that might modulate
the development of children’s attentional capacities, c) early learning
experiences, exposure to television or other factors that may up-regulate
or shape children’s attention styles, and d) early intervention
strategies designed to remediate attentional and executive functional
capacities in children at risk.
Future
studies of ADHD risk should ideally combine assessments of some of the
currently identified candidate genes with neuropsychologic and
neurophysiologic studies, coupled with direct observations. Given the
evidence the cultural factors can and do exert important effects on
defining specific behavioral characteristics of the syndrome, the search
for risk factors vis-à-vis the underlying phenotype must not rely on
behavioral descriptions alone. Methods for discriminating signal from
noise will likely depend upon multi-method and multi-informant strategies,
direct observation, the use of genetically informative designs coupled
with the examination of environmental forces, and methods for compiling
and sorting through the many characteristics of a "case" to
determine which indeed are true cases vs. non-cases (e.g., best estimate
methods with input from experienced clinicians).
REFERENCES
American
Psychiatric Association. Diagnostic and statistical manual of mental
disorders (4th ed.). Washington, DC: 1994.
Biederman
J, Milberger S, Faraone S, Kiely K, Guite J, Mick E, Ablon S, Warburton R,
Reed E (1995), Family-environmental risk factors for attention-deficit
hyperactivity disorder. Arch Gen Psychiatry 52:464-470.
Biederman
J, Wilens T, Mick E, Spencer T, Faraone SV (1999), Pharmacotherapy of
attention-deficit/Hyperactivity disorder reduces risk for substance use
disorder. Pediatrics 104:20.
Gilberg
C, Melander H, Von Knorring AL, et al: Long-term stimulant treatment of
children with attention-deficit hyperactivity disorder symptoms. Arch.Gen.Psychiatry
1997;54:857-864.
Jensen
PS, Martin D, Cantwell DP. Comorbidity in ADHD: implications for research,
practice, and DSM-V. J Am Acad Child Adolesc Psychiatry.
1997;36:1065-79.
Jensen
PS, Roper M, Fisher P, Piacentini J, Canino G, et al.: "Test-Retest
Reliability of the Diagnostic Interview Schedule for Children (DISC 2.1):
Parent, Child, and Combined Algorithms." Archives of General
Psychiatry, 52:61-71, 1995.
*Jensen
P, *Rubio-Stipec M, Dulcan M, Canino G, Bird H, Lahey B, Richters J,
Schwab-Stone M: "The NIMH Methods for the Epidemiology of Child and
Adolescent Mental Disorders (MECA) Study: Are Both Parent and Child
Informants Always Needed?", J Am Acad Child Adol Psychiatry, in press
(* designates joint first authorship).
Jensen
PS, Mrazek D, Knapp P, et al. "Evolution and Revolution in Child
Psychiatry: ADHD as a Disorder of Adaptation", J. Am. Acad. Child
Adol. Psychiatry, 36:1672-1679, 1997.
Jensen
PS, Kettle L, Roper MS, Sloan MT, Dulcan Mk, Hoven C, Bird HR,
Bauermeister JJ, Payne JD: "Are Stimulants Over-prescribed? Treatment
of ADHD in Four U.S. Communities", J. Am. Acad. Child Adol.
Psychiatry, 38:797-804, 1999.
Lahey
B, Applegate B, McBurnett K, Biederman J, Greenhill L, Hynd G, Barkley R,
Newcorn J, Jensen P, Richters J, Garfinkel B, Kerdyk L, Frick P,
Ollendick T, Perez D, Hart E, Waldman I, Shaffer D: "DSM IV Field
Trials for Attention Deficit/Hyperactivity Disorder in Children and
Adolescents," American Journal of Psychiatry, 151:1673-1685, 1994.
Levy
F, Hay DA, McStephen M, Wood C, Waldman I. Attention-deficit hyperactivity
disorder: a category or a continuum? Genetic analysis of a large-scale
twin study. Journal of American Academic Child and Adolescent
Psychiatry. 1997;36:737-744.
MTA
Cooperative Group: 14-month Randomized Clinical Trial of Treatment
Strategies for Attention Deficit Hyperactivity Disorder. Arch Gen
Psychiatry, in press.
MTA
Cooperative Group: Moderator and Mediator Challenges to the MTA Study:
Effects of Comorbid Anxiety Disorder, Family Poverty, Session Attendance,
and Community Medication on Treatment Outcome. Arch Gen Psychiatry, in
press.
National
Institutes of Health: Diagnosis and Treatment of Attention Deficit
Hyperactivity Disorder. NIH Consensus Statement 1998 Nov 16-18; 16(2). J
Am Acad Child Adolesc Psychiatry, in press.
Rappley
MD, Gardiner JC, Jetton JR, Houang RT: The use of methylphenidate in
Michigan. Arch.Pediatr.Adolesc.Med. 1995;149:675-679.
Richters
J, Arnold LEA, Jensen PS, Abikoff H, Conners CK, Greenhill L, Hechtman L,
Hinshaw S, Pelham W, Swanson J: "NIMH Collaborative Multisite,
Multimodal Treatment Study of Children with ADHD: I. Background and
Rationale," J. Am. Acad. Child Adol. Psychiatry, 34:987-1000, 1995.
Schwab-Stone
M, Dulcan M, Jensen P, Canino G, Bird H, Lahey B, Rae D: "The NIMH
Methods for the Epidemiology of Child and Adolescent Mental Disorders (MECA)
Study: Validity of the DISC - 2", J. Am. Acad. Child Adol.
Psychiatry, 35:878-888, 1996.
Shaffer
D, Fisher P, Dulcan M, Davies M, Piacentini J, Schwab-Stone M, Lahey B,
Bourdon K, Jensen P, Bird H, Canino G, Regier D: "The Second Version
of the NIMH Diagnostic Interview Schedule for Children (DISC - 2)",
J. Am. Acad. Child Adol. Psychiatry, 35:865-877, 1996.
Sloan
M, Jensen P, Kettle L: "Assessing Services for Children with ADHD:
Gaps and Opportunities." Journal of Attention Disorders, 3:13-29,
1999.
Swanson
JM, Sergeant J, Taylor E, Sonuga-Burke E, Cantwell D, Jensen P:
"Attention Deficit Hyperactivity Disorder and Hyperkinetic
Disorder," Lancet, 351:429-33, 1998.
Weisz
JR, Jensen PS: Efficacy and Effectiveness of Psychotherapy and
Pharmacotherapy with Children and Adolescents. Mental Health Services, in
press.
Zito
JM, Safer DJ, Riddle MA, Johnson RE, Speedie SM, Fox M: Prevalence
variations in psychotropic treatment of children. J.Child.Adolesc.Psychopharmacology
1998;8:99-105.
Zito
JM, Safer DJ, Dosreis S, Riddle MA: Racial disparity in psychotropic
medications prescribed for youths with medicaid insurance in Maryland. J.Am.Acad.Child
Adolesc.Psychiatry 1998;37:179-184.
Thomas M.
Achenbach, Ph.D.,
University of Vermont
Having
metamorphosed from Minimal Brain Damage (MBD) through Hyperkinetic
Reaction of Childhood and Attention Deficit Disorder (ADD), what is now
known as Attention Deficit Hyperactivity Disorder (ADHD) is receiving
enormous publicity in the popular media, as well as in the
professional literature. Its precise nature, etiology, prevalence,
developmental course, and appropriate treatment remain subject to debate.
However, existing findings can help us focus in on what we need to learn.
What We Know
1. Criteria for
ADHD.
The DSM criteria for ADHD changed in 1980 (DSM-III), 1987 (DSM-III-R), and
1994 (DSM-IV), and they differ from the ICD-10 (1992) criteria. The
different diagnostic criteria classify different groups and proportions of
children. In addition to these diagnostic criteria, there are numerous
rating scales and other procedures for assessing ADHD. Because there is no
single gold standard, data from multiple procedures must be aggregated for
both research and clinical purposes.
2. Gender and Age
Differences.
Most
research, clinical services, and diagnostic criteria focus on young boys.
Fewer girls and fewer older children meet DSM criteria for ADHD, possibly
owing to lower base rates for disruptive behaviors in these groups. Adult
ADHD has become widely publicized, but the DSM does not provide criteria
for adult ADHD.
3. Prevalence.
Studies suggest prevalence rates of about 2 to 6% for 7-10-year-old boys.
Lower prevalence rates for girls and other ages may reflect the
differential relevance of ADHD diagnostic criteria rather than real
differences in prevalence. Both epidemiologic data and the many referrals
for clinical and special education services indicate that large numbers of
boys are identified as having ADHD.
4. Correlates and
Comorbidity.
Poor school achievement, poor peer relations, and conduct problems are
commonly reported correlates of ADHD. Reports of comorbidity with other
disorders may reflect biases (e.g., Berkson’s bias) arising in the study
of clinical samples.
5. Etiology.
Hyperactive behavior was initially viewed as reflecting brain damage, but,
brain damage has not been supported as a typical cause. Indicating
moderate heritability, genetic research suggests polygenic influences on
quantitative parameters, rather than single gene causation of a disease
entity.
6. Treatment.
Research has documented the efficacy of stimulant medications, but the
effects may not be specific to ADHD, because similar effects on attention
have been found with nonADHD subjects. Parent training, behavior
modification, and special educational services are widely used, often in
conjunction with stimulant medications.
7. Long-term
Outcomes.
Follow-ups of clinical samples show that children diagnosed as ADHD
subsequently manifest a variety of problems, suggesting that ADHD leads to
multiple kinds of psychopathology. However, longitudinal studies of
general population samples indicate that attention problems per se
mainly predict poor academic achievement.
8. Cross-cultural
Comparisons.
Standardized ratings of children’s attention problems vary modestly
across epidemiologic samples from diverse cultures. U.S. children are not
exceptionally high in attentional problems.
9. Long-term
Trends.
Comparisons of U.S. general population samples over a 13-year period
showed significant increases in ADHD problems. However, the increases in
ADHD problems may reflect general increases found in many behavior
problems rather than in ADHD per se.
What We Need to
Learn
Epidemiologic
tabulations of cases meeting current diagnostic criteria are not likely to
be helpful until we learn more about the issues listed below.
1. How Can We
Take Account of Differences Related to Gender, Age, Assessment Procedures,
and Sources of Data?
The application of the same a priori diagnostic cutpoints to both
genders, all ages, different assessment procedures, and different sources
of data may obscure differences related to these factors. More research
needs to focus on females in general and on adolescents and adults of both
genders. Empirically based multi-source assessment instruments can provide
more precise data on the base rates and correlates of ADHD symptoms by
gender, age, and source.
2. How Can We
Better Aggregate Diagnostic Data?
Prevailing diagnostic procedures for ADHD fail to specify how different
kinds of data should be aggregated in determining caseness. Rather than
assuming that all procedures are co-equivalent or that diagnosticians
accurately combine all relevant data, we need to compare algorithms for
quantifying the probability and severity of ADHD.
3. How Can We Use
Epidemiological Findings Based on Better Diagnostic Criteria to Improve
Research on Etiology?
Norms and findings from epidemiological research can help us refine
etiological hypotheses and can improve ascertainment of true cases for
testing such hypotheses.
4. How Can We Use
Epidemiological and Etiological Findings to Improve Prevention and
Treatment?
More accurate identification of true cases, better documentation of
incidence and prevalence, and better knowledge of etiologies will lead to
more precisely targeted prevention and treatment efforts.
Andrew S. Rowland
Ph.D., National
Institute for Environmental Health Sciences
ADHD
is one of the most common childhood disabilities, yet its prevalence,
etiology, and developmental course remain poorly described. In part, these
data gaps exist because few population-based epidemiologic studies of ADHD
have been conducted. Several recent developments underscore the need for
additional epidemiologic studies of ADHD. First, there has been a major
paradigm shift; the literature suggests ADHD is best conceptualized as a
lifelong disability rather than as a childhood disorder. This implies that
treatment and monitoring may last a lifetime and impairment from ADHD
probably has been significantly underestimated. The evidence suggesting a
biologic basis for ADHD has become more persuasive; for example, several
studies suggest a possible etiologic role for two genes involved in
dopamine regulation and several imaging studies suggest possible
differences in brain structure and function among adults with ADHD.
Finally, at least two million children in the U.S. are being treated for
ADHD with stimulant medication and the average number of years children
are being treated is increasing. This should make epidemiologic research
to identify preventable risk factors, a public health priority.
Case
definition and Epidemiologic study of ADHD
There
have been many years of controversy over dimensional versus categorical
approaches to measuring psychopathology and over the validity of
particular screening instruments, diagnostic interviews, and
classification schemes. Although developing a clear consensus around these
issues may not be possible, this must not prevent us from making informed
choices and collecting the epidemiologic data that are needed to guide
public health decision making. Because DSM-IV is the basis of current
diagnoses and treatment, we need to collect data applying these criteria.
For
an epidemiologic study we are conducting in North Carolina, we developed a
methodology for operationalizing the DSM-IV criteria for ADHD. The key
decisions we made included using a population-based sample, using multiple
respondents including parents and teachers, and requiring moderate
agreement between respondents about symptoms and impairment. I will talk
about difficulties we faced in defining impairment, and assigning case
status to children taking medication. I will also discuss how we plan to
validate our screening diagnoses and analyze our data dimensionally as
well as categorically. Merits of alternative study designs and possible
use of national surveys like NHANES will also be addressed.
Preliminary
data from this study have been promising. The early data suggest that the
standard ADHD prevalence estimate of 3% to 5% among school-aged children
probably underestimates prevalence by a factor of two or three, that many
children with ADHD are not being identified and that many children
continue to meet ADHD criteria despite stimulant treatment. We also found
that African-American children were being treated with stimulant
medication less frequently than White children, even though the prevalence
of ADHD in the two groups was similar.
Historically,
ADHD was believed to be the sequelae of brain damage caused during
pregnancy. The concept of minimal brain dysfunction (which combined
learning disabilities and hyperactivity) was discarded long ago but
unfortunately so have many of the etiologic leads that suggested adverse
pregnancy events might be important risk factors. My perspective is that
ADHD should be viewed as a developmental disorder and the primary
hypothesis of our study is that preterm delivery is an important,
etiologic risk factor. I will summarize the data supporting this
perspective and hypothesis. We also hypothesize that exposures to
environmental toxicants like cigarette smoke, alcohol, and lead during
early brain development may be etiologic risk factors. We also collected
data on a number of potential social risk factors. Even if one is only
interested in biologic risk factors, social factors cannot be overlooked
because they probably play an important role in which children get
identified, treated, and suffer the most severe impairment.
We
plan to look whether the social characteristics and risk factors for
children who are primarily inattentive are different from children who are
primarily hyperactive or have the combined subtype. Researchers have
proposed that children who have ADHD with co-morbid anxiety or ADHD with
co-morbid conduct problems have different developmental pathways.
Examining how the social characteristics and the developmental trajectory
of children with these conditions differ from other ADHD children should
help clarify whether these sub-classifications are important.
Public Health
Priorities:
Although
some researchers have proposed creating age, gender and race-based norms
for ADHD, examining how prevalence varies between sub-groups is a critical
step in understanding possible risk factors, and for understanding gaps in
who is being identified and who is being treated. Collecting better
epidemiologic data on ADHD is critical for setting public health
priorities.