Книги по МРТ КТ на английском языке / Functional Neuroimaging in Child Psychiatry Ernst 1 ed 2000

274F. B. Wood and D. L. Flowers

compensation. In comparing nonaVected controls with dyslexic men on a rhyme detection task, they found the dyslexic men to have lower activation in the left superior posterior perisylvian region and supramarginal/angular gyrus regions as well as in a posterior frontal region (near Broca's area); however, the two groups did not diVer in those regions during rest or tone detection conditions (Rumsey et al., 1992). The Wnding partly replicates Flowers et al. (1991) in showing posterior superior temporal hypometabolism in dyslexia. In its general implication of abnormal activation of left hemisphere language regions, this Wnding standing alone is tautologic as discussed above. However, several other features of the work preserve the larger Rumsey series from at least the narrowest forms of tautology.

In the Wrst place, in this work, normal versus dyslexic diVerences also extend to right hemisphere regions and to rest conditions. Normal left hemisphere, but lower right temporal and posterior frontal activation in dyslexic subjects during a tonal memory task (Rumsey et al., 1994a) suggest that ineYciency of nonverbal functions may also be involved in developmental dyslexia (Rumsey et al., 1994b), possibly related to rapid serial processing. Resting state conditions showed group diVerences as well, as when Rumsey et al. (1992) reported that during an eyes closed rest condition the dyslexic subjects had lower blood Xow in the right perisylvian area and concomitant higher blood Xow in the medial frontal region. Notably, in a Wnding strongly suggestive of heterogeneity, many but not all of the same subjects who were hypoactive at rest were also hypoactive in right anterior areas during the tonal memory task. While this Wnding by itself might seem to implicate a generalized deWcit notion for dyslexia, already criticized above as unenlightening, it also escapes from that pitfall, as shown by the following Wnding.

Rumsey et al. (1994b) showed that dyslexic men do not diVer from controls in activation of the left temporal and posterior frontal areas during performance of a sentence comprehension task (a Wnding interpretable as consistent with the typical case of intact comprehension in the compensated dyslexic individual who remains phonologically impaired). In fact, on this task the dyslexic subjects showed normal levels of activation of the same left perisylvian region that was abnormally hypoactive during the rhyme detection task, described above. A generalized deWcit explanation is thereby contradicted, nor is it possible to dismiss this Wnding as tautologic, since the task that had elicited the deWcient activation was narrow (rhyme detection) and a cognitively ªhigherº use of language (for sentence comprehension) evoked normal activation in this region. Finally, Rumsey et al. (1997b) showed a range of

abnormalities that included a replication of the earlier Gross-Glenn et al. (1991) Wnding of lingual gyrus excess in dyslexic adults on a pseudoword reading task. This tends to secure a major locus suYciently diVerent from the traditional language activation areas and implies the prospect of true heterogeneity within dyslexics. Though the Rumsey series has never explicitly investigated the heterogeneity issue, the very multiplicity and scope of their Wndings across studies, sometimes replicating earlier work, plausibly suggests the possibility of heterogeneity.

A number of studies, therefore, implicate a considerable variety of brain regions, not only the left posterior perisylvian cortex but also the extrastriate visual pathway including lingual gyri and the frontal lobes, and they imply the possibility of relevant individual diVerences within dyslexia. It is their very imprecision that preserves the current Wndings from tautology. Their demonstration of a variety of brain regional abnormalities, sometimes involving abnormalities of excess as well as of deWcit, in some studies but not others suggests something considerably more complex is at work than ªsimpleº language failure.

Summary and future directions

If the complexity of the data described above suggests that a strong heterogeneity paradigm or at least a converging pathway paradigm would be viable, then they are conversely antithetical to any simple notion of a single underlying information-processing malfunction that characterizes all dyslexia. As this review indicates, and as others have argued (Poeppel, 1996), the prospect of isolating speciWc phonologic or other cognitive mechanisms is beset with numerous complexities, confounds, and conXicting Wndings. While the eVort to isolate cognitive mechanisms is nonetheless defensible and well worth pursuing (Demonet et al., 1996), it is clear that its pursuit ± like any other in science ± is a cost±beneWt question that must take the larger research goals into account. In the case of dyslexia, there is abundant evidence in the above review that brain metabolism or blood Xow in dyslexia can be abnormal in a wide variety of task or control conditions, even at rest. The evidence also strongly suggests that there may be heterogeneity across dyslexic subjects in the functional neuroanatomy of these abnormalities. On the assumption that the tautology paradigm can be excluded and that some degree of neuroanatomic speciWcity can plausibly be sought using the strong heterogeneity paradigm or at least a converging pathway paradigm, we may then suggest the following directions for functional neuroimaging research in dyslexia.

To account for valid genetic or behavioral subtypes

Genetic heterogeneity is at least so probable (if not certain) that it no longer suYces to deWne dyslexic groups simply by their reading disability. Designs that consider all dyslexics as a single group expected to share a single deWcit run the serious risk of the tautology paradigm (by considering only those neurofunctional characteristics that are similar across genetic variance). Furthermore, by ignoring the genetically related variance the true underlying mechanisms will not even be investigated, much less clariWed. Subtyping need not be explicitly genetic (though that is desirable when possible): subtyping can be based upon behavioral phenotypes, so long as they have plausible genetic validity. (See Grigorenko et al. (1997) for operational deWnitions of four such candidate phenotypes.)

To test the anatomic model

The assumption that dyslexia is ªnothing butº a circumscribed cognitive deWcit is at least questionable. If a given brain region, no matter how subtle or diVusely represented, is implicated in dyslexia, then it is naive to think that dyslexia would be the only consequence. Others, including those within the psychiatric domain, should at least be sought.

To avoid exclusive reliance on subtraction methods

Subtraction methodology compares a given cognitive activation task with a control condition in an eVort to isolate a particular cognitive mechanism that is impaired. Usually this means that a given comparison of two tasks is found to generate metabolic activation in a given region in normal controls but not in dyslexic subjects. Beyond the obvious problem of the overriding of subtypes, this procedure inherently risks tautology if the task diVerence is plausibly related to reading or language and if it activates brain regions generally accepted to be substantially involved in language. If an anatomic region is shown to be hypoactive during a languageor reading-related task, but not during a control task, then at the very least certain further clariWcations should be sought. Task accuracy, eVort, anxiety, and related confounds should be explicitly equalized between the two conditions (since the essence of the subtraction method depends on the isolation of a particular cognitive task diVerence). Other subtractions or single task conditions, which on a priori grounds are likely to reXect dysfunction in the relevant anatomic area, should also be attempted in the eVort to test whether the brain region in question is dysfunctional on

Dyslexia 275

tasks other than dyslexia-related ones. (The noninvasive and relatively inexpensive nature of fMRI can be exploited to permit multiple studies using a variety of tasks, thus capitalizing on what is otherwise a limitation, i.e., fMRI's virtual dependence on subtraction methods.) Finally, with some technologies, chieXy PET studies of glucose metabolism, and in selected situations where task activation diVerences have already been demonstrated, the strongest test of chronic dysfunction in a localized region can be attempted: comparisons between two groups during rest. Rest conditions oVer a particularly stringent test, since the only demand that is made on the brain is that it should comply with the rest instruction. Failure to demonstrate group diVerences at rest is, therefore, not particularly compelling; however, successful demonstration of diVerences at rest oVers particularly strong evidence of a group diVerence that is anatomically based.

To consider longitudinal studies in children

Longitudinal studies are possible with fMRI and may allow data to be acquired early enough in childhood to capture the genetic heterogeneity at a time when it is less subtle and more readily demonstrable. Variance in readingrelated mechanisms often occurs surprisingly early in the school career, changing rapidly thereafter. (See Meyer et al. (1998) for an illustration involving rapid automatized naming ability, an important underlying ability in reading.)

Summary

Present data acquired by functional neuroimaging in dyslexia suggests a diverse range of neurofunctional deWcits, not well described by current models stressing an isolated cognitive deWcit in dyslexia. Since it is likely that part of the variance across this range is genetically based, the next generation of studies in dyslexia will need to take this variance into account in experimental design and also will need to test the convergence of varied functional activation deWcits on given brain regions or processes. As always, the expectation of a single experimentum crucis is unrealistic: only a systematic, programmed eVort in these directions is likely to be productive.

iReferencesi

Bench, C. J., Frith, C. D., Grasby, P. M. et al. (1993). Investigations of the functional anatomy of attention using the Stroop test.

Neuropsychologia, 31, 907±22.

Bookheimer, S. Y., ZeYro, T. A., Blaxton, T., Gaillard, W. and

Theodore, W. (1995). Regional cerebral blood Xow during object naming and word reading. Hum. Brain Map., 3, 93±106.

Cardon, L. R., Smith, S. D., Fulker, D. W., Kimberling, W. J., Pennington, B. F. and DeFries, J. C. (1994). Quantitative trait locus for reading disability on chromosome 6. Science, 26, 276±9.

Corbetta, M., Miezin, F. M., Dobmeyer, S., Shulman, G. L. and Petersen, S. E. (1991). Selective and divided attention during visual discriminations of shape, color, and speed: functional anatomy by positron emission tomography. J. Neurosci., 11, 2382±402.

DeFries, J. C. and Decker, S. N. (1982). Genetic aspects of reading disability: a family study. In Reading Disorders: Varieties and Treatments, ed. R. N. Malatesha and P. G. Aaron, pp. 255±79. New York: Academic Press.

DeFries, J. C., Fulker, D. W. and LaBuda, M. C. (1987). Evidence for a genetic aetiology in reading disability in twins. Nature, 329, 537±9.

DeFries, J. C., Olson, R. K., Pennington, B. F. and Smith, S. D. (1991). Colorado Reading Project: an update. In The Reading Brain: The Biological Basis of Dyslexia, ed. D. D. Duane and D. B. Gray, pp. 53±88. Parkton, MD: York Press.

Demonet, J. F., Chollett, F., Ramsay, S. et al. (1992). The anatomy of phonological and semantic processing in normal subjects.

Brain, 115, 1753±68.

Demonet, J. F., Wise, R. and Frackowiak, R. S. J. (1993). Language functions explored in normal subjects by positron emission tomography: a critical review. Hum. Brain Map., 1, 39±47.

Demonet, J. F., Price, C., Wise, R. and Frackowiak, R. S. J. (1994a). A PET study of cognitive strategies in normal subjects during language tasks. InXuence of phonetic ambiguity and sequence processing on phoneme monitoring. Brain, 117, 671±82.

Demonet, J. F., Price, C., Wise, R. and Frackowiak, R. S. J. (1994b). DiVerential activation of right and left posterior sylvian regions by semantic and phonological tasks: a positron-emission tomography study in normal human subjects. Neurosci. Lett.,

182, 25±8.

Demonet, J. F., Fiez, J., Paulesu, E., Petersen, S. and Zatorre, R. (1996). PET studies of phonological processing: a critical reply to Poeppel. Brain Lang., 55, 352±79.

Eden, G. F., van Meter, J. W., Rumsey, J. M., Maisog, J. Ma., Woods, R. P. and ZeYro, T. A. (1996). Abnormal processing of visual motion in dyslexia revealed by functional brain imaging. Nature,

382, 66±9.

Flowers, D. L., Wood, F. B. and Naylor, C. E. (1991). Regional cerebral blood Xow correlates of language processes in reading disability. Arch. Neurol., 48, 637±43.

Frith, C. D., Kapur, N., Friston, K. J., Liddle, P. F. and Frackowiak, R. S. J. (1995). Regional cerebral activity associated with the incidental processing of pseudo-words. Hum. Brain Map., 3, 153±60.

Fulker, D. W., Cardon, L. R., DeFries, J. C., Kimberling, W. J., Pennington, B. F. and Smith, S. D. (1991). Multiple regression analysis of sib-pair data on reading to detect quantitative trait loci. Reading Writing Interdisc. J., 3, 299±313.

Galaburda, A. M., Sherman, G. F., Rosen, G. D., Aboitiz, F. and Geschwind, N. (1985). Developmental dyslexia: four consecutive patients with cortical anomalies. Ann. Neurol., 18, 222±33.

Grabowski, T. J., Damasio, H., Frank, R. J. et al. (1995). Neuroanatomical analysis of functional brain images: validation with retinotopic mapping. Hum. Brain Map., 2, 134±48.

Grigorenko, E. L., Wood, F. B., Meyer, M. S. et al. (1997). Susceptibility loci for distinct components of developmental dyslexia on chromosomes 6 and 15. Am. J. Hum. Genet., 60, 27±39.

Gross-Glenn, K., Duara, R., Barker, W. W. et al. (1991). Positron emission tomographic studies during serial word-reading by normal and dyslexic adults. J. Clin. Exp. Neuropsychol., 13, 531±44.

Hagman, J. O., Wood, F., Buchsbaum, M. S., Tallal, P., Flowers, L. and Katz, W. (1992). Cerebral brain metabolism in adult dyslexic subjects assessed with positron emission tomography during performance of an auditory task. Arch. Neurol., 49, 734±9.

Hallgren, B. (1950). SpeciWc dyslexia (ªcongenital word blindnessº): a clinical and genetic study. Act. Psychiat. Neurolog. Scand., 65, 1±287.

Herholz, K., Pietrzyk, U., Karbe, H., Wurker, M., Wienhard, K. and Heiss, W. D. (1994). Individual metabolic anatomy of repeating words demonstrated by MRI-guided positron emission tomography. Neurosci. Lett., 182, 47±50.

Hohnen, B. and Stevenson, J. (1995). Genetic eVects in orthographic ability: a second look. Behav. Genet., 25, 271.

Howard, D., Patterson, K., Wise, R. et al. (1992). The cortical localization of the lexicons. Brain, 115, 1769±82.

Ingvar, M., Greitz, T., Eriksson, L., Stone-Elander, S., Trampe, P. and Euler, C. (1995). Developmental dyslexia studied with PET. Hum. Brain Map., Suppl. 1, 378.

Karbe, H., Wurker, M., Herholz, K., Ghaemi, M., Pietrzyk, U., Kessler, J. and Heiss, W. D. (1995). Planum temporale and Brodmann's area 22. Arch. Neurol., 52, 869±74.

Kinsbourne, M., Rufo, D. T., Gamzu, E., Palmer, R. L. and Berliner, A. K. (1991). Neuropsychological deWcits in adults with dyslexia.

Devel. Med. Child Neurol., 33, 763±75.

Kiyosawa, M., Inoue, C., Kawasaki, T. et al. (1996). Functional neuroanatomy of visual object naming: a PET study. Graef. Arch. Clin. Emp. Ophthalmol., 234, 110±15.

Lubs, H. A., Smith, S., Kimberling,W., Pennington, B., Gross-Glenn, K. and Duara, R. (1988). Dyslexia subtypes: genetics, behavior, and brain imaging. In Language, Communication, and the Brain, ed. F. Plum, pp. 139±47. New York: Raven Press.

Menard, M. T., Kosslyn, S. M., Thompson, W. L., Alpert, N. M. and Rauch, S. L. (1996). Encoding words and pictures: a positron emission tomography study. Neuropsychologia, 34, 185±94.

Meyer, M. S., Wood, F. B., Hart, L. and Felton, R. H. (1998). Selective predictive value of rapid automatized naming within poor readers. J. Learn. Disabil., 31, 106±17.

Olson, R., Wise, B., Conners, F., Rack, J. and Fulker, D. (1989). SpeciWc deWcits in component reading and language skills: genetic and environmental inXuences. J. Learn. Disabil., 22, 339±48.

Paulesu, E., Frith, C. D. and Frackowiak, R. S. J. (1993). The neural correlates of the verbal component of working memory. Nature,

362, 342±4.

Paulesu, E., Frith, U., Snowling, M. et al. (1996). Is developmental dyslexia a disconnection syndrome? Evidence from PET scanning. Brain, 119, 143±57.

Pennington, B. F. (1997). Using genetics to dissect cognition. [Editorial comment] Am. J. Hum. Genet., 60, 13±16.

Petersen, S. E. and Fiez, J. A. (1993). The processing of single words studied with positron emission tomography. Annu. Rev. Neurosci., 16, 509±30.

Petersen, S. E., Fox, P., Posner, M., Mintun, M. and Raichle, M. (1989). Positron emission tomographic studies of the processing of single words. J. Cogn. Neurosci., 1, 153±70.

Poeppel, D. (1996). A critical review of PET studies of phonological processing. Brain Lang., 55, 317±51.

Price, C. J. (1977). Functional anatomy of reading. In Human Brain Function, ed. R. S. J. Frackowiak, K. J. Friston, C. D. Frith, R. J. Dolan and J. C. Mazziotta, pp. 301±27. London: Academic Press.

Price, C. J., Wise, R. J. S., Watson, J. D. G., Patterson, K., Howard, D. and Frackowiak, R. S. J. (1994). Brain activity during reading. The eVects of exposure duration and task. Brain, 117, 1255±69.

Robinson, R. G. and Price, T. R. (1982). Post-stroke depressive disorders: a follow-up study of 103 outpatients. Stroke, 13, 635±41.

Rumsey, J. M., Andreason, P., Zametkin, A. J. et al. (1992). Failure to activate the left temporoparietal cortex in dyslexia. An oxygen 15 positron emission tomographic study. Arch. Neurol., 49, 527±34.

Rumsey, J. M., Andreason, P., Zametkin, A. J. et al. (1994a). Right frontotemporal activation by tonal memory in dyslexia, an O15

PET study. Biol. Psychiatry, 36, 171±80.

Rumsey, J. M., Zametkin, A. J., Andreason, P. et al. (1994b). Normal activation of frontotemporal language cortex in dyslexia, as measured with oxygen 15 positron emission tomography. Arch. Neurol., 51, 27±38.

Rumsey, J. M., Horwitz, B., Donohue, B. C., Nace, A. K., Maisog, J. M. and Andreason, P. (1997a). Phonological and orthographic components of word recognition: a PET ± rCBF study. Brain, 120, 739±59.

Rumsey, J. M., Nace, K., Donohue, B., Wise, D., Maisog, J. M. and Andreason, P. (1997b). A positron emission tomographic study of impaired word recognition and phonological processing in dyslexic men. Arch. Neurol., 54, 562±73.

Sergent, J., Ohta, S. and MacDonald, B. (1992). Functional neuroanatomy of face and object processing. Brain, 115, 16±36.

Shaywitz, S. E. et al. (1998). Functional disruption in the organization of the brain for reading in dyslexia. Proc. Natl. Acad. Sci USA,

95, 2636±41.

Shaywitz, B. A., Pugh, K. R., Constable, R. T. et al. (1995). Localization of semantic processing using functional magnetic resonance imaging. Hum. Brain Map., 2, 149±58.

Small, S. L., Noll, D. C., Perfetti, C. A., Hlustik, P., Wellington, R. and Schneider, W. (1996). Localizing the lexicon for reading aloud: replication of a PET study using fMRI. Neuroreport, 7, 961±5.

Smith, S. D., Kimberling, W. J., Pennington, B. F. and Lubs, H. A. (1983). SpeciWc reading disability: identiWcation of an inherited form through linkage analysis. Science, 219, 1345±7.

Tamas, L. B., Shibasaki, T., Horikoshi, S. and Ohye, C. (1993). General activation of cerebral metabolism with speech: a PET study. Int. J. Psychophysiol., 14, 199±208.

Vandenberghe, R., Price, C., Wise, R., Josephs, O. and Frackowiak, R. S. J. (1996). Functional anatomy of a common semantic system for words and pictures. Nature, 383, 254±5.

Warburton, E., Wise, R. J. S., Price, C. J. et al. (1996). Noun and verb retrieval by normal subjects. Studies with PET. Brain, 119, 159±79.

Watson, J. D. G., Myers, R., Frackowiak, R. S. J. and Haynal, J. V. (1993). Area V5 of the human brain: evidence from a combined study using positron emission tomography and magnetic resonance imaging. Cerebr. Cortex, 3, 79±94.

Wolff, M. and Obregon, M. (1992). Early naming deWcits, developmental dyslexia and a speciWc deWcit hypothesis. Brain Lang.,

42, 217±47.

Wolf, P. H. and Melngailis, I. (1994). Family patterns of developmental dyslexia. Am. J. Med. Genet. (Neuropsychiatr. Genet.), 54, 122±31.

Wood, F. (1990). Functional neuroimaging in neurobehavioral research. In Neuromethods, Vol. 17: Neuropsychology, ed. A. A. Boulton, G. B. Baker and M. Hiscock, pp. 107±25. Clifton NJ: Humana Press.

Wood, F., Flowers, L., Buchsbaum, M. and Tallal, P. (1991). Investigation of abnormal left temporal functioning in dyslexia through rCBF, auditory evoked potentials, and positron emission tomography. Reading Writing Interdisc. J., 3, 379±93.

Zatorre, R. J., Meyer, E., Gjedde, A. and Evans, A. C. (1996). PET studies of phonetic processing of speech: review, replication, and reanalysis. Cereb. Cortex, 6, 21±30.

16

Attention-de®cit hyperactivity disorder: neuroimaging and behavioral/cognitive probes

Julie B. Schweitzer, Carl Anderson and

Monique Ernst

Clinical phenomenology and epidemiology

Attention-deWcit hyperactivity disorder (ADHD) is characterized by persistent inattention and/or situationally excessive motor activity, and impulsive behavior (Barkley, 1990). According to the current Diagnostic and Statistical Manual of Mental Disorder (4th edition (DSM-IV); American Psychiatric Association, 1994), individuals must exhibit symptoms for at least 6 months and must express the symptoms by 7 years of age. The symptoms must be developmentally inappropriate and exhibited in at least two settings. DSM-IV speciWes three subtypes: predominantly inattentive (ADHD-I), predominantly hyperactive±impulsive (ADHD-HI), and combined type (ADHD-C). The number and nature of items endorsed within lists of inattentive and hyperactive/impulsive symptoms determines the speciWc diagnostic subtype.

ADHD is the most prevalent childhood psychiatric disorder and is estimated to aVect 3±11% of the school-age population, depending on the source of the sample (American Psychiatric Association, 1994; Wolraich et al., 1996). There is a much higher incidence rate in boys, who are 2.5±9 times more likely than girls to be diagnosed with ADHD (Szatmari et al., 1989; Barkley, 1990; Wolraich et al., 1996). The disorder often has a chronic course, with 30±50% of aVected children exhibiting ADHD symptoms into adulthood (Barkley et al., 1990; Weiss and Hechtman, 1993). Numerous problems are associated with childhood and adulthood ADHD, including poor academic performance, learning disabilities, conduct disorders, antisocial personality disorder, lower occupational success, poor social relationships, frequent car accidents, and a higher incidence of anxiety and depression (Barkley et al., 1990, 1996; Biederman et al., 1993; Weiss and Hechtman, 1993; Murphy and Barkley, 1996; Mannuzza et al., 1998).

Theoretical perspectives

Diagnostic labels

The diagnostic label for ADHD has changed several times during the twentieth century, reXecting the continuously evolving conceptualization of the disorder. The metamorphosis in the labels, from ªminimal brain dysfunctionº, to ªhyperkinetic impulse disorderº, to ªattention-deWcit disorderº, mirrors the changing emphasis on brain, motor, and attentional dysfunction. Recent theoretical models of ADHD, however, are less likely to focus on ªattention deWcitsº and more likely to highlight neurobehavioral dysfunction aVecting response inhibition, working memory, and the implementation of organizational strategies subsumed under the term of ªexecutive functionsº (Denckla, 1996; Barkley, 1997a,b). In these models, problems with sustained attention, distractibility, and reduced motor control are secondary to disrupted response inhibition and interference control. The deWcient response inhibition ultimately results in behavior that is less internally guided (i.e., less able to follow verbal rules and less goal-directed). (See Barkley (1997a) and Denckla (1996) for a more indepth discussion of these models.) These ADHD models implicate the dysfunction of discrete neural circuits the functional integrity of which can be examined in neuroimaging studies.

To examine these circuits in ADHD, it is critical to study clinically homogeneous samples, such as the ADHD-I, ADHD-HI, or ADHD-C subtypes speciWed in DSM-IV. Subjects who are free of comorbid disorders can be studied as relatively homogeneous behavioral subgroups with objectively quantiWed motor and/or attentional dysfunction. For example, as hyperactivity is primarily thought to stem from an inability to inhibit motor acts, the use of imaging paradigms that require subjects to inhibit

responses (e.g., performance of go-no-go tasks) may be exploited to reveal functional neuroanatomic diVerences between healthy controls and subtypes of ADHD.

Functional neuroanatomy

Recently, studies indicating a role for the basal ganglia in a variety of neuropsychiatric conditions involving motor and attentional dysfunctions (Mega and Cummings, 1994; Rauch and Savage, 1997; Peterson et al., 1998) have suggested the diVerential involvement in ADHD of Wve functionally interconnected subcortical structures: the caudate nucleus, putamen, globus pallidus, subthalamic nucleus, and ventral mesencephalon. In addition, Wve parallel basal ganglia±thalamocortical circuits have been proposed (Alexander et al., 1986, 1990) to convey the output of these subregions through speciWc thalamic zones to diVerent parts of the frontal cortex. These functionally segregated striato±thalamo±cortico±striatal loops contribute respectively to motor, somatosensory, oculomotor, executive, emotion, and motivation functions: (i) a motor circuit involving the arcuate premotor area, supplementary motor area, somatosensory cortex, superior parietal lobe, and motor cortex, all of which project to the posterior putamen, subserving sensorimotor integration; (ii) an oculomotor pathway involving the frontal eye Welds, as well as the supplementary eye Welds, both of which project to the head and body of the caudate and subserve the attentional control of eye movements; (iii) the prefrontal cognitive/association pathway, composed of the dorsolateral prefrontal cortex (DLPFC), the dorsolateral head of the caudate, and the anterior putamen, regions involved in working memory and the planning of voluntary movements; (iv) the orbitofrontal pathway, which is composed of the lateral orbitofrontal cortex and its projections to both the head and body of the caudate, and which is less involved in working memory and spatial processing than the prefrontal pathway and more involved with response inhibition and emotion; and (5) the ventral striatal pathway, composed of the anterior cingulate, medial orbitofrontal cortex, ventromedial caudate, and the nucleus accumbens, which subserves motivational processes and inXuences the other pathways via extensive projections back to the ventral mesencephalic dopamine system; this, in turn, modulates all of the dorsal striatum (Heimer et al., 1997). The motor and prefrontal basal ganglia±thalamocortical circuits are depicted in Fig. 16.1a (p. 242) and the regional cortico±striatal pathways are illustrated in Fig. 16.1b (p. 242).

Since the late 1980s, a large number of functional imaging studies (see Tables 16.1±16.3) have implicated

various components of these pathways in the pathophysiology of ADHD. Functional imaging studies of ADHD tend to follow a dimensional rather than a categorical approach and focus on a combination of behavioraland circuit-speciWc paradigms to elucidate neural abnormalities speciWc to behavioral subtypes.

How do these functionally segregated circuits exchange information to enable the smooth Xow of motor and attentional functions? One view is that regions of the striatum such as the putamen contain a combinatorial map of body regions allowing extensive overlap and interdigitation between body-part elements (Brown et al., 1998). The functional connectivity of this cortico-striatal gridwork appears to be regulated by dopamine levels. The putamen projects somatotopically to both the external and internal segments of the globus pallidus and to the substantia nigra pars reticulata (SNR). In general, striatal projections to the external segment are inhibitory on movement and considered part of the ªindirect pathway,º whereas projections to the internal segment facilitate movement and are part of the ªdirect pathwayº to the SNR (Smith et al., 1998). The SNR, in turn, regulates the dopaminergic inputs from the substantia nigra pars compacta back to the putamen and other striatal regions. In addition, both the SNR and the internal segment of the globus pallidus project directly to the thalamus, which somatotopically innervates the frontal motor cortical areas (supplementary motor area and motor cortex). Normal dopaminergic tone in the putamen facilitates independent Wring patterns of globus pallidus neurons (Bergman et al., 1998).

The hallmark of the loss of dopaminergic tone in the basal ganglia is Parkinson's disease, which is characterized by diYculty initiating and executing movements and by a breakdown of independent Wring within the globus pallidus. In Parkinson's disease, contiguous neurons in the internal segment of the globus pallidus Wre in highly synchronized oscillations in phase with behavioral tremor (Bergman et al., 1998), and inhibitory output from the external segment predominates. Therefore, dopamine appears to modulate the cross-linkage between diVerent pathways in the gridwork of cortico±striatal projections regulating the balance between indirect and direct striato-pallidal outXow to thalamic and frontal cortical networks.

To summarize, the striatum (caudate nucleus, putamen, and ventral striatum or nucleus accumbens) is divided into functional domains based on cortical projections (Heimer et al., 1997). Prefrontal cortical and associative areas including primary and secondary visual and auditory cortical regions project to the dorsolateral head of the caudate and the anterior putamen. The sensorimotor cortex projects to

large regions of the putamen (Graybiel et al., 1994). The ventral striatum or nucleus accumbens receives extensive projections from the anterior cingulate and medial orbitofrontal cortex. How do the Wndings of recent functional imaging studies of ADHD Wt with this current model of basal ganglia function?

Neuroimaging studies linked to functional neuroanatomy

As a general caveat to the interpretation of the following studies, the reader must keep in mind that most pediatric

studies are conducted in children who have a history (past or current) of exposure to stimulant treatment, whereas adult studies usually involve individuals who did not have the opportunity to be treated for ADHD as children and have no prior history of exposure to stimulant treatment. Generally, stimulant treatment is suspended between 48h to 2 weeks prior to an imaging study, depending on the study protocol and the requirements of Institutional Review Boards. The contribution of past or current exposure of stimulants to deviance in brain function is not clear. Information about stimulant treatment is not always available and not systematically indicated in this review of the literature.

The same studies may be mentioned several times throughout the text, according to the structure being discussed, but the description of their methodology will not be repeated.

Prefrontal cortex

The starting point of functional basal ganglia loops are motor commands initiated by frontal regions. The frontal lobe is divided into posterior motor regions and anterior prefrontal regions, particularly the DLPFC. Evidence for dysfunction in the DLPFC circuit comes from the growing number of studies showing abnormalities in the prefrontal cortex of subjects with ADHD. Cognitive and higher level processes are thought to be dependent on the functional connections between frontal regions and basal ganglia, particularly the caudate nucleus (Alexander et al., 1990). The frontal lobes receive input relayed by the thalamus from the subcortical nuclei of the basal ganglia, including the caudate and globus pallidus. The largest morphometric MRI study (subjects with ADHD: n5 57; mean age 11.7 years, range 5.8±17.8; controls: n5 55; mean age 12.0 years, range 5.5±17.8) (Castellanos et al., 1996) reported that boys with ADHD had a signiWcantly smaller right, but not left, prefrontal volumes than control boys.

Following work by Lou et al. (1984, 1989) reporting regional cerebral blood Xow (rCBF) abnormalities in the striatum of children with ADHD (small and heterogeneous samples including children with history of neonatal neuro-

logic insults), Zametkin et al. (1990) demonstrated functional abnormalities in the brain of adults with ADHD. This study used positron emission tomography (PET) and [18F]- Xurodeoxyglucose (FDG), and was remarkable for its thorough assessment of the subjects with ADHD, and its clean (no comorbidity) and relatively large sample sizes (25 adults with ADHD, 50 controls). Zametkin and colleagues (1990) reported lower absolute metabolic rate of glucose (CMRGlu) in many brain regions, including the prefrontal cortex, in adults with ADHD compared with controls. PET measures were collected while subjects performed a continuous performance test (CPT) that was used to standardize the mental and motoric state of the subjects.

Given this strong Wnding, Zametkin initiated studies in adolescents after reWning the PET methodology to minimize radiation exposure (Chapter 6) (Zametkin et al., 1993; Ernst et al., 1994a). In contrast to the extensive reduction of CMRGlu in adults with ADHD compared with controls, adolescents (ADHD: n5 20; mean age 14.7"1.7 years; healthy controls: n5 19; mean age 14.4"1.4 years) showed regionally limited diVerences between boys with ADHD (15) and control boys (13). In contrast, girls with ADHD (5) showed a statistically signiWcant 15% reduction of global CMRGlu compared with that in control girls (6) (Ernst et al., 1994a). When regional CMRGlu (rCMRGlu) was normalized (divided by whole-brain CMRGlu), the left frontal cortex was the region most aVected in adults (Zametkin et al., 1990) and adolescents with ADHD (left anterior middle prefrontal gyrus) (Ernst et al., 1994a).

Because of the encouraging results in the girl sample, and the diYculty in generalizing them owing to small sample size, a larger independent sample of girls with (10) and without (11) ADHD was studied (Ernst et al., 1997a). This study did not replicate the abnormally low CMRGlu in girls with ADHD. However, although of similar age, the girls of the initial study (Ernst et al., 1994a) were more sexually mature than those of the second study (Ernst et al., 1997a). Because the data showed a negative association between sexual maturation and CMRGlu (more mature, lower

CMRGlu), the absence of diVerences between ADHD and control girls may have been masked in sexually immature girls. The role of sexual maturation on CMRGlu, as well as potential CMRGlu deviance in girls with ADHD compared with controls, remains to be determined. The diYculty in conducting such studies makes it diYcult to examine these questions in a timely fashion.

The Wndings of decreased CMRGlu in adults with ADHD were partially replicated in a more recent study using FDG PET (Flowers et al., 1997). This study also used a CPT per-

formance to standardize the mental and motoric state of the subjects (11 subjects with attention-deWcit disorder (ADD) with and without hyperactivity based on DSM-III; 10 nonADD with reading disability; no data on gender available). Similar to Wndings reported by Zametkin et al. (1990), Flowers et al. (1997) reported lower absolute rCMRGlu in the frontal lobes in adults with ADD compared with a control group of adults without ADD but with a reading disability. In contrast to the report of abnormally low frontal rCMRGlu (Zametkin et al., 1990), Flowers et al. (1997) found higher normalized rCMRGlu in the left inferior frontal cortex relative to controls with dyslexia. The diVerences between Zametkin's study and Flower's study in the control samples (healthy versus reading disabled) and in the subtypes of ADHD included in the patient sample may have accounted for the discrepant Wndings.

Further evidence for the presence of functional diVerences between ADHD and control subjects in frontal lobe comes from activation studies of PET and H215O using a cognitive challenge in the form of an auditory working memory task in adult males. In two independent studies (Schweitzer et al., 1998, 2000) control subjects (nine in the 1998 study, six in that of 2000), showed activation of the right dorsolateral prefrontal cortex (Brodmann areas 46 and 9), which has been reported to be active during retrieval in memory (Tulving et al., 1994; Andreasen et al., 1995). In contrast, adults with ADHD-C type and no comorbidity (12 in the 1988 study, and six in that of 1999) showed no activation of these regions, suggesting a disruption in the use of retrieval function. While the ADHD groups evidenced diminished activation in frontal regions, they displayed increased activation in other brain regions (e.g., occipital and cerebellar), suggesting that these nonfrontal regions may be activated to compensate for the underfunctioning frontal regions. Such hypothesis could be tested using connectivity analysis with larger samples.

Studies using PET and FDG have detected only limited eVects of stimulants on the frontal lobes in ADHD. Stimulant administration (acute oral, chronic oral, and intravenous) did not raise prefrontal rCMRGlu as was expected given the abnormally low CMRGlu of untreated adults with ADHD (Matochik et al., 1993, 1994; Ernst et al., 1994b, 1997b). Such negative Wnding suggests that CMRGlu may not be the appropriate variable to examine.

In contrast, blood Xow studies using fMRI have been more sensitive to the eVects of stimulant medications. Using steady-state fMRI relaxometry, Teicher et al. (1996) found that low oral doses (0.5mg/kg) methylphenidate in children with ADHD (13 males, two females; mean age 9.93"0.45 years; all subtypes) increased perfusion in the left dorsolateral prefrontal cortex and the right caudate and putamen

during a resting state condition. Using fMRI with a response inhibition task (go-no-go task), Vaidya et al. (1998) studied the acute eVects of methylphenidate on frontal lobe cerebral blood Xow (CBF) in 10 boys with ADHD (aged 10.5"1.4 years; eight combined type, two inattentive type) and six control boys (aged 9.3"1.5 years). The authors state, however, that ratings of three (not including the two diagnosed as ADHDI) of the ten did not reach signiWcance on the hyperactivity index on an ADHD rating scale. In patients, methylphenidate was administered at the regular prescribed dose (7.5±30mg) and was discontinued for 36h prior to scanning for the oV-drug session; controls were scanned oV-drug and 2.0±2.5h after the administration of a 10mg dose. Methylphenidate improved task performance and increased perfusion in frontal regions in children with ADHD, as well as in normal controls (Vaidya et al., 1998). However, the eVects of methylphenidate diVered in the striatum as a function of group (increase in ADHD, decrease in controls). The authors concluded that the stimulant modulates frontal activation in a similar way in ADHD and controls and suggested that dysfunction in ADHD more likely involves the striatum and striato±frontal connections.

Caudate nuclei

The strongest evidence of disruption of the DLPFC loop in ADHD comes from a growing body of structural MRI Wndings. Decreased volume and altered asymmetries of the caudate nucleus in children with ADHD have been reported in several studies (Hynd et al., 1993; Castellanos et al., 1996; Filipek et al., 1997). However, less consistent are the directions of lateralization observed in the caudate nucleus of children with and without ADHD. In the Wrst study to examine the caudate volumes in ADHD, Hynd and colleagues (1993) studied 11 children with ADHD (eight males and three females; ages not reported) and 11 normal controls (six males and Wve females; ages not reported) and found a decrease in the size of the head of the left caudate nucleus. The control children evidenced a left-larger-than- right pattern of asymmetry, whereas the children with ADHD evidenced a right-greater-than-left asymmetry. Filipek and colleagues (1997) compared the volume of total caudates and caudate heads in 15 boys with ADHD (aged 12.4"3.4 years) without comorbid diagnosis with 15 healthy control subjects (aged 14.4"3.4 years). The subjects with ADHD had smaller left total caudate and caudate head volumes, with a loss of the normal left-predominant asymmetry seen in controls. An exploratory analysis suggested that the stimulant responders in the ADHD group evidenced the smallest bilateral caudate volumes. In the currently largest structural MRI study of ADHD,