The overlap between attention-deficit/ hyperactivity disorder and alcohol or
drug abuse or dependence (referred to here as substance use disorders [SUDs]) in adolescents has been an area of increasing
clinical, research and public health interest. Appearing in early childhood,
ADHD affects from 6% to 9% of children and adolescents worldwide (Anderson et
al., 1987) and up to 5% of adults (Kessler, in press). Longitudinal data
suggest that childhood ADHD persists into adolescence in 75% of cases and into
adulthood in approximately one-half of cases (for review, see Weiss, 1992).
Substance use disorders usually appear in adolescence or early adulthood and
affect between 10% to 30% of U.S. adults and a less defined, but
sizable, number of juveniles (Kessler, 2004). The study of comorbidity between SUDs and ADHD
is relevant to both research and clinical practice in developmental pediatrics,
psychology and psychiatry with implications for diagnosis, prognosis, treatment
and health care delivery.

Overlap Between ADHD and SUD

Structured psychiatric diagnostic interviews assessing ADHD and other
disorders in substance-abusing groups have indicated that from one-third to
one-half of adolescents with SUDs have ADHD (DeMilio, 1989; Milin et al.,
1991). For example, aggregate data from government-funded studies of mainly
cannabis-abusing youth indicate that ADHD is the second most common comorbidity with from 40% to 50% of both girls and boys
manifesting full criteria for ADHD. Data largely ascertained from adult groups
with SUDs also show an earlier onset and more severe
course of SUD associated with ADHD (Carroll and Rounsaville,
1993; Levin and Evans, 2001).

ADHD as a Risk Factor or Precursor
for SUD

The association of ADHD and SUDs is particularly
compelling from a developmental perspective as ADHD appears to manifest itself
earlier than the SUD; therefore, the SUD is an unlikely risk factor for ADHD.
Thus, it is important to evaluate to what extent ADHD is a precursor of SUDs. Prospective studies of children with ADHD have
provided evidence that the group with conduct or bipolar disorders co-occurring
with ADHD have the poorest outcome with respect to developing SUDs and major morbidity (Biederman
et al., 1997; Mannuzza et al., 1993). As part of an
ongoing prospective study of ADHD, it was found that differences in the risk
for SUDs in adolescents with ADHD (mean age=15)
compared to controls without ADHD were accounted for by comorbid
conduct or bipolar disorders (Biederman et al.,
1997). However, it also has been shown that the age of risk for SUD onset in
adolescents without comorbid ADHD is approximately 17
years in girls and 19 years in boys (Biederman et
al., in press-a; Milberger et al., 1997b). These
findings were confirmed by Katusic and associates
(2005) and Molina and Pelham (2003), who have shown elevated risk of SUDs in adolescents with ADHD.

ADHD treatment and SUD.
Clarification of the critical influence of ADHD treatment in youth on later SUDs remains hampered by methodological issues. Since
prospective studies in youth with ADHD are naturalistic, and hence not
randomized for treatment, attempts to disentangle positive or deleterious
effects of treatment from the severity of the underlying condition(s) are
hampered by serious confounds. Whereas concerns of the abuse liability and
potential kindling of specific types of abuse (e.g., cocaine) secondary to early
stimulant exposure in children with ADHD have been raised (Drug Enforcement
Administration, 1995; Vitiello, 2001), the
preponderance of clinical data do not appear to support such a contention.

To reconcile findings in this important area, my group completed a
meta-analysis of the literature (Faraone and Wilens, 2003; Wilens et al.,
2003). We included studies examining the later risk of SUDs
in children exposed to stimulant pharmacotherapy, identifying two studies into
adolescence and five studies into adulthood. We found that stimulant
pharmacotherapy did not increase the risk for later SUDs.
In fact, we found that stimulant pharmacotherapy protected against later SUDs (odds ratio of 1.9) and that the effect was stronger
in adolescents relative to adults (Wilens et al.,
2003). It is notable that the magnitude of risk reduction (e.g., 50% reduction
in risk) indicated that the ultimate risk of SUDs in
treated individuals with ADHD may approximate the level of risk in individuals
without ADHD (general population).

SUD Pathways Associated With ADHD

An increasing body of literature shows an intriguing association between
ADHD and cigarette smoking. It has been previously reported that ADHD is a
significant predictor for early initiation of cigarette smoking (before age 15)
and that conduct and mood disorders comorbid with
ADHD put youth at particularly high risk for early-onset smoking (Milberger et al., 1997a) (Figure). Data also suggest that one-half of smokers with ADHD go on
to later SUDs (Biederman et
al., in press-b). This is not surprising given that not only does smoking lead
to peer group pressures and availability of illicit substances, but that
nicotine exposure may make the brain more susceptible to later behavioral
disorders and SUDs (Trauth
et al., 2000). Furthermore, nicotinic-modulating agents are increasingly being
evaluated for the treatment of ADHD (Wilens et al.,
in press-b). Of interest, prospective data funded by the National Institute on
Drug Abuse suggest that stimulant treatment of ADHD reduces not only the time
to onset but also the incidence of cigarette smoking (Monuteaux,
2004).

The precise mechanism(s) mediating the expression of SUDs
in ADHD remains to be seen. The self-medication hypothesis is compelling in
ADHD considering that the disorder is chronic and often associated with
demoralization and failure, factors frequently associated with SUDs in adolescents. Moreover, it has been found that among
substance-abusing adolescents with and without ADHD, adolescents with ADHD
reported using substances more frequently to attenuate their mood and to help
them sleep. No evidence of differences in types of substances has emerged
between substance-abusing teen-agers with or without ADHD (Biederman
et al., 1997). In addition, the potential importance of self-medication needs
to be tempered against more systematic data showing the strongest association
between ADHD and SUDs is comorbidity
and familial contributions, such as exposure to parental SUDs
during vulnerable developmental phases.