Functional abnormalities in symptomatic concussed athletes: an fMRI study J.-K. Chen, a K.M. Johnston, b S. Frey, a M. Petrides, a K. Worsley, a,c and A. Ptito a, * a Montreal Neurological Institute, McGill University, Montreal, PQ, Canada b Department of Neurosurgery, Kinesiology and Physical Education, McGill University, Montreal, PQ, Canada c Department of Mathematics and Statistics, McGill University, Montreal, PQ, Canada Received 24 April 2003; revised 23 October 2003; accepted 30 December 2003 Our aim was to quantify with functional magnetic resonance imaging (fMRI) changes in brain activity in concussed athletes and compare the results with those of normal control subjects. Regional brain activations associated with a working memory task were obtained from a group of concussed athletes (15 symptomatic, 1 asymptomatic) and eight matched control subjects, using blood oxygen level dependent (BOLD) fMRI. The average percent signal change from baseline to working memory condition in each region of interest was computed. Symptomatic concussed athletes demonstrated task-related activations in some but not all the regions of interest, even when they performed as well as the control subjects. Furthermore, several concussed athletes had additional increases in activity outside the regions of interest, not seen in the control group. Quantitative analysis of BOLD signals within regions of interest revealed that, in general, concussed athletes had different BOLD responses compared to the control subjects. The task- related activation pattern of the one symptom-free athlete was comparable to that of the control group. We also repeated the study in one athlete whose symptoms had resolved. On the first study, when he was still symptomatic, less task-related activations were observed. On follow-up, once his symptoms had disappeared, the task-related activations became comparable to those of the control group. These results demonstrate the potential of fMRI, in conjunction with the working memory task, to identify an underlying pathology in symptomatic concussed individuals with normal structural imaging results. D 2004 Elsevier Inc. All rights reserved. Keywords: Athlete; Concussion; fMRI Introduction Most studies agree that, of all cases of head trauma, approxi- mately 80% to 90% fall within the category of mild head injury (MHI) or concussion (Kraus and Nourjah, 1988; Nell and Brown, 1991; Vazquez-Barquero et al., 1992). Concussion used to be considered as a temporary fluctuation in consciousness without enduring effects on cognition. As such, no long-term sequelae were believed to ensue. Recent cases of athletes and individuals with persistent cognitive deficits following concussion (the post-con- cussion syndrome) suggest that this is not always the case. Despite being considered as ‘mildly’ injured, individuals who sustain a concussion often exhibit diverse symptomatology and abnormal neuropsychological profiles, such as deficits in working memory, attention, information processing speed, and more generally, in executive function, known to be linked to the frontal lobe (Bohnen et al., 1992; Collins et al., 1999; Hinton-Bayre et al., 1997; Leininger et al., 1990; Matser et al., 1999; Newcombe and Briggs, 1994). Although the exact pathophysiological changes after concus- sion are not known, there is now increased acceptance of the idea that concussion results mainly in functional disturbance rather than structural damage. This view is reflected in a more compre- hensive definition of concussion that was recently proposed by the Concussion in Sport Group (Aubry et al., 2002). According to this updated definition, concussion may be caused either by a direct blow to the head or elsewhere on the body with an ‘impulsive’ force transmitted to the head. Such injury typically results in the rapid onset of short-lived impairment of neurological function that usually resolves spontaneously. Importantly, al- though concussion may result in neuropathological changes, it is typically associated with grossly normal structural neuroimag- ing examinations and the acute clinical symptoms largely reflect a functional deficit rather than a structural injury. Thus, functional neuroimaging techniques that examine the metabolic/physiologi- cal state of the brain may have great potential for demonstrating brain abnormalities that may be undetectable by morphological imaging methods. Existing functional imaging studies of individuals with head trauma are mainly based on resting metabolic measurements using positron emission tomography (PET) and single-photon emission computerized tomography (SPECT). Data acquired from these studies have shown perfusion deficits that extend beyond any structural damage shown by CT or MRI (Abdel-Dayem et al., 1987; Alavi et al., 1987; Jansen et al., 1996; Langfitt et al., 1986; Newton et al., 1992). The most commonly reported finding is frontal hypometabolism or a decrease in frontal cerebral blood flow. One major disadvantage of these imaging techniques, however, is the requirement of a radioactive tracer, which greatly 1053-8119/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.neuroimage.2003.12.032 * Corresponding author. Neuropsychology Department/Cognitive Neu- roscience Unit, Montreal Neurological Institute, 3801 University Street, Montreal, Quebec, Canada H3A 2B4. Fax: +1-514-398-1338. Available online on ScienceDirect (www.sciencedirect.com.) www.elsevier.com/locate/ynimg NeuroImage 22 (2004) 68 – 82