Implicit learning deficits in dyslexic adults: An fMRI study
Introduction
Developmental dyslexia (DD) is defined as a specific reading disability resulting in unexpected, specific, and persistent low reading achievement despite conventional instruction, adequate intelligence and sociocultural opportunity (Shaywitz, 1998). The neuronal correlates of DD were extensively investigated in previous neuroanatomical and neuroimaging studies, and a brain network critically involved in this developmental disorder was found. Particularly, cortical and subcortical brain regions, including frontal, temporal and parietal cortices, as well as insula and cerebellum seem to be significantly affected in individuals with DD, supporting the phonological processing deficit model of DD (Paulesu et al., 1996, Brunswick et al., 1999, Temple et al., 2001). Although this model describes the major aspects of DD, it cannot be generalized on the basis of deficits exhibited in other domains such as visual processing, rapid processing of sensory stimuli and skill automatization. Regarding visual processing functional neuroimaging studies demonstrate visual motion deficits in DD individuals linked to anomalous activation of the magnocellular pathway using low-luminance and low-contrast stimuli (Eden et al., 1996). These magnocellular dysfunctions may extend to other modalities, such as the auditory or tactile systems. For example, some studies showed that DD individuals fail to perceive short and rapidly varying sounds, thus demonstrating a specific impairment in rapid auditory processing (Tallal, 1980, Tallal et al., 1993, van Ingelghem et al., 2001).
Another domain of DD, which has been studied less than phonological processing and magnocellular pathway models, involves specific deficits related to skill automatization. In particular, some studies demonstrated that children with DD show significant impairments in automatization of gross and fine motor skills. While no deficits were found in single-task conditions, marked deficits became evident in dual-task conditions in which a new task was introduced to test the automaticity of the first one (Nicolson and Fawcett, 1990, Fawcett and Nicolson, 1992). Given these condition-specific features the automatization deficit is probably linked to alterations in cerebellar functions, as also suggested by Nicolson et al. (1999) study. In this study neuronal activity in the cerebellum was reduced in DD individuals both during acquisition of a new sequence and during execution of over-learned tasks.
Evidence from both clinical studies and animal models confirm the role of the cerebellum in acquisition and execution of a sequence. In particular, severe impairments in new sequence learning were found in patients with diffuse cortical cerebellar degeneration (Pascual-Leone et al., 1993) and focal cerebellar lesions (Molinari et al., 1997, Gomez-Beldarrain et al., 1998). These results were corroborated in animal studies. Specifically, Nixon and Passingham (2000) reported acquisition deficits in a sequence in monkeys with cerebellar nuclear lesions. Furthermore, it was demonstrated in rats that bilateral extensive lesions of the cerebellar dentate nucleus produce deficiencies in the acquisition of egocentric-based motor sequences (Gaytan-Tocaven and Olvera-Cortes, 2004).
Correct skill automatization may take place under entirely implicit conditions (Seidler et al., 2002) or may result from repetitive trial-and-error training, as also found in normal readers (Jenkins et al., 1994, Jueptner et al., 1997a, Jueptner et al., 1997b). Both behavioral and neuroimaging methods have been used to study the latter condition resulting from repetitive trial-and-error training (Nicolson and Fawcett, 1990, Fawcett and Nicolson, 1992, Nicolson et al., 1999). Recent studies demonstrated implicit learning deficits in DD using the serial reaction time task (SRTT) initially developed by Nissen and Bullemer (1987). The SRTT results indicate a lack of sequential learning in DD subjects who display similar responses in randomized and sequenced blocks (Vicari et al., 2003, Vicari et al., 2005, Howard et al., 2006).
Given this implicit learning impairment in DD subjects, the main objective of the present study was to explore neuronal activations during a classical version of the SRTT learning task in DD adults compared to normal readers.
Section snippets
Subjects
Fourteen adults with developmental dyslexia (DD, mean age = 42.1, range 34–55 years; 10 females and 4 males) and 14 normal readers (C, mean age = 37.2, range 28–47 years; 10 females and 4 males) matched for chronological age and sex were selected for the present study. Participants' educational level was also similar in the two groups: 9 dyslexic and 7 controls had a secondary school diploma, and 5 dyslexic and 7 controls an university degree.
A clinical psychologist diagnosed the DD people
Behavioral measures
Average values for the median reaction times obtained by the two groups in the 7 blocks of the SRTT are shown in Fig. 1.
The presence of implicit learning was verified by analyzing the performances of both groups throughout the experiment (R1–R2). While the C group performance was modulated throughout blocks (R1–R2) (one-way ANOVA: F(6,78) = 3.1, P < 0.01), the DD subjects' response pattern was not modulated by block presentation order (one-way ANOVA: F(6,78) = 1.1, P = 0.4).
A post hoc analysis (Tukey's
Discussion
The present findings document the absence of implicit sequence learning in DD adults, thus confirming previous behavioral findings obtained in children and young people affected by DD (Vicari et al., 2003, Vicari et al., 2005, Howard et al., 2006). In particular, the contrast between the last sequenced block (S5) and the last random block (R2) confirmed the impairment of implicit sequence learning in the DD group. While C subjects showed marked implicit learning that modulated their responses
References (82)
- et al.
Role of the cerebellum in spatial orientation in the rat
Behav. Neural. Biol.
(1992) - et al.
Neuroimaging studies of the cerebellum: language, learning and memory
Trends Cogn. Sci.
(1998) - et al.
Distinct contribution of the cortico-striatal and cortico-cerebellar systems to motor skill learning
Neuropsychologia
(2003) - et al.
Analysis of fMRI time-series revisited
NeuroImage
(1995) - et al.
Classical and Bayesian inference in neuroimaging: applications
NeuroImage
(2002) - et al.
Classical and Bayesian inference in neuroimaging: theory
NeuroImage
(2002) - et al.
Bilateral lesion of the cerebellar-dentate nucleus impairs egocentric sequential learning but not egocentric navigation in the rat
Neurobiol. Learn. Mem.
(2004) - et al.
Dyslexics are impaired on implicit higher-order sequence learning, but not on implicit spatial context learning
Neuropsychologia
(2006) - et al.
Anatomical risk factors that distinguish dyslexia from SLI predict reading skill in normal children
J. Commun. Disord.
(2002) - et al.
Automaticity: a new framework for dyslexia research?
Cognition
(1990)