You are on page 1of 10


JADXXX10.1177/1087054716680074Journal of Attention DisordersMolina et al.

Journal of Attention Disorders

Antisocial Alcoholism in Parents of

© The Author(s) 2016
Reprints and permissions:
Adolescents and Young Adults With
DOI: 10.1177/1087054716680074

Childhood ADHD

Brooke S. G. Molina1, Elizabeth M. Gnagy2, Heather M. Joseph1,

and William E. Pelham Jr2

Objective: Test the hypothesis that alcoholism, including antisocial alcoholism, is more prevalent among mothers and
fathers of children with versus without ADHD. Method: Mothers (312 ADHD group, 235 non-ADHD group) and
fathers (291 ADHD group, 227 non-ADHD group) in the Pittsburgh ADHD Longitudinal Study were interviewed along
with their adolescent and young adult offspring. Results: Maternal and paternal alcoholism, with and without comorbid
antisociality, was more prevalent in the ADHD group. Paternal alcoholism without antisociality was only marginally higher
for probands after controlling for paternal ADHD. Offspring conduct disorder comorbidity was associated with parental
antisociality but not parental antisocial alcoholism. Conclusion: Our findings that 44% of proband fathers and 25% of
proband mothers experienced alcohol problems with or without antisociality are further evidence of increased alcoholism
prevalence in families affected by ADHD. Maternal alcoholism and antisociality are prominent contributors to this family-
level vulnerability. These findings indicate the need to assess long-term offspring outcomes as a function of parental alcohol
and externalizing comorbidities, and perhaps other indicators of parental alcoholism phenotype, as familial vulnerability
unfolds across development. (J. of Att. Dis. XXXX; XX(X) XX-XX)

alcoholism, ADHD, family, antisocial, personality disorder, alcohol

Several decades of research have identified a robust relation (1990) found that 34.2% of biological fathers of children
between parent alcoholism and offspring externalizing with ADHD had alcohol abuse histories, compared with
behavior (Pelham & Lang, 1993; Sher, 1991; Waldron, 14.1% of controls. Biederman and colleagues (1992)
Martin, & Heath, 2009; Zucker, Heitzeg, & Nigg, 2011). reported higher rates of alcohol dependence in first-degree
Genetic, and to a lesser extent environmental (e.g., child relatives of children with ADHD (18%) than in controls
rearing, peer socialization), factors have been supported as (11%). August and colleagues (1999) found that 65% to
contributory (Haber, Jacob, & Heath, 2005; Hicks, Foster, 85% of children with ADHD had a parent with a Diagnostic
Iacono, & McGue, 2013; Kendler, Ohlsson, Sundquist, & and Statistical Manual of Mental Disorders (4th ed.;
Sundquist, 2016). ADHD, which includes the highly heri- DSM-IV; American Psychiatric Association, 1994) alcohol-
table core symptom of impulsivity and is an early precursor ism diagnosis as compared with 48% of controls. However,
to conduct and alcohol-related problems (Barkley, 2015; despite these reports, other studies have failed to detect
Lee, Humphreys, Flory, Liu, & Glass, 2011), should be higher levels of alcoholism in biological relatives of chil-
associated with parental alcoholism. However, a limited dren clinically diagnosed with ADHD (Biederman et al.,
number of studies have tested the specific relation between 2008; Cadoret & Stewart, 1991; Lahey et al., 1988;
parental alcoholism and offspring ADHD. Using Swedish Milberger, Faraone, Biederman, Chu, & Wilens, 1998).
population-based registries of children and adults with
ADHD, Skoglund and colleagues found that 6.5% had par-
ents with alcohol use disorders (AUDs) versus only 3% of 1
University of Pittsburgh, PA, USA
controls (Skoglund, Chen, Franck, Lichtenstein, & Larsson, 2
Florida International University, Miami, USA
2015). Rates of parental alcohol problems based on clinic
Corresponding Author:
(Barkley, Fischer, Edelbrock, & Smallish, 1990; Biederman Brooke S. G. Molina, Western Psychiatric Institute & Clinic, University
et al., 1992) and community (August, Realmuto, Tamara, & of Pittsburgh, 3811 O’Hara St., Pittsburgh, PA 15213, USA.
Hektner, 1999) samples are higher. Barkley and colleagues Email:
2 Journal of Attention Disorders 

Little research on this topic (the association between reported. Determining whether parental alcohol problems
childhood ADHD and parental alcoholism) has directly occur at higher rates in families affected by ADHD, rela-
considered the well-known heterogeneity of alcoholism and tive to the co-occurrence of antisociality, will aide future
its impact on the association with ADHD. Alcoholism is research efforts focused on distinct causal mechanisms
variable on multiple dimensions including age of onset, associated with these phenotypes.
severity, chronicity, and mental health comorbidities To our knowledge, one study has tested the specific rela-
(Babor, Hesselbrock, Meyer, & Shoemaker, 1994). Such tion between childhood ADHD diagnosis and parent
heterogeneity may account for the variable expression of antisocial-alcoholism. Using the well-known multi-site
behavior disorders among offspring, as not all children of Collaborative Study on Genetics of Alcoholism sample, off-
alcoholic parents develop behavior problems. Some spring ADHD was examined for its association with alco-
researchers suggest that far less than half of children of holism among parents treated for alcohol dependence
alcoholics (COAs) develop these difficulties, which might (Kuperman, Schlosser, Lidral, & Reich, 1999). Compared
be connected to the heterogeneity of parent alcoholism with children of non-alcoholics (5.9%), risk of offspring
(Sher, 1991) and concomitant psychobiological underpin- ADHD was significantly higher in children of alcoholic only
nings (Zucker et al., 2011). Thus, the extent to which alco- families (13.2%) and higher still in children of alcoholic par-
holism has a deleterious effect on offspring may not be ents with ASPD (19.0%). However, only parental alcohol-
determined by parent’s AUD per se, but by the pattern of ism (not comorbid antisociality) was associated with
risk factors, including parental alcoholism comorbidity, that offspring ADHD when tested in a multivariate model. Thus,
are associated with onset and severity of course. these initial data, confined to the study of treatment-seeking
Antisocial personality disorder (ASPD) is one of the alcoholic adults, suggest an association between parental
most common phenotypic comorbidities of alcoholism to alcoholism and offspring ADHD but leave open the question
have been associated with early onset and a chronic severe of the importance of comorbid antisociality among parents.
course. ASPD is part of the externalizing behavior contin- More recently, Chronis and her colleagues found that pater-
uum and is associated with elevated rates of offspring psy- nal antisociality and not recent drunkenness was associated
chopathology including ADHD (Herndon & Iacono, 2005). with child conduct problems in a sample of children with
Beauchaine and McNulty (2013) argue that highly herita- ADHD (LeMoine, Romirowsky, Woods, & Chronis-
ble trait impulsivity, a core symptom of ADHD, may be the Tuscano, 2015). Father drunkenness, antisociality, and
earliest visible indicator of risk of conduct problems, ADHD symptoms were correlated, but without a compari-
ASPD, and substance abuse. In line with this thinking, the son sample of children without ADHD, their joint prediction
combination of parent alcoholism and ASPD (antisocial of childhood ADHD could not be evaluated.
alcoholism) is a particularly strong indicator of offspring Another important question that has been minimally
risk for such maladaptive outcomes that may be due to evaluated is whether or not the gender of the alcoholic par-
greater severity of parental psychopathology, polygenetic ent explains part of the variability in the association between
risk of undercontrolled behaviors, and an especially disad- parent alcoholism and offspring disorders such as ADHD.
vantageous rearing environment for the child. Zucker, There is evidence that maternal alcoholism is associated
Ellis, and Fitzgerald (1993) found that antisocial alcohol- with offspring externalizing disorders such as ADHD (e.g.,
ism was associated with earlier onset, higher overall levels Knopik et al., 2006) and substance use disorders (Chassin,
of comorbid alcohol symptoms, and a wider variety of Curran, Hussong, & Colder, 1996). Knopik and her col-
comorbid alcohol symptoms in alcoholic fathers, and sug- leagues (Knopik et al., 2006; Knopik, Jacob, Haber,
gest that aggregating parental risk in this way has increased Swenson, & Howell, 2009) used genetically informed twin
prognostic value for offspring behavior problems. In their study designs to assess the association between parental
study, sons of antisocial alcoholics were more likely to alcohol dependence and offspring ADHD using children-
develop behavior problems than were sons of non-antiso- of-female (Knopik et al., 2006) and male-twin designs
cial alcoholics and sons of non-alcoholics (Wong, Zucker, (Knopik et al., 2009). Increased risk of offspring ADHD
Puttler, & Fitzgerald, 1999). Studies of children with associated with maternal alcohol dependence was sup-
ADHD has corroborated the association between parent ported in both studies, but only the latter controlled for
ASPD and child ADHD, showing higher levels of familial maternal ADHD, which was a partial confounder (Knopik
ASPD for children with than for children without ADHD et al., 2009). In a smaller community sample of children in
(Biederman et al., 1992) and higher rates of parent greater Pittsburgh, Pennsylvania (not selected for psycho-
antisocial behaviors such as police contacts or arrests, pathology), alcoholism in first- and second-degree biologi-
fights, and assaults for children with than for children with- cal relatives was associated with male, but not female,
out ADHD (Barkley et al., 1990). The extent to which alco- offspring externalizing behavior (Molina, Donovan, &
holism, which is widely known to co-occur with ASPD in Belendiuk, 2010); parental ADHD was not controlled but
adulthood, was accounting for these findings was not mother’s general mental health and not alcohol use was a
Molina et al. 3

statistically significant mediator. Others have concluded ADHD (Pelham & Hoza, 1996), and (c) additional inclu-
that the literature is mixed with respect to gender- sion/exclusion criteria described below. Diagnostic infor-
differentiated pathways to alcoholism (Zucker et al., 2011) mation was collected in childhood using several sources,
and the complexity of these associations and need for addi- including the parent and teacher Disruptive Behavior Disor-
tional research has been nicely detailed (Knopik et al., ders (DBD) Rating Scale that assesses the DSM-III-R and
2009). Finally, it is unclear whether maternal antisocial DSM-IV symptoms of the disruptive behavior disorders
alcoholism per se is responsible for offspring behavior (Pelham, Gnagy, Greenslade, & Milich, 1992). Parents
problems. Thus, the primary goal of this article is to exam- completed a semi-structured diagnostic interview with
ine the associations between maternal and paternal alcohol- PhD-level clinicians consisting of the DSM-III-R or DSM-
ism and antisocial alcoholism with offspring ADHD IV descriptors for ADHD, Oppositional Defiant Disorder
diagnosis. We will control for parental ADHD to understand (ODD), and CD, with supplemental probe questions regard-
the extent to which these familial associations include, or ing situational and severity factors. Following DSM guide-
are independent of, ADHD among the parents. lines, diagnoses were made if a sufficient number of
The Pittsburgh ADHD Longitudinal Study (PALS) is a symptoms was endorsed (considering information from
longitudinal follow-up study of elementary school-aged both parents and teachers) to result in diagnosis. At least
children comprehensively diagnosed with Diagnostic and two PhD-level clinicians independently reviewed all ratings
Statistical Manual of Mental Disorders (3rd ed., rev.; DSM- and interviews to confirm the DSM diagnoses. Exclusion-
III-R; American Psychiatric Association, 1987) or DSM-IV ary criteria also assessed in childhood included a full scale
ADHD and followed prospectively through adolescence IQ less than 80, a history of seizures or other neurological
and early adulthood (e.g., Molina, Pelham, Gnagy, problems, and/or a history of pervasive developmental dis-
Thompson, & Marshal, 2007; Molina et al., 2014). The order, schizophrenia, or other psychotic or organic mental
PALS includes interviews with parents and provides an disorders. Mean age at initial evaluation was 9.40 years,
opportunity to test the hypothesis that antisocial alcoholism SD = 2.27 years; ADHD group participants were contacted
accounts for a significant proportion of alcoholism preva- for follow-up an average of 8.35 years later, SD = 2.79. Par-
lence among parents of children diagnosed with ADHD. ticipation rate was 70.5% (364/516), and only one of 14
The study, having collected interview data with mothers comparisons between participants and nonparticipants was
and with fathers, also provides the opportunity to test the statistically significant (mean CD symptom rating, Cohen’s
hypothesis that this association occurs among mothers, as d = 0.30). Data for the current study were taken from the
well as fathers of children with and without ADHD, and first follow-up visit when probands averaged 17.74 years of
that proband offspring with conduct problems (ADHD plus age (SD = 3.38 years). Most were boys (89.6%) and Cauca-
conduct disorder [CD]) will have the highest rates of paren- sian (84.6%). Most minorities were African American
tal antisocial alcoholism. (11.0% of probands).

Non-ADHD group. Participants without ADHD were

Method recruited from the Pittsburgh area for their demographic
similarity to the ADHD group at follow-up (age within 1
year, sex, race, parental education). They were recruited on
The PALS includes 364 children diagnosed with DSM-III-R a rolling basis to create ADHD/non-ADHD groups with
or DSM-IV ADHD and followed longitudinally through equal proportions of each demographic characteristic. Par-
adolescence into early adulthood (probands) and 240 demo- ticipants without ADHD were recruited via several large
graphically similar individuals without ADHD recruited pediatric practices in Allegheny County (40.8%), advertise-
and followed for comparison purposes (controls). The ments in local newspapers and the university hospital staff
PALS methods are described briefly and the reader is newsletter (27.5%), local universities, and colleges (20.8%),
referred elsewhere for additional details (Molina et al., and other methods (e.g., word of mouth). A telephone
2007). screening interview administered to parents gathered basic
demographic characteristics, history of diagnosis and treat-
ment for ADHD and other behavior problems, presence of
exclusionary criteria as previously listed for probands, and
ADHD group.  Children with ADHD were diagnosed at the a checklist of ADHD symptoms. Individuals who met DSM-
ADD Clinic at the Western Psychiatric Institute and Clinic III-R criteria for ADHD—either currently or historically—
in Pittsburgh, Pennsylvania, during the years 1987 to 1996. were excluded. The two groups did not differ in age, sex,
They were eligible for longitudinal follow-up on the basis race, or parent education. A higher percentage of ADHD
of (a) their ADHD diagnosis in childhood, (b) participation than non-ADHD group participants were from single parent
in an 8-week summer treatment program for children with households, 33.2% versus 23.6%, p < .05; ADHD group
4 Journal of Attention Disorders 

parents’ incomes were lower, M = US$63K versus US$76K, Parent’s ASPD.  ASPD was assessed using the SCID-II ASPD
p < .01; and more probands than controls had been adopted, Module (SCID-II; Spitzer, Williams, & Gibbon, 1987)
8.0% versus 0.4%, p < .01. administered to biological parents about themselves and
about the other biological parent. ASPD diagnoses were
Subsample for the current study. The primary goal of the assigned if either biological parent’s report met diagnostic
current study was to compare rates of biological parent criteria. Childhood CD was not required for diagnosis of
alcoholism and antisocial alcoholism between the ADHD ASPD in adulthood due to concerns about retrospective
and non-ADHD groups. Thus, a subsample of the larger recall of spouses’ childhood behaviors, and because empiri-
PALS sample was used for the current study to ensure that cal evidence suggests that this criterion is not needed to
alcoholism and ASPD information described biological capture impairing ASPD and associated offspring risk for
parents only. After excluding families for whom no bio- psychiatric problems (Compton, Conway, Stinson, Colliver,
logical parent data were available and families for whom & Grant, 2005).
only partial data were available, the subsample size for
analysis included 312 ADHD group mothers, 291 ADHD Parental ADHD.  Parental ADHD was included in the anal-
group fathers, 235 non-ADHD group mothers, and 227 yses to control for its possible role in the association
non-ADHD group fathers. These numbers reflected 90% between parental alcoholism and offspring ADHD. Par-
(n = 547/604) of the mothers and 86% (n = 518/604) of the ents were assigned a lifetime ADHD diagnosis if they
fathers of participants in the larger PALS sample. Average endorsed at least one of the following during their first
ages were 45.6 (SD = 5.7) for ADHD group mothers, 49.5 wave of participation in the study: (a) six or more child-
(SD = 6.6) for ADHD group fathers, 46.3 (SD = 5.4) for hood symptoms of hyperactivity/impulsivity or inatten-
non-ADHD group mothers, and 49.4 (SD = 6.6) for non- tion on the Disruptive Behavior Disorders Rating Scale
ADHD group fathers. (Pelham et al., 1992), reporting retrospectively; (b) four
or more current symptoms of hyperactivity/impulsivity or
inattention (Barkley, unpublished measure) that included
Interview Procedures the DSM-IV ADHD symptoms; or (c) a self-reported life-
Follow-up interviews with parents and offspring were com- time diagnosis of ADHD. The lower threshold for symp-
pleted in the ADD program offices by post-baccalaureate toms in adulthood, relative to the Diagnostic and
research staff. Interviewers were not blind to recruitment Statistical Manual of Mental Disorders (5th ed., DSM-5;
source (i.e., presence or absence of ADHD in childhood), American Psychiatric Association, 2013) threshold of
but they were trained to avoid bias in data collection. five or more symptoms, was based on recent findings
Informed consent was obtained and all participants were with the PALS data set and an effort to be over-, rather
assured confidentiality of all disclosed material except in than under-, inclusive when testing this variable as a pos-
cases of impending danger or harm to self or others (rein- sible covariate (Sibley et al., 2012).
forced with a US Department of Health and Human Services
Certificate of Confidentiality). Proband CD.  At follow-up, a diagnosis of CD in the past
year for adolescents was made using a combination of
parent, teacher, and self-report on the DBD adapted for
Measures adolescents, and parent and self-report on the Diagnos-
Parent alcoholism.  A lifetime history of alcohol problems in tic Interview Schedule for Children (DISC-IV; Shaffer,
the biological parents was assessed using the Structured Fisher, Lucas, Dulcan, & Schwab-Stone, 2000). If each
Clinical Interview for the DSM-IV with parents (SCID-IV; symptom of CD was endorsed by any rater, it was
First, Spitzer, Gibbon, & Williams, 1998). This interview, counted and adolescents who exceeded DSM-IV symp-
which has excellent reliability and validity and reflected the tom cutoffs were diagnosed with CD. For young adults
version of the DSM in use at the time of data collection, (18+), CD diagnosis was approximated using selected
yielded lifetime diagnoses of alcohol abuse or dependence items from the Self-Reported Delinquency Question-
for mothers and for fathers. In the absence of direct inter- naire (Elliot, Huizinga, & Ageton, 1985) and the SCID-
view, alcohol problems of the non-interviewed biological II ASPD Module (SCID-II; Spitzer et al., 1987)
parent were measured with spousal report on the Short administered to self and parents. From these measures,
Michigan Alcoholism Screening Test (SMAST; Selzer, young adults who endorsed two or more of nine items
Vinokur, & van Rooijen, 1975). For the SMAST, alcohol- that overlapped with CD diagnostic criteria were
ism was considered present upon endorsement of three or assigned a CD diagnosis. (A lower threshold of two,
more alcohol-related problems or one of three diagnostic rather than three, CD symptoms was used for diagnosis
items indicating receipt of treatment or help for drinking to adjust for the smaller number of CD symptoms avail-
(Selzer et al., 1975). able for this age range.)
Molina et al. 5

Table 1.  Rates of Antisocial Alcoholism in Biological Parents of Children With/Without Childhood ADHD and by CD Comorbidity
at Follow-Up.

ADHD probands ADHD probands

  Non-ADHD group ADHD group Without CD With CD

Mothers (%)
  No alcoholism or ASPD 198 (84) 211 (68) 162 (71) 49 (61)
  Alcoholism only 28 (12) 56 (18) 37 (16) 18 (22)
  ASPD only 6 (3) 22 (7) 10 (4) 9 (11)
  Both alcoholism and ASPD 3 (1) 23 (7) 19 (8) 4 (5)
  Total 235 312 228 80
Fathers (%)
  No alcoholism or ASPD 142 (63) 124 (43) 100 (47) 24 (32)
  Alcoholism only 32 (15) 49 (17) 38 (18) 11 (15)
  ASPD only 22 (10) 39 (13) 22 (10) 16 (21)
  Both alcoholism and ASPD 31 (14) 79 (27) 54 (25) 24 (32)
  Total 227 291 214 75

Note. CD = conduct disorder; ASPD = antisocial personality disorder.

Results with CD (43%; 92/214), χ2(1, N = 289) = 0.30. Rates of mater-

nal antisocial alcoholism were comparable for offspring with
Bivariate Regression Analyses Comparing ADHD and without comorbid CD (5% vs. 8%, respectively). Rates of
and Non-ADHD Groups paternal antisocial alcoholism were also similar for offspring
with and without comorbid CD (32% vs. 25%).
Rates of parental alcoholism, ASPD, and parental antisocial
alcoholism are reported in Table 1 for mothers and for
fathers, separately for the ADHD and non-ADHD groups, Multivariate Logistic Regression Analyses
and separately for probands with and without CD at follow- Comparing ADHD and Non-ADHD Groups
up. Using bivariate analyses, rates of maternal alcoholism
(with or without ASPD) were higher in the ADHD (25%; A series of planned pairwise contrasts (SPSS GLM) were
79/312) versus non-ADHD group (13%; 31/235), χ2(1, N = performed to examine whether parent alcoholism with and
547) = 12.3, p < .01, and rates of paternal alcoholism were without ASPD were each independently associated with
higher in the ADHD (44%; 128/291) versus non-ADHD offspring ADHD. First, alcoholism only, ASPD only, and
group (28%; 63/227), χ2(1, N = 518) = 14.4, p < .01. Also, antisocial alcoholism were compared with no psychopa-
rates of maternal antisocial alcoholism were higher in the thology. Then, antisocial alcoholism was compared with
ADHD (7%; 23/312) versus non-ADHD group (1%; 3/235), alcoholism only and to ASPD only. Offspring ADHD (no/
χ2(1, N = 545) = 11.1, p < .01, and rates of paternal antiso- yes) was the binary dependent variable. A similar set of
cial alcoholism were higher in the ADHD (27%; 79/291) pairwise comparisons were conducted within the ADHD
versus non-ADHD group (14%; 31/227), χ2(1, N = 518) = group to examine the associations of the parent diagnoses
13.9, p < .01. ADHD group fathers had the highest rate of with offspring ADHD plus CD diagnosis. We estimated all
antisocial alcoholism (27%), nearly double the rate for non- models controlling for parental ADHD, which accounted
ADHD group fathers (14%), and nearly four times the rate for 18.4% of the proband mothers with alcoholism and
for ADHD group mothers (7%). 18.7% of the proband fathers with alcoholism, versus none
of the non-ADHD mothers with alcoholism and 1% of the
non-ADHD fathers with alcoholism.
Bivariate Analyses Comparing ADHD Groups
Results are shown in Table 2. They show that, control-
With and Without Offspring CD Comorbidity ling for parental ADHD, mothers of offspring with ADHD
For ADHD group mothers, rates of alcoholism were not sig- were more likely than mothers of offspring without ADHD
nificantly different between those with (28%; 22/80) versus to have alcoholism, with or without ASPD. They were also
without offspring with CD (25%; 56/228), χ2(1, more likely to have ASPD without alcoholism. They were
N = 308) = 0.27. This was also the case for ADHD group not, however, more likely to have antisocial alcoholism
fathers: Rates of alcoholism were not significantly different than either disorder alone. Findings for fathers were the
between those with (47%; 35/75) versus without offspring same with the exception that likelihood of alcoholism alone
6 Journal of Attention Disorders 

Table 2.  Relation Between Maternal and Paternal Antisocial Alcoholism and Offspring ADHD and CD Among Offspring With

ADHD group (no/yes) Offspring CD (no/yes)

  Full sample Within ADHD group

  Contrast estimate (SE) p Contrast estimate (SE) p

  Alcoholism only vs. no diagnosis .129 (.06) .007 .080 (.07) NS
  Antisocial only vs. no diagnosis .259 (.10) .007 .209 (11) .05
  Antisocial alcoholism vs. no diagnosis .317 (.10) .002 .100 (.10) NS
  Alcoholism only vs. antisocial alcoholism −.158 (.11) NS .180 (.11) NS
  Antisocial alcoholism vs. antisocial only .058 (.13) NS .309 (.14) .027
  Alcoholism only vs. no diagnosis .114 (.06) .067 .001 (.08) NS
  Antisocial only vs. no diagnosis .170 (.07) .014 .234 (.08) .005
  Antisocial alcoholism vs. no diagnosis .227 (.06) .000 .104 (.06) .10
  Alcoholism only vs. antisocial alcoholism −.112 (.07) NS .118 (.09) NS
  Antisocial alcoholism vs. antisocial only .057 (.08) NS .129 (.09) NS

Note. Total ns for maternal models were 526 for the ADHD group analysis and 297 for the offspring CD analysis (within ADHD group), and for pa-
ternal models, they were 505 and 290, respectively. Models were estimated controlling for parental ADHD, offspring race/ethnicity and gender, family
socioeconomic status (SES; parent education), and presence of one or two biological parents currently in the family. With the exception of parental
ADHD, none of the covariates was significant and did not significantly change the interpretation of the results, and they were, therefore, trimmed from
the final models. CD = conduct disorder.

was only marginally elevated for fathers of offspring with mothers of children with ADHD, versus 28% of the fathers
ADHD. and 13% of the mothers of children without ADHD, experi-
enced a problem with alcohol at some time in their lives.
These group differences were statistically significant and
Multivariate Logistic Regression Analyses based on large sample sizes relative to most studies of this
Comparing ADHD Groups With and Without question in the literature (about 300 ADHD and just above
Offspring CD Comorbidity 200 non-ADHD families in our study vs. below 200 per
Conducted within the ADHD group only, these same pair- group in other widely cited clinic samples). Although our
wise contrasts resulted in fewer associations. Mothers of percentages appear high relative to some prior reports (e.g.,
probands with CD were more likely to have ASPD than Barkley et al., 1990; Biederman et al., 1992), they are lower
mothers of probands without CD, and mothers of probands than others (August et al., 1999) and lower than might be
with CD were more likely to have ASPD than antisocial expected based on national data collected within our win-
alcoholism compared with mothers of probands without dow of study (42% of adult men and 19.5% of adult women
CD. Fathers of probands with CD were also more likely to had a DSM-IV lifetime AUD in the 2001-2002 National
have ASPD than fathers of probands without CD. Fathers of Epidemiologic Survey on Alcohol and Related Conditions
probands with CD were only marginally more likely to have [NESARC]; Hasin, Stinson, Ogburn, & Grant, 2007). Thus,
antisocial alcoholism than fathers of probands without CD. our rates for index parents are high but not out of line with
those expected in the lifetime of adults in the United States.
Our methods did not permit distinctions between lifetime
Discussion and current diagnoses or abuse versus dependence; such
The results of this study are consistent with a long line of distinctions can lower rates dramatically and also hint at the
studies demonstrating associations between parent alcohol- likely possibility that alcohol problems pre-dated offspring
ism and offspring externalizing behaviors and support our ADHD in many of these families. For example, only 3.8%
hypothesis that alcoholism is more prevalent among parents of mid-aged adults (30-44 years old) in the NESARC sur-
of children with, versus without, childhood ADHD. The vey were diagnosable with alcohol dependence (vs. abuse)
current findings extend prior research by demonstrating the in the past year (Hasin et al., 2007). Taken together, our
associations for a large sample of children rigorously diag- findings and those of others (e.g., August et al., 1999;
nosed with ADHD in childhood and for mothers and fathers Barkley et al., 1990; Biederman et al., 1992; Skoglund
separately. Forty-four percent of the fathers and 25% of the et al., 2015) provide further evidence for the importance of
Molina et al. 7

screening for a positive family history of alcohol problems et al., 2008) and small numbers of antisocial mothers may
among individuals being assessed for ADHD, and this risk have led to the removal of negative socialization effects by
should include maternal history despite the generally higher the most impaired parents (fathers in particular). We note,
rate of alcohol problems among men versus women more however, that ASPD among parents—among either fathers
generally (Hasin et al., 2007). As articulated within a devel- or mothers—was a statistically significant correlate with
opmental psychopathology perspective of the ontogeny of proband CD as an adolescent or young adult. These findings
externalizing disorders (Beauchaine & McNulty, 2013), confirm and extend those of LeMoine et al. (2015) to moth-
although at any given assessment a patient’s parents may ers, by showing an association between parental (both
not be actively experiencing alcohol-related difficulties, mother and father) antisociality and child conduct problem
inherited, prenatal, and other environmental liabilities comorbidity in offspring ADHD.
toward alcoholism may still be at play. Moreover, although We were especially interested in effects pertaining to
trends are shifting, mothers continue to assume much of the gender of parent and hypothesized that mothers (like
parenting responsibilities (Yogman, Garfield, & Committee fathers) of children with ADHD would also have higher
on Psychosocial Aspects of Child and Family Health, 2016), rates of alcoholism than mothers of children without
and the possibility of maternal alcohol use contributing to ADHD. The results suggest that maternal (not just pater-
environmental risk should be considered. nal) alcoholism and ASPD might be important factors in
Our findings are partially in line with studies suggesting the transmission of risk for externalizing disorders. While
a genetic link between parent alcoholism and offspring the vast majority of literature to date has focused on
ADHD (Knopik et al., 2006; Knopik et al., 2005) that may paternal alcoholism risk, partly due to the lower rates of
be particularly prominent in fathers. After controlling for alcoholism among women compared with men, our find-
paternal ADHD, the association for fathers was only mar- ings encourage ongoing consideration of maternal alco-
ginally significant for alcoholism without antisociality and holism as a risk factor for offspring adjustment. There is
statistically significant for antisocial alcoholism. In fact, evidence to suggest that maternal alcoholism is associ-
62% of the 128 proband fathers with alcohol problems had ated with offspring substance use disorders (e.g., Chassin
comorbid antisociality. Thus, we may be seeing what et al., 1996) and externalizing disorders such as ADHD
Knopik and colleagues describe as cross-generational (e.g., Knopik et al., 2006). There is also evidence to sug-
genetic covariation (Knopik et al., 2009) in which parents gest that psychosocial mediators such as adolescent stress
transmit genetic vulnerability to offspring for multiple dis- mediate the association between maternal alcoholism and
orders that include alcoholism, ADHD, and antisociality. offspring substance use (Chassin et al., 1996), and that
Investigators have written about familial risk of alcoholism this mediated pathway (for maternal or paternal alcohol-
being transmitted via multiple mechanisms reflecting the ism) may be stronger in offspring with ADHD than in
heterogeneity of alcoholism more generally (Zucker et al., controls (Marshal, Molina, Pelham, & Cheong, 2007).
2011). Our findings support the hypothesis that antisocial Our data indicated a nearly two-fold higher likelihood of
alcoholism may be particularly important in the familial maternal alcoholism among mothers of children with
transmission of risk for an aggregated vulnerability to anti- ADHD than among mothers of children without ADHD
social alcoholism and ADHD among males. Twenty-seven (25% vs. 13%, respectively), indicating the possibility
percent of the probands had a father with antisocial alcohol- (which remains to be tested) that maternal alcoholism
ism (the highest parental diagnostic subgroup across analy- partly represents a path of vulnerability toward alcohol-
ses for mothers and for fathers). ism among some children with ADHD. Indeed, we previ-
Our hypothesis that parental antisocial alcoholism would ously found that children trained to enact behaviors
be most prevalent for probands with CD was not supported. characteristic of children with externalizing disorders
Five percent of the probands with CD had an antisocial (uncooperative, distractible, etc.) cause elevated levels of
alcoholic mother versus 8% of the probands without CD. stress in, and alcohol consumption by, mothers in a labo-
Among fathers, substantial minorities of offspring were ratory setting (Pelham et al., 1997). Maternal alcohol con-
captured in both sides of this comparison and the difference sumption also exacerbated maladaptive parenting (Lang,
was only marginally significant. Thirty-two percent of the Pelham, Atkeson, & Murphy, 1999). Parenting may be
probands with CD had antisocial alcoholic fathers versus another important mediator of maternal (as well as pater-
25% of the probands without CD. One possible explanation nal) alcoholism in families with a child with ADHD. This
for our unexpected findings, aside from small cell sizes, is an important hypothesis to test in future longitudinal
may lie in the contribution of child rearing to the ultimate analyses, as well as conducting an exploration of the
behavioral (i.e., conduct-related) outcome of the child. potential mediators and moderators of maternal alcohol-
Although this serious parental comorbidity should theoreti- ism among youth with ADHD in relation to the long-term
cally lead to the worst offspring outcome (Zucker et al., adult outcomes of ADHD persistence, antisociality, and
1993), relatively high divorce rates in our sample (Wymbs alcoholism among offspring.
8 Journal of Attention Disorders 

In sum, this study provides evidence of an association Journal of the American Academy of Child & Adolescent
between parental alcoholism, antisocial alcoholism, and Psychiatry, 38, 1262-1270.
offspring ADHD, for fathers and for mothers. There are Babor, T. F., Hesselbrock, V., Meyer, R. E., & Shoemaker, W.
some limitations to our findings including the clinic- (1994). Types of alcoholics: Evidence from clinical, experi-
mental, and genetic research. New York: New York Academy
referred nature of the sample and the timing of our assess-
of Sciences.
ments. We were not able to study the course of parental
Barkley, R. A. (Ed.). (2015). Attention-deficit hyperactivity disor-
alcoholism or ADHD prospectively, which may ultimately der: A handbook for diagnosis and treatment (4th ed.). New
provide more detailed and veridical information about alco- York, NY: Guilford Press.
holism and ADHD chronicity over time and within families. Barkley, R. A., Fischer, M., Edelbrock, C. S., & Smallish, L.
Our assessment of offspring outcome was limited to CD at (1990). The adolescent outcome of hyperactive children
initial follow-up for a wide age range. Examination of off- diagnosed by research criteria: I. An 8-year prospective fol-
spring outcome beyond late adolescence and early adult- low-up study. Journal of the American Academy of Child &
hood, to mid-adulthood when developmentally limited Adolescent Psychiatry, 29, 546-557.
expressions of behavior have subsided, and chronic alcohol Beauchaine, T. P., & McNulty, T. (2013). Comorbidities and
problems are most apparent, will be important. Future stud- continuities as ontogenic processes: Toward a develop-
mental spectrum model of externalizing psychopathology.
ies that include a larger proportion of female offspring with
Developmental Psychopathology, 25, 1505-1528.
ADHD, and that examine mediators and moderators of the
Biederman, J., Faroane, S. V., Keenan, K., Benjamin, J., Krifcher,
association between parents’ antisocial alcoholism and all B., Moore, C., . . . Tsuang, M. T. (1992). Further evidence
offspring outcomes into adulthood, will improve under- for family-genetic risk factors in attention deficit hyperactiv-
standing of the ADHD–alcohol link in families. In the ity disorder: Patterns of comorbidity in probands and rela-
meantime, the results of the current study highlight the tives in psychiatrically and pediatrically referred samples.
importance of considering multi-layered risk within fami- Archives of General Psychiatry, 49, 728-738. doi:10.1001/
lies that may include some combination of ADHD, alcohol- archpsyc.1992.01820090056010
ism, and antisociality, for both males and females. Biederman, J., Petty, C. R., Wilens, T. E., Fraire, M. G., Purcell,
C. A., Mick, E., . . . Faraone, S. V. (2008). Familial risk analy-
Declaration of Conflicting Interests ses of attention deficit hyperactivity disorder and substance
use disorders. American Journal of Psychiatry, 165, 107-115.
The author(s) declared no potential conflicts of interest with Cadoret, R. J., & Stewart, M. A. (1991). An adoption study of
respect to the research, authorship, and/or publication of this attention deficit/hyperactivity/aggression and their rela-
article. tionship to adult antisocial personality. Comprehensive
Psychiatry, 32, 73-82.
Funding Chassin, L., Curran, P. J., Hussong, A. M., & Colder, C. R. (1996).
The author(s) disclosed receipt of the following financial support The relation of parental alcoholism to adolescent substance
for the research, authorship, and/or publication of this article: This use: A longitudinal follow-up study. Journal of Abnormal
research was supported by Grants AA011873, AA00202, and Psychology, 105, 70-80.
AA015100. The research was also partly supported by Grants Compton, W. M., Conway, K. P., Stinson, F. S., Colliver, J. D., &
AA08746, AA12342, AA0626, DA12414, DA05605, DA034731, Grant, B. F. (2005). Prevalence, correlates, and comorbidity
F31DA017546, MH12010, MH4815, MH47390, MH45576, of DSM-IV antisocial personality syndromes and alcohol and
MH50467, MH53554, MH092466, MH083692, MH069614 specific drug use disorders in the United States: Results from
MH0099030, MH097819, IES R324A120169, IES R324B060045, the national epidemiologic survey on alcohol and related con-
IES L03000665A and ES0515, and MH018951. ditions. Journal of Clinical Psychiatry, 66, 677-685.
Elliot, D. Z., Huizinga, D., & Ageton, S. S. (1985). Explaining
delinquency and drug use. Beverley Hills, CA: SAGE.
First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W.
American Psychiatric Association. (1987). Diagnostic and statis- (1998). Structured Clinical Interview for DSM-IV Axis I
tical manual of mental disorders (3rd ed., rev.). Washington, Disorders—Non-patient edition (SCID-I/NP, Version 2.0—
DC: Author. 8/98 revision). New York: Biometrics Research Department,
American Psychiatric Association. (2000). Diagnostic and statis- New York State Psychiatric Institute.
tical manual of mental disorders (4th ed.). Washington, DC: Haber, J. R., Jacob, T., & Heath, A. C. (2005). Paternal alcoholism
Author. and offspring conduct disorder: Evidence for the “common
American Psychiatric Association. (2013). Diagnostic and sta- genes” hypothesis. Twin Research and Human Genetics, 8,
tistical manual of mental disorders (5th ed.). Arlington, VA: 120-131.
American Psychiatric Publishing. Hasin, D. S., Stinson, F. S., Ogburn, E., & Grant, B. F. (2007).
August, G. J., Realmuto, G. M., Tamara, J., & Hektner, J. M. Prevalence, correlates, disability, and comorbidity of DSM-IV
(1999). Persistence and desistence of oppositional defiant alcohol abuse and dependence in the United States: Results
disorder in a community sample of children with ADHD. from the National Epidemiologic Survey on Alcohol and
Molina et al. 9

Related Conditions. Archives of General Psychiatry, 64, Milberger, S., Faraone, S. V., Biederman, J., Chu, M. P., & Wilens,
830-842. T. (1998). Familial risk analysis of the association between
Herndon, R. W., & Iacono, W. G. (2005). Psychiatric disorder in attention-deficit/hyperactivity disorder and psychoactive sub-
the children of antisocial parents. Psychological Medicine, stance use disorders. Archives of Pediatrics & Adolescent
35, 1815-1824. Medicine, 152, 945-951.
Hicks, B. M., Foster, K. T., Iacono, W. G., & McGue, M. (2013). Molina, B. S. G., Donovan, J. E., & Belendiuk, K. A. (2010).
Genetic and environmental influences on the familial trans- Familial loading for alcoholism and offspring behavior:
mission of externalizing disorders in adoptive and twin off- Mediating and moderating influences. Alcoholism: Clinical
spring. JAMA Psychiatry, 70, 1076-1083. & Experimental Research, 34, 1972-1984.
Kendler, K. S., Ohlsson, H., Sundquist, K., & Sundquist, J. Molina, B. S. G., Pelham, W. E., Gnagy, E. M., Thompson, A., &
(2016). Cross-generational transmission from drug abuse in Marshal, M. P. (2007). ADHD risk for drinking and alcohol
parents to attention-deficit/hyperactivity disorder in children. use disorder is age specific in adolescence and young adult-
Psychological Medicine, 46, 1301-1309. hood. Alcoholism: Clinical & Experimental Research, 31,
Knopik, V. S., Heath, A. C., Jacob, T., Slutske, W. S., Bucholz, 643-654.
K. K., Madden, P. A. F., . . . Martin, N. G. (2006). Maternal Molina, B. S. G., Walther, C. A. P., Cheong, J., Pedersen,
alcohol use disorder and offspring ADHD: Disentangling S., Gnagy, E. M., & Pelham, W. E. (2014). Heavy alco-
genetic and environmental effects using a children-of-twins hol use in early adulthood as a function of childhood
design. Psychological Medicine, 36, 1461-1472. doi:10.1017/ ADHD: Developmentally-specific mediation by social
S0033291706007884 impairment and delinquency. Experimental and Clinical
Knopik, V. S., Jacob, T., Haber, J. R., Swenson, L. P., & Howell, Psychopharmacology, 22, 110-121.
D. N. (2009). Paternal alcoholism and offspring ADHD prob- Pelham, W. E., Gnagy, E., Greenslade, K. E., & Milich, R. (1992).
lems: A children of twins design. Twin Research and Human Teacher ratings of DSM-III-R symptoms for the disruptive
Genetics, 12, 53-62. behavior disorders. Journal of the American Academy of
Knopik, V. S., Sparrow, E. P., Madden, P. A. F., Bucholz, Child & Adolescent Psychiatry, 31, 210-218.
K. K., Hudziak, J. J., Reich, W., . . . Heath, A. C. (2005). Pelham, W. E., & Hoza, B. (1996). Intensive treatment: A summer
Contributions of parental alcoholism, prenatal substance treatment program for children with ADHD. In E. D. Hibbs
exposure, and genetic transmission to child ADHD risk: A & P. S. Jensen (Eds.), Psychosocial treatments for child
female twin study. Psychological Medicine, 35, 625-635. and adolescent disorders: Empirically based strategies for
doi:10.1017/S0033291704004155 clinical practice (pp. 311-340). Washington, DC: American
Kuperman, S., Schlosser, S. S., Lidral, J., & Reich, W. (1999). Psychological Association.
Relationship of child psychopathology to parental alcoholism Pelham, W. E., & Lang, A. R. (1993). Parental alcohol consump-
and antisocial personality disorder. Journal of the American tion and deviant child behavior: Laboratory studies of recipro-
Academy of Child & Adolescent Psychiatry, 38, 686-692. cal effects. Clinical Psychology Review, 13, 763-784.
Lahey, B. B., Piacentini, J. C., McBurnett, K., Stone, P., Hartdagen, Pelham, W. E., Lang, A. R., Atkeson, B., Murphy, D. A., Gnagy,
S., & Hynd, G. (1988). Psychopathology in the parents of E. M., Greiner, A. R., . . . Greenslade, K. E. (1997). Effects
children with conduct disorder and hyperactivity. Journal of of deviant child behavior on parental distress and alcohol
the American Academy of Child & Adolescent Psychiatry, 27, consumption in laboratory interactions. Journal of Abnormal
163-170. Child Psychology, 25, 413-424.
Lang, A. R., Pelham, W. E., Atkeson, B. M., & Murphy, D. A. Selzer, M. L., Vinokur, A., & van Rooijen, L. (1975). A self-
(1999). Effects of alcohol intoxication on parenting in inter- administered Short Michigan Alcoholism Screening Test
actions with child confederates exhibiting normal or devi- (SMAST). Journal of Studies on Alcohol, 36, 117-126.
ant behaviors. Journal of Abnormal Child Psychology, 27, Shaffer, D., Fisher, P., Lucas, D. P., Dulcan, M. K., & Schwab-
177-189. Stone, M. E. (2000). NIMH Diagnostic Interview Schedule
Lee, S. S., Humphreys, K. L., Flory, K., Liu, R., & Glass, K. for Children Version IV (NIMH DISC-IV): Description, dif-
(2011). Prospective association of childhood attention-deficit/ ferences from previous versions, and reliability of some com-
hyperactivity disorder (ADHD) and substance use and abuse/ mon diagnoses. Journal of the American Academy of Child &
dependence: A meta-analytic review. Clinical Psychology Adolescent Psychiatry, 39, 28-38.
Review, 31, 328-341. Sher, K. J. (1991). Children of alcoholics: A critical appraisal
LeMoine, K. A., Romirowsky, A. M., Woods, K. E., & Chronis- of theory and research. Chicago, IL: University of Chicago
Tuscano, A. (2015). Paternal antisocial behavior (but not Press.
paternal ADHD) is associated with negative parenting and Sibley, M. H., Pelham, W. E., Molina, B. S. G., Gnagy, E.
child conduct problems. Journal of Attention Disorders. M., Waxmonsky, J. G., Derefinko, K., . . . Kuriyan, A. B.
Advance online publication. doi:10.1177/1087054715604361 (2012). When diagnosing ADHD in young adults emphasize
Marshal, M. P., Molina, B. S. G., Pelham, W. E., & Cheong, J. informant reports, DSM items, and impairment. Journal of
(2007). ADHD moderates the life stress pathway to alcohol Consulting and Clinical Psychology, 80, 1052-1061.
problems in children of alcoholics. Alcoholism: Clinical & Skoglund, C., Chen, Q., Franck, J., Lichtenstein, P., & Larsson, H.
Experimental Research, 31, 564-574. (2015). Attention-deficit/hyperactivity disorder and risk for
10 Journal of Attention Disorders 

substance use disorders in relatives. Biological Psychiatry, Zucker, R. A., Heitzeg, M. M., & Nigg, J. T. (2011). Parsing the
77, 880-886. undercontrol-disinhibition pathway to substance use disorders:
Spitzer, R., Williams, J., & Gibbon, B. (1987). Structured Clinical A multilevel developmental problem. Child Development
Interview for DSM-III-R Personality Disorders (SCID-II). New Perspectives, 5, 248-255.
York: New York State Psychiatric Institute, Biometrics Research.
Waldron, M., Martin, N. G., & Heath, A. C. (2009). Parental Author Biographies
alcoholism and offspring behavior problems: Findings in Brooke S. G. Molina, PhD, is a professor of Psychiatry and
Australian children of twins. Twin Research and Human Psychology and director of the Youth and Family Research
Genetics, 12, 433-440. Program/ADHD Research Program at the University of Pittsburgh.
Wong, M. M., Zucker, R. A., Puttler, L. I., & Fitzgerald, H. E. Dr. Molina’s research focuses on the longterm course, etiology,
(1999). Heterogeneity of risk aggregation for alcohol prob- and treatment of ADHD and substance abuse with a focus on ado-
lems between early and middle childhood: Nesting structure lescence and young adulthood.
variations. Development and Psychopathology, 11, 727-744.
Wymbs, B. T., Pelham, W. E., Jr., Molina, B. S., Gnagy, E. M., Elizabeth M. Gnagy, BS, is a senior research support specialist at
Wilson, T. K., & Greenhouse, J. B. (2008). Rate and predic- the Florida International University Center for Children and
tors of divorce among parents of youths with ADHD. Journal Families. She has served as a senior data manager and analyst for
of Consulting and Clinical Psychology, 76, 735-744. the Pittsburgh ADHD Longitudinal Study since its inception.
Yogman, M., Garfield, C. F., & Committee on Psychosocial
Heather M. Joseph, DO, is a postdoctoral scholar and a clinical
Aspects of Child and Family Health. (2016). Fathers’ roles in
assistant professor at the University of Pittsburgh. Her research
the care and development of their children: The role of pedia-
focuses on the familial transmission of ADHD and medication
tricians. Pediatrics, 138, e20161128.
management in primary care.
Zucker, R. A., Ellis, D. A., & Fitzgerald, H. E. (1993).
Developmental evidence for at least two alcoholisms: I. William E. Pelham, Jr., PhD, ABPP, is a distinguished pro-
Biopsychosocial variation among pathways into symptomatic fessor of Psychology and Psychiatry, director, Center for
difficulty. In T. F. Babor, V. M. Hesselbrock, R. E. Meyer, Children and Families (CCF), and chair of the Department of
& W. Shoemaker (Eds.), Annals of the New York Academy of Psychology at Florida International University. CCF is a mul-
Sciences: Vol. 708. Types of alcoholics: Evidence from clini- tidisciplinary research center focusing on the development and
cal, experimental, and genetic research (pp. 134-146). New dissemination of evidence-based treatments for childhood
York: New York Academy of Sciences. behavioral disorders.