- Email: [email protected]
Regional cerebral blood flow (rCBF) in normal readers: Bilateral activation with narrative text
Archrves of Clmrcal Neurops.vchology, Vol 4. pp 71-78, 1989 Pruned I” the USA All nghtr reserved
0887-6177189 $3 00 + 00 CopyrIght L 1989 Natmnal Academy of Neuropsychologlsts
Brief Report
Regional Cerebral in Normal Activation Marion
I. S. Huettner,
Blood Flow (rCBF)
Readers: Bilateral with
Narrative Text
Becky L. Rosenthal, and George W. Hynd Unlverslty
0T Ceorgla
Based on recent evidence that suggests a more active participatron of the right cerebral hemisphere in reading, this exploratory study examined changes rn regional cerebral blood flow (rCBF) in three normal subjects during the readmg of controlled narrative text. Narrative text presumably activates all neurocognitive processes rmportant in reading including the semantic, pragmatrc, emottonal, and imagery components. Because of the small number of subjects, percentage change over a baseline at rest condition in rCBF durrng readmg was compared to test/retest variability in control subjects. Effect size was also considered. The results indicated stattstically significant bilateral central posterior activation during reading. These findings are consistent wrth an evolving model of brlateral language representation m which subcortical structures and rtght hemrsphenc systems are functionally and anatomtcally tied to the dominant left hemtspheric language ten ters.
Contemporary conceptualizations of reading disorders are largely based on a neurolinguistic model advocated by Geschwind (1974, 1984), and others (e.g., Campbell & Whitaker, 1986). Generally, this model, which originated in ideas advanced by Wernicke and Dejerine nearly a century ago (Hynd & Hynd, 1984; Mayeux & Kandel, 1985), documents the significant contribution of the classical perisylvian language areas, the posterior visual, and left tertiary cortices in fluent reading, and is generally consistent with knowledge about aphasia and alexia with and without agraphia. Evidence regarding the validity of this conceptualization includes the case studies of patients with acquired surface dyslexia (Coltheart, Masterson,
Requests for reprints tal Neuropsychology,
should be sent to George W. Hynd, Center for Clinical and Developmen325 Aderhold Hall, University of Georgia, Athens, GA 30602. 71
72
M. I. S. Huettner, B. L. Rosenthal, and G. W Hynd
Byng, Prior, & Riddoch, 1983; Temple, 1984) and acquired deep dyslexia (Friedman & Perlman, 1982; Healey, 1985), the topographic mapping studies of brain electrical activity in normals and developmental dyslexics during associated perceptual and linguistic tasks (Duffy, Denckla, Bartels, & Sandini, 1980; Duffy, Denckla, Bartels, Sandini, & Kiessling, 1980), and regional cerebral blood flow (rCBF) studies involving word learning and semantic recall tasks (Maxmilian, Prohovnik, Risberg, & Hakansson, 1978; Wood, Taylor, Penny, & Stump, 1980). Collectively, these studies suggest that in cases of documented or presumed pathology, deficiencies exist on behavioral and metabolic correlates of the classical perceptual and language areas proposed by the Geschwind-Wernicke model (Mayeux & Kandel, 1985). Contrary to the Geschwind-Wernicke model, however, and perhaps more consistent with that proposed by Bastian (1898), recent neurometabolic studies employing both rCMRglc and rCBF methodologies reveal significant and equal bihemispheric central-posterior activation during reading of words and text in normal and dyslexic subjects (Gross-Glenn et al., 1987; Hynd, Hynd, Sullivan, & Kingsbury, 1987). A principle methodological difference between these studies and those documenting left perisylvian activity is their employment of word and text reading versus semantic recall tasks (e.g., Wood et al., 1980). Given that these conflicting findings may stem from the type of reading task employed, this exploratory investigation sought to examine the metabolic correlates of fluent reading of carefully controlled narrative text. Test/retest controls were employed to determine if activation increases in rCBF reflected changes over normal test/retest variability. Effect size was also considered. METHOD Subjects Subjects were five successful male graduate students at a major university. Three participated in the reading task. Two served as test/retest controls to permit comparison of activation increases in rCBF with normal test/retest variability. The age of the subjects ranged from 23 to 39 years (mean=30.8). For those who read the narrative text, an at-rest blood flow study was conducted prior to the reading study. Procedure A two study paradigm was employed. The subjects rested in a prone position with their heads positioned in a helmet with 10 sodium iodide detectors arranged on each side. The detectors were positioned such that perfusion could be assessed bilaterally in the prefrontal region, over the motor and sensory cortex, perisylvian region and parietal and occipital cortex. A blood sample was drawn immediately prior to each study so that
Bilateral rCBF Activation in Normal Readers
73
the gray matter blood partition coefficient could be determined using the subject’s hematocrit. A capnograph was used to monitor and record pC0, throughout each study and respiration rate was similarly monitored. Each subject inhaled through a fitted mouthpiece trace amounts of ‘33Xenon for one minute following a one minute period in which background data were recorded. This inhalation period was followed by a 10 minute desaturation period during which time the subject breathed room air. The details of the data analysis, hardware configuration, and associated procedures are discussed in detail elsewhere (Hynd et al., 1987). For those subjects who participated in the reading trial, an unfamiliar narrative passage, “The Summer of the Beautiful White Horse” by Saroyan, was employed. Subjects read silently while they breathed ‘33Xenon during the one minute inhalation and 10 minute desaturation periods. All subjects achieved 85% or greater correct on a comprehension test to verify reading activity. RESULTS The small numbers of subjects in this exploratory investigation necessitated charting normal test/retest variation. The mean percentage change in blood perfusion for the left and right hemispheres for controls and reading activated subjects are presented in Figure 1. For the normal controls who took part in the test/retest condition, 3% and 7% increases in rCBF perfusion were noted over the first test in the left and right hemispheres, respectively. Compared to their at-rest study, reading activated subjects evidenced 12.67% and 14.67% increased perfusion during reading in the left and right hemispheres, respectively. Independent sample t-tests (one-tailed) indicate significantly more activation in both left (t=6.91, p< .OOl) and right hemispheres (t=3.81, p< .OOl) in the reading activated subjects than in the test/retest subjects. Furthermore, t-tests (one-tailed) for dependent samples reveal significantly greater bihemispheric rCBF during reading than activated subjects’ own at-rest perfusion (t= 10.16, p< .005-left; t= 10.74, p-c .005-right). Finally, there was no statistical difference in the percentage change in perfusion in the right versus left hemisphere of the reading activated subjects (t=1.22, p= .12), suggesting that both hemispheres were equally activated by reading the narrative text. Detector locations and mean percent changes in perfusion for control and reading activated subjects are summarized in Figure 2. Table 1 displays the results of independent sample t-tests on the mean percentage change of test/retest versus reading activated subjects at each detector site. In general, perfusion changes during reading were greater in the central and posterior regions of both hemispheres. The only frontal detector showing significant increased activation was in the left prefrontal area.
M. I. S. Huettner, B. L. Rosenthal, and G. W Hynd LEFT HEMISPHERE
RIGHT
HEMISPHERE
I8 16 14
I
Reading
Activated
12 IO 8 6 4 2 Test-Retest Contfol
Readmg
Test-Retest
Activated
COfltrOl
SUBJECTS
FIGURE 1. Mean percentage change in gray matter blood flow in the left and right hemispheres for subjects in the test/retest condition and the at rest/reading activated condition.
The exploratory nature of this study suggested that it was important to consider the practical significance of these findings. The magnitude of the perfusion increases during reading was calculated for each of the significant detector sites using the delta (ES) procedure (Glass & Hopkins, 1984). Delta yields an estimate of effect size expressed in standard deviation units. Generally, .8 ES (standard deviation unit) is considered a large effect. The changes at the significant detector sites ranged from 2.24 to 5.09 ES (see Table 1). While all statistical analyses performed on these data should be interpreted cautiously due to the small sample size, the large ES during reading activation at all significant detector sites suggests that these findings may reflect actual localized perfusion changes associated with the reading of narrative text. DISCUSSION The results of this study are consistent with recent PET (Gross-Glenn et al., 1987) and rCBF (Hynd et al., 1987) findings which suggest more bihemispheric involvement during reading of narrative text than predicted by the Geschwind-Wernicke and similar left hemispheric neurolinguistic models (Hynd & Hynd, 1984; Mayeux & Kandel, 1985; Satz, in press).
Bilateral rCBF Activation in Normal Readers
75
In contrast, previous rCBF studies that report left frontal or left perisylvian activation during reading have employed either disconnected single word or semantic recall tasks (Maximilian et al., 1978; Wood et al., 1980) that may not reflect the multiple neurocognitive processes activated during the reading of actual connected narrative text. Reading of narrative text may activate an ecologically more valid neurocognitive system of reading that encompasses multiple components of reading, including the semantic, pragmatic, emotional, and imagery aspects. The present results are not inconsistent with evolving theory regarding involvement of the right hemisphere in language. Both Ross (1981) and particularly Bhatnagar and Andy (1983) provide clinical and experimental evidence that the right hemisphere is important in simpler language functions including naming and prosody and that this system is anatomically
LEFT CEREBRAL
RIGHT CEREBRAL
HEMISPHERE
HEMISPHERE
a, 18 16 14
12 Ql 8 6 4 2 0 -2 -4
“-i , , , , , , , , , , 1
2
3
4
5
ETECTORS
6
7
6
3
10
(
,
,
,
11
12
13
14
,
,
,
15 16 17 DETECTORS
,
,
,
16
19
20
1’.
FIGURE 2. Mean percentage change in gray matter blood flow according to detectors for the test/retest versus reading activated subjects. Detectors 1, 2 (left), 11 and 12 (right) assessed perfusion in the prefrontal cortex; detectors 3, 4 (left), 13 and 14 (right) covered the motor cortex; detectors 7 (left) and 17 (right) were placed over the sensory cortex; detectors 5, 6, 8 (left), 15, 16, & 18 (right) were over the temporal-perisylvian region; detectors 9 (left) & 19 (right) assessed perfusion in the parietal region while detectors 10 (left) and 20 (right) were placed over the occipital cortex, *p < .05 **p < .Ol.
76
M. I. S. Huettner, El. L. Rosenthal, and G. K Hynd TABLE 1 Mean Percentage of Change in Blood Perfusion at Each Detector Site
Test/Retest (N=2) Detector
#
1 2 3 4 5 6 I 8 9 10 11 12 13 14 15 16 17 18 19 20 Left Hemi. Right Hemi
At Rest/ Activated (N=3)
Mean
SD
Mean
SD
2.00 3.20 7.55 1.50 - 1.75 2.00 0 4.50 4.00 4.50 5.00 4.95 19.75 7.00 2.50 6.00 -1.75 9.00 5.00 7.50
4.24 5.31 11.95 19.09 15.91 0 5.66 6.36 9.90 .71 7.07 7.14 6.01 7.07 6.36 8.49 1.77 7.07 4.24 6.36
13.67 4.67 14.67 12.00 11.67 11.67 12.00 12.23 19.67 14.33 8.17 11.33 17.00 12.00 11.83 19.00 15.33 16.00 15.00 22.33
3.19 9.71 2.52 4.17 8.02 5.03 6.56 2.08 3.79 2.31 4.75 4.93 8.19 2.65 7.52 3.61 4.93 5.29 4.58 4.04
3.24* .19 1.08 .98 1.30 2.58* 2.10 2.12 2.64* 5.58* .62 1.21 .40 1.19 1.43 2.49* 4.50** 1.29 2.45* 3.29*
2.95
2.75 6.50
2.61 5.52
12.66 14.80
3.71 4.13
6.91*** 3.81***
3.09 1.70
aES indicates effect size in standard *p<.o5. **p<.o1. ***p<.oo1.
deviation
t
ESa
2.37
2.41 5.09
2.27 4.11 2.24 3.00
units.
and functionally tied to and activated by the left hemispheric language system. Some limited evidence reported by Ross (1981) suggests that the emotional components of language, or prosody, are anatomically represented in a right hemispheric system that mirrors the organization of cognitivelinguistic language processes in the left hemisphere. Such a bilateral system may well be activated in the reading of connected narrative text as in this study. Further, there is increasing evidence that bilateral thalamic nuclei may play a vital role in verbal-vigilance (Ojeman, 1975, 1979), verbal-expressive, (Sioltis et al., 1986) and other neurolinguistic processes related to cognition and reading acquisition (Hynd & SemrudClikeman, 1988). Thus, in this context, this exploratory study suggests that by systematically altering and controlling the varic& semantic and linguistic components of reading, it may be possible to derive a more accurate bihemispheric model of the interaction of the perceptual, cognitive, and affec-
Bilateral rCBF Activation
in Normal Readers
tive processes involved in fluent reading based on metabolic brains of normal, noncompromised subjects.
77
changes
in the
REFERENCES Bastian, H. C. (1898). Aphasra andotherspeech defects. London: H. K. Lewis. Bhatnager, S., &Andy, 0. J. (1983). Language m the nondominant right hemisphere. Archives of Neurology, 40, 128-73 1. Campbell, S., & Whitaker, H. (1986). Corttcal maturation and developmental neurolinguistics. In .I. E. Obrzut & G. W. Hynd (Eds.), Child neuropsychology (Vol I): Theory and research (pp. 55-72). Orlando, FL: Academic Press. Coltheart, M., Masterson, J., Byng, S., Prior, M., & Riddoch. J. (1983). Surface dyslexia.
Quarterly Journal of Experimental Psychology, 35, 469-496. Duffy, F., Denckla, M. B., Bartels, P. H., & Sand%, G. (1980). Regional differences in bram electrical activity by topographic mapping. Anna/s of Neurology, 7, 412-420. Duffy, F., Denckla, M. B., Bartels, P. H., Sandini, G., & Kiessling, L. S. (1980). Dyslexia: Automated diagnosis by computerized classification of brain electrtcal activtty. Annals of
Neurology, 7,421-428. Friedman, R. B., & Perlman, M. B. (1982). On the underlying causes of semantic paralexias m a patient with deep dyslexia. Neuropsychologta, 20, 559-568. Geschwind, N. (1974). Anatomical foundations of language and dominance. In C. L. Ludlow & M. E. Doran-Quine (Eds.), The neurologtcal basis of language disorders in chrldren: Methods and drrection for research (NIH Pub. No. 79-440). Bethesda, MD: U.S. Department of Health, Education and Welfare. Geschwind, N. (1984). Cerebral dominance in biologtcal perspective. Neuropsychologia, 22,
675-683. Glass, G. V., & Hopkins, K. D. (Eds.). (1984). Statistical methods rn educatron andpsychology (2nd ed.). Englewood Cliffs, NJ: Prentice-Hall. Gross-Glenn, K., Duara, R., Yoshii, F., Barker, W. W., Chang, J. Y., Apicella, A., Boothe, T., & Lubs, H. A. (1987). Pet-scan studies during reading in dyslexic and non-dyslexic adults.
Neuroscience Abstracts. J. M. (1985, February). Developmental deep dyslexia: Implicatrons for the “rtght hem/sphere” hypothesis. Paper presented at the annual meeting of the International
Healey,
Neuropsychological Hynd, G. W., & Hynd,
Society, San Diego, CA. C. R. (1984). Dyslexia: NeuroanatomicaUneurolinguistic
perspectives.
Reading Research Quarterly, 19, 482-498. Hynd, G. W., Hynd, C. W., Sullivan, H. G., & Kingsbury, T. B. (1987). Regional cerebral flood flow (rCBF) in developmental dyslexia: Activation during reading in a surface and deep dyslexic. Journal of Learning Disabilitres, 20, 294-300. Hynd, G. W., & Semrud-Clikeman, M. (in press). Dyslexia and neurodevelopmental pathology: Relationships to cognition, intelligence and reading skill acquisition. Journal of Learnrng
Disabilitres. Maximilian, changes
V. A., Prohovnik, in the left cerebral
I., Risberg, hemisphere
J., & Hakansson, K. (1978). Regional blood flow during word pair learning and recall. Brain and
Language, 6,22-31. Mayeux, R., & Kandel, E. R. (1985). Natural language, localizable disorders of cognitive functtoning. In E. R. Prrnciples of neuralscrence (2nd ed., pp. 688-703). New OJeman, G. A. (1975). Language and the thalamus: Object stimulation. Brain and Language, 2, 101-120.
disorders of language, and other Kandel & J. H. Schwartz (Eds.), York: Elsevier. naming and recall after thalamic
78
M. I. S. Huettner,
B. L. Rosenthal,
and G. W! Hynd
Ojeman, G. A. (1979). Altering memory wtth human ventrolateral thalamic stimulation. In E. R. Hitchcock, H. T. Ballantine, Jr., &B. A. Meyerson (Eds.), Modern concepts inpsychiatTICsurgery. Amsterdam: Elsevier. Ross, E. D. (1981). The aprosodias: Functional-anatomic organization of the affecttve components of language in the right hemisphere. Archives of Neurology, 38, 561-569. Satz, P. (in press). Developmental dyslexia: An etiological reformulation. In G. Pavlidts (Ed.), D_vsle.ura: Neuropsychologlcal and learrung perspectives. New York: John Wiley. Sioltis, J. J., Portenoy, R. K., Jarden, J. 0.. Lipton, R. B., Kurtzberg, D., Rottenberg, D. A., & Foley, M. (1986, February). Behar,roral. metabolrc, and electrophyslologic correlates of compulsrve thalamic self-stunulation. Paper presented at the annual convention of the International Neuropsychological Society, Denver, CO. Temple, C. M. (1984). Surface dyslexia in a child with epilepsy. Neuropsvchologra, 22, 569-
576. Wood, R., Taylor, B., Penny, R., & Stump, D. (1980). Regional cerebral blood flow response to recognitton memory versus semantic classification tasks. Brarn and Language, 9, 113-122.
0887-6177189 $3 00 + 00 CopyrIght L 1989 Natmnal Academy of Neuropsychologlsts
Brief Report
Regional Cerebral in Normal Activation Marion
I. S. Huettner,
Blood Flow (rCBF)
Readers: Bilateral with
Narrative Text
Becky L. Rosenthal, and George W. Hynd Unlverslty
0T Ceorgla
Based on recent evidence that suggests a more active participatron of the right cerebral hemisphere in reading, this exploratory study examined changes rn regional cerebral blood flow (rCBF) in three normal subjects during the readmg of controlled narrative text. Narrative text presumably activates all neurocognitive processes rmportant in reading including the semantic, pragmatrc, emottonal, and imagery components. Because of the small number of subjects, percentage change over a baseline at rest condition in rCBF durrng readmg was compared to test/retest variability in control subjects. Effect size was also considered. The results indicated stattstically significant bilateral central posterior activation during reading. These findings are consistent wrth an evolving model of brlateral language representation m which subcortical structures and rtght hemrsphenc systems are functionally and anatomtcally tied to the dominant left hemtspheric language ten ters.
Contemporary conceptualizations of reading disorders are largely based on a neurolinguistic model advocated by Geschwind (1974, 1984), and others (e.g., Campbell & Whitaker, 1986). Generally, this model, which originated in ideas advanced by Wernicke and Dejerine nearly a century ago (Hynd & Hynd, 1984; Mayeux & Kandel, 1985), documents the significant contribution of the classical perisylvian language areas, the posterior visual, and left tertiary cortices in fluent reading, and is generally consistent with knowledge about aphasia and alexia with and without agraphia. Evidence regarding the validity of this conceptualization includes the case studies of patients with acquired surface dyslexia (Coltheart, Masterson,
Requests for reprints tal Neuropsychology,
should be sent to George W. Hynd, Center for Clinical and Developmen325 Aderhold Hall, University of Georgia, Athens, GA 30602. 71
72
M. I. S. Huettner, B. L. Rosenthal, and G. W Hynd
Byng, Prior, & Riddoch, 1983; Temple, 1984) and acquired deep dyslexia (Friedman & Perlman, 1982; Healey, 1985), the topographic mapping studies of brain electrical activity in normals and developmental dyslexics during associated perceptual and linguistic tasks (Duffy, Denckla, Bartels, & Sandini, 1980; Duffy, Denckla, Bartels, Sandini, & Kiessling, 1980), and regional cerebral blood flow (rCBF) studies involving word learning and semantic recall tasks (Maxmilian, Prohovnik, Risberg, & Hakansson, 1978; Wood, Taylor, Penny, & Stump, 1980). Collectively, these studies suggest that in cases of documented or presumed pathology, deficiencies exist on behavioral and metabolic correlates of the classical perceptual and language areas proposed by the Geschwind-Wernicke model (Mayeux & Kandel, 1985). Contrary to the Geschwind-Wernicke model, however, and perhaps more consistent with that proposed by Bastian (1898), recent neurometabolic studies employing both rCMRglc and rCBF methodologies reveal significant and equal bihemispheric central-posterior activation during reading of words and text in normal and dyslexic subjects (Gross-Glenn et al., 1987; Hynd, Hynd, Sullivan, & Kingsbury, 1987). A principle methodological difference between these studies and those documenting left perisylvian activity is their employment of word and text reading versus semantic recall tasks (e.g., Wood et al., 1980). Given that these conflicting findings may stem from the type of reading task employed, this exploratory investigation sought to examine the metabolic correlates of fluent reading of carefully controlled narrative text. Test/retest controls were employed to determine if activation increases in rCBF reflected changes over normal test/retest variability. Effect size was also considered. METHOD Subjects Subjects were five successful male graduate students at a major university. Three participated in the reading task. Two served as test/retest controls to permit comparison of activation increases in rCBF with normal test/retest variability. The age of the subjects ranged from 23 to 39 years (mean=30.8). For those who read the narrative text, an at-rest blood flow study was conducted prior to the reading study. Procedure A two study paradigm was employed. The subjects rested in a prone position with their heads positioned in a helmet with 10 sodium iodide detectors arranged on each side. The detectors were positioned such that perfusion could be assessed bilaterally in the prefrontal region, over the motor and sensory cortex, perisylvian region and parietal and occipital cortex. A blood sample was drawn immediately prior to each study so that
Bilateral rCBF Activation in Normal Readers
73
the gray matter blood partition coefficient could be determined using the subject’s hematocrit. A capnograph was used to monitor and record pC0, throughout each study and respiration rate was similarly monitored. Each subject inhaled through a fitted mouthpiece trace amounts of ‘33Xenon for one minute following a one minute period in which background data were recorded. This inhalation period was followed by a 10 minute desaturation period during which time the subject breathed room air. The details of the data analysis, hardware configuration, and associated procedures are discussed in detail elsewhere (Hynd et al., 1987). For those subjects who participated in the reading trial, an unfamiliar narrative passage, “The Summer of the Beautiful White Horse” by Saroyan, was employed. Subjects read silently while they breathed ‘33Xenon during the one minute inhalation and 10 minute desaturation periods. All subjects achieved 85% or greater correct on a comprehension test to verify reading activity. RESULTS The small numbers of subjects in this exploratory investigation necessitated charting normal test/retest variation. The mean percentage change in blood perfusion for the left and right hemispheres for controls and reading activated subjects are presented in Figure 1. For the normal controls who took part in the test/retest condition, 3% and 7% increases in rCBF perfusion were noted over the first test in the left and right hemispheres, respectively. Compared to their at-rest study, reading activated subjects evidenced 12.67% and 14.67% increased perfusion during reading in the left and right hemispheres, respectively. Independent sample t-tests (one-tailed) indicate significantly more activation in both left (t=6.91, p< .OOl) and right hemispheres (t=3.81, p< .OOl) in the reading activated subjects than in the test/retest subjects. Furthermore, t-tests (one-tailed) for dependent samples reveal significantly greater bihemispheric rCBF during reading than activated subjects’ own at-rest perfusion (t= 10.16, p< .005-left; t= 10.74, p-c .005-right). Finally, there was no statistical difference in the percentage change in perfusion in the right versus left hemisphere of the reading activated subjects (t=1.22, p= .12), suggesting that both hemispheres were equally activated by reading the narrative text. Detector locations and mean percent changes in perfusion for control and reading activated subjects are summarized in Figure 2. Table 1 displays the results of independent sample t-tests on the mean percentage change of test/retest versus reading activated subjects at each detector site. In general, perfusion changes during reading were greater in the central and posterior regions of both hemispheres. The only frontal detector showing significant increased activation was in the left prefrontal area.
M. I. S. Huettner, B. L. Rosenthal, and G. W Hynd LEFT HEMISPHERE
RIGHT
HEMISPHERE
I8 16 14
I
Reading
Activated
12 IO 8 6 4 2 Test-Retest Contfol
Readmg
Test-Retest
Activated
COfltrOl
SUBJECTS
FIGURE 1. Mean percentage change in gray matter blood flow in the left and right hemispheres for subjects in the test/retest condition and the at rest/reading activated condition.
The exploratory nature of this study suggested that it was important to consider the practical significance of these findings. The magnitude of the perfusion increases during reading was calculated for each of the significant detector sites using the delta (ES) procedure (Glass & Hopkins, 1984). Delta yields an estimate of effect size expressed in standard deviation units. Generally, .8 ES (standard deviation unit) is considered a large effect. The changes at the significant detector sites ranged from 2.24 to 5.09 ES (see Table 1). While all statistical analyses performed on these data should be interpreted cautiously due to the small sample size, the large ES during reading activation at all significant detector sites suggests that these findings may reflect actual localized perfusion changes associated with the reading of narrative text. DISCUSSION The results of this study are consistent with recent PET (Gross-Glenn et al., 1987) and rCBF (Hynd et al., 1987) findings which suggest more bihemispheric involvement during reading of narrative text than predicted by the Geschwind-Wernicke and similar left hemispheric neurolinguistic models (Hynd & Hynd, 1984; Mayeux & Kandel, 1985; Satz, in press).
Bilateral rCBF Activation in Normal Readers
75
In contrast, previous rCBF studies that report left frontal or left perisylvian activation during reading have employed either disconnected single word or semantic recall tasks (Maximilian et al., 1978; Wood et al., 1980) that may not reflect the multiple neurocognitive processes activated during the reading of actual connected narrative text. Reading of narrative text may activate an ecologically more valid neurocognitive system of reading that encompasses multiple components of reading, including the semantic, pragmatic, emotional, and imagery aspects. The present results are not inconsistent with evolving theory regarding involvement of the right hemisphere in language. Both Ross (1981) and particularly Bhatnagar and Andy (1983) provide clinical and experimental evidence that the right hemisphere is important in simpler language functions including naming and prosody and that this system is anatomically
LEFT CEREBRAL
RIGHT CEREBRAL
HEMISPHERE
HEMISPHERE
a, 18 16 14
12 Ql 8 6 4 2 0 -2 -4
“-i , , , , , , , , , , 1
2
3
4
5
ETECTORS
6
7
6
3
10
(
,
,
,
11
12
13
14
,
,
,
15 16 17 DETECTORS
,
,
,
16
19
20
1’.
FIGURE 2. Mean percentage change in gray matter blood flow according to detectors for the test/retest versus reading activated subjects. Detectors 1, 2 (left), 11 and 12 (right) assessed perfusion in the prefrontal cortex; detectors 3, 4 (left), 13 and 14 (right) covered the motor cortex; detectors 7 (left) and 17 (right) were placed over the sensory cortex; detectors 5, 6, 8 (left), 15, 16, & 18 (right) were over the temporal-perisylvian region; detectors 9 (left) & 19 (right) assessed perfusion in the parietal region while detectors 10 (left) and 20 (right) were placed over the occipital cortex, *p < .05 **p < .Ol.
76
M. I. S. Huettner, El. L. Rosenthal, and G. K Hynd TABLE 1 Mean Percentage of Change in Blood Perfusion at Each Detector Site
Test/Retest (N=2) Detector
#
1 2 3 4 5 6 I 8 9 10 11 12 13 14 15 16 17 18 19 20 Left Hemi. Right Hemi
At Rest/ Activated (N=3)
Mean
SD
Mean
SD
2.00 3.20 7.55 1.50 - 1.75 2.00 0 4.50 4.00 4.50 5.00 4.95 19.75 7.00 2.50 6.00 -1.75 9.00 5.00 7.50
4.24 5.31 11.95 19.09 15.91 0 5.66 6.36 9.90 .71 7.07 7.14 6.01 7.07 6.36 8.49 1.77 7.07 4.24 6.36
13.67 4.67 14.67 12.00 11.67 11.67 12.00 12.23 19.67 14.33 8.17 11.33 17.00 12.00 11.83 19.00 15.33 16.00 15.00 22.33
3.19 9.71 2.52 4.17 8.02 5.03 6.56 2.08 3.79 2.31 4.75 4.93 8.19 2.65 7.52 3.61 4.93 5.29 4.58 4.04
3.24* .19 1.08 .98 1.30 2.58* 2.10 2.12 2.64* 5.58* .62 1.21 .40 1.19 1.43 2.49* 4.50** 1.29 2.45* 3.29*
2.95
2.75 6.50
2.61 5.52
12.66 14.80
3.71 4.13
6.91*** 3.81***
3.09 1.70
aES indicates effect size in standard *p<.o5. **p<.o1. ***p<.oo1.
deviation
t
ESa
2.37
2.41 5.09
2.27 4.11 2.24 3.00
units.
and functionally tied to and activated by the left hemispheric language system. Some limited evidence reported by Ross (1981) suggests that the emotional components of language, or prosody, are anatomically represented in a right hemispheric system that mirrors the organization of cognitivelinguistic language processes in the left hemisphere. Such a bilateral system may well be activated in the reading of connected narrative text as in this study. Further, there is increasing evidence that bilateral thalamic nuclei may play a vital role in verbal-vigilance (Ojeman, 1975, 1979), verbal-expressive, (Sioltis et al., 1986) and other neurolinguistic processes related to cognition and reading acquisition (Hynd & SemrudClikeman, 1988). Thus, in this context, this exploratory study suggests that by systematically altering and controlling the varic& semantic and linguistic components of reading, it may be possible to derive a more accurate bihemispheric model of the interaction of the perceptual, cognitive, and affec-
Bilateral rCBF Activation
in Normal Readers
tive processes involved in fluent reading based on metabolic brains of normal, noncompromised subjects.
77
changes
in the
REFERENCES Bastian, H. C. (1898). Aphasra andotherspeech defects. London: H. K. Lewis. Bhatnager, S., &Andy, 0. J. (1983). Language m the nondominant right hemisphere. Archives of Neurology, 40, 128-73 1. Campbell, S., & Whitaker, H. (1986). Corttcal maturation and developmental neurolinguistics. In .I. E. Obrzut & G. W. Hynd (Eds.), Child neuropsychology (Vol I): Theory and research (pp. 55-72). Orlando, FL: Academic Press. Coltheart, M., Masterson, J., Byng, S., Prior, M., & Riddoch. J. (1983). Surface dyslexia.
Quarterly Journal of Experimental Psychology, 35, 469-496. Duffy, F., Denckla, M. B., Bartels, P. H., & Sand%, G. (1980). Regional differences in bram electrical activity by topographic mapping. Anna/s of Neurology, 7, 412-420. Duffy, F., Denckla, M. B., Bartels, P. H., Sandini, G., & Kiessling, L. S. (1980). Dyslexia: Automated diagnosis by computerized classification of brain electrtcal activtty. Annals of
Neurology, 7,421-428. Friedman, R. B., & Perlman, M. B. (1982). On the underlying causes of semantic paralexias m a patient with deep dyslexia. Neuropsychologta, 20, 559-568. Geschwind, N. (1974). Anatomical foundations of language and dominance. In C. L. Ludlow & M. E. Doran-Quine (Eds.), The neurologtcal basis of language disorders in chrldren: Methods and drrection for research (NIH Pub. No. 79-440). Bethesda, MD: U.S. Department of Health, Education and Welfare. Geschwind, N. (1984). Cerebral dominance in biologtcal perspective. Neuropsychologia, 22,
675-683. Glass, G. V., & Hopkins, K. D. (Eds.). (1984). Statistical methods rn educatron andpsychology (2nd ed.). Englewood Cliffs, NJ: Prentice-Hall. Gross-Glenn, K., Duara, R., Yoshii, F., Barker, W. W., Chang, J. Y., Apicella, A., Boothe, T., & Lubs, H. A. (1987). Pet-scan studies during reading in dyslexic and non-dyslexic adults.
Neuroscience Abstracts. J. M. (1985, February). Developmental deep dyslexia: Implicatrons for the “rtght hem/sphere” hypothesis. Paper presented at the annual meeting of the International
Healey,
Neuropsychological Hynd, G. W., & Hynd,
Society, San Diego, CA. C. R. (1984). Dyslexia: NeuroanatomicaUneurolinguistic
perspectives.
Reading Research Quarterly, 19, 482-498. Hynd, G. W., Hynd, C. W., Sullivan, H. G., & Kingsbury, T. B. (1987). Regional cerebral flood flow (rCBF) in developmental dyslexia: Activation during reading in a surface and deep dyslexic. Journal of Learning Disabilitres, 20, 294-300. Hynd, G. W., & Semrud-Clikeman, M. (in press). Dyslexia and neurodevelopmental pathology: Relationships to cognition, intelligence and reading skill acquisition. Journal of Learnrng
Disabilitres. Maximilian, changes
V. A., Prohovnik, in the left cerebral
I., Risberg, hemisphere
J., & Hakansson, K. (1978). Regional blood flow during word pair learning and recall. Brain and
Language, 6,22-31. Mayeux, R., & Kandel, E. R. (1985). Natural language, localizable disorders of cognitive functtoning. In E. R. Prrnciples of neuralscrence (2nd ed., pp. 688-703). New OJeman, G. A. (1975). Language and the thalamus: Object stimulation. Brain and Language, 2, 101-120.
disorders of language, and other Kandel & J. H. Schwartz (Eds.), York: Elsevier. naming and recall after thalamic
78
M. I. S. Huettner,
B. L. Rosenthal,
and G. W! Hynd
Ojeman, G. A. (1979). Altering memory wtth human ventrolateral thalamic stimulation. In E. R. Hitchcock, H. T. Ballantine, Jr., &B. A. Meyerson (Eds.), Modern concepts inpsychiatTICsurgery. Amsterdam: Elsevier. Ross, E. D. (1981). The aprosodias: Functional-anatomic organization of the affecttve components of language in the right hemisphere. Archives of Neurology, 38, 561-569. Satz, P. (in press). Developmental dyslexia: An etiological reformulation. In G. Pavlidts (Ed.), D_vsle.ura: Neuropsychologlcal and learrung perspectives. New York: John Wiley. Sioltis, J. J., Portenoy, R. K., Jarden, J. 0.. Lipton, R. B., Kurtzberg, D., Rottenberg, D. A., & Foley, M. (1986, February). Behar,roral. metabolrc, and electrophyslologic correlates of compulsrve thalamic self-stunulation. Paper presented at the annual convention of the International Neuropsychological Society, Denver, CO. Temple, C. M. (1984). Surface dyslexia in a child with epilepsy. Neuropsvchologra, 22, 569-
576. Wood, R., Taylor, B., Penny, R., & Stump, D. (1980). Regional cerebral blood flow response to recognitton memory versus semantic classification tasks. Brarn and Language, 9, 113-122.