|
 Reviews
of the literature leave no doubt that genes influence the etiology of
attention-deficit/hyperactivity disorder (ADHD) (Faraone et al., 1998).
Notably, twin studies show the heritability of ADHD to be about 0.80,
indicating that the effect of genes is substantial. These genetic epidemiological
studies have motivated molecular genetic studies of ADHD that have produced
intriguing but conflicting results (Faraone and Biederman, 1998). Researchers
have focused on genes in dopamine pathways because animal models, theoretical
considerations, and the effectiveness of stimulant treatment implicate
dopaminergic dysfunction in the pathophysiology of the disorder. Two genes
that have been intensively studied are the dopamine transporter gene (DAT)
and the dopamine D4 receptor gene (DRD4). Some studies of these genes
strongly suggest that they influence susceptibility to ADHD. There are,
however, several negative studies for each gene.
The inconsistent results from molecular
genetic studies could mean that rather than being a unitary disorder,
ADHD comprises several disorders having different genetic and nongenetic
etiologies. If this were so, then the power to detect genetic effects
would be small and we would expect to observe an inconsistent pattern
of replication. What, then, is the evidence for genetic heterogeneity
in ADHD? In the next column in this series, Todd will review the evidence
for genetic heterogeneity based on the type of ADHD symptoms seen in twin
pairs. This column focuses on evidence suggesting 2 other clinical features
that may be useful for parsing the genetic heterogeneity of ADHD: psychiatric
comorbidity and long-term outcome.
Epidemiological studies have documented
high rates of psychiatric comorbidity among children with psychiatric
disorders. These data confirm the adult epidemiological literature that
suggests that comorbidity is the rule rather than the exception for psychiatric
disorders. Researchers and clinicians have known for decades about ADHD’s
comorbidity with conduct disorder (CD) and learning disabilities. More
recently, researchers have documented its comorbidity with mood and anxiety
disorders.
To examine the familial heterogeneity
of ADHD, my colleagues and I have tested competing hypotheses about the
association of ADHD with other psychiatric disorders. Our analyses from
independent studies of DSM-III attention deficit disorder (ADD) and DSM-III-R
ADHD suggested that ADHD with CD or bipolar disorder (BPD) may be a distinct
familial subtype of ADHD (Faraone et al., 1998). Stratification of ADHD
patients by the presence of CD and/or BPD appears to cleave the universe
of ADHD children into familially homogeneous subgroups. Put simply, there
seem to be 2 types of ADHD families: those in which CD and/or BPD occur
comorbidly with ADHD and those in which ADHD occurs without these disorders.
We have also shown that ADHD and major
depression share common familial vulnerabilities, but our data cannot
separate distinct familial types of ADHD based on the presence of depression
in the family. Instead, depression seems to be a nonspecific manifestation
of the familial predisposition to ADHD. Whereas CD or BPD appears to be
a marker for genetic heterogeneity in ADHD, with different subforms having
different familial (and presumably genetic) causes, major depression appears
to be a marker of phenotypic heterogeneity. In ADHD families, it is one
of several manifestations of the genes that cause ADHD.
In contrast to our findings for CD,
BPD, and depression, our data suggest that anxiety disorders and learning
disabilities are not good candidates for resolving either genetic or phenotypic
heterogeneity. These disorders are only weakly associated with ADHD in
families, which suggests that they do not share genetic causes with ADHD.
Notably, a meta-analysis of several studies supports the above conclusions
about depression (Faraone and Biederman, 1997), but more work is needed
to reach similar conclusions regarding anxiety disorders and learning
disabilities.
Many groups have reported systematic
differences between the families of ADHD children with and without CD.
For example, compared with other ADHD children, fathers of ADHD+CD children
have a high prevalence of substance abuse, depression, childhood CD, and
adult antisocial personality disorder. Notably, Szatmari et al. (1993)
confirmed the familial coaggregation of ADHD and CD in a population-based
epidemiological family study, as did Silberg et al. (1996) in a population-based
twin study. The latter investigators concluded that their results were
consistent with the existence of a biologically based group of children
who manifest both ADHD and conduct disturbances.
These studies of ADHD+CD are compelling,
but they did not address whether BPD is also associated with the ADHD+CD
phenotype. Our meta-analysis documented a link between ADHD and BPD by
showing an increased prevalence of ADHD among children of BPD parents
and an increased prevalence of BPD among relatives of ADHD children. Moreover,
our family data show that ADHD, CD, and BPD tend to be transmitted together
in families (Faraone et al., 1998). Although no other groups have examined
the familial association between ADHD, BPD, and CD, there are reports
indicating high levels of comorbidity in youth between ADHD and BPD and
between CD and BPD.
Studies of long-term outcome provide
another window on clinically meaningful variability in ADHD that may have
implications for genetic heterogeneity. Several medium- and long-term
follow-up studies have examined the natural history of ADHD. Despite methodological
variability in sample characteristics, diagnostic criteria, age of subjects,
assessment instruments, and the frequency and timing of follow-up reassessments,
these studies have consistently documented that only a subgroup of children
with ADHD have a disorder that persists into adolescence and young adulthood
(Barkley, 1998). Could there be biological and, perhaps, genetic differences
between persistent and remitting forms of ADHD?
Several studies suggest that genes
influence persistent ADHD more than they influence remitting ADHD (Faraone
et al., 2000). A prospective 4-year follow-up study found that by mid-adolescence,
85% of boys with ADHD continued to have ADHD while 15% had remitted. The
prevalence of ADHD among parents was 16.3% for the persistent ADHD probands
and 10.8% for the remitted ADHD probands. For sibs, the respective prevalence
rates were 24.4% and 4.6%. These data suggest that children with persistent
ADHD have a more familial form of ADHD than those whose ADHD remits by
adolescence.
Two retrospective studies provided
additional evidence for the increased familiality of persistent ADHD.
One showed that children of parents with childhood-onset ADHD were at
high risk for meeting diagnostic criteria for ADHD: 84% of the adults
with ADHD who had children had at least 1 child with ADHD and 52% had
2 or more children with ADHD. The 57% rate of ADHD among children of adults
with ADHD was much higher than the more modest 15% risk for ADHD in siblings
of children with ADHD. These findings were consistent with a prior DSM-III—based
study which found that 41% of siblings of adult ADD probands had ADHD
compared with none of the non-ADD comparison siblings.
Another retrospective study compared
children with ADHD and ADHD adolescents who retrospectively reported childhood
onset. It found that the relatives of adolescent probands had higher rates
of ADHD compared with the relatives of child probands. Thus, a prospective
study of children and retrospective studies of adolescents and adults
all suggest that persistent ADHD is highly familial and thus may be more
strongly influenced by genes compared with remitting ADHD.
To summarize, prior work suggests
2 clinical features that might be useful for parsing the genetic heterogeneity
of ADHD: comorbidity with CD or BPD and persistence of ADHD into adolescence.
These inferences about genetic heterogeneity are limited by the fact that
much of the data are from family studies, which cannot disentangle genetic
from environmental sources of familial transmission (Faraone et al., 1999).
We need twin studies of genetic heterogeneity to show whether genes mediate
differences among these putative subtypes.
This column’s discussion of genetic
heterogeneity adopts a categorical as opposed to a dimensional view of
the nature of ADHD. A categorical view sees ADHD as a distinct condition.
In contrast, a dimensional view sees ADHD as a continuous trait. Some
people have no or few ADHD symptoms, while others have moderate or severe
ADHD symptoms. In a dimensional framework, the clinical category of ADHD
is seen as resulting from the imposition of an arbitrary threshold on
the continuous dimension of ADHD symptoms. Despite the importance of categories
for clinical work, we must recognize that a dimensional view might also
explain the apparent genetic heterogeneity of the disorder. Population-based
twin studies suggest that the clinical syndrome of ADHD is influenced
by the same set of genes that influences the expression of subclinical
forms of the disorder (Levy et al., 1997). They support the idea that
there is a set of genes that influence ADHD symptoms. People with many
of these genes develop ADHD, people with few are asymptomatic, and those
in between show some ADHD symptoms but do not meet diagnostic criteria
for the disorder. Could it be possible that ADHD with CD or BPD is simply
a genetically severe form of ADHD? Could the same be said of persistent
ADHD?
Ultimately, these questions must await
large-scale molecular genetic studies for their answers. Meanwhile, the
identification of highly familial subtypes of ADHD could prove useful
for researchers seeking to optimize the statistical power of genetic association
and linkage studies. Statistical power increases with the magnitude of
risk ratios computed by dividing the prevalence of a disorder among biological
relatives by the prevalence in the population. Table
1 shows the risk ratios of ADHD
in relatives when different subtypes of ADHD are used to select families
(see Faraone et al., 2000, for details). Because these risk ratios increase
dramatically with narrower definitions of ADHD, these narrower definitions
may prove useful for selecting cases for molecular genetic studies.
top of page
|
