Abstract

Attention deficit/hyperactivity disorder (ADHD) is posited to be the result of a deficit in executive functions (EFs). The presence of EF deficits in adults with ADHD is not consistent, and EF deficits are not unique to ADHD, thus adding a level of complexity to differential diagnoses. The current study used three measures of EF to discriminate between college-level adults diagnosed with ADHD and reading disability (RD). The RD group performed below ADHD on all EF tasks and logistic regression analyses demonstrated poor sensitivity and adequate specificity of the EF measures to categorize the clinical groups. Results suggest that clinicians should be cognizant of the limitations of measures of EF in the differential diagnosis of ADHD.

Attention deficit/hyperactivity disorder (ADHD) is marked by the three primary symptom domains of inattention, hyperactivity, and/or impulsivity. It emerges in childhood and has been shown to persist into adulthood (Faraone et al., 2000). There have been a variety of theories over the years regarding the core cognitive deficit(s) that characterize ADHD (Barkley, 1997; Brown, 2000, 2005; Douglas, 1983; Quay, 1988a, 1988b).

Barkley (1997) proposed a unified theory of ADHD that identified executive dysfunction (specifically, deficient inhibitory control) as the primary source of impairment that leads to secondary neuropsychological deficits and tertiary behavioral dysregulation. Executive function (EF) deficits have been documented in children and adolescents with ADHD in numerous cross-sectional and longitudinal studies, and in referred and non-referred samples. A recent meta-analysis included 83 studies comparing children and adolescents with ADHD to controls and found medium effect sizes (0.46–0.69) across EF measures (Willcutt, Doyle, Nigg, Faraone, & Pennington, 2005).

Compared with the literature in child samples, research on EF deficits in adult samples is more limited. The majority of existing studies support significant differences between adults with ADHD and controls on EF measures. A recent meta-analysis (Boonstra, Oosterlaan, Sergeant, & Buitelaar, 2005) found medium effect sizes (0.62–0.89) for measures of EF across 13 studies comparing adults with ADHD with controls. In contrast, another recent meta-analysis that explored a broader range of cognitive domains did not report significant effect sizes for EF measures (Schoechlin & Engel, 2005). However, this study did not classify the “… more basic aspects of executive functions such as working memory and inhibition…” as “executive,” restricting this domain to measures of higher level reasoning and problem solving (i.e., Wisconsin Card Sorting Test [WCST], Tower of Hanoi). The authors concede that if they “… had included all measures that are discussed as measures of executive functioning, effect size would have reached a much higher value” (p. 740).

Barkley, Murphy, and Fischer (2008) also noted that the WCST has consistently demonstrated null results when comparing the performance of individuals with ADHD and non-ADHD controls. Barkley and colleagues conclude that set switching is not negatively impacted by a diagnosis of ADHD, although the WCST may tap other higher level problem-solving skills in addition to set switching. Although the WCST has been shown consistently in the literature not to differ significantly between those diagnosed with ADHD and controls, Barkley and colleagues reported that inhibition, interference control (or resistance to distraction), verbal working memory and fluency, among other neuropsychological abilities, are negatively affected in persons with ADHD. Other researchers have also demonstrated that those diagnosed with ADHD perform more poorly than controls on neuropsychological measures of response inhibition planning and set shifting (e.g., Riccio, Wolfe, Romine, Davis, & Sullivan, 2004; Stavro, Ettenhofer, & Nigg, 2007; Young, Morris, Toone, & Tyson, 2007), as well as impulsivity (Malloy-Diniz, Fuentes, Borges Leite, Correa, & Bechara, 2007). A study by Carr, Nigg, and Henderson (2006) suggests that these findings may be due to poor motoric inhibition, rather than attentional inhibition.

Additional studies have assessed the predictive utility of one or more EF measures in discriminating adults with ADHD from control groups. Schreiber, Javorsky, Robinson and Stern (1999) found that measures from the Boston Quantitative Scoring System for the Rey-Osterrieth Complex Figure (BQSS; Configural Accuracy, Planning, Perseveration and Neatness) differentiated between adults with ADHD and without, with 81% specificity and 75% sensitivity. Lovejoy and colleagues (1999) reported specificity rates ranging from 92% to 100% for measures of verbal fluency (VF; Controlled Oral Word Association; COWA), verbal list learning (California Verbal Learning Test; CVLT), inhibition and impulsivity (Stroop Neuropsychological Screening Test; SNST), and sequencing, shifting set, and working memory (Trails A and B). Estimates of sensitivity were much lower, however, with the COWA being the only measure to rise above the 50% mark. Woods, Lovejoy, Stutts, Ball, and Fals-Stewart (2002) reanalyzed the data from Lovejoy and colleagues' (1999) study, but employed a discrepancy approach. They calculated the discrepancy between each measure and the participant's intelligence score, as measured by the WAIS-R, and evaluated the diagnostic classification rates for those scores. They demonstrated variable discriminative abilities with somewhat better specificity (56%–100%) than sensitivity (56%–81%). In short, a number of EF measures have been utilized to discriminate those with ADHD from control samples, with generally better specificity than sensitivity.

The majority of extant studies compare ADHD with controls, but very few studies have attempted to discriminate ADHD from other clinical samples. Such studies are important because many of the EF difficulties associated with ADHD may be associated with other diagnoses. Furthermore, significant comorbidity exists between ADHD and several other common disorders (e.g., anxiety disorders, mood disorders, learning disabilities) (Biederman, Monuteaux et al., 2006; Biederman, Petty et al., 2006). Measures capable of discriminating between ADHD and other clinical disorders will significantly aid in the type of decision-making that is required in clinical settings (i.e., differential diagnosis).

A meta-analysis conducted by Homack and Riccio (2004) investigated EF differences—specifically, performance on the Stroop test—between children and adolescents diagnosed with ADHD and other clinical groups. Discrimination accuracy differed depending on the clinical group being compared with the ADHD sample. Specificity was greater when discriminating youth with ADHD from those with emotional and behavioral disturbances and weaker when attempting to discriminate ADHD from other developmental disabilities (i.e., learning disability, Autism, or Tourette syndrome).

Katz, Woods, Goldstein, Auchenbach, and Geckle (1998) investigated the discriminative power of several measures believed to tap into EF skills in a sample of adults with ADHD and depression. They included variables from the CVLT, the Paced Auditory Serial Addition Test (PASAT), and the Stroop Test and found that while almost the entire ADHD group was correctly classified, much of the depression group was misclassified as having ADHD. In other words, the measures were highly sensitive to ADHD (sensitivity = 96.6%), but not specific (specificity = 40%). Taylor and Miller (1997) also found poor predictive utility for EF measures when comparing a psychiatric diagnosis group to a comorbid ADHD + psychiatric diagnosis group. These studies suggest that deficits on EF measures may not be unique to ADHD and may characterize other psychiatric disorders as well.

Of particular interest in both clinical and educational settings is the differential diagnosis of ADHD and reading disability (RD). It is unclear whether EF measures may be useful in the differential diagnosis of ADHD and RD. Some studies have indicated that EF deficits are associated with ADHD but not RD in children (Hall, Halperin, Schwartz, & Newcorn, 1997; Pennington, Groisser, & Walsh, 1993), while others have suggested that RD is associated with EF deficits (e.g., children: Levin, 1990; adults: Weyandt, Rice, Linterman, Mitzlaff, & Emert, 1998). A recent meta-analysis indicated that RD is associated with executive dysfunction in children, but the effect sizes varied across specific measures and were moderated by task modality (Booth, Boyle, & Kelly, 2010). Tasks with language components evidenced a greater effect size. Nonverbal tasks also exhibited significant effect sizes, but EF deficits (i.e., effect sizes) were more pronounced when language demands were increased. Thus, impairments in the non-EF components of EF measures, such as language, may account for or contribute to the observed deficits in RD.

Measures of EF necessarily include non-EF processes (e.g., language, spatial perception, memory, and motor speed) and therefore may be sensitive to deficits in component or “ingredient” (Wodka et al., 2008a, 2008b) non-EF processes, complicating differential diagnosis with learning disabilities. For example, poor performance on the Trail Making (TM) test, for which one is required to connect letters and numbers in sequence while shifting between sets, might be attributable to an EF deficit in cognitive flexibility or to a deficit in specific language processes (e.g., verbal automaticity) that are impaired in RD.

Previous studies have addressed the potential contribution of these “ingredient” processes in various ways. Marzocchi and colleagues (2008) attempted to control for non-EF processes in children with ADHD or RD by including measures that tap into more “basic” processes as covariates when comparing these groups on EF measures. However, this study often used completely different measures to account for the non-EF processes. For example, a visual short-term memory test, the Benton Visual Retention Test, was included as a covariate to account for the basic, non-EF aspect that is one component skill in a visual working memory task, the Self-Ordered Pointing Task. From the same laboratory, Boonstra, Kooij, Oosterlaan, Sergeant, and Buitelaar (2010) repeated the same study design in adults and reported that adults with ADHD exhibited significantly more difficulties with inhibition and set shifting, after controlling for non-EF measures and IQ. Three possible problems with this method are that (i), in some cases, the non-EF covariate chosen was not directly comparable with the chosen EF measure on multiple dimensions, (ii) the non-EF and EF measures used different norms, and (iii) controlling for IQ is problematic, as it reflects an attribute of the disorder and is not unrelated to the independent variable (Dennis et al., 2009).

Another approach to controlling for basic processes could be facilitated by the Delis–Kaplan Executive Function System (D-KEFS; Delis, Kaplan, & Kramer, 2001). This test battery takes a process approach to classic neuropsychological measures that purport to measure EF and includes contrast tasks that attempt to tease apart EF from non-EF abilities. Contrast tasks are subtests that measure one aspect or component of the higher level EF task and are utilized to account for the impact of the more fundamental or “ingredient” skill (e.g., attention, perception, language). For example, the TM test in the D-KEFS includes subtests evaluating an examinee's ability to scan, rapidly connect letters or numbers in sequence, and their motor speed. Once these underlying skills are accounted for, the examinee's true ability to switch sets can be assessed more reliably. Contrast scores can be generated by computing the scaled score difference between component and EF tasks. This scaled score difference is then plotted on a normal distribution. Thus, the D-KEFS may be well suited to isolate EF skills while removing the impact of “ingredient” processes, such as language.

One study to date has investigated performance on the D-KEFS in an ADHD sample. Wodka and colleagues (2008b) found that children with ADHD performed significantly below controls on all Color-Word Interference (CWI) subtests and on the Tower test (Total Achievement), but found no group differences on the TM or VF tests. On the CWI test, the ADHD group performed below the control group on both the component process subtests (e.g., Color Naming, Word Reading), and on both more complex EF subtests (i.e., Inhibition, Inhibition Switching). This may be because the component subtests include executive components, or because ADHD is also associated with additional non-EF cognitive deficits. Contrast scores were also compared between groups, with no differences noted.

D-KEFS contrast scores have a restricted range and may not be optimal for discriminating clinical groups. Furthermore, Crawford, Sutherland, and Garthwaite (2008) have demonstrated that the reliability of these scores is very low and the use of such scores in diagnostic decision-making is discouraged. An alternate approach to isolate pure EF abilities while controlling for component processes is to use hierarchical logistic regression. Component measures may be entered as potential predictors of group membership in earlier blocks to control for their impact, followed by more complex measures thought to measure EF. In this way, it is possible to statistically control for performance on directly related, co-normed measures of fundamental abilities. The ability to more precisely isolate EF abilities may enhance classification accuracy in discriminating developmental disabilities that have varying profiles of cognitive deficits.

The current investigation explored whether performance on select D-KEFS subtests can discriminate between adults with ADHD and RD. Hierarchical logistic regression was used to control for non-EF components of EF tests and assess the ability of “purer” EF measures to discriminate between these two groups. Three subtests from the D-KEFS were included: TM, VF, and CWI. We hypothesized that these three measures of EF would discriminate ADHD from RD, that the ADHD group would be more accurately classified than the RD group, and that adults with ADHD would perform more poorly on the EF measures (after the “ingredient” component measures were covaried).

Methods

Participants

This was an archival study that was approved by the institutional review board. Two hundred twenty-six college students, representing fourteen 2- and 4-year institutions, constituted the initial sample. All potential participants had been referred to an on-campus assessment center for evaluation of academic difficulties and had completed a comprehensive neuropsychological battery that included selected subtests of the D-KEFS.

Sixty-four participants, who met criteria for either ADHD or RD, were selected from this larger sample. Participants with comorbid ADHD and RD were not included (n = 17). The study sample was older than the initial sample but comparable on other demographic variables, but otherwise did not significantly differ from the initial sample in terms of sex, ethnicity, and comorbid mood or anxiety diagnoses. The ADHD group (n = 26) had a mean age of 24.08 (SD = 4.98), a mean Full Scale IQ (FSIQ) of 112.57 (SD = 11.86), was 23% women and 88.46% Caucasian. The RD group (n = 38) had a mean age of 20.29 (SD = 2.61), a mean Wechsler Full Scale IQ (FSIQ) of 100.69 (SD = 11.82), was 50% women and 68.42% Caucasian (Table 1).

Table 1.

Demographic characteristics of ADHD and RD groups

 RD (%), n = 38 ADHD (%), n = 26 χ2 p 
Age 20.56 (0.519)a 24.08 (4.98)a −3.969b <.005 
FSIQ 100.69 (11.82)a 112.57 (11.86)a −3.757b <.005 
Women (%) 50 23.10 4.701 .03 
Ethnicityc (%)   3.455 .063 
 European American 68.42 88.46   
 African American 23.68 3.85   
 Asian   
 Hispanic 2.63 3.85   
 Other   
Anxiety disorder 15.38 6.236 .013 
Mood disorder 10.53 19.23 0.97 .325 
 RD (%), n = 38 ADHD (%), n = 26 χ2 p 
Age 20.56 (0.519)a 24.08 (4.98)a −3.969b <.005 
FSIQ 100.69 (11.82)a 112.57 (11.86)a −3.757b <.005 
Women (%) 50 23.10 4.701 .03 
Ethnicityc (%)   3.455 .063 
 European American 68.42 88.46   
 African American 23.68 3.85   
 Asian   
 Hispanic 2.63 3.85   
 Other   
Anxiety disorder 15.38 6.236 .013 
Mood disorder 10.53 19.23 0.97 .325 

Note:aMean (SD). bt-Values. cEuropean American versus all other groups used for chi-square tests due to small n. RD = Reading Disability; ADHD = Attention deficit/hyperactivity disorder.

Participants with RD met low achievement and/or ability-achievement regression-corrected discrepancy criteria. Low achievement was determined by a standard score of 85 or below, or 1 SD below the normative mean, on the Basic Reading composite score of the Woodcock-Johnson III Tests of Achievement (WJ-3; Woodcock, McGrew, & Mather, 2001). The discrepancy criterion was met if a student's score on the Basic Reading composite was one standard error of prediction below the predicted value based on the correlation between the initial sample's FSIQ and Basic Reading Composite. Multiple criteria were used to identify poor readers in order to include both those students with poor reading abilities that are commensurate with their cognitive abilities, as well as those that have below-expected reading abilities for their intellectual level (Fletcher et al., 1994; Stanovich & Siegel, 1994).

Participants were included in the ADHD group if they met two requirements: (i) assignment of a DSM (American Psychiatric Association, 2000) diagnosis of ADHD by a licensed psychologist based on the results of the comprehensive evaluation, clinical interview, collateral report of significant childhood symptoms of inattention and/or hyperactivity/impulsivity (Barkley & Murphy, 2006), behavioral observations, and questionnaires (i.e., Conners Adult ADHD Rating Scales, Behavior Rating Inventory of Executive Function) and (ii) a significant level of self-reported current symptomatology on the ADHD Behavior Checklist for Adults (Barkley & Murphy, 2006), defined as a summary score of at least 1.5 standard deviations above the age-based normative mean (Murphy & Barkley, 1995) on either the Inattentive and/or the Hyperactive/Impulsive subscale. Self-report was not used in isolation to make a diagnosis. Due to the nature of this archival study, the clinicians did have access to the D-KEFS measures when making their diagnoses; however, a diagnosis of ADHD based on the DSM-IV-TR relies primarily on information provided during a clinical interview, behavioral observations, and behavioral rating questionnaires (both self-report and informant report, and childhood and current ratings), and the ultimate clinical diagnosis was based primarily on these sources of information. Cognitive and academic performance was never used in isolation as a diagnostic tool for identifying ADHD.

Exclusions

Participants with a history of a traumatic brain injury, seizure disorder, or other neurological conditions were excluded. In addition, those 40 years of age or older, those diagnosed with past or present psychotic disorders or pervasive developmental disorders, and those with both a Wechsler Adult Intelligence Scale–III (WAIS-III; Wechsler, 1997) VIQ and PIQ below 80 were excluded (n = 10 ). Participants with comorbid DSM diagnoses of mood or anxiety disorders were not excluded from the sample.

The ADHD group was significantly older (t = −3.491, df = 57, p < .005) and had a significantly higher mean FSIQ (t = −3.610, df = 61, p < .005) than the RD group. In addition, the ADHD group had more than three times as many men as women, compared with an almost equal number in the RD group (p < .05). The ADHD group was more likely to have a comorbid diagnosis of anxiety disorder. There was no statistically significant difference observed in the ethnic composition of the two groups, although the ADHD group comprised a notably larger proportion of European American students than students of color compared with the RD group (Table 1).

Measures

Selected subtests from the Delis-Kaplan Executive Function System (D-KEFS; Delis, et al., 2001) served as dependent measures. Additional measures used to classify participants as RD or ADHD are also described below. Recent studies suggest that secondary gain, such as accommodations and medications, may lead to feigning of attention symptoms, thus effort testing should regularly be implemented (Harrison, Edwards, & Parker, 2007; Suhr, Hammers, Dobbins-Buckland, Zimak, & Hughes, 2008; Sullivan, May, & Galbally, 2007). Due to the archival nature of this study, effort testing was not included; however, since the time of this study, effort tests have been included in the standard battery and future research will reflect this added level of confidence in the results.

Classification Measures

Woodcock-Johnson III Tests of Achievement (WJ-3)

The WJ-3 (Woodcock et al., 2001) is a well-standardized and commonly used assessment instrument. Basic Reading Skills, which is a composite score based on two subtests, Word Attack and Letter-Word Identification, that assess word and nonword decoding abilities, was used to identify poor readers who may have difficulties in reading accuracy.

Wechsler Adult Intelligence Scale (WAIS-III)

The WAIS-III (Wechsler, 1997) is an individually administered standardized test of general intellectual functioning. Several domains are evaluated, including verbal comprehension, perceptual organization, working memory, and processing speed. The FSIQ score reflects performance across all of these domains and is a representation of one's global intelligence. The FSIQ was utilized to identify a significant discrepancy between intellectual ability and reading achievement.

ADHD Behavior Checklist for Adults

This checklist (Barkley & Murphy, 2006) consists of the 18 symptoms included in the DSM-IV criteria for ADHD, with 9 items addressing current inattention and 9 addressing current hyperactive/impulsive symptoms. Responses range from 0 to 3 (Rarely or Never, Sometimes, Often, or Very Often). Thus, summary scores for inattentive and hyperactive/impulsive symptoms range from 0 to 27. Summary scores, rather than symptom counts, were utilized in the proposed study. Utilization of the summary score allows the use of age-based normative data to determine whether the overall severity of symptoms is significantly greater than that reported by adult age peers.

Executive Function

Delis–Kaplan Executive Function System (D-KEFS)

The D-KEFS (Delis et al., 2001) is a battery of nine tests that assesses key components of EF and has the benefit of a national normative sample to compare across tests and age groups. Three tests were included in this study: TM, VF, and CWI tests. TM measures flexibility of thinking and the ability to shift set on a visual-motor task. Three conditions were selected: number sequencing, letter sequencing, and number-letter switching. The first two conditions require participants to connect numbers and letters, respectively, in numerical and alphabetical order as fast as they can. The last condition is similar, but requires one to switch sets by drawing a line from number to letter, letter to number, and so on, in numerical and alphabetical order. The VF task assesses fluent productivity in the verbal domain and requires the rapid generation of words beginning with a specific letter (letter fluency), within a specific semantic category (category fluency) or, in the final condition, alternating between two categories (category switching). Finally, the CWI task assesses verbal inhibition. All four conditions from this test were utilized: Color Naming, Word Reading, Inhibition, and Inhibition/Switching. In the first two conditions, participants are asked to name colors and read color words, respectively. The third condition requires inhibition of the automatic response to read the color word and instead state the ink color in which the word is written (e.g., the word “red” written in blue ink). Finally, the last condition requires one to switch between naming the dissonant ink colors and reading the conflicting words.

Analysis Overview

Variables were examined for skewness, kurtosis, and outliers. Variables that evidenced significant skewness were transformed and all analyses were completed with both transformed and untransformed variables. As results did not differ, analyses based on untransformed raw scores are reported. Each test was analyzed separately in hierarchical logistic regressions in order to first control for the contribution of component (nonexecutive) skills, then determine the relative ability of each EF condition to correctly predict and categorize individuals with RD or ADHD. The first block for each analysis was age. The reasons for this were two-fold. First, raw scores were used rather than scaled scores to increase variance. Second, the ADHD group was significantly older than the RD group. Sex was not included as a covariate because it was not related to any measure. Despite significant group differences, FSIQ was not covaried, as it reflects an attribute of these disorders and is not unrelated to the independent variable (Dennis et al., 2009).

Results

Mean raw scores for each group are displayed in Table 2. Scores on the TM and CWI subtests represent response times, while scores on the VF subtest represent number of words generated. Hierarchical logistic regression analyses were conducted for each subtest of the D-KEFS (Table 3). Across each model, the RD group was more accurately classified compared with the ADHD group. In the first model, which included age, the component measures, and the switching task of the TM test, 78% of the entire sample was correctly identified. Eighty-four percent of the RD group was correctly classified (specificity), while only 69% of the ADHD group was accurately classified (sensitivity). Of the D-KEFS measures included in the model, only the final block change (i.e., adding Number-Letter Switching to the model) significantly improved the model (Step: χ2 = 6.658, df = 1, p = .01). Age (B = 0.375, SEB = 0.113, Wald = 11.05, 95% CI [1.166, 1.814]) and Number-Letter Switching (B = −0.039, SEB = 0.019, Wald = 4.19, 95% CI [0.927, 0.998]) were significant predictors of group membership in the final model. The Nagelkerke R2 value, a measure of fit that evaluates the proportion of the total variability of the outcome that is accounted for by the model, was 45.3% for the final model (Hosmer and Lemeshow test: χ2 = 6.480, df = 8, p = .594).

Table 2.

Mean raw and standard scores for ADHD and RD groups on dependent measures

 ADHD raw scores, mean (SDRD raw scores, mean (SDF(1,63) ADHD standard scores, mean (SDRD standard scores, mean (SDF(1,63) 
TM Number Sequencing 26.12 (10.03) 29.69 (11.26) 1.46 11.08 (2.73) 10.19 (2.57) 1.4 
TM Letter Sequencing 24.58 (8.48) 31.94 (14.91) 0.03* 11.42 (2.44) 9.42 (3.85) 5.25* 
TM Letter-Number Switching 66.54 (26.16) 86.17 (38.95) 0.04* 10.15 (2.48) 8.14 (3.37) 6.06* 
VF Letter Fluency 38.08 (10.44) 31.25 (10.07) 7.51* 10.50 (3.39) 8.61 (3.14) 5.79* 
VF Category Fluency 39 (10.12) 41.78 (8.30) 1.45 10.38 (4.22) 11.69 (3.35) 1.93 
VF Category Switching 13.62 (2.37) 14.03 (2.52) 0.55 10.23 (2.90) 10.89 (3.03) 0.46 
CWI Color Naming 27.54 (6.49) 30.83 (6.19) 4.12* 10.19 (2.91) 8.39 (2.75) 6.18* 
CWI Word Reading 21.31 (4.99) 24.53 (5.48) 5.61* 10.50 (2.87) 8.36 (3.12) 7.58* 
CWI Interference 51.69 (15.14) 58.19 (17.87) 2.27 9.85 (3.26) 8.56 (3.24) 2.39 
CWI Inhibition/Switching 59.92 (14.46) 64.75 (16.34) 1.45 9.42 (3.09) 8.50 (3.23) 1.28 
 ADHD raw scores, mean (SDRD raw scores, mean (SDF(1,63) ADHD standard scores, mean (SDRD standard scores, mean (SDF(1,63) 
TM Number Sequencing 26.12 (10.03) 29.69 (11.26) 1.46 11.08 (2.73) 10.19 (2.57) 1.4 
TM Letter Sequencing 24.58 (8.48) 31.94 (14.91) 0.03* 11.42 (2.44) 9.42 (3.85) 5.25* 
TM Letter-Number Switching 66.54 (26.16) 86.17 (38.95) 0.04* 10.15 (2.48) 8.14 (3.37) 6.06* 
VF Letter Fluency 38.08 (10.44) 31.25 (10.07) 7.51* 10.50 (3.39) 8.61 (3.14) 5.79* 
VF Category Fluency 39 (10.12) 41.78 (8.30) 1.45 10.38 (4.22) 11.69 (3.35) 1.93 
VF Category Switching 13.62 (2.37) 14.03 (2.52) 0.55 10.23 (2.90) 10.89 (3.03) 0.46 
CWI Color Naming 27.54 (6.49) 30.83 (6.19) 4.12* 10.19 (2.91) 8.39 (2.75) 6.18* 
CWI Word Reading 21.31 (4.99) 24.53 (5.48) 5.61* 10.50 (2.87) 8.36 (3.12) 7.58* 
CWI Interference 51.69 (15.14) 58.19 (17.87) 2.27 9.85 (3.26) 8.56 (3.24) 2.39 
CWI Inhibition/Switching 59.92 (14.46) 64.75 (16.34) 1.45 9.42 (3.09) 8.50 (3.23) 1.28 

Note: *p < .05 in the Bonferroni comparison. RD = Reading Disability; ADHD = Attention deficit/hyperactivity disorder; TM = Trail Making; VF = Verbal Fluency; CWI = Color-Word Interference.

Table 3.

Final models for hierarchical logistic regressions predicting group membership (ADHD or RD)

Variables B SEB Wald df Exp (BMeasure of fit 
Trail making 
 1. Age** 0.375 0.113 11.05 1.46 45.3% 
 2. Number Sequencing 0.023 0.047 0.244 1.02 
 Letter Sequencing −0.032 0.049 0.422 0.97 
 3. Number-Letter −0.039 0.019 4.19 0.96 
 Switching* 
Verbal Fluency 
 1. Age* 0.259 0.097 7.20 1.30 49.9% 
 2. Letter Fluency** 0.156 0.049 10.05 1.17 
 Category Fluency −0.097 0.059 2.73 0.91 
 3. Category Switching −0.204 0.183 1.25 0.82 
Color-Word Interference 
 1. Age** 0.325 0.102 10.23 1.38 38.4% 
 2. Color Naming −0.135 0.098 1.90 0.87 
Word Reading 0.046 0.079 0.34 1.05 
 3. Inhibition −0.050 0.032 2.42 0.95 
 4. Inhibition/Switching 0.012 0.028 0.17 1.01 
Variables B SEB Wald df Exp (BMeasure of fit 
Trail making 
 1. Age** 0.375 0.113 11.05 1.46 45.3% 
 2. Number Sequencing 0.023 0.047 0.244 1.02 
 Letter Sequencing −0.032 0.049 0.422 0.97 
 3. Number-Letter −0.039 0.019 4.19 0.96 
 Switching* 
Verbal Fluency 
 1. Age* 0.259 0.097 7.20 1.30 49.9% 
 2. Letter Fluency** 0.156 0.049 10.05 1.17 
 Category Fluency −0.097 0.059 2.73 0.91 
 3. Category Switching −0.204 0.183 1.25 0.82 
Color-Word Interference 
 1. Age** 0.325 0.102 10.23 1.38 38.4% 
 2. Color Naming −0.135 0.098 1.90 0.87 
Word Reading 0.046 0.079 0.34 1.05 
 3. Inhibition −0.050 0.032 2.42 0.95 
 4. Inhibition/Switching 0.012 0.028 0.17 1.01 

Note. The numbers in the first column represent the block in which the variables were entered. Values listed in this table are for the final model only. *p < .05, **p < .005.

The second model, which included component measures of the VF test, correctly identified 71% of the sample. Just over half of the ADHD group was correctly classified (58%), while 81% of the RD group was accurately identified. Of the D-KEFS measures included in the model, only the first block (i.e., Letter Fluency and Category Fluency) significantly improved the model (Step: χ2 = 14.461, df = 2, p = .001). Age (B = 0.259, SEB = .097, p = < .05, Wald = 7.20, 95% CI [1.072, 1.566]) and Letter Fluency (B = 0.156, SEB = 0.049, Wald = 10.05, 95% CI [1.061, 1.287]) were significant predictors of group membership in the final model. The measure of fit of the final model was 49.9% (Hosmer and Lemeshow test: χ2 = 18.262, df = 8, p = .019).

The third model correctly classified 77% of the sample. It included component measures of the CWI test. As with the other two models, the CWI model accurately classified a greater portion of the RD group (86%) than the ADHD group (65%). Age (B = 0.325, SEB = 0.102, Wald = 10.23, 95% CI [1.123, 1.652]) was the only significant predictor of group membership; no other component of the CWI test improved correct group classification and no other block yielded a significant change in the model's ability to discriminate the two groups after age. The measure of fit of the final model was 38.4% (Hosmer and Lemeshow test: χ2 = 5.617, df = 8, p = .690).

Discussion

The current study investigated the ability of EF measures from the D-KEFS to discriminate between adults with ADHD and RD after controlling for more basic or “ingredient” processes. Our primary hypothesis posited that each of the three D-KEFS EF measures would discriminate between the two groups. This hypothesis was partially supported, as the TM EF measure, Number-Letter Switching, did add predictability of group membership, after controlling for age and performance on the associated “ingredient” tasks (Number Sequencing and Letter Sequencing). However, the EF components of VF and CWI were not significant contributors in their respective models, although the overall prediction models utilizing VF and CWI tasks were both significant. After accounting for age, only an “ingredient” task from the VF test (Letter Fluency) significantly discriminated the ADHD and RD groups in the final model. None of the CWI subtests contributed to the prediction of group membership.

The hypothesis that these D-KEFS subtests would more accurately classify adults identified as having ADHD than adults with RD was not supported for this sample. Results from three hierarchical logistic regressions demonstrated strong specificity (81%–86%) but weaker sensitivity (58%–69%) and, in fact, each of these D-KEFS measures proved to be more accurate in classifying the RD group than the ADHD group.

Lastly, our hypothesis that the ADHD group would perform below the RD group on all EF measures also was not supported. Contrary to our expectations, participants in the RD group performed more slowly than the ADHD group on Number-Letter Switching, the one EF task that did discriminate the groups. In fact, across all measures, the ADHD group performed as well as or better than the RD group. More specifically, statistically significant differences were observed on Letter Sequencing and Letter-Number Switching from the Trailmaking Test, Letter Fluency from the VF Test, and Color Naming and Word Reading from the CWI Test, with those diagnosed with RD performing below those with ADHD. Finally, it is noteworthy that the group mean performance for both groups fell in the average range compared with the normative sample on all of these D-KEFS subtests.

High levels of heterogeneity exist in the ADHD population, such that not all individuals with ADHD are impaired on every test of EF. One reason for this may be the particular EF measures that are employed. The measures included in this study assessed a range of EF processes that have evidenced significant deficits in some studies of adults with ADHD including the ability to switch sets, self-monitor, and inhibit automatic responses. However, measures tapping other components of the multidimensional construct of EF (e.g., planning, problem solving, and/or motor inhibition) might have been more sensitive and resulted in better identification of ADHD.

Characteristics of our study sample may also have contributed to the poor sensitivity of these measures to identify individuals with ADHD. Although this sample was comprised of college students from a broad range of both 2- and 4-year institutions, it does not represent the full spectrum of adults with ADHD. Most notably, it was characterized by higher intellectual function and consisted primarily of adults diagnosed with the Predominantly Inattentive subtype of ADHD (Inattention = 16, Hyperactive/Impulsive = 2, Combined = 8). Such samples may be higher functioning, allowing for progress to the college level, whereas other persons with ADHD with significant EF deficits may be more likely to have poor outcomes, that is, criminality, chronic drug abuse, or more significant psychiatric comorbidity; thus, they are less likely to be included in research samples (Schoechlin & Engel, 2005). All these characteristics may serve to diminish the extent of EF deficits in this sample of individuals with ADHD (although, see Brown, Reichel, & Quinlan, 2009, who argue that EF deficits are still observable in individuals with high IQ's).

Although there are compelling arguments against covarying IQ in populations with developmental disabilities, we explored whether a diagnosis of RD or ADHD had a significant impact on EF performance, even after accounting for intellectual ability. Post-hoc exploratory analyses were conducted that first controlled for FSIQ, then age, then each ingredient measure, and included the EF task in the final step. As expected, FSIQ was a significant predictor of group membership for each analysis. For the VF and the CWI analyses, the results were consistent with the initial analyses. After inclusion of FSIQ in the TM analysis, the Number-Letter Sequencing task's contribution to predicting group membership was no longer significant. Overall, our analyses suggest that while IQ is affected by the nature of the developmental disorder and shares variance with many EF measures, intelligence alone does not account for the differences observed in performance on EF tasks.

The results of our study also highlight the complexity inherent in the differential diagnosis of ADHD and RD. Our results are consistent with previous reports of deficits on EF measures in RD samples, and also draw attention to the critical importance of taking “ingredient” processes into account when interpreting performance on measures of EF. Across most D-KEFS measures, the RD group performed comparably or below the ADHD group, and the RD group was more accurately classified in each of the logistic regression models. In the model assessing VF, only an ingredient subtest, Letter Fluency, contributed significantly to the model. Although it was the executive task (Letter-Number Switching) that contributed significantly to the classification of RD in the model assessing TM, suggesting some EF deficit on this particular measure, it is not possible to definitively exclude the contribution of lower-level cognitive processes (e.g., letter sequencing) on this task. The superior classification of RD based on D-KEFS performance likely reflects, at least in part, the significant contribution of ingredient skills that are frequently affected in RD (i.e., speed and automaticity of verbal retrieval) to successful performance on these D-KEFS subtests.

These findings serve as a reminder to clinicians not to rely on performance on EF measures alone to diagnose individuals with ADHD. EF deficits exist in only a minority of those diagnosed with ADHD, and are not strongly correlated with functional outcomes, such as impairments in occupational functioning (Barkley & Murphy, 2011). Executive dysfunction symptoms may not also be as pronounced in vocationally and/or academically successful adults, who may be more likely to seek an evaluation. Furthermore, ADHD is as much a behavioral disorder as it is cognitive. While cognitive deficits due to neural abnormalities may be the source of dysfunction, these deficits often manifest as complex behavioral patterns over time that lead to functional impairments (Barkley, 1997) and may be best captured in detailed clinical interviews, and self- and informant-reports. This does not mean that cognitive data cannot assist clinical diagnosis and treatment monitoring, but rather it is encouraged (Gualtieri, 2005).

Barkley and Murphy (2010, 2011) presented detailed discussions on the utility of EF ratings versus EF tests in predicting occupational outcomes in adults with ADHD. Across outcome measures, EF ratings significantly predicted more domains compared to and beyond EF tests, consistent with the notion that EF is a multi-level construct and that EF tests may have poor ecological validity and assess a lower-level domain, or merely one aspect, of that complex construct that do not always correlate strongly with daily functioning or ADHD symptoms (Barkley & Fischer, 2011). Barkley and Murphy (2010) noted that, “These results can arise when measures of the lowest level of a complex and multi-level domain are necessary, yet not sufficient, to represent higher-level functions or abilities utilized in daily life situations and done to meet strategic goals” (p. 169). In summary, EF tests may be useful in providing supportive evidence of underlying cognitive impairments that contribute to functional impairments in daily living related to ADHD; however, clinicians must be aware that not all EF abilities are necessarily impaired in all individuals, and rating scales are evidenced to be more predictive of deficits in daily functioning and functioning over time. Future studies may investigate the added impact of rating scales when discriminating between individuals with ADHD and other clinical groups.

Strengths of this study include the large sample size and the use of hierarchical logistic regression to control for component processes in order to more effectively isolate the EF component of measures. This is also one of the few studies comparing the ability of EF measures to discriminate individuals with ADHD from another clinical sample, rather than normally developing controls. This is a crucial question, as it is faced on a daily basis by clinicians faced with making differential diagnoses. It is clear that EF deficits may characterize a number of clinical populations, including those with ADHD. Currently, available measures do not provide the necessary specificity although it is possible that groups may have distinct profiles of executive impairments that could potentially be discriminated by more fine-grained measurement of a range of executive abilities.

Several limitations should be considered in the interpretation of the current findings. The three measures selected from the D-KEFS only assess a limited number of EF abilities, thus reflecting a restricted range of possible EF components. EF is a multifaceted construct comprising a multitude of cognitive processes; thus, future studies may benefits from assessing a broader range of EF components and to explore methods for combining results from numerous individual measures. As an example, Biederman and colleagues (2004) have utilized multiple measures of EF to create an index of executive impairment. Secondly, the generalizability of these results may be limited because the sample comprises a clinical sample of college students and is not representative of the population of adults with ADHD and RD. Future studies should explore the utility of the D-KEFS in differential diagnosis in more heterogeneous samples of persons diagnosed with ADHD and RD. In addition, the current study did not include a control group, which would help characterize the specificity and sensitivity of the D-KEFS measures for use in a clinical setting and potentially validate the existence of EF deficits in adults with ADHD and RD in comparison with their typically developing peers.

In summary, the present study highlighted the lack of sensitivity of select measures assessing cognitive inhibition and set shifting for identifying adults with ADHD, and for discriminating those diagnosed with ADHD from those diagnosed with RD in a clinical sample. These results emphasize the need for clinicians not to rely solely on neuropsychological evidence of EF deficits to diagnose ADHD and to carefully consider the contribution of lower-level component processes to performance on measures of EF.

Funding

This research was supported in part by a Georgia State University Research on the Challenges of Acquiring Language and Literacy Fellowship and the Regents Center for Learning Disorders.

Conflict of Interest

None declared.

References

American Psychiatric Association
Diagnostic and statistical manual of mental disorders
 , 
2000
4th ed. text revision
Washington, DC
American Psychiatric Association
Barkley
R. A.
Behavioral inhibition, sustained attention, and executive functions: Constructing a Unifying Theory of ADHD
Psychological Bulletin
 , 
1997
, vol. 
121
 (pg. 
65
-
94
)
Barkley
R. A.
Fischer
M.
Predicting impairment in major life activities and occupational functioning in hyperactive children as adults: Self-reported executive function (EF) deficits versus EF tests
Developmental Neuropsychology
 , 
2011
, vol. 
36
 
2
(pg. 
137
-
161
)
Barkley
R. A.
Murphy
K. R.
Clinical workbook
 , 
2006
3rd ed.
New York
The Guilford Press
Barkley
R. A.
Murphy
K. R.
Impairment in occupational functioning and adult ADHD: The predictive utility of executive function (EF) ratings versus EF tests
Archives of Clinical Neuropsychology
 , 
2010
, vol. 
25
 (pg. 
157
-
173
)
Barkley
R. A.
Murphy
K. R.
The nature of executive function (EF) deficits in daily life activities in adults with ADHD and their relationship to performance on EF tests
Journal of Psychopathology and Behavioral Assessment
 , 
2011
, vol. 
33
 (pg. 
137
-
158
)
Barkley
R. A.
Murphy
K. R.
Fischer
M.
ADHD in adults: What the science says
2008
New York, NY
Guilford Press
Biederman
J.
Monuteaux
M. C.
Doyle
A. E.
Seidman
L. J.
Wilens
T. E.
Ferrero
F.
, et al.  . 
Impact of executive function deficits and attention-deficit/hyperactivity disorder (ADHD) on academic outcomes in children
Journal of Consulting and Clinical Psychology
 , 
2004
, vol. 
72
 (pg. 
757
-
766
)
Biederman
J.
Monuteaux
M. C.
Mick
E.
Spencer
T.
Wilens
T. E.
Silva
J. M.
, et al.  . 
Young adult outcome of attention deficit hyperactivity disorder: A controlled 10-year follow-up study
Psychological Medicine
 , 
2006a
, vol. 
36
 (pg. 
167
-
179
)
Biederman
J.
Petty
C.
Fried
R.
Fontanella
J.
Doyle
A. E.
Seidman
L. J.
, et al.  . 
Impact of psychometrically defined deficits of executive functioning in adults with attention deficit hyperactivity disorder
The American Journal of Psychiatry
 , 
2006b
, vol. 
163
 (pg. 
1730
-
1738
)
Boonstra
A. M.
Kooij
J. J. S.
Oosterlaan
J.
Sergeants
J. A.
Buitelaar
J. K.
To act or not to act, that's the problem: Primarily inhibition difficulties in adult ADHD
Neuropsychology
 , 
2010
, vol. 
24
 
2
(pg. 
209
-
221
)
Boonstra
A. M.
Oosterlaan
J.
Sergeant
J. A.
Buitelaar
J. K.
Executive functioning in adult ADHD: A meta-analytic review
Psychological Medicine
 , 
2005
, vol. 
35
 (pg. 
1097
-
1108
)
Booth
J.
Boyle
J.
Kelly
S.
Do tasks make a difference? Accounting for heterogeneity of performance of children with reading difficulties on tasks of executive function: Findings from a meta-analysis
British Journal of Developmental Psychology
 , 
2010
, vol. 
28
 (pg. 
133
-
176
)
Brown
T. E.
Brown
T.
Emerging understandings of attention deficit disorders and comorbidities
Attention deficit disorders and comorbidities in children, adolescents and adults
 , 
2000
Washington, DC
American Psychiatric Press
(pg. 
3
-
55
)
Brown
T. E.
Attention deficit disorder: The unfocused mind in children and adults
 , 
2005
New Haven, CT
Yale University Press
Brown
T. E.
Reichel
P. C.
Quinlan
D. M.
Executive function impairments in high IQ adults with ADHD
Journal of Attention Disorders
 , 
2009
, vol. 
13
 (pg. 
161
-
167
)
Carr
L. A.
Nigg
J. T.
Henderson
J. M.
Attentional versus motor inhibition in adults with attention-deficit/hyperactivity disorder
Neuropsychology
 , 
2006
, vol. 
20
 
4
(pg. 
430
-
441
)
Crawford
J. R.
Sutherland
D.
Garthwaite
P. H.
On the reliability and standard errors of measurement of contrast measures from the D-KEFS
Journal of the International Neuropsychological Society
 , 
2008
, vol. 
14
 
6
(pg. 
1069
-
1073
)
Delis
D. C.
Kaplan
E.
Kramer
J. H.
Delis-Kaplan Executive Function System (D-KEFS)
 , 
2001
San Antonio, TX
The Psychological Corporation
Dennis
M.
Francis
D. J.
Cirino
P. T.
Schachar
R.
Barnes
M. A.
Fletcher
J. M.
Why IQ is not a covariate in cognitive studies of neurodevelopmental disorders
Journal of International Neuropsychological Society
 , 
2009
, vol. 
15
 (pg. 
1
-
13
)
Douglas
V. I.
Rutter
M.
Attentional and cognitive problems
Developmental neuropsychiatry
 , 
1983
New York
Guilford Press
(pg. 
280
-
329
)
Faraone
S. V.
Biederman
J.
Spencer
T.
Wilens
T.
Seidman
L. J.
Mick
E.
, et al.  . 
Attention-deficit/hyperactivity disorder (ADHD) in adults: An overview
Society of Biological Psychiatry
 , 
2000
, vol. 
48
 (pg. 
9
-
20
)
Fletcher
J. M.
Shaywitz
S. E.
Shankweiler
D. P.
Katz
L.
Liberman
I. Y.
Stuebing
K. K.
, et al.  . 
Cognitive profiles of reading disability: Comparisons of discrepancy and low achievement definitions
Journal of Educational Psychology
 , 
1994
, vol. 
86
 (pg. 
6
-
23
)
Gualtieri
C.
A practical approach to objective attention deficit/hyperactivity disorder diagnosis and management
Psychiatry
 , 
2005
, vol. 
2
 (pg. 
17
-
25
)
Hall
S.
Halperin
J.
Schwartz
S.
Newcorn
J.
Behavioral and executive functions in children with attention-deficit hyperactivity disorder and reading disability
Journal of Attention Disorders
 , 
1997
, vol. 
1
 (pg. 
235
-
247
)
Harrison
A. G.
Edwards
M. J.
Parker
K. C. H.
Identifying students faking ADHD: Preliminary findings and strategies for detection
Archives of Clinical Neuropsychology
 , 
2007
, vol. 
22
 (pg. 
577
-
588
)
Homack
S.
Riccio
C. A.
A meta-analysis of the sensitivity and specificity of the Stroop Color and Word Test with children
Archives of Clinical Neuropsychology
 , 
2004
, vol. 
19
 (pg. 
725
-
743
)
Katz
L. G.
Woods
D. S.
Goldstein
G.
Auchenbach
R. C.
Geckle
M.
The utility of neuropsychological tests in evaluation of attention-deficit/hyperactivity disorder (ADHD) versus depression in adults
Assessment
 , 
1998
, vol. 
5
 (pg. 
45
-
51
)
Levin
B.
Organizational deficits in dyslexia: Possible frontal lobe dysfunction
Developmental Neuropsychology
 , 
1990
, vol. 
6
 (pg. 
95
-
110
)
Lovejoy
D. W.
Ball
J. D.
Keats
M.
Stutts
M. L.
Spain
E. H.
Janda
L.
, et al.  . 
Neuropsychological performance of adults with attention deficit hyperactivity disorder (ADHD): Diagnostic classification estimates for measures of frontal lobe/executive functioning
Journal of the International Neuropsychological Society
 , 
1999
, vol. 
5
 (pg. 
222
-
233
)
Malloy-Diniz
L.
Fuentes
D.
Borges Leite
W.
Correa
H.
Bechara
A.
Impulsive behavior in adults with attention deficit/hyperactivity disorder: Characterization of attentional, motor and cognitive impulsiveness
Journal of the International Neuropsychological Society
 , 
2007
, vol. 
13
 (pg. 
693
-
698
)
Marzocchi
G. M.
Oosterlaan
J.
Zuddas
A.
Cavolina
P.
Geurts
H.
Redigolo
D.
, et al.  . 
Contrasting deficits on executive functions between ADHD and reading disabled children
Journal of Child Psychology and Psychiatry
 , 
2008
, vol. 
49
 (pg. 
543
-
552
)
Murphy
K.
Barkley
R. A.
Preliminary normative data on DSM-IV criteria for adults
The ADHD Report
 , 
1995
, vol. 
3
 (pg. 
6
-
7
)
Pennington
B. F.
Groisser
D.
Welsh
M. C.
Contrasting cognitive deficits in attention deficit hyperactivity disorder versus reading disability
Developmental Psychology
 , 
1993
, vol. 
29
 (pg. 
511
-
523
)
Quay
H. C.
Bloomingdale
L. M.
Sergeant
J.
Attention deficit disorder and the behavioral inhibition system: The relevance of the neuropsychological theory of Jeffrey A. Gray
Attention deficit disorder: Criteria, cognition, intervention
 , 
1988a
New York
Pergamon Press
(pg. 
117
-
126
)
Quay
H. C.
Bloomingdale
L. M.
The behavioral reward and inhibition systems in childhood behavior disorder
Attention deficit disorder: Vol. 3. New research in treatment, psychopharmacology, and attention
 , 
1988b
New York
Pergamon Press
(pg. 
176
-
186
)
Riccio
C. A.
Wolfe
M. E.
Romine
C.
Davis
B.
Sullivan
J. R.
The Tower of London and neuropsychological assessment of ADHD in adults
Archives of Clinical Neuropsychology
 , 
2004
, vol. 
19
 (pg. 
661
-
671
)
Schoechlin
C.
Engel
R. R.
Neuropsychological performance in adult attention-deficit hyperactivity disorder: Meta-analysis of empirical data
Archives of Clinical Neuropsychology
 , 
2005
, vol. 
20
 (pg. 
727
-
744
)
Schreiber
H. E.
Javorsky
D. J.
Robinson
J. E.
Stern
R. A.
Rey-Osterrieth Complex Figure performance in adults with attention deficit hyperactivity disorder: A validation study of the Boston Qualitative Scoring System
The Clinical Neuropsychologist
 , 
1999
, vol. 
13
 (pg. 
509
-
520
)
Stanovich
K. E.
Seigel
L. S.
Phenotypic performance profile of children with reading disabilities: A regression-based test of the phonological-core variable difference model
Journal of Educational Psychology
 , 
1994
, vol. 
86
 (pg. 
24
-
53
)
Stavro
G. M.
Ettenhofer
M. L.
Nigg
J. T.
Executive functions and adaptive functioning in young adult attention-deficit/hyperactivity disorder
Journal of the International Neuropsychological Society
 , 
2007
, vol. 
13
 (pg. 
324
-
334
)
Suhr
J.
Hammers
D.
Dobbins-Buckland
K.
Zimak
E.
Hughes
C.
The relationship of malingering test failure to self-reported symptoms and neuropsychological findings in adults referred for ADHD evaluation
Archives of Clinical Neuropsychology
 , 
2008
, vol. 
23
 (pg. 
521
-
530
)
Sullivan
B. K.
May
K.
Galbally
L.
Symptom exaggeration by college adults in attention-deficit hyperactivity disorder and learning disorder assessments
Applied Neuropsychology
 , 
2007
, vol. 
14
 
3
(pg. 
189
-
207
)
Taylor
C. J.
Miller
D. C.
Neuropsychological assessment of attention in ADHD adults
Journal of Attention Disorders
 , 
1997
, vol. 
2
 (pg. 
77
-
88
)
Wechsler
D.
Wechsler adult intelligence scale
 , 
1997
3rd ed.
San Antonio, TX
Psychological Corp
Weyandt
L.
Rice
J.
Linterman
I.
Mitzlaff
L.
Emert
E.
Neuropsychological performance of a sample of adults with ADHD, Developmental Reading Disorder, and controls
Developmental Neuropsychology
 , 
1998
, vol. 
14
 (pg. 
643
-
656
)
Willcutt
E. G.
Doyle
A. E.
Nigg
J. T.
Faraone
S. V.
Pennington
B. F.
Validity of the executive function theory of attention-deficit/hyperactivity disorder: A meta-analytic review
Society of Biological Psychiatry
 , 
2005
, vol. 
57
 (pg. 
1336
-
1346
)
Wodka
E. L.
Loftis
C.
Mostofsky
S. H.
Prahme
C.
Larson
J. C. G.
Denckla
M. B.
, et al.  . 
Prediction of ADHD in boys and girls using the D-KEFS
Archives of Clinical Neuropsychology
 , 
2008a
, vol. 
23
 (pg. 
283
-
293
)
Wodka
E. L.
Mostofsky
S. H.
Prahme
C.
Larson
J. C. G.
Loftis
C.
Denckla
M. B.
, et al.  . 
Process examination of executive function in ADHD: Sex and subtype effects
The Clinical Neuropsychologist
 , 
2008b
, vol. 
22
 (pg. 
826
-
841
)
Woodcock
R. W.
McGrew
K.
Mather
N.
The Woodcock-Johnson Tests of Achievement: Third edition
 , 
2001
Itasca, IL
Riverside
Woods
S. P.
Lovejoy
D. W.
Stutts
M. L.
Ball
J. D.
Fals-Stewart
W.
Comparative efficiency of a discrepancy analysis for the classification of attention-deficit/hyperactivity disorder in adults
Archives of Clinical Neuropsychology
 , 
2002
, vol. 
17
 (pg. 
351
-
369
)
Young
S.
Morris
R.
Toone
B.
Tyson
C.
Planning ability in adults with attention- deficit/hyperactivity disorder
Neuropsychology
 , 
2007
, vol. 
21
 
5
(pg. 
581
-
589
)