|
|||||||||||||
|
|||||||||||||
This is a beginning to intermediate level course. After completing this course, mental health professionals will be able to:
This course provides an overview of the nature of ADHD, briefly considers its history, describes its developmental course and outcomes, and discusses its causes. Current critical issues related to these matters will be raised along the way. Given the thousands of scientific papers on this topic, this course must, of necessity, concentrate on the most important topics in this literature. Readers interested in more detail can pursue other sources (Accardo, Blondis, Whitman, & Stein, 2001; Barkley, 1998; Weiss & Hechtman, 1993). The author’s theoretical model of ADHD also will be presented, providing a more parsimonious accounting of the many cognitive and social deficits in the disorder that points to numerous promising directions for future research while rendering a deeper appreciation for the developmental significance and seriousness of ADHD. This theory shows that ADHD involves more than just attention deficits – it is also a central problem with inhibition, self-regulation, and the cross-temporal organization of social behavior.
(This course was adapted with permission from the chapter by R. A. Barkley on ADHD in the text by E. J. Mash and R. A. Barkley (2003), Child Psychopathology (2nd ed.). New York: Guilford Publications, Guilford.com).
Literary references to individuals having serious problems with inattention, hyperactivity, and poor impulse control date back to Shakespeare, who made reference to a malady of attention in King Henry VIII. A hyperactive child was the focus of a German poem, “Fidgety Phil,” by physician, Heinrich Hoffman (see Stewart, 1970). William James (1890), in his Principles of Psychology, described a normal variant of character that he called the “explosive will” that resembles the difficulties experienced by those who today are called ADHD. But, more serious clinical interest in ADHD children first occurred in three lectures of the English physician George Still (1902) before the Royal Academy of Physicians.
Still reported on a group of 20 children in his clinical practice whom he defined as having a deficit in “volitional inhibition” (p. 1008) that led to a “defect in moral control” (p. 1009) over their own behavior. Described as aggressive, passionate, lawless, inattentive, impulsive, and overactive, many of these children today would be diagnosed not only as ADHD but also as having oppositional defiant disorder (ODD). Still’s observations were quite astute, describing many of the associated features of ADHD that would come to be corroborated in research over the next century: (1) an overrepresentation of male subjects (ratio of 3:1 in Still’s sample); (2) high comorbidity with antisocial conduct and depression; (3) an aggregation of alcoholism, criminal conduct, and depression among the biological relatives; (4) a familial predisposition to the disorder, likely of hereditary origin; (5) yet with the possibility of the disorder also arising from acquired injury to the nervous system.
Interest in these children arose in North America around the time of the great encephalitis epidemics of 1917–1918. Children surviving these brain infections had many behavioral problems similar to those comprising contemporary ADHD (Ebaugh, 1923; Hohman, 1922; Stryker, 1925). These cases and others known to have arisen from birth trauma, head injury, toxin exposure, and infections (see Barkley, 1998) gave rise to the concept of a brain-injured child syndrome (Strauss & Lehtinen, 1947), often associated with mental retardation, that would eventually become applied to children manifesting these same behavior features but without evidence of brain damage or retardation (Dolphin & Cruickshank, 1951; Strauss & Kephardt, 1955). This concept evolved into that of minimal brain damage, and eventually minimal brain dysfunction (MBD), as challenges were raised to the label in view of the dearth of evidence of obvious brain injury in most cases (see Kessler, 1980, for a more detailed history of MBD).
By the 1950-70s, focus shifted away from etiology and toward the more specific behavior of hyperactivity and poor impulse control characterizing these children, reflected in labels such as “hyperkinetic impulse disorder” or “hyperactive child syndrome” (Burks, 1960; Chess, 1960). The disorder was thought to arise from cortical overstimulation due to poor thalamic filtering of stimuli entering the brain (Knobel, Wolman, & Mason, 1959; Laufer, Denhoff, & Solomons, 1957). Despite a continuing belief among clinicians and researchers of this era that the condition had some sort of neurological origin, the larger influence of psychoanalytic thought held sway. And so, when the second edition of Diagnostic and Statistical Manual of Mental Disorders (DSM-II) appeared, all childhood disorders were described as “reactions,” and the hyperactive child syndrome became “hyperkinetic reaction of childhood” (American Psychiatric Association, 1968).
The recognition that the disorder was not caused by brain damage seemed to follow a similar argument made somewhat earlier by the prominent child psychiatrist Stella Chess (1960). It set off a major departure between professionals in North America and those in Europe that continues, to a lessening extent, to the present. Europe continued to view hyperkinesis for most of the latter half of this century as a relatively rare condition of extreme overactivity often associated with mental retardation or evidence of organic brain damage. This discrepancy in perspectives has been converging over the last decade as evident in the similarity of the DSM-IV criteria (see below) with those of ICD-10 (World Health Organization, 1994). Nevertheless, the manner in which clinicians and educators view the disorder remains quite disparate; in North America, Canada, and Australia, such children have ADHD, a developmental disorder, whereas in Europe they are viewed as having conduct problem or disorder, a behavioral disturbance believed to arise largely out of family dysfunction and social disadvantage.
By the 1970s, research emphasized the problems with sustained attention and impulse control in addition to hyperactivity (Douglas, 1972). Douglas (1980, 1983) theorized that the disorder had four major deficits: (1) the investment, organization, and maintenance of attention and effort; (2) the ability to inhibit impulsive behavior; (3) the ability to modulate arousal levels to meet situational demands; and (4) an unusually strong inclination to seek immediate reinforcement. Douglas’s emphasis on attention along with the numerous studies of attention, impulsiveness, and other cognitive sequelae that followed (see Douglas, 1983; and Douglas & Peters, 1978, for reviews) eventually led to renaming the disorder as attention deficit disorder (ADD) in 1980 (DSM-III; American Psychiatric Association, 1980). Significant, historically, was the distinction in DSM-III between two types of ADD: those with hyperactivity and those without it. Little research existed at the time on the latter subtype that would have supported such a distinction being made in an official and increasingly prestigious diagnostic taxonomy. Yet, in hindsight, this bald assertion led to valuable research on the differences between these two supposed forms of ADD that otherwise would never have taken place. That research may have been fortuitous, as it may be leading to the conclusion that a subset of those having ADD without hyperactivity may actually a separate, distinct, and qualitatively unique disorder rather than a subtype of ADHD (Milich, Ballantine & Lynam, 2001).
Even so, within a few years of the creation of the label ADD, concern arose that the important features of hyperactivity and impulse control were being de-emphasized when in fact they were critically important to differentiating the disorder from other conditions and to predicting later developmental risks (Barkley, 1998; Weiss & Hechtman, in press). In 1987, the disorder was renamed as attention-deficit hyperactivity disorder in DSM-III-R (American Psychiatric Association, 1987), and a single list of items incorporating all three symptoms was specified. Also important here was the placement of the condition of ADD without hyperactivity, renamed undifferentiated attention-deficit disorder, in a separate section of the manual from ADHD with the specification that insufficient research existed to guide in the construction of diagnostic criteria for it at that time.
During the 1980s, reports focused instead on problems with motivation generally, and an insensitivity to response consequences specifically (Barkley, 1989a; Glow & Glow, 1979; Haenlein & Caul, 1987). Research was demonstrating that under conditions of continuous reward, the performances of ADHD children were often indistinguishable from normal children on various lab tasks but when reinforcement patterns shifted to partial reward or to extinction (no reward) conditions, children with ADHD showed significant declines in their performance (Douglas & Parry, 1983, 1994; Parry & Douglas, 1983). It was also observed that deficits in the control of behavior by rules characterized these children (Barkley, 1989a).
Over the past decade, researchers employed information-processing paradigms to study ADHD and found that problems in perception and information processing were not so evident as were problems with motivation and response inhibition (Barkley, Grodzinsky, & DuPaul, 1992; Schachar & Logan, 1990; Sergeant, 1988; Sergeant & Scholten, 1985a, 1985b). The problems with hyperactivity and impulsivity also were found to form a single dimension of behavior (Achenbach & Edelbrock, 1983; Goyette, Conners, & Ulrich, 1978; Lahey et al., 1988), which others described as “disinhibition” (Barkley, 1990). All of this led to the creation of two separate lists of items and thresholds for ADHD when the DSM-IV was published later in the decade (American Psychiatric Association, 1994); one for inattention and another for hyperactive–impulsive behavior. Unlike its predecessor, DSM-III-R, the establishment of the inattention list once again permitted the diagnosis of a subtype of ADHD that consisted principally of problems with attention (ADHD predominantly inattentive type). It also permitted, for the first time, the distinction of a subtype of ADHD that consisted chiefly of hyperactive–impulsive behavior without significant inattention (ADHD, predominantly hyperactive–impulsive type). Children having significant problems from both item lists were titled ADHD, combined type. The specific criteria from DSM-IV are discussed in more detail below (see “Diagnostic Criteria and Related Issues”).
Healthy debate continues to the present over the core deficit(s) in ADHD with increasing weight being given to problems with behavioral inhibition, self-regulation, and the related domain of executive functioning (Barkley, 1997a, 1997b, 2000; Douglas, 1999; Nigg, 2001; Quay, 1997). The symptoms of inattention may actually be evidence of impaired working memory and not perceptual, filtering, or selection (input) problems (Barkley, 1997b, 2006). Likewise, controversy continues to swirl around the place of a subtype composed primarily of inattention within the larger condition of ADHD (see Clinical Psychology: Science and Practice, 2001, Vol. 8 (4) for a debate on this issue), with some arguing for it being a new, unique disorder from ADHD (Barkley, 2001; Milich et al., 2001) and others arguing that this distinction may be premature (Hinshaw, 2001; Lahey, 2001) or not especially important to treatment planning (Pelham, 2001). Relatively consistent across viewpoints, however, is the opinion that a subset of children having only high levels of inattention probably represents a qualitatively different problem in attention (deficient selective attention and sluggish cognitive processing) than is seen in ADHD (poor persistence, inhibition, and resistance to distraction).
Research employing factor analysis has repeatedly identified two distinct behavioral dimensions underlying the various behavioral problems (symptoms) thought to characterize ADHD in both children (Barkley, 2006; Burns, Boe, Walsh, Sommers-Flanagan, & Teegarden, 2001; DuPaul et al., 1997; Lahey et al., 1994; Pillow, Pelham, Hoza, Molina, & Stultz, 1998) and adults (Barkley, Murphy, & Fischer, 2008). These two dimensions have been identified across various ethnic and cultural groups, including Native American children (Beiser, Dion, & Gotowiec, 2000).
Attention represents a multidimensional construct (Bate, Mathias, & Crawford, 2001; Mirsky, 1996; Strauss, Thompson, Adams, Redline, & Burant, 2000) implying that several qualitatively distinct problems with attention may be evident in children (Barkley, 2001a). The dimension impaired in ADHD reflects an inability to sustain attention or persist at tasks or play activities, remember and follow through on rules and instructions, and resist distractions while doing so. I have elsewhere argued that this dimension more likely reflects problems with the executive function of working memory than poor attention, per se (Barkley, 1997b) and evidence is becoming available to support this contention (Oosterlan, Sheres, & Sergeant, in press; Seguin, Boulerice, Harden, Tremblay, & Pihl, 1999; Wiers, Gunning, & Sergeant, 1998). Parents and teachers frequently complain that these children do not seem to listen as well as they should for their age, cannot concentrate, are easily distracted, fail to finish assignments, are forgetful, and change activities more often than others (DuPaul, Powers, Anastopolous, & Reid, 1999). Research employing objective measures corroborates these complaints through observations of more “off-task” behavior, less work productivity, greater looking away from assigned tasks (including television), less persistence at tedious tasks, such as continuous performance tasks, being slower and less likely to return to an activity once interrupted, less attentive to changes in the rules governing a task, and less capable of shifting attention across tasks flexibly (Borger & van der Meere, 2000; Hoza, Pelham, Waschbusch, Kipp, & Owens, 2001; Lorch et al., 2000; Luk, 1985; Newcorn et al., 2001; Seidman, Biederman, Faraone, Weber, & Ouellette, 1997; Shelton et al., , 1998). This inattentive behavior distinguishes these children from those with learning disabilities (Barkley, DuPaul, & McMurray, 1990) or other psychiatric disorders (Chang et al., 1999; Swaab-Barneveld et al., 2000) and does not appear to be a function of other disorders often comorbid with ADHD (anxiety, depression, or oppositional and conduct problems) (Murphy, Barkley, & Bush, 2001; Klorman et al., 1999; Newcorn et al., 2001; Nigg, 1999; Seidman et al., 1997).
As with attention, inhibition is a multidimensional construct (Nigg, 2000; Olson, Schilling, & Bates, 1999) and thus various, qualitatively distinct forms of inhibitory impairments may eventually be found in children. The problems with inhibition seen in ADHD are thought to involve voluntary or executive inhibition of prepotent responses rather than impulsiveness that may be more motivationally controlled, as in a heightened sensitivity to available reward (reward seeking) or to excessive fear (Nigg, 2001). Some evidence suggests that an excess sensitivity to reward or to sensation seeking may be more associated with severity of conduct disorder or psychopathy than with severity of ADHD (Beauchaine, Katkin, Strassberg, & Snarr, 2001; Dougherty & Quay, 1991; Fischer, Barkley, Smallish, & Fletcher, 2002; Matthys, van Goozen, de Vries, Cohen-Kettenis, & van Engeland, 1998). Evidence is less clear about deficits in automatic or involuntary inhibition, as in eyeblinking or negative priming, being associated with ADHD (Nigg, 2001).
More specifically, ADHD children manifest difficulties with excessive activity level and fidgetiness, less ability to stay seated when required, greater touching of objects, moving about, running, and climbing than other children, playing noisily, talking excessively, acting impulsively, interrupting others’ activities, and being less able than others to wait in line or take turns in games (American Psychiatric Association, 1994). Parents and teachers describe them as acting as if driven by a motor, incessantly in motion, always on the go, and unable to wait for events to occur. Research objectively documents them to be more active than other children (Barkley & Cunningham, 1979a; Dane, Schachar, & Tannock, 2000; Luk, 1985; Porrino et al., 1983; Shelton et al., 1998), to have considerable difficulties with stopping an ongoing behavior (Schachar, Tannock, & Logan, 1993; Milich, Hartung, & Haigler, 1994; Nigg, 1999, 2001; Oosterlaan, Logan, & Sergeant, 1998), to talk more than others (Barkley, Cunningham, & Karlsson, 1983), to interrupt others’ conversations (Malone & Swanson, 1993), to be less able to resist immediate temptations and delay gratification (Anderson, Hinshaw, & Simmel, 1994; Barkley, Edwards, Laneiri, & Metevia, 2001; Olson et al., 1999; Rapport, Tucker, DuPaul, Merlo, & Stoner, 1986; Solanto et al., 2001), and to respond too quickly and too often when they are required to wait and watch for events to happen, as is often seen in impulsive errors on continuous performance tests (Losier, McGrath, & Klein, 1996; Newcorn et al., 2001). Although less frequently examined, the differences in activity and impulsiveness have been found between children with ADHD and those with learning disabilities (Barkley, DuPaul, & McMurray, 1990; Bayliss & Roodenrys, 2000; Klorman et al., 1999; Willcutt et al., 2001). Mounting evidence further shows that these inhibitory deficits are not a function of other psychiatric disorders that may overlap with ADHD (Barkley, Edwards, Laneiri, & Metevia, 2001; Halperin, Matier, Bedi, Sharpin, & Newcorn, 1992; Fischer et al., in press; Murphy, Barkley, & Bush, 2001; Nigg, 1999; Oosterlaan et al., 1998; Seidman et al., 1997).
Interestingly, recent research shows that the problems with inhibition arise first (at age 3–4 years) ahead of those related to inattention (at age 5–7 years), or than the sluggish cognitive tempo that characterizes the predominantly inattentive subtype that may arise even later (ages 8-10) (Hart et al., 1995; Loeber, Green, Lahey, Christ, & Frick, 1992; Milich et al., 2001). Whereas the symptoms of disinhibition in the DSM item lists seem to decline with age, perhaps owing to their heavier weighting with hyperactive than impulsive behavior, those of inattention remain relatively stable during the elementary grades (Hart et al., 1995). But eventually decline by adolescence (Fischer, Barkley, Fletcher, & Smallish, 1993a), though not to normal levels. Why the inattention arises later than the disinhibitory symptoms and does not decline when the latter do over development remains an enigma. As noted above, it may simply reflect the different weightings of symptoms in the DSM. Those of hyperactivity may be more typical of preschool to early school-age children and are over-represented in the DSM list while those reflecting inattention may be more characteristic of school-age children. Another explanation comes from the theoretical model described below (Barkley, 1997b) in which inhibition and the two types of working memory (nonverbal and verbal) emerge at separate times in development.
The symptoms comprising ADHD are greatly affected in their level of severity by a variety of situational and task-related factors (Barkley, 2006). Douglas (1972) commented on the greater variability of task performances made by ADHD compared to control children. Many others since then have found that when the ADHD child must perform multiple trials within a task assessing attention and impulse control, the range of scores around that child’s own mean performance is frequently greater than in normal children (see Douglas, 1983). The finding is especially common in measures of reaction time (Chee, Logan, Schachar, Lindsay, & Wachsmuth, 1989; Fischer et al., 2002; Kuntsi, Oosterlaan, & Stevenson, 2001; Murphy et al., 2001; Scheres, Oosterlaan, & Sergeant, 2001).
A number of other factors influence the ability of children with ADHD to sustain their attention to task performance, control their impulses to act, regulate their activity level, and/or to produce work consistently. The performance of ADHD children is worse: (1) later in the day than earlier (Dane et al., 2000; Porrino et al., 1983; Zagar & Bowers, 1983); (2) in greater task complexity such that organizational strategies are required (Douglas, 1983); (3) when restraint is demanded (Barkley & Ullman, 1975; Luk, 1985); (4) under low levels of stimulation (Antrop, Roeyers, Van Oost, & Buysse, 2000; Zentall, 1985); (5) under more variable schedules of immediate consequences in the task (Carlson, Mann, & Alexander, 2000; Carlson & Tamm, 2000; Douglas & Parry, 1983, 1994; Slusarek, Velling, Bunk, & Eggers, 2001; Tripp & Alsop, 1999); (6) under longer delay periods prior to reinforcement availability (Solanto et al., 2001; Sonuga-Barke, Taylor, & Heppinstall, 1992; Tripp & Alsop, 2001); and (7) in the absence of adult supervision during task performance (Draeger, Prior, & Sanson, 1986; Gomez & Sanson, 1994).
Besides the aforementioned factors, which chiefly apply to task performance, variability has also been documented across more macroscopic settings. For instance, children with ADHD are most problematic in their behavior when persistence in work-related tasks is required (i.e., chores, homework, etc.) or where behavioral restraint is necessary, especially in settings involving public scrutiny (i.e., in church, in restaurants, when a parent is on the phone, etc.) than in free play situations (Altepeter & Breen, 1992; Barkley & Edelbrock, 1987; DuPaul & Barkley, 1992). Although they will be more disruptive when their fathers are at home than during free play, children with ADHD are still rated as much less problematic when the father is at home than in most other contexts. Fluctuations in the severity of ADHD symptoms have also been documented across a variety of school contexts (Barkley & Edelbrock, 1987; DuPaul & Barkley, 1992). In this case, contexts involving task-directed persistence and behavioral restraint (classroom) are the most problematic, with significantly fewer problems posed by contexts involving less work and behavioral restraint (i.e., at lunch, in hallways, at recess, etc.), and even fewer problems being posed during special events (i.e., field trips, assemblies, etc.) (Altepeter & Breen, 1992).
Children with ADHD often demonstrate deficiencies in many other cognitive abilities. Among these, are difficulties with: (1) physical fitness, gross and fine motor coordination, and motor sequencing (Breen, 1989; Denckla & Rudel, 1978; Harvey & Reid, 1997; Kadesjo & Gillberg, 1999; Mariani & Barkley, 1997); (2) speed of color naming (Tannock, Martinussen, & Frijters, 2000); (3) verbal and nonverbal working memory and mental computation (Barkley, 1997; Mariani & Barkley, 1997; Murphy et al., 2001; Zentall & Smith, 1993); (4) story recall (Lorch et al., 2000; Sanchez, Lorch, Milich, & Welsh, 1999); (5) planning and anticipation (Grodzinsky & Diamond, 1992; Klorman et al., 1999); (6) verbal fluency and confrontational communication (Grodzinsky & Diamond, 1992; Zentall, 1988); (5) effort allocation (Douglas, 1983; Nigg et al., 1998; Sergeant & van der Meere, 1994; Voelker, Carter, Sprague, Gdowski, & Lachar, 1989); (6) developing, applying, and self-monitoring organizational strategies (Clark et al., 2000; Hamlett, Pellegrini, & Connors, 1987; Purvis & Tannock, 1997; Zentall, 1988); (7) the internalization of self-directed speech (Berk & Potts, 1991; Copeland, 1979; Winsler, 1998; Winsler, Diaz, Atencio, McCarthy, & Chabay, 2000); (8) adhering to restrictive instructions (Danforth, Barkley, & Stokes, 1991; Roberts, 1990; Routh & Schroeder, 1976); and (9) self-regulation of emotion (Braaten & Rosen, 2000; Hinshaw, Buhrmeister, & Heller, 1989; Maedgen & Carlson, 2000). The latter difficulties with emotional control may be especially salient in children having ADHD with comorbid oppositional defiant disorder (ODD) (Melnick & Hinshaw, 2000). Several studies have also demonstrated that ADHD may be associated with less mature or diminished moral development (Hinshaw, Herbsman, Melnick, Nigg, & Simmel, 1993; Nucci & Herman, 1982; Simmel & Hinshaw, 1993). Many of these cognitive difficulties appear to be specific to ADHD and are not a function of its commonly comorbid disorders, such as learning disabilities, depression, anxiety, or oppositional/conduct disorder (Barkley, Edwards, Lanieri, & Metevia, 2001; Clark, Prior, & Kinsella, 2000; Klorman et al., 1999; Murphy et al., 2001; Nigg, 1999; Nigg et al., 1998).
The commonality among most or all of these seemingly disparate abilities is that all have been considered to fall within the domain of “executive functions” in the field of neuropsychology (Barkley, 1997b; Denckla, 1996) or “metacognition” in developmental psychology (Flavell, 1970; Torgesen, 1994; Welsh & Pennington, 1988), or to be affected by these functions. All seem to be mediated by the frontal cortex, and particularly the prefrontal lobes (Fuster, 1997; Stuss & Benson, 1986). Executive functions have been defined as those neuropsychological processes that permit or assist with human self-regulation (Barkley, 1997b, 2001a, 2001b), which itself has been defined as any behavior by a person that modifies the probability of a subsequent behavior by that person so as to alter the probability of a later consequence (Kanfer & Karoly, 1972). By classifying cognitive actions or thinking as private behavior, one can understand how these private self-directed, cognitive (executive) actions fall within the definition of human self-regulation—they are private behaviors (cognitive acts) that modify other behaviors so as to alter the likelihood of later consequences for the individual. And, by appreciating the role of the frontal lobes generally, and the prefrontal cortex particularly, in these executive abilities, it is easy to see why researchers have repeatedly speculated that ADHD probably arises out of some disturbance or dysfunction of this brain region (Barkley, 1997b; Heilman, Voeller, & Nadeau, 1991; Levin, 1938; Mattes, 1980; Nigg, 2006).
The most recent diagnostic criteria for ADHD as defined in DSM-IV (American Psychiatric Association, 1994) are set forth in Table 1.1. These diagnostic criteria are some of the most rigorous and most empirically derived criteria ever available in the history of clinical diagnosis for this disorder. They were derived from a committee of some of the leading experts in the field, a literature review of ADHD, an informal survey of empirically derived rating scales assessing the behavioral dimensions related to ADHD by the committee, and from statistical analyses of the results of a field trial of the items using 380 children from 10 different sites in North America (Lahey et al., 1994).
|
Table 1.1: DSM-IV Criteria for ADHD |
|
A. Either (1) or (2):
B. Some hyperactive–impulsive or inattentive symptoms that caused impairment were present before age 7 years. C. Some impairment from the symptoms is present in two or more settings (e.g., at school [or work] and at home). D. There must be clear evidence of clinically significant impairment in social, academic, or occupational functioning. E. The symptoms do not occur exclusively during the course of a Pervasive Developmental Disorder, Schizophrenia, or other Psychotic Disorder and are not better accounted for by another mental disorder (e.g., Mood Disorder, Anxiety Disorder, Dissociative Disorder, or a Personality Disorder). Code based on type:
Note. From American Psychiatric Association (1994, pp. 83–85). Copyright 1994 by the American Psychiatric Association. Reprinted by permission. |
Despite its empirical basis, the DSM criteria have some problems. As noted earlier, evidence is mounting that the predominantly inattentive type of ADHD (ADHD-PI) may be comprised of a rather heterogeneous mix of children, a subset of whom have a qualitatively different disorder of attention and cognitive processing (Milich et al., 2001). This subset is probably not a subtype of ADHD but may represent a separate disorder (Barkley, 1998, 2006; Milich et al., 2001), manifesting a sluggish cognitive style and selective attention deficit, having less comorbidity with oppositional and conduct disorder, demonstrating a more passive style of social relationship, having memory retrieval problems, and, owing to their lower level of impulsiveness, probably having a different, more benign, developmental course. Other children consigned to this subtype may be children who formerly met Combined Type status but with age had a sufficient decline in their hyperactive symptoms to no longer qualify for this subtype. For example, in our follow-up study of hyperactive children, all of whom likely had the Combined Type in childhood, we found that 16 percent of these cases (or 27% of persistent cases) now met criteria only for the Inattentive Type as young adults (Barkley, Fischer et al., 2002). Such cases might best be thought of as residual Combined Types rather than as having the Inattentive subtype. Likewise some children in the latter subtype place just a single symptom or two short of Combined Type status yet resemble the Combined Type, albeit in milder form, in all other respects. Mixing these formerly Combined and subthreshold Combined Type children together into the Inattentive Type is likely to constrain research on the distinctive features of this subtype, its etiology, response to treatments, and developmental course. In agreement with Milich et al. (2001), I believe the subset having hypoactivity, lethargy, and sluggish cognitive tempo should be set aside as a separate disorder from ADHD (Barkley, 2001a).
It is also unclear whether the predominantly hyperactive–impulsive type (ADHD-PHI) is really a separate type from the Combined Type (ADHD-C) or simply an earlier developmental stage of it. The field trial found that ADHD-PHI was primarily comprised of preschool aged children, whereas ADHD-C was primarily school-aged children. As noted above, this is what one would expect to find given that the hyperactive–impulsive symptoms appear first, followed within a few years by those of inattention. If one is going to require that inattention symptoms be part of the diagnostic criteria, then the age of onset for such symptoms will necessitate that ADHD-C have a later age of onset than ADHD-PHI. It seems that these two types may actually be developmental stages of the same type of ADHD.
Are the two separate symptom lists important rather than the one combined list used in DSM-III-R? . Perhaps, but this is unsure because these dimensions are highly correlated with each other and also have substantial shared genetic overlap (McLoughlin et al., 2007).. In the field trial (Lahey et al., 1994), significant levels of inattention mainly predicted additional problems with completing homework that were not as well predicted by the hyperactive–impulsive behavior. Otherwise, the latter predicted most of the other areas of impairment studied in this field trial. Other studies find that childhood symptoms of hyperactivity are related to adverse adolescent outcomes, such as antisocial behavior, substance abuse, and school disciplinary actions, such as suspensions/expulsions (Babinski, Hartsough, & Lambert, 1999). Symptoms of inattention seem to be primarily predictive of impairment in academic achievement, particularly reading, and school performance (DuPaul, Power, Anastopoulos, & Reid, 1998; Fischer, Barkley, Fletcher, & Smallish, 1993b; Weiss & Hechtman, 1993; Rabiner, Coie, and the Conduct Problem Prevention Research Group, 2000). Severity of hyperactive-impulsive behavior is often found to be the dimension of ADHD that more strongly predicts later conduct disorder and so risk for various forms of substance use and abuse (Molina, Smith, & Pelham, 1999). A recent study suggests that adolescent inattention, however, may contribute further to the risk for tobacco use beyond that risk contributed by severity of conduct disorder alone (Burke, Loeber, & Lahey, 2001).
Another critical issue deserving consideration is how well the diagnostic thresholds set for the two symptom lists apply to age groups outside of those used in the field trial (ages 4–16 years, chiefly). This concern arises out of the well-known findings that the behavioral items comprising these lists, particularly those for hyperactivity, decline significantly with age (DuPaul et al., 1998; Hart et al., 1996). Applying the same threshold across such a declining developmental slope could produce a situation where a larger percentage of young preschool aged-children (ages 2–3 years) would be inappropriately diagnosed as ADHD (false positives), whereas a smaller than expected percentage of adults would meet the criteria (false negatives). Support of just such a problem with using these criteria for adults was found in a study (Murphy & Barkley, 1996a) collecting norms for DSM-IV item lists on a large sample of adults, ages 17 to 84 years. The threshold needed to place an individual at the 93rd percentile for that person’s age group declined to four of nine inattention items and five of nine hyperactive–impulsive items for ages 17 to 29 years, then to four of nine on each list for the 30- to 49-year age group, then to three of nine on each list for those 50 years and older. Additional evidence also supports the use of a lower threshold for diagnosis with adults (typically 4 symptoms per list) (Barkley et al., 2008). Studies of the utility of the diagnostic thresholds to preschool children younger than 4 years remain to be done. Until then, it seems prudent to utilize the recommended symptom list thresholds only for children ages 4 to 16 years.
The issue of selecting symptom cutoff scores raises a related conceptual problem for ADHD as well. Is ADHD a static psychopathology, the symptoms of which remain essentially the same regardless of age? Or is it a developmental disorder (delay in rate)? In the latter case it must always be determined by comparison to same-age peers. While the DSM criteria imply that ADHD is a developmental disorder (symptoms must be developmentally inappropriate), it also treats the disorder as a relatively static category by using fixed symptom cutoff scores across all age groups. Available research indicates that ADHD is most likely a dimensional disorder (Levy & Hay, 2001; Nigg, 2006), representing the extreme of or delay in a normal trait(s), and so is akin to other developmental disorders, such as mental retardation. If so, then like all developmental disorders, ADHD reflects a delay in the rate at which a normal trait is developing, not an absolute loss of function, failure to develop, or pathological state. It needs to be diagnosed as a developmentally relative deficit, say for instance the 93rd or 98th percentile in severity of symptoms for age (DuPaul et al., 1999).
This notion of changing symptom thresholds with age raises another critical issue for developing diagnostic criteria for ADHD, and this is the appropriateness of the content of the item set for different developmental periods. Inspection of the item lists suggests that the items for inattention may have a wider developmental applicability across school-age ranges of childhood and even into adolescence and young adulthood. Those for hyperactive-impulsive behavior, in contrast, seem much more applicable to young children and less appropriate or not at all to older teens and adults. Recall from above (Hart et al., 1996) that the symptoms of inattention remain stable across middle childhood into early adolescence, whereas those for hyperactive–impulsive behavior decline significantly over this same course. Although this may represent a true developmental decline in the severity of the latter symptoms, and possibly in the severity and prevalence of ADHD itself, it could also represent an illusory developmental trend. That is, it might be an artifact of using more preschool focused items for hyperactivity and more school age focused items for inattention.
An analogy using mental retardation may be instructive. Consider the following items that might be chosen to assess developmental level in preschool-aged children: being toilet-trained, recognizing colors, counting to 10, repeating 5 digits, buttoning snaps on clothing, recognizing simple geometric shapes, and using a vocabulary repertoire of at least 50 words. Evaluating whether or not a child is able to do these things may prove to be very useful in distinguishing mental deficiency in preschoolers. However, if one continued to use this same item set to assess children with mental deficiency as they grew older, one would find a decline in the severity of the retardation in such children as progressively more items were achieved with age. One would also find that the prevalence of retardation would decline markedly with age as many formerly delayed children “outgrew” these problems. But we know this would be illusory because mental retardation represents a developmentally relative deficit in the achievement of mental and adaptive milestones.
Returning to the diagnosis of ADHD, if the same developmentally restricted item sets are applied throughout development with no attempt to adjust either the thresholds or, more importantly, the types of items developmentally appropriate for different periods, we might see the same results as with the analogy to mental retardation shown here. Similar results are found in ADHD (see below) giving one pause before one interpreting the observed decline in symptom severity (and even the observed decline in apparent prevalence!) as being accurate. As it now stands, ADHD is being defined mainly by one of its earliest developmental manifestations (hyperactivity) and one of its later (school-age) yet secondary sequelae (goal-directed persistence) and only minimally by its central features (inhibition and executive functioning).
Also of concern is the absence of any requirement in the DSM for the symptoms to be corroborated by someone that has known the patient well, such as a parent, sibling, long-time friend, or partner. Most likely, this arises from the focus on children throughout much of the history of the ADHD diagnostic category. Children routinely come to professionals with someone who knows them well (parents). But, in the case of adults who are self-referred to professionals, this oversight could prove potentially problematic (Barkley et al., 2008). For instance, available evidence suggests that ADHD children (Henry, Moffitt, Caspi, Langley, & Silva, 1994) and teens and young adults significantly under-report the severity of their symptoms relative to the reports of parents (Barkley, Fischer et al., 2002; Edwards, Barkley, Laneri, & Metevia, 2001; Fischer et al., 1993b; Gittelman & Mannuzza, 1986; Romano, Tremblay, Vitaro, Zoccolillo, & Pagani, 2001). If this occurs in older adults having ADHD as well, it would mean that self-referred patients might under-estimate the severity of their disorder resulting in a sizable number of false negative decisions being made by clinicians. There is good reason that self-awareness might be limited by this disorder. Neuropsychological research indicates that self-awareness is relatively localized to the prefrontal lobes with disorders affecting this region, such as Alzheimer’s disease, markedly reducing self-awareness (Fuster, 1997; Stuss & Benson, 1986). As evidence reviewed below suggests, under-activity and under-development in these same regions of the brain are likely to be involved in ADHD and so the disorder ought to restrict self-awareness.
These issues are not merely academic. My colleagues and I have been involved in follow-up research with ADHD children into their adulthood and have been impressed at the chronicity of impairments created by the disorder despite an apparent decline in the percentage of cases continuing to meet diagnostic criteria and an apparent decline in the severity of the symptoms used in these criteria (Barkley, Fischer, Edelbrock, & Smallish, 1990; Barkley, Fischer, Smallish, & Fletcher, 2002; Fischer et al., 1993a). Recently, we found that if the formerly ADHD children, who are now adults, are interviewed using the DSM criteria, just 5 percent of them report sufficient symptoms to receive the diagnosis (Barkley, Fischer et al., 2002), a figure nearly identical to that for the New York longitudinal studies (Mannuzza, Klein, Bessler, Malloy, & LaPadula, 1993, 1998). If instead the parents are interviewed, this figure rises to 46% – a nine-fold difference in persistence of disorder as a function of reporting source. If instead of the recommended DSM symptom threshold, one substitutes a developmentally referenced criterion (the 98th percentile) based on same-age control adults, then 12% of the probands now have the disorder as adults based on self-reports while the figure climbs to 66% based on parental reports. Whose reports are more valid of current functioning? We addressed this by examining the relationship of self-reports and parent-reports to various domains of major life activities and outcomes (education, occupational functioning, friendships, crime, etc.). Parent reports made a substantially larger contribution to nearly all outcome domains and did so for more such domains than did self-reports, suggesting that parental reports probably have greater validity. The higher rates of disorder they reported at outcome are probably the more accurate ones. Such adjustments for age and source of reporting, however, do not correct for the potentially increasing inappropriateness of the item sets for this aging sample, and so it is difficult to say how many of those not meeting these adjusted criteria may still have had the disorder.
A different issue pertains to whether or not the criteria should be adjusted for the gender of the children being diagnosed. Research evaluating these and similar item sets demonstrates that male youngsters display more of these items and to a more severe degree than do female youngsters in the general population (Achenbach, 1990; DuPaul et al., 1999). Given that the majority of children in the DSM field trial were boys (Lahey et al., 1994), the symptom threshold chosen in the DSM is most appropriate to males. This results in girls having to meet a higher threshold relative to other girls to be diagnosed as ADHD than do boys relative to other boys. Gender-adjusted thresholds would seem to be in order to address this problem, yet this would evaporate the currently disproportionate male-to-female ratio of 3:1 found across studies (see below).
The requirement of an age of onset for ADHD symptoms (7 years) in the diagnostic criteria has also come under attack from its own field trial (Applegate et al., 1997), a longitudinal study (McGee, Williams, & Feehan, 1992), and a review of this criterion from historical, empirical, and pragmatic perspectives (Barkley & Biederman, 1997). Such a criterion for age of onset suggests that there may be qualitative differences between those who meet the criterion (early onset) and those who do not (late onset). Some results do suggest that those with an onset before age 6 years may have more severe and persistent conditions and more problems with reading and school performance more generally (McGee et al., 1992). But these were matters of degree and not kind in this study. The DSM-IV field trial also was not able to show any clear discontinuities in degree of ADHD or in the types of impairments it examined between those meeting and those not meeting the 7-year age of onset. It remains unclear at this time as to just how specific an age of onset may need to be for distinguishing ADHD from other disorders. And more recent studies have shown that adults reporting an age of onset of ADHD before 7 do not differ in any important respects from those reporting an onset after age 7 (Barkley et al., 2008). Suffice to say that no other mental disorder in the DSM-IV has so precise an age of onset, arguing that ADHD should not as well. I have suggested that age 16 might serve as a better age of onset criterion for diagnosis (Barkley et al., 2008).
A related potential problem for these criteria occurs in their failure to stipulate a lower bound age group for giving the diagnosis below which no diagnosis should be made. This is important because research on preschool children has shown that a separate dimension of hyperactive–impulsive behavior from aggression or defiant behavior does not seem to emerge until about 3 years of age (Achenbach & Edelbrock, 1987; Campbell, 1990). Below this age, these behaviors cluster together to form what has been called behavioral immaturity, externalizing problems, or an under-controlled pattern of conduct. This implies that the symptoms of ADHD may be difficult to distinguish from other early behavioral disorders until at least 3 years of age, and so this age might serve as a lower bound for diagnostic applications.
Similarly, research implies that a lower bound of IQ might also be important (IQ>50) below which the nature of ADHD may be quite different. Minimal research seems to exist that speaks to the issue of a discontinuity or qualitative shift in the nature of ADHD in individuals below IQs of 50. Some indirect evidence implies that this may occur, however. Rutter and colleagues (Rutter, Bolton, et al., 1990; Rutter, Macdonald, et al., 1990) have concluded that children who fall below this level of IQ may have a qualitatively different form of mental retardation. This is inferred from findings that this group is over represented for its position along a normal distribution and from findings that genetic defects contribute more heavily to this subgroup. Given this shift in the prevalence and causes of mental retardation below this level of IQ, a similar state of affairs might exist for the form of ADHD associated with it necessitating its distinction from the type of ADHD that occurs in individuals above this IQ level. Consistent with such a view have been findings that the percentage of positive responders to stimulant medication falls off sharply below this threshold of IQ (Demb, 1991).
Another issue pertinent to the above is the problem of the duration requirement being set at 6 months. This has been chosen mainly out of tradition (earlier DSMs) with no research support for selecting this particular length of time for symptom presence. It is undoubtedly important that the symptoms be relatively persistent if we are to view this disorder as a developmental disability rather than arising purely from context or out of a transient, normal developmental stage. Yet specifying a precise duration is difficult in the absence of much research to guide the issue. Research on preschool-aged children might prove helpful here, however. Such research has shown that many children aged 3 years (or younger) may have parents or preschool teachers who report concerns about the activity level or attention of the children, yet these concerns have a high likelihood of remission within 12 months (Beitchman, Wekerle, & Hood, 1987; Campbell, 1990; Lerner, Inui, Trupin, & Douglas, 1985; Palfrey, Levine, Walker, & Sullivan, 1985). It would seem for preschoolers that the 6–month duration specified in the DSM-IV may be too brief, resulting in overidentification of ADHD children at this age (false positives). However, this same body of research found that for those children whose problems lasted at least 12 months or beyond age 4 years, the behavior problems were highly persistent and predictive of continuance into the school-age range. Such research suggests that the duration of symptoms be set at 12 months or more.
The DSM requirement that the symptoms be demonstrated in at least two of three environments so as to establish pervasiveness of symptoms is new to this edition and problematic. The DSM implies that two of three sources of information (parent, teacher, employer) must agree on the presence of the symptoms. This confounds settings with sources of information. The degree of agreement between parents and teacher for any dimension of child behavior is modest, often ranging between .30 and .50 (Achenbach, McConaughy, & Howell, 1987). This sets an upper limit on the extent to which parents and teachers are going to agree on the severity of ADHD symptoms and, thus, on whether or not the child has the disorder in that setting. Such disagreements among sources certainly reflect differences in the child’s behavior as a function of true differential demands of these settings. But they also reflect differences in the attitudes and judgments between different people. Insisting on such agreement may reduce the application of the diagnosis to some children unfairly as a result of such well-established differences between parent and teacher opinions. It may also create a confounding of the disorder with issues of comorbidity with oppositional defiant disorder (ODD) (Costello, Loeber, & Stouthamer-Loeber, 1991). Parent-only identified ADHD children may have predominantly ODD with relatively milder ADHD, whereas teacher-only identified ADHD children may have chiefly ADHD and minimal or no ODD symptoms. Children identified by both parents and teachers as ADHD may, therefore, carry a higher likelihood of ODD. They may also simply reflect a more severe condition of ADHD than do the home- or school-only cases, being different in degree rather than in kind. Research is clearly conflicting on the matter (Cohen & Minde, 1983; Rapoport, Donelly, Zametkin, & Carrougher, 1986; Schachar, Rutter, & Smith, 1981; Taylor, Sandberg, Thorley, & Giles, 1991). Considering that teacher information on children is not always obtainable or convenient, that parents can convey the essence of that information to clinicians, and that diagnosis based on parents’ reports will lead to a diagnosis based on teacher reports 90% of the time (Biederman, Keenan, & Faraone, 1990), all imply that parent reports could suffice for diagnostic purposes for now. However, more recent evidence suggests that the best discrimination of ADHD children from other groups might be achieved by blending the reports of parents and teachers such that one counts the number of different symptoms endorsed across both sources of information (Crystal, Ostrander, Chen, & August, 2001; Mitsis, McKay, Schulz, Newcorn, & Halperin, 2000).
Many of these problematic issues are likely to be addressed in future editions of the DSM. Even so, the present criteria are actually some of the best ever advanced for the disorder and represent a vast improvement over the state of affairs that existed prior to 1980. The various editions of DSM also have spawned a large amount of research into ADHD, its symptoms, subtypes, criteria, and even etiologies that likely would not have occurred had such criteria not been set forth for professional consumption and criticism. The most recent criteria provide clinicians with a set of guidelines more specific, reliable, empirically based or justifiable, and closer to the scientific literature on ADHD than earlier editions. With some attention to the above issues, the DSM criteria could be made to be even more rigorous, valid, and useful.
Social critics (Breggin, 1999; Kohn, 1989; Schrag & Divoky, 1975) and the Church of Scientology via its Citizens Commission on Human Rights have charged that professionals have been too quick to label energetic and exuberant children as having a mental disorder. They also assert that educators may be using these labels as an excuse for simply poor educational environments. In other words, children who are hyperactive or ADHD are actually normal but are being labeled as mentally disordered because of parent and teacher intolerance (Kohn, 1989) or lack of love at home (Breggin, 1999). If this were actually true, then we should find no differences of any cognitive, neurological, genetic, behavioral, or social significance between children so labeled and normal children. We should also find that the diagnosis of ADHD is not associated with any significant risks later in development for maladjustment within any domains of adaptive functioning, or social, occupational, or school performance. Furthermore, research on potential etiologies for the disorder should, likewise, come up empty-handed. This is hardly the case, as evidence reviewed in this chapter attests. Differences between ADHD and normal children are too numerous to take these assertions of normality seriously. As will be shown later, substantial developmental risks await the child meeting clinical diagnostic criteria for the disorder, and certain potential etiological factors are becoming consistently noted in the research literature.
Conceding all of this, however, does not automatically entitle ADHD to be placed within the realm of valid (“real”) disorders. Wakefield (1999) has argued that disorders must meet two criteria to be viewed as valid: (1) engender substantial harm to the individual or those around him or her and (2) incur dysfunction of natural and universal mechanisms that have been selected in an evolutionary sense (have survival value). The latter criterion is simply the definition of an adaptation as used in evolutionary biology. Disorders are failures in adaptations that produce harm. In the case of psychology, these universal mechanisms are psychological ones possessed by all normally developing humans, regardless of culture. ADHD handily meets both criteria. Those with ADHD, as described in the theory presented below, have significant deficits in behavioral inhibition and inattention (the executive functions) that are critical for effective self-regulation. And, those with ADHD experience numerous domains of impairment (risks of harm) over development, as will become evident below.
The prevalence of ADHD varies across studies, at least in part, due to different methods of selecting samples, the nature of the populations from which they are drawn (nationality or ethnicity, urban vs. rural, community vs. primary care settings, etc.), the criteria used to define ADHD (DSM criteria vs. rating scale cutoff), and certainly to the age range and sex composition of the samples. When only the endorsement of the presence of the behavior of hyperactivity (not the clinical disorder) is required from either parent or teacher rating scales, prevalence rates can run as high as 22% to 57% (Lapouse & Monk, 1958; McArdle, O’Brien, & Kolvin, 1995; Werry & Quay, 1971). This underscores the point made earlier that being described as inattentive or overactive by a parent or teacher does not in and of itself constitute a disorder in a child.
Szatmari (1992) reviewed the findings of six large epidemiological studies that identified cases of ADHD within these samples. The prevalences found in these studies ranged from a low of 2% to a high of 6.3% with most falling within the range of 4.2% to 6.3%. Other subsequent studies have also found similar prevalence rates in elementary school-aged children ([4-5.5%] Breton, Bergeron, Valla, Berthiaume, Gaudet, Lambert, St. Georges, Houde, & Lepine, 1999; [7.9%] Briggs-Gowan, Horwitz, Schwab-Stone, Leventhal, & Leaf, 2000; [5–6%] DuPaul, 1991; [2.5–4%] Pelham, Gnagy, Greenslade, & Milich, 1992). Lower rates result from using complete DSM criteria and parent reports [2-6%](Breton et al., 1999) and higher ones if just a cutoff on teacher ratings are used ([up to 23%] DuPaul et al., 1999; [15.8%] Nolan, Gadow, & Sprafkin, 2001; [14.3%] Trites, Dugas, Lynch, & Ferguson, 1979). Sex and age differences in prevalence are routinely found in research. For instance, prevalence rates may be 4% in girls and 8% in boys in the preschool age group (Gadow, Sprafkin, & Nolan, 2001) yet fall to 2% to 4% in girls and 6% to 9% in boys during the 6- to 12-year-old age period based on parent reports (Breton et al., 1999; Szatmari et al., 1989). The prevalence will decrease again to 0.9 to 2% in girls and 1% to 5.6% in boys by adolescence (Breton et al., 1999; Lewinsohn, Hops, Roberts, Seeley, & Andrews, 1993; McGee et al., 1990; Romano et al., 2001; Szatmari et al., 1989). Even then, if both a symptom threshold and the requirement for impairment are used, the prevalence may decrease by 20-60% from that figure based on symptom thresholds alone (Breton et al., 1999; Romano et al., 2001; Wolraich, Hannah, Baumgaertel, & Feurer, 1998). As noted above, prevalence rates are routinely higher (sometimes more than double) when teacher reports are used in comparison to parent reports (Breton et al., 1999; DuPaul et al., 1999; Gadow et al., 2001). Switching from DSM-III-R criteria used before 1994 to DSM-IV in use since that time may have resulted in a near doubling in prevalence owing to the inclusion of the new inattentive subtype that was not included in DSM-III-R (Wolraich, Hannah, Pinnock, Baumgaertel, & Brown, 1996). Some segments of the population may also have greater levels of ADHD than others. For instance, Jensen and colleagues found a prevalence of 12% for ADHD among the children of military personnel using DSM-III-R criteria (Jensen et al., 1995) – a figure more than double that found in other studies using these same criteria with general population samples (Szatmari, 1992). A recent meta-analysis of worldwide prevalence studies reported an average of 5.5% of children (Polanczyk et al., 2007) while a epidemiological study of U.S. adults placed the prevalence at 4.4%.(Kessler et al., 2006).
Szatmari et al. (1989) found that the prevalence of ADHD in a large sample of children from Ontario, Canada also varied as a function of young age, male gender, chronic health problems, family dysfunction, low socioeconomic status, presence of a developmental impairment, and urban living. Others have found similar conditions being associated with the risk for ADHD (Lavigne et al., 1996; Velez, Johnson, & Cohen, 1989). Important, however, was the additional finding in the Szatmari et al. (1989) study that when comorbidity with other disorders was statistically controlled in the analyses, gender, family dysfunction, and low socioeconomic status were no longer significantly associated with prevalence. Health problems, developmental impairment, young age, and urban living remained significantly associated with prevalence, however.
As noted above in discussing DSM criteria, it may be that the declining prevalence of ADHD with age is partly artifactual. This could result from the use of items in the diagnostic symptom lists that are chiefly applicable to young children. This could create a situation where individuals remain impaired in the construct(s) comprising ADHD as they mature while outgrowing the symptom list for the disorder, resulting in an illusory decline in prevalence as was noted in my Milwaukee follow-up study discussed above. Until more age-appropriate symptoms are studied for adolescent and adult populations, this issue remains unresolved.
As noted above, sex appears to play a significant role in determining prevalence of ADHD within a population. On average, male children are between 2.5 and 5.6 times more likely than female children to be diagnosed as ADHD within epidemiological samples, with the average being roughly 3:1 (Breton et al., 1999; DuPaul et al., 1999; Lewinsohn et al., 1993; McGee et al., 1990; Szatmari, 1992). Within clinic-referred samples, the sex ratio can be considerably higher, suggesting that boys with ADHD are far more likely to be referred to clinics than girls. This is probably because boys are more likely to have a comorbid oppositional or conduct disorder. Szatmari (1992) found that sex differences were no longer associated with the occurrence of ADHD once other comorbid conditions were controlled for in their statistical analyses implies that this may be the case. The sex ratio could also be an artifact of applying a set of diagnostic criteria developed primarily on males to females, as discussed above.
Studies of clinic-referred girls often find that they are as impaired as clinic-referred boys with ADHD, have as much comorbidity, and may even have greater deficits in intelligence, according to meta-analytic reviews of sex differences in ADHD (Gaub & Carlson, 1997; Gershon, 2001). Some studies suggest these clinic-referred girls, at least as adolescents, may have more internalizing symptoms, such as depression, anxiety, and stress, greater problems with teacher relationships, and poorer verbal abilities (vocabulary) in comparison to ADHD boys (Rucklidge & Tannock, 2001). Like the boys, girls with ADHD also manifest more conduct, mood, and anxiety disorders, have lower intelligence, and have greater academic achievement deficits than do control samples (Biederman, Faraone, et al., 1999; Rucklidge & Tannock, 2001). Males with ADHD may have greater problems with cognitive processing speed than females but these differences were no longer significant after controlling for severity of ADHD (Rucklidge & Tannock, 2001). No sex differences have been identified in executive functioning, with both sexes being more impaired than control samples on such measures (Barkley, 2006; Barkley et al., 2008; Castellanos et al., 2000; Murphy et al., 2001). In contrast, studies drawing their ADHD samples from the community find that girls are significantly less likely to have comorbid ODD and CD than boys with ADHD, do not have greater intellectual deficits than ADHD boys, yet may be as socially and academically impaired as boys with the disorder (Gaub & Carlson, 1997; Gershon, 2001; Green et al., 2001).
Few studies have examined the relationship of ADHD to social class, and those that have are not especially consistent. Lambert et al. (1978) found only slight differences in the prevalence of hyperactivity across social class when parent, teacher, and physician all agreed on the diagnosis. However, social class differences in prevalence did arise when only two of these three sources had to agree, with there generally being more ADHD children in lower than higher social classes. For instance, when parent and teacher agreement (but not physician) was required, 18% of those identified as hyperactive were in the high social class, 36% in the middle, and 45% in the low social class. Where only the teacher’s opinion was used, the percentages were 17%, 41%, and 41%, respectively. Trites (1979), and later Szatmari (1992), both found that rates of ADHD tended to increase with lower socioeconomic class. However, in his own study Szatmari (Szatmari et al., 1989) found that low socioeconomic status was no longer associated with rates of ADHD when other comorbid conditions, such as conduct disorder, were controlled. For now, it is clear that ADHD occurs across all socioeconomic levels. Variations across social classes may be artifacts of the source used to define the disorder or of the comorbidity of ADHD with other disorders related to social class, such as oppositional and conduct disorder.
Early studies of the prevalence of hyperactivity, relying principally on teacher ratings, found significant disparities across four countries (United States, Germany, Canada, and New Zealand), ranging from 2% in girls and 9% in boys in the United States to 9% in girls and 22% in boys in New Zealand (Trites et al., 1979). Similarly, O’Leary, Vivian, and Nisi (1985) found rates of hyperactivity to be 3% in girls and 20% in boys in Italy using this same teacher rating scale and cutoff score. However, this may have resulted from the use of a threshold established on norms collected in the United States across these other countries, where the distributions were quite different from those found in the United States.
Later studies, especially those using DSM criteria, have found the disorder across numerous countries. Among a Japanese sample (Kanbayashi, Nakata, Fujii, Kita, & Wada, 1994) using parent ratings of items from DSM-III-R, a prevalence rate of 7.7% of the sample was found. Baumgaertel (1994) used teacher ratings of DSM-III, DSM-III-R, and DSM-IV symptom lists in a large sample of German elementary school children and found rates of 4.8% for ADHD-C, 3.9% for ADHD-PHI, and 9% for ADHD-PI subtypes based on DSM-IV. In India, among over 1,000 children screened at a pediatric clinic, 5.2% of children ages 3 to 4 years were found to have ADHD by DSM-III-R criteria, whereas the rate rose to over 29% for ages 11 to 12 years (Bhatia, Nigam, Bohra, & Malik, 1991). This was not a true epidemiological sample, however. Differences in prevalence across ages could simply reflect cohort effects—children are referred to this clinic for different reasons at different ages. Prevalence rates found in other countries more recently are:
Cultural differences in the interpretations given to symptoms of ADHD by teachers or parents and in expectations for child behavior undoubtedly exist and likely contributed to the higher rates of disorder found in some of these countries compared to North American rates. Also, most of these studies used teacher or parent ratings rather than clinical diagnostic criteria. As already noted above, prevalence rates of hyperactivity or ADHD are typically higher when simply a threshold on a rating scale is the only criterion for establishing a case of the disorder. Where clinical criteria are employed, rates are more conservative. Nevertheless, these studies show that hyperactivity or ADHD is present in all countries studied to date. Although it may not receive the same diagnostic label in each, the behavior pattern comprising the disorder appears to universal.
Differences among ethnic groups in rates of hyperactivity within the United States have been reported. Langsdorf, Anderson, Walchter, Madrigal, and Juarez (1979) reported that almost 25% of African-American children and 8% of Latino-Americans met a cutoff score on a teacher rating scale commonly used to define hyperactivity, whereas Ullmann (cited in O’Leary et al., 1985) reported rates of 24% for African-American children and 16% of white Americans on a teacher rating scale. Lambert et al. (1978) found higher rates of hyperactivity among African-American than white American children only when the teachers were the only ones reporting the diagnosis; Latino-American children were not found to differ from white American children in this respect. Such differences, however, may arise in part because of socioeconomic factors that are differentially associated with these ethnic groups in the United States. Such psychosocial factors are strongly correlated with aggression and conduct problems. As noted above, those factors no longer make a significant contribution to the prevalence of ADHD when comorbidity for other disorders is controlled (Szatmari, 1992). Doing the same within studies of ethnic differences might well reduce or eliminate these differences in prevalence among them. Thus, it would seem that ADHD arises in all ethnic groups studied so far. Whether the differences in prevalence across these ethnic groups are real or are a function of the source of information about the symptoms of ADHD and, possibly, socioeconomic factors remains to be determined.
Major follow-up studies of clinically referred hyperactive children have been ongoing during the last 25 years at five sites: (1) Montreal (Weiss & Hechtman, in press), (2) New York City (Gittelman, Mannuzza, Shenker, & Bonagura, 1985; Mannuzza, Gittelman-Klein, Bessler, Malloy, & Lapadulo, 1993), (3) Iowa City (Loney, Kramer, & Milich, 1981), (4) Los Angeles (Satterfield, Hoppe, & Schell, 1982), and (5) Milwaukee (Barkley, Fischer, et al., 1990; Barkley et al., 2008). Follow-up studies of children identified as hyperactive from a general population have also been conducted in the United States (Lambert, 1988), New Zealand (McGee, Williams, & Silva, 1984; Moffitt, 1990), and England (Taylor et al., 1991), among others.
But before embarking on a summary of their results, some cautionary notes are in order. First, the limited number of follow-up studies does not permit a great deal of certainty to be placed in the specificity of the types and degrees of outcomes likely to be associated with ADHD. Even so, more can likely be said about the outcomes of ADHD children than about most other childhood mental disorders. Second, the discontinuities of measurement that exist in these follow-up studies between their different points of assessments of their subjects make straightforward conclusions about developmental course difficult. Third, the differing sources of children greatly affect the outcomes to be found, with children drawn from clinic-referred populations having two to three times the occurrence of some negative outcomes and more diverse negative outcomes than those drawn from population screens (i.e., Barkley, Fischer, et al., 1990, vs. Lambert, 1988). Fourth, the differing entry/diagnostic criteria across follow-up studies must be kept in mind in interpreting and cross-referencing their outcomes. Most studies selected for children known at the time as “hyperactive.” Such children are most likely representative of the course of the ADHD Combined Type from the current DSM taxonomy. Even then, the degree of deviance of the samples on parent and teacher ratings of these symptoms was not established at the entry point in most of these studies. These studies also cannot be viewed as representing the Inattentive subtype, for which no follow-up information is currently available. The descriptions of clinic-referred ADHD children who are of similar age groups to those in the follow-up studies but who are not followed over time may help understand the risks associated with different points in development. However, these may also be contaminated by cohort effects at the time of referral and so can only be viewed as suggestive. Such cohort effects may be minor; that is, adolescents with ADHD referred to clinics seem to have similar types and degrees of impairment as ADHD children followed up to adolescence (Barkley, Anastopoulos, Guevremont, & Fletcher, 1991 vs. Barkley, Fischer, et al., 1990). In painting the picture of the developmental outcome of ADHD, then, broad strokes are permissible but the finer details await more and better-refined studies. I concentrate here on the course of the disorder itself, returning to the comorbid disorders and associated conditions likely to arise in the course of ADHD in a later section of this chapter (“Comorbid Psychiatric Disorders”).
The average onset of ADHD symptoms, as noted earlier, is often in the preschool years, typically at ages 3 to 4 years (Applegate et al., 1997; Loeber et al., 1992; Taylor et al., 1991) and, more generally, by entry into formal schooling. Yet onset is heavily dependent on the type of ADHD under study. First to arise is the pattern of hyperactive–impulsive behavior and, in some cases, oppositional and aggressive conduct, giving that subtype the earliest age of onset. The Combined Type has an onset within the first few grades of primary school (ages 5-8)(Hart et al., 1995), most likely due to the requirement that both hyperactivity and inattention be present to diagnose this subtype. The Inattentive Type appears to emerge a few years later (ages 8-12) than the other types (Applegate et al., 1997).
Preschool-aged children who are perceived as difficult and resistant to control or who have inattentive and hyperactive behavior that persists for at least a year or more are highly likely to have ADHD and to remain so into elementary school years (Beitchman et al., 1987; Campbell, 1990; Palfrey et al., 1985) and even adolescence (Olson, Bates, Sandy, & Lanthier, 2000). Persistent cases seem especially likely to occur where parent-child conflict, greater maternal directiveness and negativity, and greater child defiant behavior exist (Campbell, March, Pierce, & Ewing, 1991; Olson et al., 2000; Richman, Stevenson, & Graham, 1982). More negative temperament and greater emotional reactivity to events are also more common in preschool ADHD children (Barkley, DuPaul, & McMurray, 1990; Campbell, 1990). It is little wonder that greater parenting stress is associated with preschool ADHD children and seems to be at its highest relative to later age groups (Mash & Johnston, 1983a, 1983b). Within the preschool setting, ADHD children will be found to be more often out of their seats, wandering the classroom, being excessively talkative and vocally noisy, and disruptive of other children’s activities (Campbell, Schleifer, & Weiss, 1978; Schleifer et al., 1975).
By the time ADHD children move into the elementary-age range of 6 to 12 years, the problems with hyperactive–impulsive behavior are likely to continue and to be joined now by difficulties with attention (executive functioning and goal-directed persistence. Difficulties with work completion and productivity, distraction, forgetfulness related to what needs doing, lack of planning, poor organization of work activities, trouble meeting time deadlines associated with home chores, school assignments, and social promises or commitments to peers are now combined with the impulsive, heedless, and disinhibited behavior typifying these children since preschool age. Problems with oppositional and socially aggressive behavior may emerge at this age in at least 40% to 70% of ADHD children (Barkley, 1998; Loeber et al., 1992; Taylor et al., 1991).
By ages 8 to 12 years, these early forms of defiant and hostile behavior may evolve further into symptoms of conduct disorder in 25-45% or more of all children with ADHD (Barkley, Fischer, et al., 1990; Gittelman et al., 1985; Loeber et al., 1992; Mannuzza et al., 1993; Taylor et al., 1991). Certainly by late childhood most or all of the deficits in the executive functions related to inhibition in the model presented earlier are likely to be arising and interfering with adequate self-regulation (Barkley, 1997b). Not surprisingly, the overall adaptive functioning (self-sufficiency) of many ADHD children (Stein, Szumowski, Blondis, & Roizen, 1995) is significantly below their intellectual ability. This is also true of preschoolers with high levels of these externalizing symptoms (Barkley et al., 1998). The disparity between adaptive functioning and age appropriate expectations (or IQ) may itself be a predictor of greater severity of ADHD as well as risk for oppositional and conduct problems in later childhood (Shelton, Barkley, et al., in press). The disorder takes its toll on self-care, personal responsibility, chore performance, trustworthiness, independence, appropriate social skills, and timeliness, specifically, and moral conduct generally (Barkley, 1998; Hinshaw et al., 1993).
If ADHD is present in clinic-referred children, the likelihood is that 50% to 80% will continue to have their disorder into adolescence, with most studies supporting the higher figure (August, Stewart & Holmes, 1983; Claude & Firestone, 1995; Barkley, Fischer, et al., 1990; Gittelman et al., 1985; Mannuzza et al., 1993). Using the same parent rating scales at both the childhood and adolescent evaluation points, Fischer et al. (1993a) were able to show that inattention, hyperactive-impulsive behavior, and home conflicts declined by adolescence. The hyperactive group showed far more marked declines than the control group, mainly because the former were so far from the mean of the normative group to begin with in childhood. Nevertheless, even at adolescence, the groups remained significantly different in each domain with the mean for the hyperactives remaining two standard deviations or more above the mean for the controls. This emphasizes the point made earlier that simply because severity levels of symptoms are declining over development does not mean hyperactive children are necessarily outgrowing their disorder relative to normal children. Like mental retardation, ADHD may need to be defined by a developmentally relative deficiency, rather than an absolute one, that persists in most children over time.
The persistence of ADHD symptoms across childhood as well as into early adolescence appears, again, to be associated with initial degree of hyperactive–impulsive behavior in childhood, the coexistence of conduct problems or oppositional hostile behavior, poor family relations and specifically conflict in parent–child interactions, as well as maternal depression, and duration of mental health interventions (Fischer et al., 1993b; Taylor et al., 1991). These predictors have also been associated with the development and persistence of oppositional and conduct disorder into this age range (12–17 years; Fischer et al., 1993b; Loeber, 1990; Mannuzza & Klein, 1992; Taylor et al., 1991).
Studies following large samples of clinic-referred children with hyperactivity, or (ADHD), into adulthood are few in number. Only four follow-up studies have retained at least 50 percent or more of their original sample into adulthood and reported on the persistence of symptoms to that time. These are the Montreal study by Weiss, Hechtman, and their colleagues (see Weiss & Hechtman, in press), the New York City study by Mannuzza, Klein, and colleagues (see Mannuzza, Klein, Bessler, Malloy, & LaPadula, 1993, 1998), the Swedish study by Rasmussen and Gillberg (2001), and my research with Mariellen Fischer in Milwaukee (Barkley et al., 2008; Barkley, Fischer et al., 2002Fischer et al., 2002). The results regarding the persistence of disorder into young adulthood (middle 20s) are mixed but can be better understood as being a function of reporting source and the diagnostic criteria used (Barkley et al., 2002).
The Montreal study (N=103) found that two-thirds of their original sample (N=64; mean age of 25 years) claimed to be troubled as adults by at least one or more disabling core symptoms of their original disorder (restlessness, impulsivity, or inattention) and that 34% had at least moderate to severe levels of hyperactive, impulsive, and inattentive symptoms (Weiss & Hechtman, 1993, p. 73). In Sweden (N=50), Rasmussen and Gillberg (2001) obtained similar results, with 49% of probands reporting marked symptoms of ADHD at age 22 years compared to 9% of controls. Formal diagnostic criteria for ADHD, as in DSM-III or later editions, were not employed at any of the outcome points in either study, however. In contrast, the New York study has followed two separate cohorts of hyperactive children using DSM criteria to assess persistence of disorder. That study found that 31% of their initial cohort (N=101) and 43% of their second cohort (N=94) met DSM-III criteria for ADHD by ages 16-23 (mean age=18.5 years) (Gittelman, et al. 1985; Mannuzza et al., 1991). Eight years later, (mean age 26 years), however, these figures fell to 8% and 4%, respectively (now using DSM-III-R criteria) (Mannuzza et al., 1993, 1998). Those results might imply that the vast majority of hyperactive children no longer qualify for the diagnosis of ADHD by adulthood.
The interpretation of the relatively low rate of persistence of ADHD into adulthood, particularly for the New York study, is clouded by at least two issues apart from differences in selection criteria. One is that the source of information about the disorder changed in all of these studies from that used at the childhood and adolescent evaluations to that used at the adult outcome. At study entry and at adolescence, all studies used the reports of others (parents and typically teachers). By mid-adolescence, all found that the majority of hyperactive participants (50-80%) continued to manifest significant levels of the disorder (see above). In young adulthood (approximately age 26 years), both the New York and Montreal studies switched to self-reports of disorder.
The rather marked decline in persistence of ADHD from adolescence to adulthood could stem from this change in source of information. Indeed, the New York study found this to be likely when, at late adolescence (mean age of 18-19 years), they interviewed both the teenagers and their parents about psychiatric status of the teens (Mannuzza & Gittelman, 1986). There was a marked disparity between the reports of parents and teens concerning the presence of ADHD (11% vs. 27%; agreement 74%, Kappa=.19). Other research also suggests that the relationship between older children’s (age 11) self-reports of externalizing symptoms, such as those involved in ADHD, and those of parents and teachers is quite low (r=.16-.32; Henry, Moffitt, Caspi, Langley, & Silva, 1994). Thus, changing sources of reporting in longitudinal studies on behavioral disorders could be expected to lead to marked differences in estimates of persistence of those disorders.
The question obviously arises as to whose assessment of the proband is more accurate. This would depend on the purpose of the assessment, but the prediction of impairment in major life activities would seem to be an important one in research on psychiatric disorders. Our Milwaukee study examined these issues by interviewing both the participants and their parents about ADHD symptoms at the young adult follow-up (age 21 years). It then examined the relationship of each source’s reports to significant outcomes in major life activities (education, occupation, social, etc.) after controlling for the contribution made by the other source. As noted earlier, another limitation in the earlier studies may reside in the DSM criteria in that they grow less sensitive to the disorder with age. Using a developmentally referenced criterion (age comparison) to determine diagnosis may identify more cases than would the DSM approach. As discussed earlier, the Milwaukee study found that the persistence of ADHD into adulthood was heavily dependent on the source of the information (self or parent) and the diagnostic criteria (DSM or developmentally referenced). Self-report identified just 5-12 percent of probands as currently ADHD (DSM-III-R) while parent reports placed this figure at 46-66 percent. Using the DSM resulted in lower rates of persistence (5 percent for proband reports and 46 percent for parents) while using a developmentally referenced cutoff (98th percentile) yielded higher rates of persistence (12% by self-report and 66 percent by parent reports). Parental reports appeared to have greater validity in view of their greater contribution to impairment and to more domains of current impairment than did self-reported information (Barkley et al., 2002). We concluded that past follow-up studies grossly under-estimated the persistence of ADHD into adulthood by relying solely on the self-reports of the probands.
Individuals diagnosed with ADHD are often found to have a number of other disorders besides their ADHD (Barkley, 2006). What is known about comorbidity is largely confined to the Combined Type of ADHD. In community derived samples, up to 44% of ADHD children have at least one other disorder and 43% have at least two or more additional disorders (Szatmari et al., 1989a). The figure is higher, of course, for children drawn from clinics. As many as 87 percent of clinically diagnosed ADHD children may have at least one other disorder and 67% have at least two other disorders (Kadesjo & Gillberg, 2001). The disorders likely to co-occur with ADHD are briefly described below.
The most common comorbid disorders with ADHD (Combined Type) are oppositional defiant disorder (ODD) and, to a lesser extent, conduct disorder (CD). Indeed, the presence of ADHD increases the odds of ODD/CD by 10.7 fold (95% Confidence Interval [CI] = 7.7-14.8) in general population studies (Angold, Costello, & Erkanli, 1999). Studies of clinic-referred ADHD children find that between 54% and 67% will meet criteria for a diagnosis of ODD by 7 years of age or later. ODD is a frequent precursor to CD, a more severe and often (though not always) later occurring stage of ODD (Loeber, Burke, Lahey, Winters, & Zera, 2000). The co-occurrence of CD with ADHD may be 20% to 50% in children and 44-50% in adolescence with ADHD (Barkley, 1998; Barkley et al., 1990; Biederman et al., 1992; Lahey, McBurnett, & Loeber, 2000; Waschbusch, 2002). By adulthood, up to 26% may continue to have CD while 12-21% will qualify for antisocial personality disorder (ASPD) (Biederman et al., 1992; Barkley, Fischer, Smallish, & Fletcher, 2002; Mannuzza & Klein, 1992; Rasmussen & Gillberg, 2001; Weiss & Hechtman, in press). Similar or only slightly lower degrees of overlap are noted in studies using epidemiologically identified samples rather than those referred to clinics. ADHD, therefore, has a strong association with conduct problems and antisocial disorders, such as ODD, CD, and ASPD, and has been found to be one of the most reliable early predictors of these disorders (Fischer et al., 1993b; Hinshaw & Lee, 2002 this volume; Lahey et al., 2000). Recent longitudinal research suggests that severity of early ADHD is actually a contributing factor to risk for later ODD regardless of severity of early ODD (Burns & Walsh, 2001), perhaps due to the problems with poor emotion (anger) regulation in ADHD noted above. Familial associations among the disorders have also been consistently found, whether across boys and girls with ADHD or across Caucasian and African-American samples (Biederman et al., 1992; Faraone, Biederman, Mick, et al., 2000; Samuel et al., 1999). This suggests some underlying causal connection among these disorders. Evidence from twin studies indicates a shared or common genetic contribution to the three disorders, particularly between ADHD and ODD (Coolidge, Thede, & Young, 2000; Silberg et al., 1996). When CD occurs in conjunction with ADHD, it may represent simply a more severe form of ADHD having a greater family genetic loading for ADHD (Thapar, Harrington, & McGuffin, 2001). Other research, however, also suggests a shared environmental risk factor may also account for the overlap of ODD and CD with ADHD beyond their shared genetics (Burt, Krueger, McGue, & Iacono, 2001), that risk factor likely being family adversity generally and impaired parenting specifically (Patterson, Degarmo, & Knutson, 2000). To summarize, ODD and CD have a substantial likelihood of co-occuring with ADHD with the risk for ODD/CD being mediated in large part by severity of ADHD and its family genetic loading and in part by adversity in the familial environment.
One of the strongest predictors of risk for substance use and abuse disorders (SUDs) among ADHD children upon reaching adolescence and adulthood is prior or co-existing CD or ASPD (Burke, Loeber, & Lahey, 2001; Chilcoat & Breslau, 1999; Molina & Pelham, 1999; White, Xie, Thompson, Loeber, & Stouthamer-Loeber, in press). Given the heightened risk for ODD/CD/ASPD in ADHD children as they mature, one would naturally expect a greater risk for SUDs as well. While an elevated risk for alcohol abuse has not been documented in follow-up studies, the risk for other SUDs among hyperactive children followed to adulthood ranges from 12-24% (Fischer et al., in press Gittelman et al., 1985; Mannuzza et a.., 1993, 1998; Rasmussen & Gillberg, 2001). One longitudinal study of hyperactive children suggested that childhood treatment with stimulant medication may predispose toward greater substance use and abuse disorders (Lambert, in press; Lambert & Hartsough, 1998). Most longitudinal studies, however, find no such elevated risk and, in some cases, a protective effect if stimulant treatment is continued for a year or more or into adolescence (Barkley, Fischer et al., in press-b; Biederman, Wilens, Mick, Spencer, & Faraone, 1999; Chilcoat & Breslau, 1999; Loney, Kramer, & Salisbury, in press). The basis for the conflicting findings in the Lambert study was likely due to not examining or statistically controlling for severity of ADHD and CD at adolescence and young adulthood (Barkley et al., in press-b).
The overlap of anxiety disorders with ADHD has been found to be 10% to 40% in clinic-referred children, averaging to about 25% (see Barkley, 2006; Biederman, Newcorn, & Sprich, 1991; Tannock, 2000, for reviews). In longitudinal studies of ADHD children, however, the risk of anxiety disorders is no greater than in control groups at either adolescence or young adulthood (Fischer, Barkley et al., in press; Mannuzza et al., 1993, 1998; Russo & Biedel, 1994; Weiss & Hechtman, in press). The disparity in findings is puzzling. Perhaps some of the overlap of ADHD with anxiety disorders in children is due to referral bias (Biederman, Faraone, & Lapey, 1992; Tannock, 2000). General population studies of children, however, do suggest an elevated odds ratio of having an anxiety disorder in the presence of ADHD of 3.0 (95%CI = 2.1-4.3), with this relationship being significant even after controlling for comorbid ODD/CD (Angold et al., 1999). This implies that the two disorders may have some association apart from referral bias, at least in childhood. The co-occurrence of anxiety disorders with ADHD has been shown to reduce the degree of impulsiveness relative to those ADHD children without anxiety disorders (Pliszka, 1992). Some research suggests that the disorders are transmitted independently in families and so are not linked to each other in any genetic way (Biederman et al., 1991; Last, Hersen, Kazdin, Orvaschel, & Perrin, 1991). This may not be the case for the inattentive type of ADHD, where higher rates of anxiety disorders have been noted in some studies of these children (see Milich et al., 2001 for a review; Russo & Biedel, 1994), though not always (Barkley, DuPaul, & McMurray, 1990), and in their first- and second-degree relatives (Barkley, DuPaul, & McMurray, 1990; Biederman et al., 1992) again though not always (Lahey & Carlson, 1992; Milich et al., 2001). Regrettably, research on the overlap of anxiety disorders with ADHD has generally chosen to collapse across the types of anxiety disorders in evaluating this issue. Greater clarity and clinical utility from these findings might occur if the types of anxiety disorders present were to be examined separately.
The evidence for the co-occurrence of mood disorders, such as major depression or dysthymia (a milder form of depression), with ADHD is now fairly substantial (see Faraone & Biederman, 1997; Jensen, Martin, & Cantwell, 1997; Spencer, Wilens, Biederman, Wozniak & Crawford, 2000 for reviews). Between 15 and 75% of those with ADHD may have a mood disorder, though most studies place the association between 20 and 30% (Biederman et al., 1992; Cuffe et al, 2001; Fischer et al., in press). The odds of having depression given the presence of ADHD in general population samples is 5.5 (95% CI 3.5-8.4)(Angold et al., 1999). Some evidence also suggests that these disorders may be related to each other in that familial risk for one disorder substantially increases the risk for the other (Biederman et al., 1991, 1992; Faraone & Biederman, 1997), particularly where ADHD is comorbid with CD. Similarly, a recent follow-up study (Fischer et al., in press) found a 26 percent risk of major depression among ADHD children by young adulthood but this risk was largely mediated by the co-occurrence of CD. Likewise, a meta-analysis of general population studies indicated that the link between ADHD and depression was entirely mediated by the linkage of both disorders to CD (Angold et al., 1999). In the absence of CD, ADHD was not more likely to be associated with depression.
The comorbidity of ADHD with bipolar (manic-depressive) disorder is controversial (Carlson, 1990; Geller & Luby, 1997). Some studies of ADHD children indicate that 10-20% may have bipolar disorder (Milberger, Biederman, Faraone, Murphy, & Tsuang, 1995; Spencer et al., 2000; Wozniak et al., 1995) – a figure substantially higher than the 1% risk for the general population (Lewinsohn, Klein, & Seeley, 1995). Follow-up studies of hyperactive children, however, have not documented any significant increase in risk of bipolar disorder in children with ADHD followed into adulthood (Fischer et al., 2002; Mannuzza et al., 1993, 1998; Weiss & Hechtman, in press). However that risk would have to exceed 7 percent or more for these studies to have sufficient power to detect any comorbidity. A 4-year follow-up of ADHD children reported that 12% met criteria for bipolar disorder in adolescence (Biederman, Faraone, Mick, Wozniak et al., 1996). ADHD children without bipolar disorder do not have an increased prevalence of bipolar disorder among their biological relatives (Biederman et al., 1992; Faraone, Biederman, & Monuteaux, 2001; Lahey et al., 1988) while ADHD children with bipolar disorder do (Faraone, Biederman, Wozniak, Mundy, Mennin, & O’Donnell, 1997; Faraone et al., 2001), suggesting that where the overlap occurs it may represent a familially distinct subset of ADHD. Children and adolescents diagnosed with childhood bipolar disorder often have a significantly higher lifetime prevalence of ADHD, particularly in their earlier childhood years (Carlson, 1990; Geller & Luby, 1997; Strober et al., 1988). Where the two disorders co-exist, the onset of bipolar disorder may be earlier than in bipolar disorder alone (Faraone et al., 1997; Faraone et al., 2001; Sachs, Baldassano, Truman, & Guille, 2000). Some of this overlap with ADHD may be partly an artifact of similar symptoms comprising the symptom lists used for both diagnoses (i.e., hyperactivity, distractibility, poor judgment, etc.) (Geller & Luby, 1997). In any case, the overlap of ADHD with bipolar disorder appears to be unidirectional—a diagnosis of ADHD seems not to increase the risk for bipolar disorder, whereas a diagnosis of childhood bipolar disorder seems to dramatically elevate the risk of a prior or concurrent diagnosis of ADHD (Geller & Luby, 1997; Spencer et al., 2000).
Up to 18 percent of children may develop a motor tic in childhood that declines to a base rate of about 2% by mid-adolescence and less than 1% by adulthood (Peterson, Pine, Cohen, & Brook, 2001). Tourette’s Disorder (TD), a more severe disorder involving multiple motor and vocal tics, occurs in less than 0.4% of the population (Peterson et al., 2001). A diagnosis of ADHD does not appear to necessarily elevate these risks for a diagnosis of tics or Tourette’s Disorder, at least not in childhood or adolescence (Peterson et al., 2001). Among clinic-referred adults diagnosed with ADHD, there may be a slightly greater occurrence of tic disorders (12%; Spencer et al., 2001). In contrast, individuals with obsessive-compulsive disorder (OCD) or TD have a marked elevation in risk for ADHD, averaging 48% or more (range 35–71%; Comings, 2000). Complicating matters is the fact that the onset of ADHD often seems to precede that of TD in cases of comorbidity (Comings, 2000). Yet Pauls, Hurst, Kidd, Krueger, and Cohen (1986) have shown that TD and ADHD occur independently among relatives of each disorder, suggesting that a Berkson’s bias (comorbidity with ADHD leads to clinic referral) may be operating in clinical referrals for TD such that comorbid cases are more likely to get referred.
Apart from an increased risk for various psychiatric disorders, children and teens with ADHD (Combined Type) are also more likely to experience a substantial array of developmental and health risks, discussed below. Far less is known about the extent to which these correlated problems are evident in the Inattentive Type, particularly the subgroup having problems with sluggish cognitive tempo described above. The developmental and social problems most likely to occur with the Combined Type are briefly listed in Table 1.2.
|
Table 1.2: Summary of Impairments Likely to be Associated with ADHD |
Cognitive
Language
Adaptive Functioning
Motor Development
Emotion
School Performance
Task Performance
Medical/Health Risks
|
As a group, as many as 60 percent of ADHD, compared to up to 35 percent of normal children, may have poor motor coordination or developmental coordination disorder (Barkley, DuPaul, & McMurray, 1990; Hartsough & Lambert, 1985; Kadesjo & Gillberg, 1999; Szatmari et al., 1989b; Stewart et al., 1966). Neurological examinations for "soft" signs related to motor coordination and motor overflow movements find ADHD children to demonstrate more such signs as well as generally sluggish gross motor movements than control children, including those with purely learning disabilities (Carte, Nigg, & Hinshaw, 1996; Denckla & Rudel, 1978; Denckla, Rudel, Chapman, & Krieger, 1985; McMahon & Greenberg, 1977; Shaywitz & Shaywitz, 1985; Werry et al., 1972). These overflow movements have been interpreted as indicators of delayed development of motor inhibition (Denckla et al., 1985).
Studies using tests of fine motor coordination, such as balance, fine motor gestures, electronic or paper-and-pencil mazes, and pursuit tracking often find children with ADHD to be less coordinated in these actions (Hoy et al., 1978; Mariani & Barkley, 1997; McMahon & Greenberg, 1977; Moffitt, 1990; Shaywitz & Shaywitz, 1985; Ullman et al., 1978). Simple motor speed, as measured by finger-tapping rate or grooved pegboard tests, does not seem to be as affected in ADHD as is the execution of complex, coordinated sequences of motor movements (Barkley, Murphy, & Kwasnik, 1996a; Breen, 1989; Grodzinsky & Diamond, 1992; Mariani & Barkley, 1997; Marcotte & Stern, 1997; Seidman et al., 1995, 1996). The bulk of the available evidence, therefore, supports the existence of deficits in motor control, particularly when motor sequences must be performed, in those with ADHD.
The vast majority of clinic-referred children with ADHD have difficulties with school performance, most often under-productivity of their work. ADHD children frequently fall below normal or control groups of children on standardized achievement tests (Barkley, DuPaul, & McMurray, 1990; Fischer et al., 1990; Hinshaw, 1992, 1994). These differences are likely to be found even in preschool-age children with ADHD (Barkley, Shelton, et al., 1998; Mariani & Barkley, 1997), suggesting that the disorder may take a toll on the acquisition of academic skills and knowledge even before entry into first grade. This makes sense given that some of the executive functions believed to be disrupted by ADHD in the model presented earlier are also likely to be involved in some forms of academic achievement (i.e., working memory in mental arithmetic or spelling; internalized speech in reading comprehension; verbal fluency in oral narratives and written reports, etc.).
Between 19% and 26% of children with ADHD are likely to have any single type of learning disability, conservatively defined as a significant delay in reading, arithmetic, or spelling relative to intelligence and achievement in one of these three areas at or below the 7th percentile (Barkley, 1990). If defined as simply a significant discrepancy between intelligence and achievement, then up to 53% of hyperactive children could be said to have a learning disability (Lambert & Sandoval, 1980). Or, if the criterion of simply two grades below grade level is used, then as many as 80% of ADHD children in late childhood (age 11 years) may have learning disorders (Cantwell & Baker, 1992). Studies suggest that the risk for reading disorders among ADHD children is 16-39%, while that for spelling disorders is 24-27%, and for math disorders, the risk is 13-33% (August & Garfinkel, 1990; Barkley, 1990; Casey et al., 1996; Frick et al., 1991; Semrud-Clikeman et al., 1992).
While the finding that children with ADHD are more likely to have learning disabilities (Gross-Tsur, Shalev, & Amir, 1991; Tannock & Brown, 2000) might imply a possible genetic link between the two disorders, more recent research (Doyle, Faraone, Dure, & Biederman, 2001; Faraone et al., 1993; Gilger, Pennington, & DeFries, 1992) shows that the two sets of disorders are transmitted independently in families. Some subtypes of reading disorders associated with ADHD may share a common genetic etiology (Gilger et al., 1992). This may arise from the finding that early ADHD may predispose children toward certain types of reading problems, whereas early reading problems do not generally give rise to later symptoms of ADHD (Chadwick, Taylor, Taylor, Heptinstall, & Danckaerts, 1999; Rabiner, Coie, & The Conduct Problems Research Group, 2000; Velting & Whitehurst, 1997; Wood & Felton, 1994). The picture is less clear for spelling disorders, where a common or shared genetic etiology to both ADHD and spelling disorder has been shown in a joint analysis of twin samples from London and Colorado (Stevenson, Pennington, Gilger, DeFries, & Gillis, 1993). This may result from the fact that early spelling ability seems to be linked to the integrity of working memory (Mariani & Barkley, 1997; Levy & Hobbes, 1989) that may be impaired in those with ADHD (see theoretical model above). Writing disorders have not received as much attention in research on ADHD, though handwriting deficits are often found among children with ADHD, particularly those having the Combined type of the disorder (Marcotte & Stern, 1997).
Rapport, Scanlan, and Denney (1999) provide some evidence for a dual pathway model of the link between ADHD and academic underachievement. Briefly, ADHD may predispose to academic underachievement through its contribution to a greater risk for ODD/CD and conduct problems in the classroom more generally, the net effect of which is to adversely impact productivity and general school performance. But ADHD is also associated with cognitive deficits not only in attention, but general intelligence (see below) and working memory (see above) all of which may have a direct and adverse impact on academic achievement. Supportive of this view as well are findings that it is more the inattention dimension of ADHD is most closely associated with academic achievement problems than the hyperactive-impulsive dimension (Faraone, Biederman, Weber, 1998; Hynd et al. 1991; Marshall, Hynd, Handwaerck, et al., 1997). According to this dual pathway model, both pathways will require interventions if the marked association of ADHD with school under-achievement is to be addressed.
A higher prevalence of speech and language disorders has also been documented in many studies of ADHD children, typically ranging from 30% to 64% of the samples (Gross-Tsur et al., 1991; Hartsough & Lambert, 1985; Humphries, Koltun, Malone, & Roberts, 1994; Szatmari et al., 1989; Taylor et al., 1991). The converse is also true; children with speech and language disorders have a higher than expected prevalence of ADHD (approximately 30–58%), among other psychiatric disorders (See Tannock & Brown, 2000; Tannock & Schachar, 1996, for reviews on comorbidity with ADHD).
Clinic-referred ADHD children often have lower intelligence than control groups used in these same studies, particularly in verbal intelligence (Barkley, Karlsson, & Pollard, 1985; Mariani & Barkley, 1997; McGee et al., 1992; Moffitt, 1990; Stewart, Pitts, Craig, & Dieruf, 1966; Werry et al., 1987). Differences in IQ have also been found between hyperactive boys and their normal siblings (Halperin & Gittelman, 1982; Tarver-Behring, Barkley, & Karlsson, 1985; Welner, Welner, Stewart, Palkes, & Wish, 1977). The differences found in these studies often range from 7-10 standard score points. Studies using both community samples (Hinshaw, Morrison, Carte, & Cornsweet, 1987; McGee, Williams, & Silva, 1984; Peterson et al., 2001) and behavior-problem samples (Sonuga-Barke, Lamparelli, Stevenson, Thompson, & Henry, 1994) also have found significant negative associations between degree of ADHD and intelligence (rs = -.25-.35). In contrast, associations between ratings of conduct problems and intelligence in children are often much smaller or even nonsignificant, particularly when hyperactive–impulsive behavior is partialed out of the relationship (Hinshaw et al., 1987; Lynam, Moffitt, & Stouthamer-Loeber, 1993; Sonuga-Barke et al., 1994). This implies that the relationship between IQ and ADHD is not likely to be a function of comorbid conduct problems (see Hinshaw, 1992, for a review).
ADHD is classified in DSM-IV as a disruptive behavior disorder because of the significant difficulties it creates in social conduct and general social adjustment. The interpersonal behaviors of those with ADHD, as noted earlier, are often characterized as more impulsive, intrusive, excessive, disorganized, engaging, aggressive, intense, and emotional. And so they are “disruptive” of the smoothness of the ongoing stream of social interactions, reciprocity, and cooperation that is an increasingly important part the children’s daily life with others (Whalen & Henker, 1992).
Research finds that ADHD affects the interactions of children with their parents and, hence, the manner in which parents may respond to these children (Johnston & Mash, 2001). Those with ADHD are more talkative, negative and defiant, less compliant and cooperative, more demanding of assistance from others, and less able to play and work independently of their mothers (Barkley, 1985; Danforth et al., 1991; Gomez & Sanson, 1994; Johnston, 1997; Johnston & Mash, 2001). Their mothers are less responsive to the questions of their children, more negative and directive, and less rewarding of their children’s behavior (Danforth et al., 1991; Johnston & Mash, 2001). Mothers of ADHD children have been shown to give both more commands as well as more rewards to their ADHD sons than daughters (Barkley, 1989b; Befera & Barkley, 1984) but also to be more emotional and acrimonious in their interactions with their sons (Buhrmester, Camparo, Christensen, Gonzalez, & Hinshaw, 1992; Taylor et al., 1991). ADHD children and teens seem to be nearly as problematic for their fathers as their mothers (Buhrmester et al., 1992; Edwards et al., 2001; Johnston, 1997; Tallmadge & Barkley, 1983). Contrary to what may be seen in normal mother-child interactions, the mother–child conflicts in ADHD children and teens may actually increase when fathers join the interaction, especially in hyperactive boys (Buhrmester et al., 1992; Edwards et al., 2001). Such increased maternal negativity and acrimony toward sons in these interactions has been shown to predict greater noncompliance in classroom and play settings and greater covert stealing away from home, even when the level of the child’s own negativity and parental psychopathology are statistically controlled in the analyses (Anderson, Hinshaw, & Simmel, 1994). The negative parent–child interaction patterns also occur in the preschool age group (Cohen, Sullivan, Minde, Novak, & Keens, 1983; DuPaul, McGoey, Eckert, & VanBrakle, 2001) and may be even more negative and stressful (to the parent) in this age range (Mash & Johnston, 1982, 1990) than in later age groups. With increasing age, the degree of conflict in these interactions lessens but remains deviant from normal into later childhood (Barkley, Karlsson, & Pollard, 1985; Mash & Johnston, 1982) and adolescence (Barkley, Anastopoulos, Guevremont, & Fletcher, 1992; Barkley, Fischer, et al., 1991; Edwards et al., 2001). Negative parent–ADHD child interactions in childhood have been observed to be significantly predictive of continuing parent–teen conflicts 8 to 10 years later in adolescence in families with ADHD children (Barkley, Fischer, et al., 1991). Few differences are noted between the interactions of mothers of ADHD children with those children as compared to their interactions with the siblings of the ADHD children (Tarver-Behring et al., 1985).
It is the presence of comorbid ODD that is associated with the highest levels of interaction conflicts between parents and their ADHD children and adolescents (Barkley, Anastopoulos, et al., 1992; Barkley, Fischer, et al., 1992; Edwards et al., 2001; Johnston, 1996). In a sequential analysis of these parentR
ContinuingEdCourses.Net is approved by the American Psychological Association (APA) to sponsor continuing education for psychologists. ContinuingEdCourses.Net maintains responsibility for this program and its content.
ContinuingEdCourses.Net is approved by the Association of Social Work Boards (ASWB) to offer continuing education for social workers, through the Approved Continuing Education (ACE) program. ContinuingEdCourses.Net maintains responsibility for its courses. ASWB provider #1107.
ContinuingEdCourses.Net is approved by the National Board for Certified Counselors (NBCC) as an NBCC-Approved Continuing Education Provider (ACEP) and may offer NBCC-approved clock hours for events that meet NBCC requirements. ContinuingEdCourses.Net solely is responsible for all aspects of the program. NBCC provider #6323.