EARLY REPORTS
Early reports
Metabolic abnormalities in developmental dyslexia detected by 1H magnetic resonance spectroscopy Caroline Rae, Martin A Lee, Ruth M Dixon, Andrew M Blamire, Campbell H Thompson, Peter Styles, Joel Talcott, Alexandra J Richardson, John F Stein
Summary Background Neurological and physiological deficits have been reported in the brain in developmental dyslexia. The temporoparietal cortex has been directly implicated in dyslexic dysfunction, and substantial indirect evidence suggests that the cerebellum is also implicated. We wanted to find out whether the neurological and physiological deficits manifested as biochemical changes in the brain. Methods We obtained localised proton magnetic resonance spectra bilaterally from the temporo-parietal cortex and cerebellum of 14 well-defined dyslexic men and 15 control men of similar age. Findings We found biochemical differences between dyslexic men and controls in the left temporo-parietal lobe (ratio of choline-containing compounds [Cho] to N-acetylaspartate [NA] p¶0·01) and right cerebellum (Cho/NA, p¶0·01; creatine [Cre] to NA p¶0·05; (not significant). We found lateral biochemical differences in dyslexic men in both these brain regions (Cho/NA in temporo-parietal lobe, left vs right, p¶0·01; Cre/NA in cerebellum, left vs right, p¶0·001). We found no such lateral differences in controls. There was no significant relation between the degree of contralateral chemical difference and handedness in dyslexic or control men. Interpretation We suggest that the observed differences reflect changes in cell density in the temporo-parietal lobe in developmental dyslexia and that the altered cerebral structural symmetry in dyslexia is associated with abnormal development of cells or intracellular connections or both. The cerebellum is biochemically asymmetric in dyslexic men, indicating altered development of this organ. These differences provide direct evidence of the involvement of the cerebellum in dyslexic dysfunction.
Lancet 1998; 351: 1849–52
Introduction There is now substantial evidence to suggest that developmental dyslexia is a disorder of neurobiological origin. In addition to the well-known deficit in phonological processing,1 dyslexic individuals have altered lateral cerebral symmetry,2 impaired visual3 and auditory processing,4 disordered magnocells,5 and altered patterns of cerebral activation on verbal, visual, and auditory tasks.6–8 The area of the brain most frequently implicated is the temporo-parietal cortex2,9 and, more recently, the cerebellum.10 Magnetic resonance spectroscopy has been used extensively to characterise the biochemical profiles of brain disorders in vivo—most typically in its application to traumatic events in which some gross physical or metabolic insult has occurred (such as head injury or stroke) or in its application to disorders such as epilepsy, multiple sclerosis, inherited metabolic deficits, or mental disorders (such as schizophrenia). In 1996, investigators showed that magnetic resonance spectroscopy is capable of resolving more subtle differences in normal brain—eg, the variation in brain pH with cognitive ability in boys.11 The non-invasive nature of magnetic resonance spectroscopy makes it particularly suited to investigation of a disorder such as developmental dyslexia, when concurrent assessment of the individual’s performance is desirable. To find out whether the neurological and physiological deficits in dyslexia also manifested as biochemical changes in the brain, we obtained localised 1H magnetic resonance spectra bilaterally from the temporo-parietal lobe and cerebellum of 14 well-defined dyslexic men and 15 non-dyslexic control men of similar age. The relative concentrations of choline-containing compounds (Cho; a marker of overall cellular density), creatine-containing compounds (Cre; a marker of cellular energetics) and N-acetylaspartate (NA; a marker of neuronal density) were assessed.
Patients and methods MRC Biochemical and Clinical Magnetic Resonance Unit, John Radcliffe Hospital, Oxford, UK (C Rae PhD, R M Dixon PhD, A M Blamire PhD, C H Thompson MD, P Styles DPhil); Department of Biochemistry, University of Sydney, NSW 2006, Australia (C Rae, C H Thompson); Department of Clinical Neurology, Radcliffe Infirmary, Oxford, UK (M A Lee MRCP); and University Laboratory of Physiology, University of Oxford, Oxford, UK (J Talcott PhD, A J Richardson DPhil, J F Stein MRCP) Correspondence to: Dr Caroline Rae, Department of Biochemistry, University of Sydney, NSW 2006, Australia (e-mail:
[email protected])
THE LANCET • Vol 351 • June 20, 1998
We recruited, with informed consent, 29 adult male volunteers aged 20–41 years. 15 were normal readers and 14 were dyslexic. All the dyslexic individuals had been formally diagnosed by educational psychologists either as adults or as children, most within the previous 2 years. They were identified as dyslexic owing to an unexpected and large discrepancy between reading and spelling achievement and expected achievement based on age-defined and intelligence-defined norms. We confirmed their reading difficulties at the time of testing by measuring their ability to read nonsense words (eg, torlep) which is a sensitive indicator of the phonological-awareness deficits of dyslexic individuals.12 As a group, the dyslexic participants named significantly fewer items on both the Castles and Coltheart tests13
1849
EARLY REPORTS
Magnetic resonance images showing regions from which spectra were obtained Slices from T1-weighted multislice images. A: axial view showing placement of 23333 cm voxels in temporo-parietal lobe. B: sagittal view showing placement of voxel in temporo-parietal lobe. C: axial view showing placement of 33333 cm voxels in cerebellum.
(30 items; dyslexic mean 19·6 [SD 6·3], control mean 27·4 [2·6]; Mann-Whitney U test 13·00, p
