konstrukt ADHD u dorosłych.pdf

Journal of Abnormal Psychology

0021-843X/07/$12.00 DOI: 10.1037/0021-843X.116.3.519

Axis I and II Comorbidity in Adults With ADHD

Torri W. Miller and Joel T. Nigg

Michigan State University

Stephen V. Faraone

State University of New York Upstate Medical University

Ongoing debate over the validity of the attention-deficit/hyperactivity disorder (ADHD) construct in

adulthood is fueled in part by uncertainty regarding implications of potentially extensive yet incompletely

described comorbid Axis I and II psychopathology. Three hundred sixty-three adults ages 18 to 37

completed semistructured clinical interviews; informants were also interviewed, and best estimate

diagnoses were obtained. Results were as follows: First, ADHD combined type (ADHD–C) had an

excess of externalizing and internalizing Axis I disorders, suggesting a gradient-of-severity relationship

between it and ADHD inattentive type (ADHD–I). Second, ADHD–C and ADHD–I did not differ in

frequency of Axis II disorders. Third, however, ADHD overall was associated with increased rates of

Axis II disorders, compared with rates in non-ADHD control participants, including both Cluster B

(primarily borderline personality disorder) and Cluster C disorders. Fourth, ADHD incrementally

accounted for clinician-rated global assessment of functioning scores above and beyond comorbid

conditions or symptoms on either Axis I or Axis II. Results further inform nosology of ADHD in adults.

Keywords: attention-deficit/hyperactivity disorder (ADHD), comorbidity, DSM–IV subtypes, personality

disorders, functional impairment

Attention-deficit/hyperactivity disorder (ADHD) is character-

ized by a persistent pattern of inattentive, hyperactive, and impul-

sive behaviors that begin in early childhood, often persist through-

out development, and interfere with adaptive functioning

(American Psychiatric Association, 2000). Population surveys es-

timate the prevalence of ADHD in adulthood to be about 5%

(Faraone & Biederman, 2005; Kessler et al., 2006), and neurobio-

logic and genetic findings from adults with ADHD are similar to

results seen in children (Faraone, 2004). Although interest in

ADHD in adults has been long-standing (Wender, 1974) and has

intensified in recent years (Faraone, 2000), adult ADHD remains

relatively underinvestigated. Historically, ADHD was primarily

conceptualized as a disorder of childhood. Subsequently, long-

term follow-up studies of children with ADHD established that

ADHD persists into adulthood in a substantial proportion of cases

(Barkley, Fischer, Smallish, & Fletcher, 2004; Mannuzza, Klein,

& Moulton, 2003; Weiss & Hechtman, 1993). A meta-analysis of

follow-up studies found that, although estimates of persistence

vary with how the diagnosis is defined, 65% of children with

ADHD will show impairing symptoms of ADHD in adulthood

(Faraone, Biederman, & Mick, 2006).

Patterns of clinical comorbidity, both Axis I and Axis II, are

critical to evaluating the clinical validity of ADHD in adults for

multiple reasons. First, comorbidity is likely very common. This is

the case in children with ADHD (Biederman et al., 1993) and

likely to also be true in adults. Although description of Axis I

comorbidity has begun in adult ADHD (Barkley, 2002; Barkley,

Fischer, Fletcher, & Smallish, 2002; Barkley et al., 2004; Marks,

Newcorn, & Halperin, 2001; Murphy & Barkley, 1996; Young,

Toone, & Tyson, 2003), Axis II comorbidity is inadequately

mapped as we describe later. Even for Axis I psychopathology,

findings in adults with ADHD have been somewhat inconsistent,

perhaps because of relatively small sample sizes in many studies;

predominantly male samples; and in some studies, reliance on

rating scales or self-report to assess ADHD. With regard to gender,

results of a prospective follow-up study of girls with ADHD

showed continued impairment into adolescence but found only

limited differences between ADHD subtypes (Hinshaw, Owens,

Sami, & Fargeon, 2006).

Second, developmental change was not addressed in the Diag-

nostic and Statistical Manual of Mental Disorders (4th ed.; DSM–

IV ; American Psychiatric Association, 1994) criteria, which are the

same regardless of age. Yet substantial developmental changes

occur in attention and activity level from childhood to adulthood

(Faraone et al., 2006). For example, hyperactive behaviors appar-

ently decline with development relative to inattentive symptoms

(Hart, Lahey, Loeber, & Hanson, 1994). Adults with ADHD

consequently may experience relatively more impairment from

inattention, disorganization, and subjective restlessness rather than

hyperactivity. Consequently, the validity and appropriateness of

the DSM–IV ADHD subtype structure (combined, ADHD–C; pri-

marily inattentive, ADHD–I; primarily hyperactive–impulsive,

ADHD–H) for clinical characterization of adults is unclear. As a

result, clinical comorbidity, which tends to differ in ADHD sub-

types in children, may not show similar patterns in adults, at least

when DSM–IV criteria are used.

In children, clinical data support the validity of distinguishing

ADHD–I from ADHD–C (Carlson & Mann, 2000). ADHD–C is

associated with more externalizing problems (Eiraldi, Power, &

Nezu, 1997; Gaub & Carlson, 1997), whereas at least some studies

Torri W. Miller and Joel T. Nigg, Department of Psychology, Michigan

State University; Stephen V. Faraone, Departments of Psychiatry and

Neuroscience and Physiology, State University of New York Upstate

Medical University.

Correspondence concerning this article should be addressed to Torri W.

Miller, Department of Psychology, Michigan State University, 43 Psychol-

ogy Building, East Lansing, MI 48824. E-mail: torri@msu.edu

519

2007, Vol. 116, No. 3, 519–528

520

MILLER, NIGG, AND FARAONE

suggest either that ADHD–I is associated with relatively more

internalizing disorders or problems (Lahey et al., 1994; Weiss,

Worling, & Wasdell, 2003) or that there are equivalent levels of

internalizing in the two subtypes (Eiraldi, Power, Karustis, &

Goldstein, 2000). Therefore, a strong prediction from the hypoth-

esis that ADHD–C and ADHD–I are distinct disorders (Milich,

Balentine, & Lynam, 2001) would be that in adults, ADHD–C

would be associated with an excess of externalizing disorders,

whereas ADHD–I would be associated with an excess of internal-

izing disorders. However, if ADHD–I is hypothesized to be a mild

form of the more severe ADHD–C (Faraone, Biederman, Weber,

& Russell, 1998), then it would be predicted that by adulthood,

ADHD–C would show excess comorbidity across all domains

relative to ADHD–I.

Only a handful of studies have examined this question in ADHD

subtypes in adults, with mixed results. Millstein, Wilens, Bieder-

man, and Spencer (1997) found more bipolar, oppositional, and

substance use disorders in ADHD–C than ADHD–I or ADHD–H.

Murphy, Barkley, and Bush (2002) found that both ADHD–C and

ADHD–I had excess dysthymia, alcohol and drug dependence–

abuse, learning disorders, and psychological distress, but

ADHD–C was further associated with oppositional defiant disor-

der (ODD), suicide attempts, arrests, and interpersonal hostility

and paranoia. In a study of children and adolescents, Volk, Hen-

derson, Neuman, and Todd (2006) found that impairment among

ADHD subtypes was not increased by comorbidity with conduct

disorder (CD), ODD, or major depressive disorder.

Third, aside from antisocial personality disorder, which is often

associated with ADHD in adulthood (Biederman et al., 1993;

Downey, Stelson, Pomerleau, & Giordani, 1997; Faraone et al.,

2000; Levin, Evans, & Kleber, 1998; Loeber, Burke, & Lahey,

2002), surprisingly little research has considered Axis II comor-

bidity. None, to our knowledge, have considered whether Axis II

comorbidity can account for ADHD-related impairment. Data on

normal personality traits suggest that ADHD may be associated

with extreme standing on personality (Nigg et al., 2002). Such

findings, as well as the aforementioned stability and chronicity of

the impulsive and dysregulated behaviors that compose ADHD,

suggest a theoretical connection between ADHD symptoms and

personality traits and, by extension, personality disorders —which

are defined as chronic, maladaptive personality traits (American

Psychiatric Association, 2000). This connection emerges in vari-

ous ways. One possibility is that ADHD alters personality and thus

increases risk for personality disorder later in development. An-

other possibility is that both ADHD and certain personality disor-

ders are related to the same extreme personality diathesis. In either

case, the question arises whether both ADHD and personality

disorder diagnoses add incremental validity in predicting impair-

ment.

Cluster B personality disorders are conceptually the most sim-

ilar to ADHD (Rey, Morris-Yates, Singh, Andrews, & Stewart,

1995) and therefore are the group we expected to be most likely to

co-occur with adult ADHD. They are characterized by inability to

control or regulate behavior, affect, and cognition and by social

and interpersonal problems that are at least superficially similar to

those seen in ADHD (Akiskal et al., 1985; Tzelepis, Schubiner, &

Warbasse, 1995; Weiss, Hechtman, & Weiss, 1999). Indeed, the

idea that ADHD may be a precursor to borderline personality

disorder (BPD) has been long-standing (Akiskal et al., 1985),

although research on this overlap is scarce. Yet a small number of

studies suggest that ADHD is associated with BPD (Dowson et al.,

2004; Rey et al., 1995), may be superimposed on the personality

difficulties of those patients with BPD (Weiss et al., 1999), and

appears more often in the childhood history of patients with adult

BPD (Fossati, Novella, Donati, Donini, & Maffei, 2002). How-

ever, these studies were relatively small and/or did not use

DSM–IV criteria or structured diagnostic interviews, leaving their

conclusions in some doubt. Other personality traits and disorders

that may be related to adult ADHD include histrionic and narcis-

sistic traits (Fischer, Barkley, Smallish, & Fletcher, 2002; May &

Bos, 2000), obsessive–compulsive personality disorder (OCPD),

and histrionic personality disorder (Modestin, Matutat, &

Wuermle, 2001), as well as personality dimensions such as novelty

or sensation seeking (Downey et al., 1997).

Fourth, the extensive comorbidity associated with ADHD un-

derscores divergent conceptualizations of how ADHD is taxonom-

ically related to comorbid conditions (Faraone, 2000). Within a

hierarchical framework, a patient meeting criteria for two or more

disorders would only be diagnosed with the higher ranking disor-

der; DSM–IV follows this approach to some degree for ADHD by

requiring that the ADHD symptoms not be better accounted for by

some other mental disorder. The comorbidity framework allows

for the assessment of all disorders; diagnostic overlap is viewed

more as the rule than the exception. From a hierarchical perspec-

tive, it remains unclear whether there is additional clinical utility in

diagnosing ADHD in the presence of co-occurring disorders or

their symptoms. One way to assess this is to examine whether

ADHD adds to the statistical prediction of impairment after co-

morbid disorders have been covaried. Lahey et al. (2004) found

this relationship in children: Children with ADHD continued to

display significant impairment relative to control participants

when comorbid psychopathology was statistically controlled. This

critical question of ADHD’s incremental validity relative to clin-

ical impairment remains essentially untested in adults, although

some data show that ADHD is a risk factor for substance use

disorders in adults, even after the researchers controlled for a

history of conduct or bipolar disorders (Biederman et al., 1995). In

a controlled study of community adults, ADHD (without regard to

comorbid conditions) was associated with impairment in terms of

lower degree attainment, less employment, more job changes,

more arrests and divorce, and lower personal satisfaction (Bieder-

man et al., 2006), but comorbid conditions were not controlled.

Aims of the Present Study

The present study sought to clarify clinical validity of the

ADHD construct in adults by examining independent predictors of

adaptive impairment and subtype differences in relation to Axis I

and Axis II comorbidity. Key questions were (a) whether

ADHD–C accrues more comorbid externalizing disorders than

ADHD–I and controls and whether the comorbid profile differs

across ADHD subtypes (interaction of Comorbid Domain

Sub-

type), (b) which domains of Axis II personality disorders are

associated with ADHD regardless of subtype, and (c) whether

ADHD or its symptoms have an influence on impairment above

and beyond that which is accounted for by Axis I and Axis II

psychopathology.

COMORBIDITY IN ADULT ADHD

521

Method

a retrospective K–SADS ADHD module. The second informant,

who knew the participant currently (usually a spouse or friend),

completed the Conners peer rating, the Barkley and Murphy peer

ratings on adult symptoms, a brief screen of antisocial behavior

and drug and alcohol use, and a structured interview about the

participant’s current ADHD symptoms, using the modified

K–SADS for current symptoms.

Establishment of best estimate diagnosis for ADHD. A diag-

nostic team that included a licensed clinical social worker, a

licensed clinical psychologist, and a board certified psychiatrist

then arrived at a “best estimate” diagnosis (Faraone, 2000). Each

team member reviewed all available information from the semi-

structured interviews, and rating scales to arrive at a judgment

about ADHD present or absent, ADHD subtype, and comorbid

disorders. In the case of disagreement, consensus was reached by

discussion. Interrater agreement on presence or absence of ADHD

(any type) was satisfactory ( k .80), and agreement on ADHD

subtype (combined, inattentive, or hyperactive) in childhood and

adulthood was also adequate (ranging from k .74 to .85). In view

of disagreement about whether the subtype classification should be

represented by childhood or adult symptom profiles (Faraone,

2000), we relied on the adult subtype here.

Assessment of comorbid Axis I disorders. The Structured Clin-

ical Interview for DSM–IV Axis I Disorders (SCID–I; First,

Spitzer, Gibbon, & Williams, 1997) was administered by a trained

master’s level clinician after extensive training in the interview

and checkout of taped interviews for validity by First and col-

leagues at Columbia. Diagnoses examined in the current study

included major depressive episode, dysthymic disorder, bipolar

disorder, substance abuse and dependence, psychotic disorders,

obsessive–compulsive, panic disorder, agoraphobia, simple pho-

bia, social phobia, and eating disorders. Autistic disorder was

screened by added symptom questions and was a rule out. Twenty-

five SCID interviews were videotaped and coded by two qualified

interviewers to assess reliability of the interview procedures. Re-

liability for comorbid disorders was acceptable (e.g., substance use

disorder, k .83; mood disorder, k .80; anxiety disorder, k

.71; antisocial personality disorder, k .84).

Assessment of Axis II disorders. Participants completed the

SCID–II prescreening form. Any disorder for which at least one

symptom was endorsed was followed up with the SCID–II inter-

view module for that disorder. This procedure results in a very low

rate of false negatives (participants who endorse zero symptoms on

the questionnaire virtually never have a disorder on administration

of the full SCID–II interview) while capturing potential personal-

ity disorders and assessing them after the screen (First, Gibbon,

Spitzer, Williams, & Benjamin, 1997).

Participants

Overview. The sample included 363 adults (185 men and 178

women) ages 18 to 37 years. Participants’ race closely mirrored the

surrounding community from which they were obtained, with 86%

Caucasian. Following procedures described later, participants were

grouped into ADHD (any subtype; n

152) and non-ADHD

211). For some analyses, participants

with ADHD were further divided into ADHD–I ( n 69) and

ADHD–C ( n 64). We had only 19 participants with ADHD–H;

this group was judged too small for reliable analysis and so was

omitted from subtype analyses.

Multistage recruitment process. Participants were recruited by

a broad net of public advertisements, including radio, newspaper,

movie theaters, and mailings to local clinics, in an effort to obtain

as broadly representative a volunteer sample as possible. We

advertised for ADHD with separate ads targeted at “individuals

that have been diagnosed or suspect they have an attention deficit

disorder (ADHD or ADD), or other attention problems.” We

advertised for the non-ADHD comparison group with ads seeking

“volunteers in good health who do not have attention problems.”

These efforts resulted in 623 applicants contacting the project

office. These prospective participants underwent a phone screen-

ing to check rule outs (inclusionary criteria were ages 18–40, no

sensorimotor disability, no neurological illness, native English

speaking, and currently prescribed antidepressant, antipsychotic,

or anticonvulsant medications). At this stage, 533 participants

were coded as eligible and went on to Stage 2. At Stage 2, eligible

participants completed semistructured clinical interviews and nor-

mative rating scales to assess ADHD and comorbid Axis I and

Axis II disorders as detailed next. Those data were then reviewed

by the best estimate clinical team to determine final diagnostic

assignment and study eligibility. One hundred seventy participants

were excluded at this stage (because of lack of cross-informant

convergence enabling clear classification as ADHD or non-

ADHD, current major depression, current severe substance use,

psychosis, or brain–head injury), yielding the final N 363.

Diagnostic instruments. A retrospective Kiddie Schedule for

Affective Disorders and Schizophrenia (K–SADS; Puig-Antich &

Ryan, 1986) was administered to assess ADHD. This well-

established procedure (Biederman et al., 1992; Biederman, Fara-

one, Keenan, Knee, & Tsuang, 1990) included the diagnoses of

childhood DSM–IV ADHD, CD, and ODD. Current symptoms

were assessed by structured interview and included K–SADS

ADHD questions worded appropriately for current adult symptoms

(Biederman et al., 1992). Respondents also completed the Barkley

and Murphy (1998) Current ADHD Symptoms rating scale. We

obtained normative, standardized dimensional ratings of attention

problems as well as other current symptoms by having participants

complete the Conners, Erhardt, and Sparrow (1999) Adult ADHD

Rating Scale, the Achenbach (1991) Young Adult Self Report

scale, and the Brown (1996) Adult ADHD rating scale.

To address potential reporting bias in self-report interviews of

ADHD (Barkley et al., 2002), two informants who knew the

participant well were interviewed. One informant, who knew the

participant as a child (usually a parent) reported on the target

participant’s childhood behaviors via an ADHD Rating Scale and

Assessment of Impairment

Global assessment of functioning (GAF; American Psychiatric

Association, 2000) scores were assigned by the interviewing cli-

nician at the end of the structured clinical interviews. This score of

overall functional adjustment was used as an index of impairment;

high scores indicate better functioning and low scores indicate

more impairment. To evaluate reliability of the impairment scores,

20 SCID–KSAD interviews were taped and reviewed by a second

clinical rater, blind to the GAF rating or diagnoses of the first rater.

GAF scores were assigned by the second rater and were compared

control participants ( n

522

MILLER, NIGG, AND FARAONE

with those assigned by the first rater. The interclass correlation

(absolute agreement) of .714 reflected adequate interrater reliabil-

ity.

disorders (all mood and anxiety disorders). Because these were

ordinal count variables, we analyzed group effects with multino-

mial logistic regression. To obtain adequate cell sizes, we catego-

rized number of externalizing and internalizing disorders as fol-

lows: 0 (none), 1, and 2 or more. Table 2 shows the resultant

frequencies. We checked all models for Sex Group interactions

and they were not significant, so we simply covaried sex.

For externalizing disorders, the three-group multinomial logistic

regression model (with sex covaried) indicated a significant model

overall,

Data Analytic Plan

We tested a series of questions regarding subtype effects and

Axis I disorders with multinomial and binomial logistic regression,

with sex effects covaried. We tested a series of questions concern-

ing impairment using linear multiple regression and multiple lo-

gistic regression models. Statistical power for all analyses ex-

ceeded .90 to detect Cohen’s (1992) medium-sized effects ( r

.15), with the exception of some pairwise post hoc tests.

(6, N 344) 27.1, p .001, –2LL 54.3; the main

effect of group was significant as well,

(4, N 344) 20.8, p

.001, –2LL 75.0. Using the control group as the reference,

presence of one externalizing disorder was more likely for partic-

ipants with ADHD–C (odds ratio [OR] 2.05, p .050), but not

for ADHD–I ( p .766). Presence of two or more externalizing

disorders was more likely for ADHD–C (OR 5.12, p .001)

and ADHD–I (OR 2.12, p .038). Thus, ADHD–C conferred

a fivefold increase in risk for two or more externalizing disorders

whereas ADHD–I conferred a doubling of such risk versus the

control group. The OR was significantly higher for ADHD–C than

ADHD–I for two or more disorders ( p .044), indicating that

ADHD–C conferred more risk of externalizing disorder than

ADHD–I.

For internalizing disorders, the three-group multinomial logistic

regression model (with sex covaried) indicated a significant model

overall,

Results

Sample Description and Review of Potential Demographic

Confounds

Demographic information for the sample is provided in Table 1.

Ratings data all showed marked clinical elevations in the ADHD

sample, indicating validity of ADHD assignments regardless of

instrument or model used. Parental household incomes were sim-

ilar in the two groups ( p .4), indicating that they came from

similar socioeconomic backgrounds. Despite this, and consistent

with prior reports (Murphy & Barkley, 1996), individuals with

ADHD were less likely to complete high school than control

participants (see Table 1). The ADHD group was less likely to

attend college than the control group (53% vs. 62%, p .05).

Those who attended a 4-year college were more likely to be

control participants than participants with ADHD (36% vs. 20%,

p .01), whereas those who attended a 2-year community college

were more likely to have ADHD than be control participants (18%

vs. 10%, p .01). Thus, the ADHD group had lower educational

attainment overall. Personal incomes were qualitatively lower for

the nonstudent ADHD ( M $32,000) than non-ADHD group

( M 40,200) though this effect was not significant. The gender

difference, with a greater proportion of male participants with

ADHD (see Table 1), occurred despite our efforts to overselect

female participants with ADHD; it is common in studies of ADHD

and in part may reflect the male preponderance of ADHD in the

population. Because some of the comorbid disorders vary by

gender, we controlled statistically for gender. Groups did not differ

significantly in percentage of minority participants, although there

were qualitatively more minorities in the ADHD group ( p .07).

We therefore checked all results with ethnicity covaried; results

were unchanged from those reported in this article. Consistent with

other studies of ADHD, participants with any ADHD were more

likely to have substance use disorder,

(6, N 344) 39.7, p .001, –2LL 49.6; the main

effect of group was again significant,

(4, N 344) 33.3, p

.001, –2LL 82.9. Using the control group as the reference, we

found that presence of one internalizing disorder was more likely

for ADHD–C (OR 4.59, p .001) and ADHD–I (OR 3.22,

p .001). Presence of two or more internalizing disorders was

significant for ADHD–C (OR 3.76, p .001) and for ADHD–I

(OR 4.02, p .001). Although the OR was slightly higher for

ADHD–I than ADHD–C for two or more disorders, the ADHD–C

versus ADHD–I effect was nonsignificant ( p .854). As can be

seen in Table 2, although the ADHD–I group had slightly more

likelihood of two or more internalizing disorders, the ADHD–C

group had more likelihood of internalizing disorder overall, al-

though this difference also was nonsignificant.

In summary, ADHD–C was associated with more comorbid

externalizing disorders than ADHD–I or non-ADHD status. How-

ever, contrary to a “distinct disorder” hypothesis, as displayed in

Table 2 ADHD–I was associated with qualitatively lower, not

higher, rates of total internalizing disorders than ADHD–C, al-

though ADHD–I was associated with a slightly but not signifi-

cantly greater chance of having two or more internalizing disorders

versus ADHD–C.

(1, N 363) 9.22, p

.01; mood disorder,

(1, N 363) 23.70, p .001; anxiety

Hypothesis 2: Axis II Comorbidity in Relation to ADHD

(1, N 363) 8.81, p .01; and antisocial person-

ality disorder,

(1, N 363) 7.32, p .01, than the non-

ADHD comparison group.

For these analyses, we created three personality disorder com-

posite variables: Cluster A disorders present or absent (presence of

one or more of paranoid, schizoid, and/or schizotypal personality

disorder), Cluster B disorders present or absent (presence of one or

more of borderline, antisocial, histrionic, and/or narcissistic per-

sonality disorder), and Cluster C disorders present or absent (pres-

ence of one or more of avoidant, dependent, and/or OCPD). Table

3 shows the frequencies of the Axis II disorders by cluster for each

group. The ADHD subtypes did not differ on any of these clusters:

Hypothesis 1: Subtype Comorbid Profiles for Axis I

Disorders

To test this question, we summed total number of lifetime (a)

externalizing disorders (lifetime ODD, CD, substance use disor-

ders, and antisocial personality disorder), and (b) internalizing

disorder,

Table 1

Description of Sample

Control

( n 211)

Any ADHD

( n 152)

ADHD–I

( n 69)

ADHD–C

( n 64)

Variable

n % MSDn % MSD n % MSDn % MSD

Male

89 42

96 63

.001 47 68

41 64

.504

White

174 83

136 90

.062 60 87

58 91

.625

Age (in years)

23.7

4.5

23.7

4.6 1.00

23.31 4.7

23.78 4.8

.828

Conners ADHD T score

47.2 11.1

62.1 10.1

.001

58.51 9.7

66.10 9.5

.526

No. of current DSM inattentive symptoms

1.81 2.2

7.0

1.7

.001

7.26 1.5

7.29 1.4

.922

No. of current DSM hyperactivity–impulsivity

symptoms

1.7

2.0

5.68 2.5

.001

3.74 2.1

7.10 1.5

.001

No. of childhood DSM inattentive symptoms

1.93 2.1

6.78 1.6

.001

6.99 1.3

7.06 1.4

.188

No. of childhood DSM hyperactivity–impulsivity

symptoms

1.64 1.7

5.13 2.2

.001

3.63 1.9

6.23 1.7

.001

Parent annual combined income a

3.73 1.1

3.64 1.2

.484

3.83 1.1

3.43 1.2

.05

Lifetime substance use disorder

84 40

88 57

.001 35 50

44 67

.112

Lifetime mood disorder

58 28

82 53

.001 36 51

38 58

.370

Lifetime any anxiety disorder

36 17

47 30

.05 24 34

18 27

.888

Lifetime antisocial personality disorder

8 4

15 11

.05

6 10

8 15

.356

Completed high school

137 94

91 84

.05 41 87

38 79

.298

Marital status

.865

.500

Married

26 14

16 13

6 10

7 14

Single

155 81

103 84

50 86

43 84

Divorced

9 5

4 3

2 4

1 2

Note. The probability value reflects the two-group comparison (control vs. any ADHD) and is based on an independent-samples t test for continuous variables or a chi-square test for dichotomous

variables. Sample sizes varied slightly for some measures because of missing data for some disorders. ADHD attention-deficit/hyperactivity disorder; ADHD–I ADHD inattentive type;

ADHD–C ADHD combined type.

a Parent annual combined income categories were as follows: 1 $25,000–$50,000; 2 $50,000–$75,000; 3 $75,000–$100,000; 4 greater than $100,000.

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: