Katrin Neumann1, Harald A. Euler2, Christine Preibisch3, Alexander Wolff von Gudenberg4
1Clinic of Phoniatry and Pedaadiology, University of Frankfurt, Theodor-Stern-Kai 7, 60590 F rankfurt/Main, Germany
2Department of Psychology, University of Kassel, Hollaend. St): 36-38, 3412 7 Kassel, Germany
3Department of Neuroradiology, University of Frankfurt, Schlensenweg 2-16, 60528 Frankfurt/Main, Germany
4Institute of the Kasseler Stottertherapie, 34308 Bad Emstal, Germany
FMRI findings of 9 male persons who stutter (PWS) compared to 16 non-stuttering controls (PWNS) are reported. Distributed, predominantly right-hernispheric overactivations in PWS during overt reading were more widespread and left-sided after fluency shaping therapy and were slightly reduced and again more right-sided two years later. These findings, together with larger activations post-treatment also in a semantic task, suggest that overactivations might reflect compensatory mechanisms. Left frontal deactivations were insensitive to therapy and therefore possibly indicate dysfunctions. Thus, fluency-inducing techniques might synchronize a disturbed signal transmission between auditory, speech motor planning, and motor areas.
A variety of hypotheses about the pathophysiology of stuttering have focused on dysfunction of speech motor control, especially in left motor areas (Webster, 1990), atypical lateralization of speech and language processes (Moore, 1984; Travis, 1978), deficiency of the language production system (Perkins, Kent, & Curlee, 1991), sensory impairments, in particular auditory (Salmelin et al., 1998), or a complex combination of speech motor and linguistic deficits (Peters et al., 2000), with no consensus over the preferred theory. Thus, neuroimage findings in PWS are expected to improve theories about stuttering, with profit for therapy. For instance, the hypothesis that the processing steps of speech production might be affected by timing disturbances of neuronal signal transmission between premotor, auditory, and speech motor (Broca) regions was recently supported by neuroimaging findings (Foundas et al., 2001; Ingham, 2001; Salmelin et al., 2000; Sommer et al., 2002) which showed (1) a failure of temporal lobe activation during speech (Ingham 2001), (2) a reversed processing sequence between the left inferior frontal cortex (articulatory programming) and left premotor and motor cortices (motor preparation) (Salmelin et al., 2000), and (3) a white matter disconnection of the left precentral cortex (premotor) with temporal and frontal language areas (motor) possibly reflecting impaired connection between these regions (Sommer et al., 2002). Accordingly, right-hemisphere overactivations could reflect compensation. Alternatively, these overactivation could have led to a subsequent dysfunction in the left hemisphere. In a recent fMRI study with male adult PWS and PWNS we detected an overactivation consistent across all PWS and absent in all PWNS in the right frontal loperculum (RFO) during overt reading, reflecting an effect specific to stuttering. As responses in the RFO were negatively correlated with the severity of stuttering we hypothesized that the overactivation in the RFO (right homologue of Broca’s area) reflected a compensation process rather tian a primary dysfunction. As this region was also overactivated during a seman;ic task we assumed a compensation mechanism during early processing steps of speech production which acts inde pendently of speech motor output demands (Preibisch, Neumann, et al., 2003).
Because fluency induction might normalize the synchronization among language regions (Braun et al., 1997; Fox et al., 1996) and because the ultimate aim of brain research in PWS is efficient therapy, we sought to document changes in brain activation patterns due to a fluency shaping therapy. EEG and PET studies had revealed a shift of brain activity to more left hemispheric regions after stuttering-reducing therapies (Boberg et al., 1983; De Nil & Kroll, 200la,b; Kroll et al., 1997; Moore, 1984), a higher activation in the anterior cingulate cortex (ACC) in PWS than in PWNS subjects during silent reading, which was reduced after therapy, and a post-treatment increased and more left-sided motor activation during overt reading. These findings were interpreted as enhanced automaticity during speech production and optimized sequencing of articulatory, phonatory, and respiratory movements (De Nil & Kroll, 2001 a,b).
We compared the fl\/[RI activation of nine PWS before and immediately after a three-week intensive fluency shaping therapy course and additionally compared these activation patterns with those of 16 PWNS (Neumann, Preibisch, et. al, 2003). Furthermore, five of the PWS underwent fl\/[RI two years after therapy (Neumann, Euler, et. al, 2003). If stuttering is caused by a disturbed neuronal synchronization during speech processing, the replacement of the automatized speech pattern of stuttering with a new one due to the therapy can be expected to cause a shift in cerebral activation patterns implicated in timing of speech processing and thus reflect compensation. If, alternatively, the "fluency shaping effect is considered to be solely due to decreased demands on the speech-motor system by rhythmization and prolongation of speech, a reduced neuronal activity should be expected after fluency treatment.
The PWS were selected from the therapy waiting list of the Institute of the Kasseler Stottertherapie, Baunatal, Germany. The Kassel Stuttering Therapy is a modified version of the Precision Fluency Shaping Program (Webster, 1974) and consists of a three-week in-patient intensive therapy and a structured maintenance program for one to two years. The main modification of the original therapy program is the utilization of a computer program (speak:gentle®, Bioservices Software, Munich, Germany) which provides biofeedback for syllable prolongation, soft voice onset, and smooth sound transitions. Details about the therapy and its short-term and long-term effects on objective and subjective fluency measures are described in Euler and Wolff von Gudenberg (2002).
Stuttering was measured as percentage of stuttered syllables in four different speech situations (talking with therapist, overt reading, calling an unknown person by telephone, and interviewing people on the street). The PWS had a mean stutter rate of 9.9 % before therapy, of .9 immediately after the intensive therapy, and of 1.7 after the follow-up period.
To compare persons who stutter severely with persons who stutter moderately in a sufficiently large sample, another 11 participants were added. These participants were of comparable age, handedness, and stuttering severity as the original sample of five PWS. Severe stuttering (n=7) was defined as more than 10% stuttered syllables, moderate stuttering (n=9) as less than l0% stuttered syllables. This larger sample of PWS was only available before therapy.
Subjects were scanned with a 1.5 T Siemens Vision Scanner (Siemens, Erlangen, Germany) using gradient echo EPI. The participants performed two tasks. In an oral reading task, which should tap speech-motor processes, they were asked to read aloud 78 short sentences. Watching meaningless signs served as a baseline. In the analysis of images acquired during speech production, we used an event-related design to reduce movement artifacts as described in Preibisch, Raab, et al. (2003). Speech output was in general fluent in all participants. This speech fluency was possibly induced by the masking effect of the scanner noise and by the segmentation of speech into short periods with pauses in between known to aid fluency (Ingham, 1984). PWS spoke more slowly immediately after therapy than during the recording before therapy. At the follow-up assessment time none of the PWS spoke remarkably slower than before therapy.
In the other task, the participants performed a silent semantic decision consisting in synomym judgements. Judgements about colors served to match the semantic task with respect to all non- specific components except linguistic processing.
In the semantic task we did not observe any overactivation in PWS compared to PWNS but deactivations in the left insula, cingulate, and frontal inferior region, which were stable over all three assessment times. After therapy PWS showed higher levels of activation in the left inferior frontal gyrus which was slightly reduced after two years (Neumann, Preibisch, et. al, 2003; Neumann, Euler, et al., 2003).
During both overt reading and semantic decisions, persons with moderate stuttering showed a larger brain activation than those with more severe symptoms. These overactivations were more widespread and more bilateral during overt reading than during semantic decision, and more left- sided in the latter one than in the speech production task (Neumann, Euler, et. al, 2003).
In the semantic task a higher activation after than before therapy was detected in the left frontral cortex. This, together with the overactivation in the RFO during semantic decision (Preibisch, Neumann, et al., 2003), indicates that a successful compensation for stuttering recruits processing steps upstream from speech motor execution.
Consistent lower activations in PWS than in PWN S were observed during overt reading in a left precentral and in bilateral occipital regions, and during semantic decision making in left inferior frontal regions and the left cingulate. Our data agree with findings of deactivations (Fox et al., 1996, 2000; Ingham et al., 2000) or deficient fiber connections in PWS in speech, language, and auditory regions (Sommer et al., 2002). These deactivations were unaffected by therapy, even after two years of follow-up and despite a controlled continued practice of the new speech pattern (Neumann, Euler, et. al, 2003). They remained unchanged even when the PWS spoke slower in their new speech pattern immediately after therapy. In particular, deactivations of the left precentral and left inferior frontal cortex are likely to reflect constitutive abnormalities whereas other deactivations in occipital and cingulate regions most likely correspond to non-specific implicit visual and emotional processing.
The left precentral deactivation during overt reading may denote an articulation difficulty. It could be related to fibre tract alterations (Sommer et al., 2002) that suggest an impaired neuronal communication between :his region and left frontal motor execution and temporal regions as a possible cause for a disturbed speech timing in stutterers. The left inferior frontal deactivation detected during semantic decisions might indicate some deficiency in semantic processing. These regions are implicated in supramodal visual and auditory semantic as well as in phonological processing (Heilman et al., 1996; Poldrack et al., 1999). This area activates invariant motor commands that program the articulators, for instance while converting letters into speech sounds. This ability is impaired in dyslexic children, who are unaware of the position of their articulators during speech (Heilman et al, 1996). Comparably, PWS might have difficulties to control the position of their articulators during speech motor programming due to feedback deficits.
An indirect comparison of the activation differences between both tasks suggests that higher activations in PWS than in PWNS were restricted to overt speech and were not evident in the covert semantic task (Neumann, Preibisch, et. al, 2003). Deactivations in PWS were task-specific but insensitive to therapy. The pre-to post-treatment differences between the patterns of overactivation were much more marked in overt reading than in semantic decision, but a shift towards more left- sided activations was common to both tasks. Differing from the view of other authors of a pure speech motor involvement in compensation processes in stuttering (De Nil & Kroll, 200la,b) we therefore conclude that compensation recruits more speech motor processes but also cognitive- linguistic functions, and that enhanced compensation after a successful therapy engages both systems to a larger extent (Neumann, Euler, etal., 2003; Neumann, Preibisch et al., 2003; Preibisch, Neumann, et al., 2003).
We assume that the overactivations reflect compensation mechanisms for a permanent deficit in PWS. The deficit could be indicated by the deactivations, and the permanence by the their temporal stability in spite of therapy. The stable deactivations contrast with comparatively variable, task-specific overactivations after therapy. The dysfunction remains stable while compensation mechanisms may change and might depend on speech pattern and tempo, and severity and idiosynchrasies of stuttering (Neumann, Euler, et. al, 2003). This View differs from that of other authors (De Nil & Kroll, 2001a) who assume that PWS are not inherently deficient in speech motor control regions. The higher activations after therapy may indicate a higher degree of control and compensation, with still insufficient automatization of speech production. Furthermore, an expected therapy-induced increase in activity which should include more left-hemispheric regions was indeed observed and included not only speech, language and auditory areas but also regions usually implicated in complex articulatory (ACC) and higher timing demands (putamen). Our data suggest that overactivations indicate a particular compensation network including a variety of regions which is active already in untreated PWS (Neumann, Euler, et. al, 2003; Preibisch, Neumann, et al., 2003), but is recruited to a higher degree and with more success after fluency shaping. This View is supported by findings that (1) subjects who stutter moderately already pre-treatment show more distributed overactivations than those who stutter severely (Neumann, Euler, et. al, 2003), (2) subjects who stuttered moderately (Neumann, Euler, et. al, 2003) activated the very same regions which were activated in all stutterers after therapy (Neumann, Preibisch, et. al, 2003), irrespective of stuttering severity and (3) both, speech planning and speaking in PWS employ similar neural systems, whereas in PWNS different systems are involved (De Nil & Bosshardt, 2001).
Interestingly, adjacent to the regions of persistent deactivation an increased post-treatment activity was detected in both tasks. Thus, we assume an increased compensation surrounding a dysfunctional region rather than its repair and a take-over of disturbed functions by neighboring regions. The therapy seems to shift the compensation for left hemispheric deactivations from right homologue areas into left hemispheric regions adjacent to the deactivated ones. This suggests that the contralateral homologue of a dysfunctioning region compensates only imperfectly. A more efficient compensation permitting accuracy of timing processes in speech production may require the restoration of a left-sided network.
The rather right-sided frontal cortex overactivation during overt reading and the deactivation in the left precentral cortex in PWS as well as higher pre-than post-treatment activation in the right hemisphere and the more distributed and increasingly left-sided overactivation in frontal regions after therapy agree with earlier neuroimaging findings (De Nil & Kroll, 200la,b, Kroll et al., 1997).
Reduced brain activity in PWS was reported when task familiarity increased (Ingham, 2001) and when PWS spoke more fluently Fox et al., 1996; Ingham et al., 2000). Indeed, we observed two years after therapy, when the new speech pattern had become more automatized or partially got lost, that activation was slightly reduced as compared to immediately after therapy (Neumann, Euler, et. al, 2003). However, a high activation level persists even after two years of practice, indicating a higher attentional and motor control demand for the new speech pattern that possibly can never become completely automatized.
Our data suggest a compensation mechanism for a deficient synchronization of language production steps (Salmelin, et al., 2000; Fox et al., 1996). In particular, the deactivated regions in the left precentral and inferior frontal cortex could be associated With disturbed White matter connections (Sommer et al., 2002) and thus be responsible for a reversed processing sequence between motor preparation and articulatory planning (Salmelin, et al., 2000). Fluency-inducing techniques might thus work as an external ‘clock generator’ and reduce stuttering by providing a pacer that synchronizes the disturbed signal transmission between auditory, speech motor planning, and motor areas. Thus, therapy seems to not eliminate deficiencies but compensate or partially circumvent them by the external pacer which establishes an internal automatisrn. Indeed, post- treatment overactivations conspicuously occur in regions which are involved in timing processes, namely frontal speech motor planning and execution areas, temporal regions, and the basal ganglia.
It seems to be unlikely that the activation changes are attributable to reduced speech-motor demands by the prolonged speech (Packman, Onslow, Richard, & van Doorn, 1996), because PWS did not show reduced but increased cerebral activity while speaking slower immediately after therapy. Moreover, the activation pattern after two years, when the speech rate on average was higher than before therapy, was comparable to the one immediately after therapy (Neumann, Euler, et. al, 2003). Thus, the post-treatment fMRI activation changes are assumed to reflect specific therapy effects, possibly cortical reorganization, rather than the mere consequence of a change in performance.
Boberg, E., Yeudall, L. T., Schopflocher, D., & Bo-Lassen, P. (1983). The effect of an intensive behavioral program on the distribution of EEG alpha power in stutterers during the processing of verbal and visuospatial information. Journal of Fluency Disorders, 8, 245-263.
Braun, A. R., Varga, M., Stager, S., Schulz, G., Selbie, S., Maisog, J. M., et al. (1997). Altered patterns of cerebral activity during speech and language production in developmental stuttering: An H2â_˜5O positron emission tomography study. Brain, 120, 761-784.
De Nil, L. F., & Bosshardt, H. G. (2001). Studying stuttering from a neurological and cognitive information processing perspective. In H. G. Bosshardt, J. S. Yaruss, & H. F. M. Peters (Eds.), Fluency disorders: theory, research, treatment and self-help. (pp. 53-58). Nijmegen, The Netherlands: Nijmegen University Press.
De Nil, L. F., & Kroll, R. M. (200la). Searching for the neural basis of stuttering treatment outcome: recent neuroimaging studies. Clinical Linguistics and Phonetics, 15, 163-168.
De Nil, L. F., & Kroll, R. M. (200lb). Understanding the neural basis of treatment using positron emission tomography. In H. G. Bosshardt, J. S. Yaruss, & H. F. M. Peters (Eds.), Fluency disorders: theory, research, treatment and self-help (pp. 43-46). Nijmegen, The Netherlands: Nijmegen University Press.
De Nil, L. F., Kroll, R. M., & Houle, S. (2001). Functional neuroimaging of cerebellar activation during single word reading and verb generation in stuttering and nonstuttering adults. Neuroscience Letters, 302, 77-80.
De Nil, L. F., Kroll, R. M., Kapur, S., & Houle, S. (2000). A positron emission tomography study of silent and oral single word reading in stut:ering and nonstuttering adults. Journal of Speech, Language, an.d Hearing Research, 43, 1038-1053.
Euler, H. A., & Wolff von Gudenberg, A. (2002). The Kassel Stuttering Therapy: Do follow-up compliance incentives help maintain fluency shaping treatment effects? In M. Gross & E. Kruse (Eds.), Aktuelle phoniatrisch-pÃ©idaudiologische Aspekte 2001/2002 (pp. 107-110). Heidelberg: Median-Verlag von Killisch-Horn.
Foundas, A. L., Bollich, A. M., Corey, D. M., Hurley, M., & Heilmann, K. M. (2001). Anomalous anatomy of speech-language areas in adults with persistent developmental stuttering. Neurology, 57, 207-215.
Fox, P. T., Ingham, R. J, Ingham, J. C., Hirsch, T. B., Downs, J. H., Martin, C., et al. (1996). A PET study of the neural systems of stuttering. Nature, 382, 158-162.
Fox, P.T., Ingham, R. J., Ingham, J. C., Zamarripa, F., Xiong, J. H., & Lancaster, J. L. (2000). Brain correlates of stuttering and syllable production. A PET performance-correlation analysis. Brain, 123, 1985-2004. Section 6. Neurological and Speech Motor Basis of Stuttering 35]
Heilman, K. M., Voeller, K., & Alexander, A. W. (1996). Developmental dyslexia: a motor- articulatory feedback hypothesis. Annals of Neurology, 39, 407-412.
Ingham, R. J. (1984). Stuttering and behavior therapy: Current status and experimental foundations. San Diego, CA: College Hill.
Ingham, R. J. (2001). Brain imaging studies of developmental stuttering. Journal of Communication Disorders, 34, 439-516.
Ingham, R. J., Fox, P. T., Ingham, J. C., & Zamarripa, F. (2000). Is overt stuttered speech a prerequisite for the neural activations associated with chronic developmental stuttering? Brain and Language, 75, 163-194.
Kroll, R. M., De Nil, L. F., Kapur, S., & Houle, S. (1997). A positron emission tomography investigation of post-treatment brain activation in stutterers. In H. F. M. Peters, W. Hulstijn, & P. H. H. M. Van Lieshout (Eds.), Speech production: Motor control, brain research and fluency disorders (pp. 307-320). Amsterdam, The Netherlands: Elsevier.
Maguire, G. A., Riley, G. D., Franklin, D. L., & Gottschalk, L. A. (2000). Risperidone for the treatment of stuttering. Journal of Clinical Psychopharmacology, 20, 479-482.
Moore, W. H. (1984). Hemispheric alpha asymmetries during an electromyographic biofeedback procedure for stuttering. Journal of Fluency Disorders, 1 7, 143-162.
Neumann, K., Euler, H. A., Wolff VO1’1 Gudenberg, A., Giraud, A. L., Lanfermann, H., Gall, V., et al. (2003). The nature and treatment of stuttering as revealed by fMRI. A within- and between- group comparison. Journal of Fluency Disorders, in press.
Neumann, K., Preibisch, C., Euler, H. A., Wolff von Gudenberg, A., Lanfermann, H., et al. (2003). The impact of fluency shaping stuttering therapy on functional neuroanatomy. Manuscript submitted for publication.
Packman, A., Onslow, M., Richard, F., & van Doom, J. (1996). Syllabic stress and variability: a model of stuttering. Clinical Linguistics and Phonetics, 10, 235-263.
Paus, T. (2001). Primate anterior cingulate cortex: where motor control, drive and cognition interface. Nature Reviews. Neuroscience, 2, 417-424.
Perkins, W. H., Kent, R., & Curlee, R. D. (1991). A theory of neuropsycholinguistic function in stuttering. Journal of Speech and Hearing Research, 34, 734-752.
Peters, H. F. M., Hulstijn, W., & van Lieshout, P. H. H. M. (2000). Recent development in speech motor research into stuttering. F olia Phoniatrica et Logopaedica, 52, 103-119.
Poldrack, R. A., Wagner, A. D., Prull, M. W., Desmond, J. E., Glover, G. H., & Gabrieli, J. D. (1999). Functional specialization for semantic and phonological processing in the left inferior prefrontal cortex. Neurolmage, 10, 15-35.
Pool, K. D., Devous, M. D., Freeman, F. J., Watson B. C., & Finitzo T., ( 1991). Regional cerebral blood flow in developmental stutterers. Archives of Neurology, 48, 509-512.
Preibisch, C., Neumann, K., Raab, P., Euler, H. A., Wolff VOI1 Gudenberg, A., Lanfermann, H., et al. (2003). Evidence for compensation for stuttering by the right frontal operculum. Neurolmage, in press.
Preibisch, C., Raab, P., Neumann, K., Euler, H. A., Wolff Von Gudenberg, A., Gall, V. et al. (2003). Event-related fMRI for the suppression of speech-associated artifacts in stuttering. Neurolmage, 19, 1076-1084.
Salmelin, R., Schnitzler, A., Schmitz, F., & Freund, H.-J. (2000). Single word reading in developmental stutterers and fluent speakers. Brain, 123, 1184-1202.
Salmelin, R., Schnitzler, A., Schn1itz,F., J tincke, L., Witte, O. W., & Freund, H.-J. (1998). Functional organization of the auditory cortex is different in stutterers and fluent speakers. Neuroreport, 9, 2225-2229.
Sommer, M., Koch, M. A., Paulus, W., Weiller, C., & Biichel, C. (2002). Disconnection of speech- relevant brain areas in persistent developmental stuttering. The Lancet, 3 60, 380-383.
Travis, L. E. (1978). The cerebral dominance theory of stuttering: 1931-1978. Journal of Speech and Hearing Disorders, 43, 278-281. 352 Theory, research and therapy in fluency disorders
Webster, R. L. (1974). The precision fluency shaping program: speech reconstructions for stutterers. Roanoke VA: Communications development cooperation.
Webster, R. L. (1990). Motor performance of stutterers: A search for mechanisms. Journal of Motor Behaviour, 22, 553-571.
Wildgruber, D., Ackermann, H., & Grodd, W. (2001). Differential contributions of motor cortex, basal ganglia, and cerebellum to speech motor control: effects of syllable repetition rate evaluated by fMRI. Neurolmage, 13, 101-109.
Wu, J. C., Maguire, G., Riley, G., Fallon, J ., LaCasse, L., Chin, S., et al. (1995). Apositron emission tomography [19F]deoxyglucose study of developmental stuttering. NeuroReport, 6, 501-505.