Journal of Learning Disabilities
Volume 31, Number 6, November/December 1998, Pages 533 – 544
Prospective Study of Tobacco Smoking and
Substance Dependencies among Samples of
ADHD and Non-ADHD Participants
Nadine M. Lambert and Carolyn S. Hartsough
Abstract
This study focused on an audience at high risk for heavy use of licit and illicit substances: young adults who as children had
attention-deficit/ hyperactivity disorder (ADHD). The participants in this study were part of a longitudinal study of the life
histories of 492 children, one third of whom were identified as hyperactive in 1974 and whose childhood symptom ratings and
medical histories were used to establish Diagnostic and Statistical Manual of Mental Disorders (3rd ed., revised; DSM-III-R)
ADHD diagnoses (American Psychiatric Association, 1987). The objectives of the study centered on describing (a)
developmental history of tobacco use among ADHD and non-ADHD participants in a longitudinal sample, (b) the characteristic
adult patterns of tobacco use from early adolescence through early adulthood, and (c) the relationship between ADHD status and
tobacco and substance dependence outcomes. Adult data were obtained for 81% of the original 492 participants (77% of the
ADHD and 86% of the controls). Lifetime and current tobacco use were assessed from child, adolescent, and adult data, yielding
eight measures of smoking status. The study showed that participants with and without ADHD did not differ in age of initiation
to smoking, but there was a significant difference in the age they began smoking regularly. By age 17, 46% of all participants
with ADHD, as contrasted with 24% of the age-mate controls, reported smoking cigarettes daily. In adulthood, the proportion of
participants with ADHD who were current smokers (42%) continued to exceed that of the age-mate controls (26%). Among
current adult smokers, 35% with ADHD smoked daily as compared to 16% of the age-mate controls. There were significantly
different lifetime tobacco dependence rates-40% compared to 19% for age-mate controls. The rates for cocaine dependence were
21% for participants with ADHD and 10% for age-mate controls. We reported a significant difference in rates of daily smoking
and tobacco dependence for those with ADHD who had used stimulant medication in childhood in contrast to controls. Results
were interpreted to support a possible link between ADHD treatment histories and levels of tobacco smoking and tobacco
dependence in adulthood.
T
he major goal of this investigation was to explore the use of tobacco by participants with and without
ADHD and to compare these usage rates with adult substance use and dependencies. Earlier research has
reported (Hartsough & Lambert, 1987; Lambert, 1988) that children identified and treated for hyperactivity
in 1974 smoked more than age-mates by the end of high school, but that by late adolescence the progression
of substance use for the hyperactive participants was similar to that of the age-mate controls. Because the
participants were not at the age of highest initiation into various substances at the time of those investigations, the
present study reclassified participants by DSM-III-R research criteria (American Psychiatric Association, 1987) as
ADHD and non-ADHD and sought to further explore the rates of tobacco use and dependency, as well as
differences in rates of other substance dependencies in adulthood.
There have been reports from other longitudinal studies of the rates of illegal drug use by adults with ADHD
(Gittelman, Mannuzza, Shenker & Bonagura, 1985; Mannuzza, Gittelman-Klein & Addalli, 1991), but these
published follow-up investigations have not presented evidence on differences in tobacco use among ADHD and
control participants. Some major longitudinal studies have reported a pattern of excessive tobacco use and illegal
substance use among either participants with concurrent childhood conduct problems (Lynskey & Ferguson, 1995)
or participants who subsequently develop adolescent conduct disorder (Barkley, Fischer, Edelbrock & Smallish, et
al., 1990).
Other investigators (Henningfield, Cohen & Slade, 1991; Henningfield, Clayton & Pollin, 1990) have noted that
tobacco dependence not only is an important addiction on its own but also is involved in the development of a
variety of other drug dependencies (Fleming, Leventhal, Glynn & Ershler, 1989; Torabi, Bailey & Madj-Jabbari,
1993). For example, studies of the development of drug use consistently show that most people who use illegal
drugs had earlier used cigarettes or alcohol (Kandel, 1980; Kandel, Kessler & Margulies, 1978; Kandel, Yamaguchi
& Chen, 1992), whereas those who had never smoked only infrequently abused illicit substances. Tobacco,
therefore, is considered to be a gateway drug to the development of other dependencies. The incidence and severity
of various drug dependencies seem to be related to tobacco use, and tobacco use, in turn, may be increased by
dependence-producing drugs. Thus, there is increasing evidence from a variety of sources that “tobacco use is
involved, possibly more than by simple association, in the use of other substances containing psychoactive
chemicals” (Henningfield et al., 1990, p. 279). The evidence for significantly higher rates of tobacco use among
individuals with ADHD suggests that this group is more susceptible to early tobacco use, which, in turn,
predisposes them to higher rates of other substance dependencies in adulthood.
Objectives
The participants in this research were involved in a longitudinal study begun in 1974 of the life histories of
community samples of 492 children, approximately one third of whom were diagnosed and treated in childhood for
hyperactivity. This study is one of only a handful of studies (Borland & Hechtman, 1976; Gittelman et al., 1985;
Livson & Leino, 1985; Loney, 1980; Weiss, Hechtman, Milroy & Perlman, 1978; Werner, 1989) that have collected
longitudinal data prospectively on this high-risk group.
To understand the onset and course of tobacco use among children, adolescents and adults in this sample, we
developed methods for locating and interviewing all of our 492 participants first identified in 1974, and devised an
adult interview protocol to produce measures of tobacco use and dependence, drug abuse and dependencies, social
and psychiatric outcomes, educational and work histories, relationships and lifestyle information, and assessment of
health status. A summary of our procedures and findings pertaining to tobacco and other substances is presented in
this report.
Using this life-history database, we investigated the following questions:
1
2
3
What are the characteristic cigarette and other drug use patterns for samples of individuals classified as ADHD
with differing medical histories, compared with those for age-mate controls, who were attaining adulthood in
the 1990s?
a Are there significant differences between participants with ADHD and non-ADHD participants in the age
at which they report initiation into smoking and becoming regular smokers?
b What are the rates of tobacco use among participants with ADHD and non-ADHD participants at the end
of high school and in adulthood?
What are the concomitant rates of tobacco dependence and other substance dependencies as measured by DSMIII-R psychiatric diagnoses in adulthood?
Do subgroups of participants, specifically those with ADHD in childhood who received stimulant medication as
contrasted with those who had not had stimulant treatment, have different tobacco use profiles and differential
rates of substance abuse and dependencies?
Method
Participants
The community samples of 492 participants for this investigation were identified in 1974 from a study of the
prevalence of hyperactivity among a representative population of 5,212 school-age children in kindergarten through
fifth grade from 191 classrooms representing the public, parochial and private schools in the East Bay Region of the
San Francisco area. Socioeconomic status of the participants was proportional to the rates in the overall census data
for the rural, suburban and urban areas in the East Bay Region of the San Francisco Metropolitan Area. Twenty
three percent of the participants were members of minority ethnic groups.
Our procedure for locating all of the hyperactive children within these sampled classrooms (Lambert, Sandoval &
Sassone, 1978; 1979) involved contacting all of the parents of children in the surveyed classrooms, their teachers,
pupil service workers and principals, as well as the physicians who were treating them. From these extensive
records we identified five samples for our prospective studies, as described next. The participants were evaluated
prospectively through the end of high school and now, as young adults.
Early in the longitudinal work, the DSM-II (American Psychiatric Association, 1968) represented the standard for
medical diagnosis. Our surveys of physicians (pediatricians, pediatric neurologists, child psychiatrists and family
practice physicians) who were treating children referred for evaluation for hyperactivity (Sandoval, Lambert &
Yandell, 1979) helped us in developing a standard medical evaluation and diagnostic system to evaluate children
who would b e identified for prevalence studies as hyperactive. The cardinal symptoms used for the diagnosis at the
time were “over-activity, restlessness, distractibility and short attention span” (American Psychiatric Association,
1968, p. 50). As the DSM criteria were changing over the years of our work, we took advantage of the participants’
home and school ratings of the participants to develop research diagnostic criteria for DSM-III (Lambert, 1988) and,
for this investigation, research diagnostic criteria for DSM-III-R ADHD so that our participant characteristics and
assignment to groups would be consistent with current diagnostic nomenclature. Parent and teacher ratings from the
Children’s Attention and Adjustment Survey (CAAS; Lambert, Hartsough & Sandoval, 1990) and medical history
data were available from the treating physicians. The home and school versions of the CAAS each provide four
scales with good psychometric properties, which we labeled “Inattention,” “Hyperactivity,” “Impulsivity” and
“Conduct Problems.” These dimensions replicated the symptoms specified in the DSM-III (Attention Deficit
Disorder With and Without Hyperactivity; American Psychiatric Association, 1980) and the DSM-III-R (Attention
Deficit Hyperactivity Disorder) criteria (American Psychiatric Association, 1987).
We recast all of our participants by research diagnostic criteria for DSM-III-R ADHD for this investigation. The
research criteria were the total ADHD ratings-that is, the sum of ratings on the Hyperactivity, Inattention and
Impulsivity items for the home and the school forms of the CAAS (alpha = .87 for Home Form and .94 for School
Form). We also utilized each participant’s medical and symptom history records. This reclassification procedure
combined information regarding onset of symptoms, medical diagnostic findings and treatment histories and criteria
for parent and teacher ratings of ADHD and resulted in the following groups.
Primary ADHD Participants: (n = 175: 152 Males, 23 Females) These individuals met the following criteria: (a)
Situational ADHD (either home or school ADHD rating on the CAAS within a selection range presented in the
CAAS manual) or Pervasive ADHD (both home and school ADHD ratings within the selection range), (b) medical
intervention and diagnosis of hyperactivity (using the diagnostic labels provided by our study of medical practices
in 1974 and without evidence of competing etiological explanations) and / or CNS treatment for more than 6
months, (c) no medical prescription of anticonvulsant medication and (d) parent report of hyperactive symptoms
before age 8.
Secondary ADHD Participants: (n = 39: 30 Males, 9 Females) The criteria for assignment to this group were (a)
Situational or Pervasive ADHD on the CAAS, (b) a medical diagnosis implicating organic factors accounting for
the presenting symptoms of hyperactivity, (c) use of anticonvulsant medication for more than 6 months and (d)
parent report of hyperactive symptoms before age 8.
ADHD Control Participants: (n = 68: 56 Males, 12 Females) These individuals met the following criteria: (a)
Situational or Pervasive ADHD on the CAAS; (b) considered to be hyperactive by two of three social systems-the
parent, the teacher or the physician; (c) parent report of the appearance of symptoms before age 8; and (d) no
evidence of medical diagnosis or CNS treatment (confirmed by subsequent parent and participant interviews).
Behavior Problem Controls: (n = 51: 43 Males, 8 Females) Although these participants initially received high
total scores on the CAAS, including ADHD symptoms, conduct problems and affective behaviors, they were
classified by (a) absence of teacher, physician or parent report that the child had ever been considered hyperactive;
(b) no evidence of medical diagnosis of hyperactivity, medical treatment for hyperactivity or medical treatment with
anticonvulsants; and (c) no evidence of the appearance of any relevant symptoms in early childhood.
Age-Mate Controls: (n = 159: 103 Males, 56 Females) These individuals were selected from the classrooms in
which the school-referred ADHD participants were enrolled and matched by birth date and gender. There was a
preponderance of boys because more boys were found in the hyperactive samples.
These five samples of participants and their life-history databases were unique in several ways. Unlike most longterm studies of ADHD, our samples were community-based, not clinic-based. They were evaluated, diagnosed, and
treated by physicians, and rated by parents and teachers on behaviors associated with hyperactivity, impulsivity, and
inattention, and other behavioral symptoms, throughout childhood and adolescence. Few longitudinal studies of
individuals with ADHD have prospective data on age-mate problem-behavior controls and age-mate problem-free
controls, and very few, if any, have life-history data that include information about the participants from the
perspectives of parents, teachers and the participants themselves. Any findings, therefore, are likely to have high
external validity.
When we completed the revised participant assignment, the next step was to determine whether there were
sufficient similarities among primary, secondary, and ADHD control groups to combine them into a single group for
some of the proposed analyses. To accomplish this, we compared the life-history data for the three groups to check
for consistencies or differences among the groups with respect to health, familial, educational, and social
background factors. The results of these analyses were consistent with earlier reports (Hartsough & Lambert, 1982,
1985; Lambert, 1982, 1988). With a few minor differences, the primary, secondary, and ADHD control groups had
similar behavior and family histories, and the ADHD groups were different from both the behavior control and the
age-mate control groups. These analyses, then, justified the further grouping of participants into large ADHD (n =
282) and non-ADHD (n = 210) groups, when indicated, for the proposed prospective study of tobacco smoking and
substance use.
Procedure
Throughout the longitudinal work from elementary through high school, and now adulthood, our periodic
interviews queried participants regarding their use of tobacco, alcohol, and various illegal substances. These data
enabled us to develop a life-history record of tobacco and other substance use that included the age at which each
substance was first tried, the rates of use during different age periods, and the frequency of use during adulthood.
Because our participants were born in the 1960s and different drug use patterns might be expected, depending on
factors associated with availability and public sanctions on drug use, we next examined cohort effects, or
differences in substance use rates by birth year, and determined that there were no differences of any magnitude
among birth-year cohorts. We then produced a continuous life-history record for each participant for tobacco and all
other substances used in childhood, adolescence, and adult-hood.
Adult Interview
The adult interview provided information that was consistent with longitudinal research describing the
developmental course of several adult outcomes, including tobacco and substance abuse (Ahlgren, Norem,
Hochhauser & Garwin, 1982; Hawkins, Catalano & Miller, 1992; Swan, Murray & Jarrett, 1991; Werner, 1989).
The interview contains eight major sections, each of which was developed after consultation with investigators
conducting major longitudinal studies and after extensive reviews of the relevant literature. The goal of this process
was to construct the best possible measures of adult functioning, arranged in a format that would take no more than
4 hours of interview time.
The questions addressing the participants’ use of tobacco as adults, as well as their attitudes and opinions regarding
smoking, were derived from two sources: (a) the California Smoking Baseline Survey: Adult Attitudes and Practices
(CSBS; Pierce & Burns, 1990), and (b) the Quick Diagnostic Interview Schedule (3rd ed., revised; QDIS-III-R;
Marcus, Robins & Bucholz, 1990). The CSBS includes questions on current smoking status, recent smoking history,
lifetime smoking history, other tobacco use (such as pipe and chewing tobacco), smoking in the workplace, and
questions about smoking and pregnancy for the female participants. The CSBS has been used in several
epidemiological studies of changes in risk factors for smoking, as well as in evaluation of the outcomes of
community health education interventions (Fortmann, Taylor, Flora & Farkas, 1997; Hovell et al., 1996).
The QDIS-III-R is a computer-administered interview based on the NIMH DIS-III-R (Robins, Helzer, Cottler &
Goldring, 1989) and provides a DSM-III-R diagnosis of tobacco abuse and dependence, as well as dependence
diagnoses of alcohol, marijuana, stimulants, and cocaine. The QDIS-III-R diagnoses and our research diagnostic
criteria for ADHD thus were consistent with the DSM-III-R criteria. The diagnostic interview scheduled for the
DSM-IV was not available at the time that the adult follow-up was being conducted. The QDIS-III-R provides
psychiatric diagnoses of the major mental disorders; however, for this investigation we have reported only those
diagnoses pertaining to substance dependence. Comparisons between the computer-administered and clinicianadministered DIS interviews (Erdman et al., 1992) show consistency in diagnoses across methods. Furthermore,
studies of the stability (Vandiver & Sher, 1991) and retest reliability (Ross, Swinson, Doumani & Larkin 1995) of
the DIS reveal good reliabilities across methods and over time.
We explored several approaches to uncovering evidence of response bias in self-report measures in our prospective
longitudinal database. The participants’ histories included interview questions about smoking and substance use
collected over several times during childhood and adolescence, making possible a comparison of adult reports of
tobacco smoking and use of other substances with the participants’ reports at an earlier period. We noted a
remarkable consistency in these reports, such as an adult report of the age at which regular smoking began with data
from childhood that they were indeed involved in smoking at that age. Our confidence in the substance use data was
supported. Recently, Babinski, Hartsough, and Lambert (in press) studied the concordance rate of self-reported
measures of criminal involvement and official arrest records for a sample that included individuals with and without
a childhood history of ADHD. For the men, kappa coefficients indicated good agreement between arrest records and
self-report measures for 8 of the 12 types of crimes, particularly for more serious crimes, such as robbery and
assault with a weapon. Poorer agreement was found for the less serious, more frequently committed crimes, such as
disorderly conduct and vandalism.
According to the evidence available, the agreement between self-report measures over time and between self-report
measures and measures from other sources for unselected samples is likely to be sufficiently high to provide reliable
measures of the outcomes of interest in this investigation.
Tobacco and Substance Use Variables
Each participant’s life-history record included age of initiation, age at which regular smoking began, use of tobacco
and other substances from childhood through adulthood, and the QDIS-III-R substance dependence diagnoses for
the major substances.
Tobacco Variables: From those sources the following tobacco-smoking variables were derived:
1
2
3
4
5
6
7
Never smoked, not even a puff;
Age at which a participant was initiated into smoking;
Age at which participant began smoking regularly;
Smoking status at age 17;
Adult smoking status;
a Reported smoking less than 100 cigarettes in lifetime;
b In adulthood, smoked more than 100 cigarettes, but not smoking currently (“former smoker”);
c Smoked more than 100 cigarettes, and currently smoking (“smoker”);
d If a smoker, smoked daily or less than daily. Participants who reported heavy use of smokeless tobacco or
other forms of nicotine were also classified as “smokers”;
Lifetime tobacco use-a composite variable using age at which regular smoking started and current smoking
status to reflect length of time a participant had been smoking;
Tobacco dependent, as rated on the QDIS-III-R.
Alcohol, Marijuana, Stimulants and Cocaine Outcome Measures: The substance abuse and dependence
measures were the QDIS-III-R diagnoses of alcohol, marijuana, stimulant and cocaine dependence.
Results
Adult Participants
The resources of the research department of the California Department of Motor Vehicles and other searching
procedures provided current addresses of 92% of the 492 participants in our database (Hartsough, Babinski, &
Lambert, 1996). We wrote sequentially to sets of participants within designated geographic regions, detailing the
goals of the project and requesting that they contact us to schedule an interview. A $50 payment was an important
incentive. Following up letters with phone calls provided an opportunity to further explain the details of our project.
Table 1 shows that we ultimately interviewed 81% of the original sample of 492. An additional 2% were deceased,
and 3% refused to be interviewed. At the end of the time set aside for the interviews, there were still some
participants for whom we had not been able to locate an updated address.
TABLE 1
Completed Interviews for Participants with ADHD and Non-ADHD Participants
Outcome of follow-up with adult participants
Group
ADHD
n = 282
Non-ADHD
n = 210
All participants
n = 492
Interviewed
Mentally
incompetent
n
%
n
%
218
77
2
182
86
400
81
Refused
Deceased
Still trying
Not found
n
%
n
%
n
%
n
%
<1
10
4
5
2
12
4
33
12
_
_
4
2
6
3
12
6
8
4
2
<1
14
3
11
2
24
5
41
8
Representativeness of Interviewed Sample: In response to any concern that loss to follow-up may be biased, that
is, highly related to the primary outcome (tobacco and other substance dependence), we compared the participating
versus non-locatable ADHD participants on initial childhood symptom data as well as demographic data from
family of origin. These analyses showed no differences of any significance. Cognizant, as well, that the differential
attrition rates might be associated with drug abuse or dependency and thereby also limit our conclusions on the
relationship between tobacco smoking and other substances, we evaluated any potential selective bias in follow-up
loss with respect to drug use using criminal justice records. These criminal justice reports consist of rap sheets
describing the nature and number of the offenses, including drug offenses (although not broken down by category
of drug involved), committed by any one of the original 492 participants in our study.
Using these rap sheets, we estimated the impact of “not found” participants on our reports of drug dependency and
abuse rates for ADHD and non-ADHD participants and produced two estimates of dependency rates that would be
achieved if we were to interview all of the “still trying” and “not found” participants. The first estimate examined
the rates of marijuana, stimulant and cocaine dependency among the participants who also had criminal histories of
drug violations. Admittedly, this estimate is a lower bound estimate. The second estimate was based on the
assumption that all of the not-found participants who had controlled substance or narcotics violations would be, in
turn, solely marijuana dependent, or stimulant dependent, or cocaine dependent. Again, this estimate purposely errs
in the opposite direction and can be considered a higher bound estimate. Comparing this range of estimates with the
rates for the participants interviewed, we determined that there was no appreciable impact on our reported rates of
substance use that could be attributed to our inability to locate and interview the “still trying” and “not found”
participants.
Representativeness of the Age-mate control Sample Smoking Rates: Another way to examine the
representativeness of our participant samples is to compare rates of “never smoked” and adult smoking to rates
reported for national samples. At the end of high school, 33% of the age-mate control participants reported that they
had never smoked. This rate is identical to the youth prevalence rate of 33% for a national sample of high school
seniors in 1987 (U.S. Department of Health and Human Services [USDHHS], 1989). These national adult
prevalence rates of “never smoked” from several surveys compare favorably to those rates for the age-mate control
participants, namely 26%, compares favorably with the prevalence data reported in the Adult Use of Tobacco
Survey in 1986 of 26.5%, and the NIDA National Household Interview Survey conducted in 1986 and 1987 of
29.4% and 29.1%, respectively (USDHHS, 1989). The 1985 California Population Survey reported a rate of 25.6%
for adults (USDHHS, 1989)—again, very close to our reported rate of smoking for the age-mate controls (27%). We
believe that it is reasonable to infer, therefore, that our age-mate control participants are representative of the
general population with respect to their smoking status.
Age of Initiation and Patterns of Smoking
Rates of Initiation: Data on age of initiation into smoking were obtained from the childhood and adolescent
interviews verified by the reports of our participants as adults in response to the question “When did you first try a
cigarette?” Eight percent of both the primary ADHD and the ADHD control participants and 2% of the age-mate
controls began smoking before 8 years of age. There were no significant differences among the mean ages of
initiation for the participant groups. The average age of initiation into tobacco, namely, 12.7 for all participants with
ADHD and 12.1 for the age-mate controls, was comparable for all groups. For female participants the age of
initiation was 12.3 for age-mate controls, and 13.2 for those with ADHD.
Ages at Which Participants Began Smoking Regularly: The average ages at which the primary ADHD,
secondary ADHD, and ADHD control participants became regular smokers were 15.2, 15.8 and 16.2, respectively.
The average age the Behavior Problem Control participants became regular smokers was 15.4. The age-mate
controls became regular smokers, on the average, at age 17.1. For male and female participants with ADHD, the age
of regular smoking was 15.5 and 15.6 years, respectively, and for age-mate controls, 16.9 and 17.5. Because the
differences between men and women in relation to the age of initiation and age of regular smoking, as well as the
rates for being adult smokers, were few, we treated the men and women as a single group for the analyses in this
study.
We then combined all of the individuals with ADHD and conducted a one-way analysis of variance to determine
whether there were significant differences in the age at which participants began regular smoking. The results
showed that both the ADHD group and the behavior controls were significantly different from the age-mate controls
in age the participants began smoking regularly, F(2, 197) = 4.1731, p < .02.
Smoking Status of Participants at Age 17: Our next examination of smoking behaviors of our participants with
ADHD and non-ADHD participants centered on smoking status at age 17. Here we were interested in the
proportions of our participants who, by age 17, had never tried a cigarette, those who had been initiated and
occasionally smoked but not daily, and those who reported daily smoking (half a pack or more every day; see Table
2). By age 17, 50% of all participants with ADHD, as contrasted with 43% of the behavior problem controls, and
27% of age-mate controls smoked cigarettes daily. The chi-square of 22.97 (4, n = 345) was significant at p < .001.
Post hoc contrasts indicated that the ADHD groups and the behavior problem control group were significantly
different from the age-mate controls, but the ADHD groups and the behavior problem control group were not
significantly different from one another. These results suggest that problem behavior in childhood reflects a general
risk factor for smoking.
TABLE 2
Rates at Age 17 of Daily Smoking, some Smoking, and Never Smoked Among
all ADHD, Behavior Control, and Age-Mate Control Participants
Group
Smoking status
All ADHD
Behavior controls
Age-mate controls
at age 17
(n = 193)
(n = 46)
(n=117)
Truly never
21
14
33
Some smoking
30
43
41
Regular daily smoking
50
43
27
note: χ2 (4, N = 345) = 22.97, p < .001
Rates of Smoking in Adulthood
Adult Smoking Status of Participants: Following the tobacco-use definitions derived from the California
Smoking Baseline survey (Pierce & Burns, 1990), we classified reports of adult smoking status as (a) never tried a
cigarette, (b) smoked less than 100 cigarettes in lifetime, (c) former smoker (smoked more than 100 cigarettes in
lifetime but not smoking at the time of the interview), and (d) smoked more than 100 cigarettes in lifetime and was a
current smoker. Table 3 shows that by adulthood, 42% of the subjects with ADHD had smoked more than 100
cigarettes and were current smokers, as compared with 36% of the behavior problem controls and 26% of the agemate controls.
Smoking status was defined by assigning participants to groups on the basis of daily smoking. A daily smoker had
used more than 100 cigarettes by adulthood and reported smoking on a daily basis. Table 4 redistributes the rates in
Table 3 according to proportions of participants who never were initiated into tobacco use, those who had smoked,
and those who smoked daily at the time of the adult interview. The results showed that participants with ADHD
who were smokers in adulthood differed from comparison groups in the amount of daily tobacco use, a finding that
suggests a second explanation for the differential rates of smoking status in adulthood, namely, that childhood
ADHD symptomatology places an individual at risk for high rates of tobacco use in adulthood.
TABLE 3
Adult Smoking Status of Participants with ADHD and Non-ADHD Participants
Adult
smoking status
Nonsmoker
0 cigarettes
< 100 cigarettes
Truly never
Some smoking
Primary
(n = 136)
ADHD groups
Secondary
(n = 33)
Control
(n = 52)
All ADHD
(n = 221)
Non-ADHD controls
Behavior
Age-mate
(n = 42)
(n = 142)
13
31
18
24
8
33
12
30
5
43
21
40
11
18
15
13
14
12
43
39
42
42
36
26
Regular daily
2
2
2
2
smoking
Note: Chi-square analysis rates of all ADHDs, behavior controls, and age-mate controls for nonsmokers, former smokers, smokers
(including other nicotine users) was χ2(8, N = 405) = 18.69, p < .02.
<1
TABLE 4
Rates of Adult Daily Smoking, Some Smoking and Never Smoked Among All ADHD, Behavior Control
and Age-Mate Control Participants
Groups
Adult Smoking Status
All ADHD
Behavior controls
Age-mate controls
Truly never
Some smoking
12
5
53
71
21
63
Regular daily smoking
35
24
16
note: χ2 (4, N = 405) = 22.52, p < .000
Substance Dependencies for ADHD and control Participants
Table 5 displays the proportions of participants in each group who received a QDIS-III-R diagnosis of substance
abuse and dependence. We used the chi-square test of homogeneity to test the statistical reliability of differences
among the ADHD groups first. Finding no differences in rates of substance dependence among the ADHD groups,
we grouped the participants with ADHD together and compared their rates with those of the behavior control and
age-mate control groups. There were no significant differences between participants with ADHD and the control
groups on alcohol and marijuana dependence. There were, however, significant differences between participants
with ADHD and controls on tobacco (p < .000), stimulant (p < .05), and cocaine (p < .05) dependence based on the
QDIS-III-R substance use diagnoses.
The QDIS-III-R diagnoses indicated that participants with ADHD were significantly more likely to be dependent on
stimulants, as opposed to dependence on either alcohol or marijuana. These findings implicate ADHD
symptomatology, rather than behavior problems, as a risk factor for tobacco, cocaine, and stimulant dependency in
adulthood.
TABLE 5
Percentage of Participants with ADHD and Age-Mate Controls with QDIS-III-R Psychiatric Diagnoses
ADHD groups
QDIS-III-R diagnoses
of substance
dependence
Primary
(n = 128)
Secondary
(n = 30)
Control
(n = 46)
All ADHD
(n = 204)
Non-ADHD controls
AgeBehavior
mate
(n = 41)
(n = 134)
Tobacco
Alcohol
Marijuana
Stimulants
Cocaine
41
42
33
33
41
28
40
38
32
39
19
30
28
27
22
27
34
29
20
23
26
22
20
11
20
27
18
21
12
10
Note: There were no significant differences in rates of dependence on substances among ADHD groups. A chi-square test of homogeneity showed
the following significant differences among the ADHD groups combined, the behavior controls, and the age-mate controls: tobacco, χ2(2, 379)
= 17.34 p < .000; stimulants, χ2(2, 379) = 6.56 p < .05; and cocaine: χ2(2, 378) = 6.78 p < .05.
Prior Use of Stimulant Medication
We next directed our attention to examining the question of whether the use of stimulant medication in childhood
was related to the use of tobacco, stimulants and cocaine in adulthood. Among participants with ADHD, 48% had
used stimulant medication for 6 months or more.
Stimulant Medication and Rates of Daily Smoking: First, we examined differential rates of adult smoking
among participants with ADHD with different stimulant medication histories. We compared the ADHD participants
who had been treated with stimulant medication for up to 1 year and those who never were treated with stimulant
medication, those who were treated with CNS stimulants for up to a year, and those who were treated for a year or
more with the rates of adult smoking for participants with ADHD who had not received stimulant medication. The
rate of 41% of daily smoking among ADHD participants who had used stimulant medication for a year or more
contrasted with a rate of 37% for participants who had no stimulant medication history. The χ2 was a significant,
and the Mantel-Haenszel, a measure of a linear trend in the data, was also significant. Stimulant medication history
of these participants with ADHD was significantly related to the rates of adult smoking. (See Table 6)
TABLE 6
Adult Smoking Rates for ADHD Smokers with Different Cerebral Stimulant Histories
Amount of adult smoking
Smoker, but not daily
Smoker, and smokes daily
Column percent
ADHD smokers with different stimulant medication histories
1 year or
Never
Up to 1 year
Row percent
more
(n = 47)
(n = 28)
(n=52)
55.3
42.9
28.8
41.7
44.7
57.1
71.2
37.0
22.0
Note: χ2 (2, N = 127) = 7.13, p < .03; Mantel-Haenszel = 7.07, p < .01.
40.9
58.3
Stimulant Medication and Substance Dependence Diagnoses: We then examined the rates of QDIS-III-R
diagnoses with respect to use of stimulant medication in childhood (see Table 7). For both cocaine and tobacco
dependence, there was a significant Mantel-Haenszel indicating a linear relationship between amount of stimulant
medication and the likelihood of becoming either tobacco dependent or cocaine dependent. The χ2 for cocaine was
significant as well but did not reach significance for tobacco dependence. The relationship between medication in
childhood and tobacco and cocaine dependence in adulthood suggests a third hypothesis to be explored: that prior
exposure to cerebral stimulants may make participants more susceptible to tobacco and cocaine dependence in
adulthood.
Additional careful analysis of the co-occurrence of tobacco and other substance use at age 17 and 18 and in
adulthood provided some intriguing findings supporting the role of tobacco smoking as a risk factor in cocaine and
stimulant dependence in early adulthood (Davidson, Lambert, Hartsough & Schenk, 1996). Breslau and colleagues
(Breslau, Kilbey & Andreski, 1991; Kilbey, Breslau & Andreski, 1992) also have reported a much lower rate for
cocaine dependence than for alcohol dependence among adults who were not tobacco dependent.
TABLE 7
Percentage of Participants with ADHD with QDIS-III-R Drug Dependence Diagnoses with Different
Cerebral Stimulant Histories
Stimulant medication histories
1 year or
Percent of ADHDs
Never
Up to 1 year
QDIS-III-R dependence
more
(n = 81)
(n = 9)
dependent
outcomes
(n=84)
Tobacco
32.1
38.5
48.8
40.2
32.1
33.3
45.2
37.7
Alcohol
22.5
23.1
32.1
26.6
Marijuana
21.0
15.4
26.2
22.1
Stimulants
15.0
17.9
27.4
20.7
Cocaine
Note: There were no significant differences for use of stimulant medication in childhood and dependence on alcohol, marijuana, or stimulants.
For tobacco dependence and cocaine dependence, the statistical findings were as follows: χ2(2, N = 204) = 4.85, p < .08, Mantel-Haenszel (1, N
= 204) =4.77, p < .03; cocaine: χ2(2, N = 204) =4.05, p < . 13, Mantel-Haenszel (1, N = 204) = 3.83, p < .05.
Discussion
In this effort to extend research on the smoking behavior of ADHD and non-ADHD participants studied
prospectively since childhood, the goals were to describe the initiation into tobacco use, age of regular smoking,
smoking status at age 17, and adult smoking status of the participants. In addition, we were interested in contrasting
the DSM-III-R substance dependencies for participants with ADHD and non-ADHD participants at adulthood.
The results showed that there were no differences among the participant groups on age of initiation into smoking,
but there were differences among the ADHD participant groups, the behavior controls and the age-mate controls in
the age at which participants began smoking regularly. This result supported hypotheses of problem behavior in
general, rather than ADHD status, as the factor explaining the higher prevalence of childhood smoking. But when
we examined the proportion of participants who were smokers in adulthood, defined as having smoked 100
cigarettes in lifetime and currently smoking, the proportion of smokers among participants with ADHD as a group
were significantly different from the controls’.
We then grouped our participants who met the “smoker” criteria into two groups-those who smoked every day, and
those who reported smoking less than daily. When we analyzed the rates of participants with these two patterns of
smoking at both age 17 (or the end of high school) and in early adulthood, the differences between the groups were
even more pronounced. On the basis of the adolescent and adult rates of daily smoking among the ADHD groups in
contrast to the behavior and age-mate controls, we inferred that the behaviors that characterize the ADHD
participant, namely, inattention, hyperactivity, and impulsivity, place a participant more at risk for late adolescent
and adult involvement with cigarettes and nicotine than being a participant with general problem behavior.
We carried this question further by investigating the adult substance use dependence of the ADHD groups and the
control groups. Using the QDIS-III-R, a computerized diagnostic interview based on the NIMH Diagnostic
Interview Schedule, we reported that the rate of substance dependence for the participants with ADHD was
significantly higher for tobacco dependence and cocaine dependence than for alcohol or marijuana. Although the
rates of stimulant dependence were higher for the participants with ADHD, they were not significantly different
from those reported for the control groups. These results suggested that participant with ADHD as a group were
more likely to become dependent on stimulants -tobacco and cocaine- in adulthood than were the behavior problem
and age-mate controls. The implication of the higher prevalence of dependence on substances with stimulating
properties is an important finding.
We explored this higher rate of dependence on stimulants for the participants with ADHD by analyzing the extent to
which the use of stimulant medication in childhood was significantly related to daily smoking in adulthood, and to
adult lifetime dependence on stimulants. Among the participants with ADHD who were smokers in adulthood, 52%
had received stimulant medication in childhood. Ninety-three percent of these were daily smokers in adulthood in
contrast to 80% of the ADHD adult smokers who had not been treated with stimulant medication. This result did
not reach significance, χ2 (1, 85) = 3.04, p < .08. Next we examined the significance of differences for the
medicated and non-medicated participants with ADHD with respect to the DSM-III-R dependence diagnoses. Table
6 displays the proportions of dependence diagnoses for the two groups. Tobacco dependence was significantly
higher for the medicated participants with ADHD. There are three major, and most likely complementary,
explanations regarding the relationship of ADHD symptomatology/treatment and tobacco and cocaine/stimulant
abuse.
The first explanation is that general behavior dysfunction in childhood and adolescence, characterized by
psychosocial unconventionality or the presence of conduct disorders, antisocial behavior, or other psychiatric
problems (all of which are also prevalent among ADHD groups), leads both to more smoking and intensive
substance use as well as to use of a variety of different substances. These results support the hypothesis put forth by
Jessor and his colleagues as well as other investigators (Jessor, 1982, 1991; Loney, 1980; Robins, 1980) that
problem behavior and conduct problems in childhood predispose to substance use in adolescence and adulthood.
Tobacco use in adolescence and adulthood, therefore, is likely to be common among all who use other substances.
Although this explanation offers a valuable perspective from which to conceptualize the issue at hand, our
preliminary analyses suggest a stronger ADHD-tobacco-cocaine connection than can be accounted for by the
general behavior dysfunction or problem behavior theories.
A second explanation proposes that tobacco serves a self-medicating role in participants with ADHD and that
initiation into, and continued use of, nicotine is sought because of its beneficial behavioral effects. Research with
human participants has suggested that both tobacco and cocaine are used for self-medication, depending on the
particular type of presenting symptomatology. For example, in Rounsaville, Anton, Carroll and Budde’s (1991)
survey of adult patients seeking treatment for cocaine abuse, 35% of them had been diagnosed with ADHD. Weiss
and Mirin (1986) reported that there were five types of cocaine abusers, among which were adults with ADHD who
used cocaine to increase their attention span and reduce motor restlessness. Cocores, Davies, Mueller and Gold
(1987) likewise have suggested that ADHD has an etiological and self-medicating role in cocaine abuse. Coming at
the ADHD/cocaine issue from another perspective, Kleber and Gawin (1986) and Gawin and Kleber (1986) treated
adult cocaine abusers with and without an ADHD diagnosis with methylphenidate and reported that the
methylphenidate treatment effected reduced cocaine use in the ADHD participants but not in those without the
ADHD symptomatology.
Although the data are only preliminary, a possible third explanation is the methylphenidate/amphetamine
sensitization hypothesis that posits the tobacco and / or amphetamine use, regardless of behavioral symptomatology,
predisposes one to stimulant and cocaine dependence in early adulthood. Thus, although the participants with
ADHD may have more risk factors predisposing them to adult cocaine abuse, the fundamental processes involved in
the sensitization hypothesis are thought to hold regardless of psychiatric symptomatology.
Noting that animal studies have shown that pre-exposure to either nicotine or amphetamines enhanced response to
cocaine, we sought evidence (Davidson, Lambert, Hartsough & Schenk, 1996) on the relative roles of
methylphenidate use and tobacco as potential sensitizing agents in lifetime use of cocaine. The model explored in
this preliminary work was derived from animal models (Horger, Giles & Schenk, 1992; Horger, Shelton & Schenk,
1991; Schenk, Lacelle, Gorman & Amit, 1987) suggesting that enhanced responsiveness to cocaine’s reinforcing
properties is a reflection of increased neurochemical sensitization. We predicted lifetime use of cocaine utilizing
hierarchical multiple regression for preliminary exploration of a model containing methylphenidate and tobacco
exposure plus a set of variables chosen to serve as controls for competing explanations of the predicted variable.
The preliminary results indicated that there was no significant explained variance contributed by gender, ADHD, or
conduct problems, but that both methylphenidate exposure and tobacco use were significant contributors to cocaine
use. Further studies of participants who become cocaine dependent in contrast to those who are only heavy users of
cocaine would be suggested by the findings reported here.
That methylphenidate was still capable of explaining a significant proportion of the variance in cocaine use even
after approximately 15 years is of great importance. Although this too, was predicted by the animal model, the
number of variables that could interact with the medication effect in humans is relatively high. Thus, both
predictions that received some support in our preliminary studies and the research herein-that of methylphenidate
exposure and of nicotine exposure as sensitizing agents in cocaine use-will require extensive further development
and testing of models of the relationship of childhood symptomatology and stimulants, including tobacco exposure,
in the prediction of cocaine use and dependence. The distinctions between these models and those developed to
explain marijuana and alcohol use must also be examined. Our preliminary analysis suggests that those who were
cocaine dependent in early adulthood as compared to those who were marijuana or alcohol dependent were more
likely to have been early and regular cigarette smokers in adolescence, and those who were diagnosed with ADHD
in childhood were more likely to smoke earlier and regularly.
Conclusions
This investigation of participants with ADHD and age-mate controls followed prospectively since childhood
confirms that ADHD is a contributory factor in adolescent and adult tobacco use and dependencies on substances
with stimulating properties among a community-based sample. Data were presented on ages of initiation into
tobacco use, ages at which participants became regular smokers, rates of smoking at age 17, and details on adult
smoking status of participants with and without ADHD. In addition, dependency rates for tobacco, alcohol,
marijuana, stimulants and cocaine were provided.
The major findings reported from this study include the following:
1.
2.
3.
4.
5.
6.
7.
The ADHD and control groups did not differ significantly on age of initiation into smoking, although 8% of the
ADHD as compared to 2% of the age-mate controls had begun smoking prior to 8 years of age.
The ADHD adolescent smokers, on average, began smoking regularly shortly after age 15, as compared to the
age-mate control smokers, who did not smoke regularly, on average, until age 17.
By age 17, 46% of all participants with ADHD, as contrasted with 24% of the age-mate controls, had, at some
point, smoked cigarettes daily.
In adulthood, the proportion of participants with ADHD who are current smokers (42%) continues to exceed
that of the age-mate controls (26%).
In adulthood, the rates of daily smoking among the adult smokers was 35% for the participants with ADHD in
contrast to 16% for the age-mate controls.
Using DSM-III-R diagnostic criteria, participants with ADHD demonstrated lifetime tobacco dependence rates
of 40%, compared to 19% for age-mate controls.
Analyses of the potential contribution of the use of stimulant medication in childhood for the participants with
ADHD suggested a possible link between stimulant medication treatment and rates of adult smoking as well as
adult diagnoses of tobacco and cocaine dependence.
ABOUT THE AUTHORS
Nadine M. Lambert, PhD, is a professor of education and the director of the School Psychology Program at the University of California at
Berkeley. Her current research interests involve studies of the developmental course of ADHD symptom patterns and their relationship to
substance use and mental health outcomes. Carolyn S. Hartsough, PhD, is research educator and the coordinator of the School Psychology
Program at the University of California at Berkeley. Her research interests include educational measurement and evaluation and longitudinal
studies of ADHD. Address: Nadine M. Lambert, Graduate School of Education, University of California at Berkeley, Berkeley, CA 94720-1670.
AUTHORS’ NOTE
The research reported in this manuscript was supported by the Tobacco-Related Disease Research Program, “Prospective Study of Cigarette
Smoking and Behavior Dysfunction,” August 1990 to June 1994.
REFERENCES
Ahlgren, A., Norem, A. A., Hochhauser, M., & Garvin, J. (1982). Antecedents of smoking among pre-adolescents.
Journal of Drug Education, 12, 325 – 340
American Psychiatric Association. (1968) Diagnostic and statistical manual of mental disorders (2nd ed.). Washington,
DC: Author.
American Psychiatric Association. (1980) Diagnostic and statistical manual of mental disorders (3rd ed.). Washington,
DC: Author.
American Psychiatric Association. (1987) Diagnostic and statistical manual of mental disorders (3rd ed. rev.).
Washington, DC: Author.
Babinski, L. M., Hartsough, C. S., & Lambert, N. M. (in press). Comparison of self-reports of criminal activity and official
arrest records for individuals with a history of behavioral problems. Psychology, Psychiatry and Allied Disciplines.
Barkley, R. A., Fischer, M., Edelbrock, C. S., & Smallish, L. (1990, July). Journal of the Academy of Child and Adolescent
Psychiatry, 29, 546-557.
Borland, B. L., & Hechtman, H. K. (1976). Hyperactive boys and their brothers: A 25-year follow-up study. Archives of
General Psychiatry, 33, 669-675.
Breslau, N., Kilbey, M. M., & Andreski, P. (1991). Nicotine dependence, major depression, and anxiety in young adults.
Archives of General Psychiatry, 48, 1069-1074.
Cocores, J. A., Davies, R. K., Mueller, P. S., & Gold, M. S. (1987). Cocaine abuse and adult attention deficit disorder.
Journal of Clinical Psychiatry, 48, 376-377.
Davidson, E. S., Lambert, N. M., Hartsough, C. S., & Schenk, S. (1996). Methylphenidate (Ritalin) and nicotine exposure
as risk factors for cocaine abuse. Manuscript submitted for publication.
Erdman, H. P.., Klein, M. H., Greist, J. H., Skare, S. S., Husted, J., Robins, L., Helzer, J. E., Goldring, E., Hamburger, M.,
& Miller, P. (1992). A comparison of two computer-administered versions of the NIMH Diagnostic Interview
Schedule. Journal of Psychiatric Research, 26, 85-95.
Fleming, R., Leventhal, H., Glynn, K., & Ershler, J. (1989). The role of cigarettes in the initiation and progression of early
substance use. Addictive Behaviors, 14, 261-272.
Fortmann, S. P., Taylor, C. B., Flora, J. A., & Jatulis, D. E. (1993). Changes in adult cigarette smoking prevalence after 5
years of community health education: The Stanford Five-City Project. American Journal of Epidemiology, 137, 8296.
Gawin, F. H., & Kleber, H. (1986). Pharmacologic treatments of cocaine abuse. Psychiatric Clinics of North America, 9,
573-583.
Gilpin, E. A., Pierce, J. P., & Farkas, A. J. (1997). Duration of smoking abstinence and success in quitting. Journal of the
National Cancer Institute, 89, 571-576.
Gittelman, R., Mannuzza, S., Shenker, R., & Bonagura, N. (1985) Hyperactive boys almost grown up: Psychiatric status.
Archives of General Psychiatry, 42, 937-947.
Hartsough, C. S., Babinski, L. M., & Lambert, N. M. (1996, September). Tracking procedures and attrition containment in
a long-term follow-up of a community-based ADHD sample. Journal of Child Psychology and Psychiatry, 37, 705713.
Hartsough, C. S., & Lambert, N. M. (1982). Some environmental and familial correlates and antecedents of hyperactivity.
American Journal of Orthopsychiatry, 52, 272-287.
Hartsough, C. S., & Lambert, N. M. (1985). Medical factors in Hyperactive and normal children: Prenatal, developmental
and health history findings. American Journal of Orthopsychiatry, 55, 190-201.
Hartsough, C. S., & Lambert, N. M. (1987). Pattern and progression of drug use among hyperactives and controls: A
prospective short-term longitudinal study. Journal of Child Psychology and Psychiatry. 28, 543-553.
Hawkins, J. D., Catalano, R. F., & Miller, J. Y. (1992). Risk and protective factors for alcohol and other drug problems in
adolescence and early adulthood: Implications for substance abuse prevention. Psychological Bulletin, 112(1), 64105.
Henningfield, J. E., Clayton, R., & Pollin, W. (1990). Involvement of tobacco in alcoholism and illicit drug use. British
Journal of Addiction, 85, 279-291.
Henningfield, J. E., Cohen, C., & Slade, J. D. (1991). Is Nicotine more addictive than cocaine? Special Issue: Future
directions in tobacco research. British Journal of Addiction, 86, 565-569.
Horger, B. A., Giles, M. K., & Schenk, S. (1992). Pre-exposure to amphetamine and nicotine predisposes rats to selfadminister a low dose of cocaine. Psychopharmacology, 107, 271-276.
Horger, B. A., Shelton, K., & Schenk, S. (1991). Pre-exposure sensitizes rats to the rewarding effects of cocaine.
Pharmacology, Biochemistry and Behavior, 37, 707-711.
Hovell, M. F., Slymen, D. J., Keating, K. J., Jones, J. A., Burkham-Kreitner, S., Hofstetter, C. R., Noel, D., & Rubin, B.
(1996). Tobacco prevention trial in California, USA. Journal of Epidemiology and community Health, 50, 340-346.
Jessor, R. (1982). Problem behavior and developmental transition in adolescence. Journal of School Health, 52, 295-300.
Jessor, R. (1991). Risk behavior in adolescence: A psychosocial framework for understanding and action. Special issue:
Adolescents at risk. Journal of Adolescent Health, 12, 597-605.
Kandel, D. B. (1980). Developmental stages in adolescent drug involvement. In D. J. Letteri, M. Sayers, & H. W. Pearson
(Eds.), Theories on drug abuse: Selected contemporary perspectives (pp. 120-127). Rockville, MD: National
Institute on Drug Abuse.
Kandel, D. B., Kessler, R. C., & Margulies, R. A. (1978). Antecedents of adolescent initiation into stages of drug use: A
developmental analysis. Journal of Youth and Adolescence, 7, 13-40.
Kandel, D. B., Yamaguchi, K, & Chen, K. (1992). Steps in the progression in drug involvement from adolescence to
adulthood: Further evidence for the gateway theory. Journal of Studies in Alcohol, 53, 447-457.
Kilbey, H., Breslau, N., & Andreski, P. (1992). Cocaine use and dependence in young adults: Associated psychiatric
disorders and personality traits. Drug and Alcohol Dependence, 29, 283-290.
Kleber, H., & Gawin, F. H. (1986) Psychopharmacological trials in cocaine abuse treatment. American Journal of Drug
and Alcohol Abuse, 12, 235-246.
Lambert, N. M. (1982). Temperament profiles of hyperactive children. American Journal of Orthopsychiatry, 52, 458467.
Lambert, N. M. (1988). Adolescent outcomes for hyperactive children: Perspectives on general and specific patterns of
childhood risk for adolescent educational, social, and mental health problems. American Psychologist, 43, 786-799.
Lambert, N. M., Hartsough, C. S., & Sandoval, J. (1990). The children’s attention and adjustment survey. Palo Alto, CA:
Consulting Psychologists Press.
Lambert, N. M., Sandoval, J., & Sassone, D. (1978). Prevalence of hyperactivity in elementary school children as a
function of social system definers. American Journal of Orthopsychiatry, 48, 446-463.
Lambert, N. M., Sandoval, J., & Sassone, D. (1979). Prevalence of treatment regimens for children considered to be
hyperactive. American Journal of Orthopsychiatry, 49, 482-490.
Livson, N., & Leino, E. V. (1985). Adolescent personality antecedents of adult cigarette smoking: A longitudinal study.
Journal of Genetic Psychology, 146, 343-355.
Loney, J. (1980). The Iowa theory of substance abuse among hyperactive adolescents. In D. J. Letteri, M. Sayers, & H. W.
Pearson (Eds.), Theories on drug abuse: Selected contemporary perspectives (pp. 131-136). Rockville, MD:
National Institute on Drug Abuse.
Lynskey, M. T., & Ferguson, D. H. (1995), Childhood conduct problems, attention deficit behaviors and adolescent
alcohol, tobacco and illicit drug use. Journal of Abnormal Child Psychology, 23, 281-302.
Mannuzza, S., Gittelman-Klein, R., & Addalli, K. A. (1991). Young adult mental status of hyperactive boys and their
brothers: A prospective follow-up study. Journal of the American Academy of Child and Adolescent Psychiatry, 30,
743-751.
Marcus, S. C., Robins, L. N., and Bucholz, K. K. (1990) Quick diagnostic interview schedule, III-R. [computer program]
St. Louis, MO: Department of Psychiatry, Washington University School of Medicine.
Pierce, J., & Burns, D. (1990). California smoking baseline survey: Adult attitudes and practices. Unpublished survey,
University of California, San Diego.
Robins, L. N. (1980). The natural history of drug abuse. In D. J. Letteri, M. Sayers, & H. W. Pearson (Eds.), Theories on
drug abuse: Selected contemporary perspectives (pp. 215-224). Rockville, MD: National Institute on Drug Abuse.
Robins, L., Helzer, J., Cottler, L., & Goldring, E. (1989). NIMH Diagnostic Interview Schedule Version III-Revised (DISIII-R). St. Louis, MO: Washington University School of Medicine.
Ross, H. E., Swinson, R., Doumani, S. & Larkin, E. J. (1995). Diagnosing comorbidity in substance abusers: A
comparison of the test-retest reliability of the two interviews. Journal of Drug and Alcohol Research, 21, 167-185.
Rounsaville, B. J., Anton, S. F., Carroll, K., & Budde, D. (1991). Psychiatric diagnosis of treatment seeking cocaine
abusers. Archives of General Psychiatry, 48, 43-51.
Sandoval, J., Lambert, N. M., & Yandell, W. (1979). Current medical practice and hyperactive children. American
Journal of Orthopsychiatry, 46, 323-324.
Schenk, S., Lacelle, G., Gorman, K., & Amit, Z. (1987). Cocaine self-administration in rats influenced by environmental
conditions: Implications for the etiology of drug abuse. Neuroscience Letters, 81, 227- 231.
Swan, A. V., Murray, M., & Jarrett. L. (1991). Smoking behavior from pre-adolescence to young adulthood. Brookfield,
VT: Gower.
Torabi, M. R., Bailey, W. J. & Majd-Jabbari, M. (1993). Cigarette smoking as a predictor of alcohol and other drug use by
children and adolescents: Evidence of the gateway drug effect. Journal of School Health, 63, 302-305.
U.S. Department of Health and Human Services. (1989). Reducing the health consequences of smoking: 25 years of
progress. A report of the Surgeon General. Washington, DC: Author.
Vandiver, T., & Sher, K. J. (1991). Temporal stability of the Diagnostic Interview Schedule. Psychological Assessment, 3,
277-281.
Weiss, G., Hechtman, L., Milroy, T., & Perlman, T. (1978). Hyperactives as young adults: School, employer, and selfrating scales obtained during the ten year follow-up evaluation. American Journal of Orthopsychiatry, 48, 438-445.
Weiss, R. D., & Mirin, S. M. (1986). Subtypes of cocaine abusers. Psychiatric Clinics of North America, 9, 491-501.
Werner, E. E. (1989). High risk children in young adulthood: A longitudinal study from birth to 32 years. American
Journal of Orthopsychiatry, 59, 72-81.
Journal of Learning Disabilities
Volume 31, Number 6, November/December 1998, Pages 533 - 544