Preseason Neurocognitive Testing 1
Preseason Neurocognitive Testing in Athletes with ADHD
Grant L. Iverson
University of British Columbia
British Columbia Mental Health & Addiction Services
Michael W. Collins & Mark R. Lovell
University of Pittsburgh Medical Center
Acknowledgements: This study was presented at the National Association of School Psychologists, Dallas, Texas, April 1, 2004. The authors thank Carrie Strangway for her comments and assistance with an earlier draft of this manuscript, and Jennifer Bernardo for assistance with manuscript preparation.
Competing Interests: Drs. Lovell and Collins have a proprietary interest in ImPACT.
Correspondences: Please address correspondence to Grant Iverson, Ph.D., Department of Psychiatry, 2255 Wesbrook Mall, University of British Columbia, Vancouver, B.C. Canada, V6T 2A1. Phone: 822-7588; Fax: 822-7756; Email: .
Abstract
Neurocognitive testing is widely cited and recommended as a component of a comprehensive concussion management program in athletics. Baseline testing is especially important for athletes who might have a developmental condition, such as a learning disability or attention-deficit hyperactivity disorder (ADHD). This is because normative data for neurocognitive tests, as a rule, are not available for youth with these conditions. The purpose of this study was to examine the effects of ADHD on the baseline preseason test performance of student athletes on a computerized neuropsychological screening battery (Immediate Postconcussion Assessment and Cognitive Testing; ImPACT). Participants were 38 adolescent athletes with ADHD and 38 control subjects matched for age, education, gender, and history of concussion. Student athletes with ADHD performed more poorly on visual memory (p < .006, Cohen’s d = .65, medium effect size), processing speed (p < .004, d = .69), and right-left orientation and inhibition (p < .003, d = .93, large effect size). These results support the recommendation of baseline testing for athletes, especially those with ADHD.
Key Words: ADHD, Concussion, Neurocognitive Testing
Introduction
Neurocognitive testing is widely cited and recommended as a component of a comprehensive concussion management program in athletics (Aubry et al., 2002; Collins & Hawn, 2002; Guskiewicz et al., 2004; McCrory et al., 2005; Putukian, 2006; Randolph, 2001). This, of course, is because neurocognitive testing is sensitive to the acute effects of concussion and can be used to serially monitor an athlete’s recovery (Barr & McCrea, 2001; Broglio, Macciocchi, & Ferrara, 2007; Collins et al., 1999; Delaney, Lacroix, Gagne, & Antoniou, 2001; Echemendia, Putukian, Mackin, Julian, & Shoss, 2001; Erlanger et al., 2003; Erlanger et al., 2001; Guskiewicz, Ross, & Marshall, 2001; Iverson, Brooks, Collins, & Lovell, 2006; Macciocchi, Barth, Alves, Rimel, & Jane, 1996; Makdissi et al., 2001; Matser, Kessels, Lezak, & Troost, 2001; McClincy, Lovell, Pardini, Collins, & Spore, 2006; McCrea, Kelly, Randolph, Cisler, & Berger, 2002; Schatz, Pardini, Lovell, Collins, & Podell, 2006; Van Kampen, Lovell, Pardini, Collins, & Fu, 2006; Warden et al., 2001). Baseline, preseason neurocognitive testing has become increasingly feasible given the availability of brief computerized batteries (Cernich, Reeves, Sun, & Bleiberg, 2007; Collie, Makdissi, Maruff, Bennell, & McCrory, 2006; Covassin et al., 2006; Erlanger et al., 2003). Regardless of whether computerized or traditional neurocognitive tests are used, preseason testing is frequently recommended as a clinical methodology in which the athlete serves as his or her own comparison if injured in the future (Collins et al., 1999; Echemendia et al., 2001; Lovell, Collins, Iverson, Johnston, & Bradley, 2004; McCrea et al., 2003; Moser, Schatz, & Jordan, 2005). Following a concussion, an athlete can be monitored to determine when he or she recovers to baseline levels of functioning.
Baseline testing is especially important for athletes who might have a developmental condition, such as a learning disability or attention-deficit hyperactivity disorder (ADHD). This is because normative data for neurocognitive tests, as a rule, are not available for youth with these conditions. In fact, having ADHD or a learning disability typically is an exclusion criterion for being included in a normative sample. Thus, if ADHD or a learning disability has an adverse effect on a particular neurocognitive test, then the post-concussion results from that test in an athlete with a pre-existing condition will be particularly difficult to interpret.
The neuropsychological problems associated with ADHD in children and adolescents have been well documented, and typically are characterized as core deficits in attention and executive functioning (e.g., Barkley, 2000; Halperin et al., 1990; Konrad, Gauggel, Manz, & Scholl, 2000; Kupietz, 1990; Loge, Staton, & Beatty, 1990; Seidel & Joschko, 1990; van der Meere & Sergeant, 1988). Those youth with comorbid learning disabilities often have more pronounced neuropsychological deficits (Korkman & Pesonen, 1994; Purvis & Tannock, 1997; Seidman et al., 1995; Seidman, Biederman, Monuteaux, Doyle, & Faraone, 2001). The purpose of this study was to determine if student athletes with self-reported ADHD perform more poorly on a commonly used computerized battery called Immediate Postconcussion Assessment and Cognitive Testing (ImPACT). It was hypothesized that athletes with ADHD would perform more poorly on at least one of the five composite scores derived from ImPACT.
Method
Participants
Thirty-eight adolescent athletes with a self-reported diagnosis of ADHD were compared to 38 carefully matched adolescent athletes with no self-reported ADHD, learning, or speech-related problems. All participants were selected from the normative sample for ImPACT (Immediate Post-Concussion Assessment and Cognitive Testing; N = 555). The students with suspected ADHD were not diagnosed through structured interviewing or testing; a psychologist or psychiatrist did not evaluate them. This is a sample of convenience, derived from a normative database. The two groups were matched precisely on age, education, gender, and number of previous concussions (self-reported number of lifetime concussions is collected as part of the health history section of ImPACT). All participants in this study were student athletes undergoing preseason testing. The average age of the students was 15.5 years (Range = 13-19) and their average number of completed years of education was 9.1 (all were in grades 8-12). The majority of the participants were boys (92%). Approximately half of the subjects with ADHD reported that they had repeated a grade in school, received special education services, and/or had learning problems (58%). The specific breakdown for each of these was as follows: repeated a grade = 21%, special education services = 24%; and learning problems = 29%. None of the control subjects reported educational problems. Most of the student athletes were football players (87% of each group). The number of previous concussions was not medically verified. These were self-reported past concussions. The breakdown for the control group was as follows: none = 76.3%, 1 = 15.8%, 2 = 5.3%, and 3 or more = 2.6%. The breakdown of past concussions for the ADHD group was none = 73.7%, 1 = 15.8%, 2 = 5.8%, and 3 or more = 5.2%. Based on the literature, we did not anticipate a need to match on number of previous concussions because 1-2 past sport-related concussions are not believed to have a lingering effect (e.g., Iverson, Brooks, Lovell, & Collins, 2006). However, because this information was collected, it was decided to use it as a matching variable to have a methodologically-tighter study.
Measure
Version 2.0 of ImPACT is a brief (20-25 minutes) computer-administered neuropsychological test battery that consists of six individual test modules that measure aspects of neurocognitive functioning including attention, memory, reaction time, and processing speed. Each test module may contribute scores to multiple composite scores. Five composite scores were used for this study. The Verbal Memory composite score represents the average percent correct for a word recognition paradigm, a symbol number match task, and a letter memory task with an accompanying interference task. These tests are conceptually similar to traditional verbal learning (word list) tasks and the auditory consonant trigrams test (i.e., the Brown-Peterson short-term memory paradigm), although the information is presented visually on the computer, not auditorily by an examiner. The Visual Memory composite score is comprised of the average percent correct scores for two tasks; a recognition memory task that requires the discrimination of a series of abstract line drawings, and a memory task that requires the identification of a series of illuminated X’s or O’s after an intervening task (mouse clicking a number sequence from 25 to 1). The first test taps immediate and delayed memory for visual designs and the second test measures short-term spatial memory (with an interference task). The Reaction Time composite score represents the average response time (in milliseconds) on a choice reaction time, a go/no-go task, and the previously mentioned symbol match task (which is similar to a traditional digit symbol task). The Processing Speed composite represents the weighted average of three tasks that are done as interference tasks for the memory paradigms. The Impulse Control composite score represents the total number of errors of omission or commission on the go/no-go test and the choice reaction time test.
In addition to the cognitive measures, ImPACT also contains a Post-Concussion Scale that consists of 22 commonly reported symptoms (e.g., headache, dizziness, “fogginess”). The dependent measure is the total score derived from this 22-item scale. The reliability (Iverson, Lovell, & Collins, 2003; Iverson, Lovell, Collins, & Norwig, 2002) and concurrent validity (Iverson, Franzen, Lovell, & Collins, 2004; Iverson, Lovell, & Collins, 2005) of the cognitive composite scores and the Post-Concussion Scale (Iverson & Gaetz, 2004; Janusz, Gioia, Gilstein, & Iverson, 2004; Lovell et al., 2006), and the sensitivity of the battery to the acute effects of concussion (Broglio et al., 2007; Collins et al., 2003; Iverson, Gaetz, Lovell, & Collins, 2002; Iverson et al., 2003; Lovell et al., 2003; Lovell et al., 2004; McClincy et al., 2006; Schatz et al., 2006; Van Kampen et al., 2006), have been examined in a number of studies.
Results
The two groups were compared on the five neuropsychological composite scores using mulivariate analysis of variance (MANOVA) followed by univariate ANOVAs. Box’s M Test was significant (p < .001), indicating that the covariance matrices differed across the dependent variables, thus violating an assumption of MANOVA. Moreover, some of the individual variables within groups had significant departures from normality as assessed by the Kolmogorov Smirnoff procedure, and one variable had unequal error variances between groups (based on Levene’s Test). MANOVA and ANOVA tend to be quite robust to these violations of underlying general linear model assumptions. Therefore, the results will be reported. There was a significant multivariate effect [Wilks’ Lambda = .71; F (5, 70) = 5.7, p < .001, eta squared = .29]. The univariate ANOVA results revealed significantly worse neuropsychological test scores for students with ADHD on the Visual Memory, Processing Speed, and Impulse Control composites. The groups did not differ on the Verbal Memory or Reaction Time composites. The students with ADHD reported more subjective symptoms on the Post-Concussion Scale. The effect sizes for the significant differences were medium to large (See Table 1). Given the violations of assumptions of the general linear model, nonparametric pairwise analyses for the five neuropsychological composite scores and the total score for the Post-Concussion Scale were conducted using Mann Whitney U tests. The results of the nonparametric analyses were identical to the parametric analyses.
_____________________
Insert Table 1 About Here
_____________________
Discussion
To our knowledge, this is the first published study examining ImPACT performance in student athletes with ADHD. The results of this study are largely consistent with the neuropsychological theories and empirical studies on ADHD in adolescents and adults. Athletes with self-reported histories of ADHD performed more poorly on computerized tests of concentration and memory, processing speed, inhibition, and in total subjective symptoms. The results of this study can be compared to a meta-analysis of neuropsychological studies in children with ADHD. Effect sizes on specific tests such as Digit Span (.52 to .76), Trails A (.26 to .54), Trails B (.46 to .72), Stroop Interference (.46 to .66), Digit Symbol Coding (.72 to .92), and WCST Categories completed (.19 to .39) were mostly in the medium to large range (Frazier, Demaree, & Youngstrom, 2004). The ImPACT tests rely heavily on attention, working memory, and speed of processing. These core abilities partially underlie each composite score. The effect sizes for ImPACT (i.e., .65 to .93) were comparable or larger than the average effect sizes reported in the ADHD literature.
This study has a number of limitations. The primary limitation is the self-report method of identifying young people with ADHD. Using a single diagnostic indicator such as a self-report with no cross-validation of history or symptoms from multiple settings or informants is not ideal. However, given the lack of a gold standard method for diagnosing ADHD many researchers have been forced to rely on an adolescent’s or adult’s self-report of ADHD symptoms in a clinical interview or on an individual’s report of a prior diagnosis (e.g., Epstein, Conners, Sitarenios, & Erhardt, 1998; Epstein, Johnson, Indira, & Conners, 2001; Murphy, 2002a, 2002b; Nigg, Butler, Huang-Pollock, & Henderson, 2002). In the present study, this problem is mitigated, to a certain degree, because this sample ranged in age from 13-19. There is a strong likelihood that the present sample received a diagnosis of ADHD within the few years prior to the study.
Because this was a sample of convenience, derived from a normative database, the individuals with ADHD were not separated into DSM-IV-TR subtypes. This is a common limitation in previous studies (e.g., Epstein et al., 2001; Kovner et al., 1998; Nigg et al., 2002; Walker, Shores, Trollor, Lee, & Sachdev, 2000), and most research to date has not consistently identified different cognitive profiles among the ADHD subtypes (e.g., Barkley, Grodzinsky, & DuPaul, 1992; Carlson, Lahey, & Neeper, 1986; Epstein et al., 1998; Murphy, Barkley, & Bush, 2001; Trommer, Hoeppner, Lorber, & Armstrong, 1988). A final limitation relates to stimulant medication. In many previous ADHD studies, participants were not taking or were taken off stimulant medications prior to testing, or the studies implemented statistical procedures to control for the possible cognitive effects of medication. In the present study, the effects of medication could not be examined because medication information was not available in the database. Despite these methodological limitations, relatively large differences were found between groups. The obvious limitations were likely offset by the methodological strengths of the study, especially the relatively narrow age band of subjects (all in grades 8 - 12) and the precise matching on age, education, gender, and number of previous concussions.