Can the Error-Monitoring System Differentiate ADHD From ADHD With Reading Disability? Reading and Executive Dysfunction as Reflected in Error Monitoring

Abstract

Objectives: ADHD and reading disability (RD) are distinct disorders that often appear together. Individuals with both disorders are currently diagnosed based on questionnaires/behavioral performance. The present study aimed to determine whether ADHD alone differs from ADHD with RD in error monitoring, which is part of the executive system, as measured while reading. Method: Event-related potentials were recorded during a lexical decision task performed by children with comorbid ADHD and RD and children with ADHD. Results: Lower executive function and reading abilities were accompanied by decreased event-related potential components in participants with ADHD and RD, compared with participants with ADHD. Conclusion: Results suggest that the error monitoring activation can be used as a possible biomarker to objectively differentiate ADHD with RD from ADHD alone.

Keywords

ADHD error monitoring reading reading disability

Introduction

ADHD is a highly prevalent disorder estimated to affect a range of 2.2% to 17.8% of school-aged children in the United States and Europe (Skounti, Philalithis, & Galanakis, 2007). These individuals suffer from excessive inattention and impulsivity, starting in childhood and persisting into adolescence and adulthood (Faraone, Sergeant, Gillberg, & Biederman, 2003; Mannuzza, Klein, Bessler, Malloy, & LaPadula, 1993). Such difficulties, which are related to deficits in executive functions (Biederman et al., 2004; Jacobson et al., 2011; Purvis & Tannock, 2000), contribute to academic failures and reading difficulties (Greven, Rijsdijk, Asherson, & Plomin, 2012). Executive function deficits manifest as difficulties in inhibition, sustained attention, working memory, temporal processing, planning, flexibility (Castellanos et al., 2002; Marzocchi et al., 2008), time perception (Nicolson, Fawcett, & Dean, 1995), and in a slowed speed of information processing (Jacobson et al., 2011).

ADHD often co-occurs with reading disabilities (RDs); children with an ADHD and RD comorbidity (ADHD + RD) represent approximately 20% to 39% of all ADHD patients (August & Garfinkel, 1990; Semrud-Clikeman, Walkowiak, Wilkinson, & Minne, 2010). Patients with ADHD + RD exhibit greater deficits than children with ADHD in most executive functions: slower naming (Rucklidge & Tannock, 2002), slower speed of processing (Willcutt, Doyle, Nigg, Faraone, & Pennington, 2005), and greater impairments in working memory, shifting, inhibition (Willcutt et al., 2001), and verbal auditory working memory (Vakil, Blachstein, Wertman-Elad, & Greenstein, 2012). They exhibit a greater reading impairment than the ADHD group as well (Willcutt et al., 2005), which is accompanied by a phonological deficit (Purvis & Tannock, 2000). In the emotional aspect, individuals with ADHD + RD show greater psychological, behavioral, and emotional problems than RD and ADHD groups separately (for review, see Sexton, Gelhorn, Bell, & Classi, 2012). Although ADHD and ADHD + RD have similar characteristics, these differ greatly in degree of severity, that several theories have attempted to find whether there is a unique etiology for ADHD + RD that is distinct from the common etiology of ADHD (de Jong, Oosterlaan, & Sergeant, 2006; Pennington, Groisser, & Welsh, 1993; Rucklidge & Tannock, 2002).

The theories dealing with the differences between ADHD + RD and ADHD, however, did not get to a clear conclusion: The Cognitive Subtype hypothesis (or the “double dissociation” theory) argued that ADHD + RD has a distinct causal mechanism and could be regarded as an independent disorder (de Jong et al., 2006; Rucklidge & Tannock, 2002). The Phenocopy hypothesis suggested that the primary deficit in ADHD + RD is RD with secondary comorbid symptoms of ADHD (Pennington et al., 1993). Alternatively, the Common Etiology hypothesis claimed that ADHD and RD originate from a common genetic origin and that the comorbid group results from “add-on” of symptoms from ADHD and RD (Gooch, Snowling, & Hulme, 2011; Willcutt et al., 2005). To date, there is no consensus as to whether ADHD + RD has an etiology distinct from or common to that of ADHD.

One of the reasons for debate between the different theories is the lack of an objective diagnostic tool for distinguishing between ADHD and ADHD + RD. These disorders are currently diagnosed based on questionnaires, teacher/parent reports, and performance tests (e.g., reading or attention normalized tests). As ADHD and ADHD + RD share common symptoms (e.g., reading difficulty, executive deficit, inattentive behavior, social problems, and academic difficulties), a differential diagnosis is often challenging (Mangina & Beuzeron-Mangina, 2009). Nevertheless, a precise diagnosis is essential to guide the appropriate therapeutic approach. For instance, individuals with ADHD + RD might suffer from reading difficulties due to a phonological deficit, while the same problem in individuals with ADHD might be due to an executive dysfunction (Purvis & Tannock, 2000; Sesma, Mahone, Levine, Eason, & Cutting, 2009), and these differences call for different types of intervention. Providing the right intervention is critically important in individuals with ADHD and ADHD + RD, especially in young children. If children are not appropriately treated, they can face serious long-term repercussions, such as emotional and social problems, juvenile delinquency, car accidents, and unstable professional and personal lives (Johnson, 2002). Taken together, these facts emphasize the need for an objective biomarker for accurately diagnosing ADHD and ADHD + RD.

Neuroimaging is an objective way to examine neurocognitive abnormalities. Most of the current neuroimaging literature on ADHD and ADHD + RD arises from the field of electrophysiology. In previous work, an ADHD + RD group showed differential activation of the left hemisphere manifested by a pronounced theta band and a less pronounced alpha band than in children with ADHD (Clarke, Barry, McCarthy, & Selikowitz, 2002). Another neuroimaging study showed that children with ADHD + RD have a lower activation of the error-detection system, a system that is in charge of learning (Van de Voorde, Roeyers, & Wiersema, 2010). This system is part of the overall executive functioning, and its level of activation reflects the level of mismatch between neural representations of the desired and actual responses (Falkenstein, Hohnsbein, Hoormann, & Blanke, 1991). The activation of this system can be measured by the event-related potentials, error-related negativities (ERNs) and correct-related negativities (CRNs), which are evoked following erroneous and correct responses, respectively. Van de Voorde et al. (2010) demonstrated a decreased ERN component in the ADHD + RD group compared with children with ADHD. The authors suggested that these children might not be aware of error commission. These results point at the ERN and CRN as possible biomarkers for differentiating between those two groups. Because children with ADHD + RD and ADHD suffer from reading problems from different sources (Purvis & Tannock, 2000), differential activation of the error-detection system could be found in reading in these groups.

Decreased activation of the error-detection system in individuals with reading difficulties after reading errors was already reported (Horowitz-Kraus, 2011; Horowitz-Kraus & Breznitz, 2008, 2011b). This decreased activity was linked to an impaired mental lexicon, which was demonstrated by a decreased N400 Event-Related Potentials (ERP) component in adults and children with dyslexia (Horowitz-Kraus, 2011; Horowitz-Kraus & Breznitz, 2008). The N400 component is evoked at approximately 400 ms following stimulus onset (Kutas & Hillyard, 1980). This component is considered to be a marker for neurophysiological processes involved in lexical/semantic access (Kutas & Federmeier, 2000). Moreover, the amplitude of the N400 response is considered by some to be an index of the intensity of cognitive processing required to gain access to a lexical mental representation (Kutas & Federmeier, 2000). Decreased N400 activity following error commission in children with severe uncompensated dyslexia was also accompanied by decreased perception abilities and decreased N100 components (Horowitz-Kraus & Breznitz, 2013).

In the present study, we aim to neurophysiologically map the differences in reading and in executive functions in children with ADHD + RD and ADHD and to find a biomarker that will objectively differentiate between the two groups (following the designs of Katz, Brown, Roth, & Beers, 2011;Trani et al., 2011). To determine the differences in brain activation between the two groups, we compared their error-monitoring activation while reading, thus combining the executive function and reading domains. We hypothesized that children with ADHD + RD will exhibit more severe reading and executive difficulties and decreased ERN amplitudes, as well as lower perception abilities (decreased N100) and lower semantic ability (decreased N400), compared with children with ADHD. To find the relationship between reading and executive abilities in these two groups and to validate the error-monitoring system as a biomarker that reflects these abilities, we correlated the behavioral and electrophysiological measures (ERN) in these two groups.

Method

Participants

The study participants were 28 age-matched participants, 14 children diagnosed with ADHD (M age 12.92 ± 0.65 years, 14 males) and 14 with ADHD + RD (i.e., the comorbid group; M age 13.19 ± .75 years, 12 males, 2 females), t(26) = −1.02, ns; all matched for nonverbal IQ (90 or above) as measured by the Raven Standard Progressive Matrices (Raven, 1969). All were native Hebrew speakers from a middle-class background, were right-handed, had normal or corrected-to-normal vision in both eyes, and were found to have normal hearing. Participants were matched for their attention/hyperactivity scores, t(26) = 0.807, ns, as measured by the Conners Rating scale for teenagers (Conners, 1989). No differences were found on attention measures between the females and the males in the study.

Participants signed assents and their parents provided written informed consent prior to the study. Both groups were diagnosed as having ADHD or ADHD + RD for at least 2 years prior to the study, with ADHD being the primary diagnosis in both groups of participants. Diagnoses were verified with operationalized criteria: a score of −1.5 or below on a standardized 1-min words/pseudowords reading test, and reading accuracy and fluency with a text read orally from a normative reading battery (Shany, Lachman, Shalem, Bahat, & Zeiger, 2006). ADHD diagnoses were verified by a score of 6/9 or above on the Conners attention questionnaire (Conners, 1989) and had characteristics of ADHD for more than 6 months. The participants in the present study were mainly characterized as having the inattentive subset type of ADHD. To rule out the effect of ADHD type on group differences in the experimental measures, independent t test was performed on the average scores for the inattentive-type and the hyperactive-type sections from the Conners questionnaire. No differences between the ADHD and ADHD + RD subgroups were found: The average score of the inattentive section for the ADHD group was 5.29 ± 2.42 and for the ADHD + RD group was 5 ± 2.18, t(26) = 0.236, ns; the average score of the hyperactive section for the ADHD group was 4.47 ± 2.12 and for the ADHD + RD group was 4.47 ± 2.12, t(26) = 0.915, ns. The majority of the students (n = 19) were receiving a long-acting stimulant medication for ADHD throughout the study, as prescribed by their community care physician. The number of medicated children was equally distributed between the ADHD (n = 10) and ADHD + RD (n = 9) groups. All participants attended a junior high school for students with learning disabilities in the center of Israel.

Procedures

Participants were assessed within their school. Tests were administered individually in one of the schoolrooms after school hours by the research team. The study was reviewed and approved by the Ethics Committee of the University of Haifa.

Behavioral and Experimental Measures

Nine tests, listed in Table 1, were administered to measure reading and executive functions in the two groups. General ability, attention, reading, and word recognition measures were administered prior to the experiment to verify a participant’s reading level, IQ, and attention ability. Several executive functions were examined using the MindFit computerized program (CogniFit Personal Coach, 2008; for more information, see Horowitz-Kraus & Breznitz, 2009). Each behavioral testing session lasted approximately 1.5 hr.

Table 1.

List of Behavioral Measures.

	Description
Reading measures	A. Decoding—1-min words/pseudowords test for children (Shany, Lachman, Shalem, Bahat, & Zeiger, 2006).
	B. Oral reading task (Shany et al., 2006).
	C. Reading comprehension task (Shatil, Nevo, & Breznitz, 2008).
Executive functions	A. Short-term memory and capacity—digit-span subtest from the WISC-III (Wechsler, 1991).
	B. Coding WAIS-III (Wechsler, 1991).
	C. Symbol search WAIS-III (Wechsler, 1991).
	D. Objects, letters, and RAS naming tests (modeled after Denckla & Rudel, 1976; Wolf, Bally, & Morris, 1986).
	E. Planning and shifting abilities (CogniFit Personal Coach, 2008).
	F. Fluency (Shany et al., 2006).

Note. WISC-III = Wechsler Intelligence Scale for Children–III; WAIS-III = Wechsler Adult Intelligence Scale–III; RAS = rapid alternating stimulus.

Electrophysiology Recordings

Reading-related brain activity was elicited during a participant’s performance of a lexical decision task (Breznitz, 1989). The participant was seated approximately 80 cm in front of an IBM-PC screen in a sound-attenuated room and presented with a lexical decision task consisting of two forms of 80 moderate-to-high frequency Hebrew words (Frost, 2001) and 80 pseudowords (created by substituting one or two letters in the real words). The stimuli were presented horizontally in the center of the screen in white on a gray background. Each stimulus consisted of four or five Hebrew letters, each letter one quarter of an inch (0.6 cm) in diameter, presented for 100 ms, with a response window of 1,900 ms.

Participants were instructed to decide whether a stimulus represented a word or a pseudoword by pushing one of two buttons on a joystick (right button for word stimulus, left button for pseudoword stimulus). They were told to do this as quickly and accurately as possible. All participants used their right hand for responding. No interaction took place between the participant and experimenter during the task. Mean reaction time (RT) was calculated separately for correct and error responses. Only responses between 300 and 2,000 ms after target onset were included in the mean RT.

The electroencephalogram was recorded continuously from 64 electrodes mounted on a custom-made cap (Bio-logic Systems Corp., San Carlos, California) according to the international 10/20 system (Jasper, 1958), sampled at a rate of 2048 Hz with an analog band pass filter of 0.1 to 70 Hz and 12-bit A/D converter, and stored for off-line analysis. An electrooculogram (EOG) was recorded by a suborbital electrode. A ground electrode was placed on the left mastoid. All electrode impedances were maintained at or below 5 kΩ.

The electroencephalogram was corrected for horizontal and vertical eye movements using the Gratton, Coles, and Donchin’s (1983) algorithm implemented in the Brain Vision Analyzer software (Version 1.05, Brain Vision Products, Munich, Germany), and filtered with a 25-Hz filter. Event-related potential epochs time-locked to each stimulus and epochs time-locked to participant responses were averaged separately. An average reference was used to reject artifacts. Response-related epochs started 100 ms before and ended 400 ms after response onset, and stimulus-related epochs started 100 ms before and ended 900 ms after stimulus onset. Both types of epochs were averaged separately for correct and incorrect trials. All epochs were subsequently inspected visually to ensure that they were free of residual artifacts.

The baseline for the ERN and CRN response-locked components was set as the period between −100 and 0 ms pre-response. The peak amplitude of these components was measured at the largest negative peak between 30 and 160 ms post-response. The baseline for the N100 and N400 stimulus-locked component was set to −100 to 0 ms pre-stimulus. Peak N100 amplitude was measured as the largest negative peak at 50 to 150 ms poststimulus onset, whereas the peak for N400 amplitude was measured as the largest negative peak between 350 and 450 ms post-stimulus. In all cases, peak amplitude for stimulus and response was computed in relation to the average prestimulus or preresponse baseline, respectively. Because the response-locked ERN and CRN were maximal at AFz and Cz electrodes, and because stimulus-locked N100 components were maximal at the FCz and Fz electrodes, and N400 at Fz, FCz, and Cz, measurements of amplitude from these corresponding electrodes were used in the analyses. Because we were interested in examining reading ability, only words were used for the analysis; whereas, pseudowords were used only to ensure lexical activity (error monitoring) in the task.

Statistical Analyses

Behavioral data (Tables 2 and 3) were analyzed using independent t-test analyses. Electrophysiological data were analyzed by three-way repeated-measures analyses of variance (RM-ANOVA). Each analysis included Group (ADHD, ADHD + RD) as the between-subjects variables, and Type of response (correct, incorrect) and Electrode (for ERN/CRN response-locked: AFz and Cz electrodes, for N100 stimulus-locked: FCz and Fz electrodes, and for N400 stimulus-locked: Fz, FCz, and Cz electrodes) as the within-subjects variable. Family-wise Type I error rates were corrected by the Bonferroni method for multiple comparisons (α < .05).

Table 2.

Means, Standard Deviations, and Results of t Test for Behavioral Measure Scores.

		ADHD	ADHD + RD	t test
Ability	Measure	M (SD)	M (SD)	t test
Reading	1-min test for words
	Accuracy (st,^a errors)	−1 (1.07)	−2.34 (1.74)	2.457*
	Speed (number of words/min, st)	−0.442 (1.22)	−1.27 (0.85)	2.083*
	1-min test for pseudowords
	Accuracy (errors, st)	−1.25 (0.89)	−1.87 (0.64)	2.1*
	Speed (number of words/min, st^a)	−1.25 (0.89)	−1.87 (0.64)	2.396*
	Oral reading
	Speed (st^a)	−0.47 (1.26)	−1.34 (0.98)	2.014, ns
	Accuracy (errors, st^a)	−1.75 (2.84)	−4.05 (3.75)	1.826, ns
	Reading comprehension
	Accuracy (st^a)	−1.51 (1.47)	−2.77 (1.6)	1.958, ns
Executive Function
A. Speed of processing	Coding
	Speed (st^a)	5.5 (2.02)	3.35 (1.73)	3.022**
	Symbol search
	Speed (st^a)	6.21 (2.11)	5.14 (2.07)	1.353, ns
	Naming letters
	Speed (letters/min)	91.36 (19.82)	78.25 (16.12)	0.192, ns
	Accuracy (errors/min)	0.07 (0.26)	0.42 (0.75)	−1.667, ns
	Naming objects
	Speed (objects/min)	73.76 (16.34)	59.96 (11.51)	1.826*
	Accuracy (errors/min)	0.28 (0.46)	0.21 (0.42)	2.583, ns
	Naming RAS
	Speed (items/min)	84.82 (19.66)	70.85 (19.16)	1.904, ns
	Accuracy (errors/min)	0.71 (0.99)	1.57 (1.94)	−1.467, ns
B. Memory	Digit span
	Accuracy (number of items forward)	7.78 (1.05)	7.35 (1.39)	0.919, ns
	Accuracy (number of items backward)	4.42 (1.28)	3.85 (1.29)	1.174, ns
	Total (st^b)	5.5 (1.95)	4.42 (1.78)	1.516, ns
C. Shifting	(st^b)	−1.5 (1.4)	−3.44 (3.56)	1.809, ns
D. Fluency	Semantic fluency
D. Fluency	Number of words/min	11 (3.62)	9.36 (3.04)	1.184, ns
E. Planning	(st^a)	0.44 (0.81)	0.62 (1.08)	−0.482, ns

Norms for these standard scores are 0 ± 0.5.

Norms for these standard scores are 10 ± 3.

p < .05. **p < .01.

Table 3.

Mean Reaction Times (in Milliseconds), Accuracy Percentage, Standard Deviations, and t Test Results for the Lexicon Decision Task.

		ADHD	ADHD + RD	t test
Parameter	Response	M (SD)	M (SD)	t test
Reaction time (ms)	Correct	738.72 (175.72)	754.1 (153.29)	−0.204, ns
Reaction time (ms)	Error	760.15 (215.13)	763.25 (217.85)	−0.038, ns
Accuracy (%)	Correct	77.14 (11.31)	66.66 (15.02)	2.084*
Accuracy (%)	Error	20.83 (10.1)	29.52 (12.46)	−2.027*

p < .05.

Results

Reading and Executive Functions in ADHD + RD and ADHD Groups

In general, children with ADHD and ADHD + RD showed moderate to severe difficulties in several reading measures (accuracy for words/pseudowords and oral reading, speed for pseudowords reading, and accuracy for reading comprehension) and executive abilities (speed of processing battery, backward and total digit-span scores, and shifting scores). Children in the ADHD + RD group made significantly more reading errors and were significantly slower when reading words and pseudowords than children with ADHD. Participants with ADHD + RD were also significantly slower in part of the speed of processing battery (speed of naming letters and in the coding test). Although in some cases the ADHD + RD group exhibited much lower scores than the ADHD group (i.e., in reading: oral reading speed and accuracy, and reading comprehension accuracy; in executive functions: memory, part of the speed of processing battery, shifting, planning, and fluency), no significant differences were found in these parameters between the groups (see Table 2 for detail).

Lexical Decision Task

Behavioral Measures

Accuracy

In the ADHD group, t-test analyses yielded a significantly higher accuracy rate compared with the ADHD + RD group (Table 3).

RT

A trend of longer RT for the ADHD + RD group and differences in RT for words answered correctly/erroneously was found. These differences did not reach statistical significance (see Table 3).

Electrophysiology Measures

ERN and CRN components

Significant main effect of Group, F(2, 25) = 3.522, p < .05, η² = .22, and Type of Response, F(2, 25) = 4.409, p < .05, η² = .245, were found with greater amplitudes for the ADHD group compared with the ADHD + RD group, and greater amplitudes for erroneous compared with correct responses. A significant Group × Type of response interaction was also found, F(2, 25) = 4.358, p < .05, η² = .259, suggesting larger differences between ERN and CRN amplitudes in the ADHD group compared with the differences in the ADHD + RD group (see Figures 1 -3 and Tables 3 and 4 for these data).

Figure 1.

ERN and CRN waveforms in participants with ADHD.

Figure 2.

ERN and CRN amplitudes in participants with ADHD + RD.

Figure 3.

Topographical maps of the ERN and CRN components for the ADHD and ADHD + RD groups.

Table 4.

Means, Standard Deviations, and t Test Results for Response-Locked (ERN and CRN) and Stimulus-Locked (N00 and N400) in the Lexical Decision Task.

			ADHD	ADHD + RD	t test
	Electrode	Response	M (SD)	M (SD)	t test
CRN	AFz	Correct	−1.4 (1.8)	−1.1 (2.26)	−0.396
ERN	AFz	Error	−4.29 (3.73)	−1.56 (2.61)	−2.241*
CRN	Fz	Correct	−0.95 (2.5)	−0.85 (1.86)	0.407
ERN	Fz	Error	−3.9 (3.1)	−0.18 (2.41)	−0.353*
N100	Fz	Correct	−4.46 (1.99)	−2.57 (1.75)	−2.659*
	Fz	Error	−5.06 (3.29)	−3.08 (3.38)	−1.564
	FCz	Correct	−4.09 (1.88)	−2.38 (1.65)	−2.551*
	FCz	Error	−5.57 (3.59)	−2.51 (2.93)	−2.465*
N400	Fz	Correct	−7.79 (2.39)	−6.42 (2.63)	−1.446
	Fz	Error	−9.15 (4.78)	−4.99 (2.04)	−2.986**
	FCz	Correct	−7.06 (3.56)	−5.88 (2.64)	−0.994
	FCz	Error	−9.12 (4.84)	−5.72 (3.18)	−2.94*
	Cz	Correct	−5.68 (2.94)	−4.14 (2.78)	−1.426
	Cz	Error	−8.18 (4.42)	−4.94 (3.41)	−2.164*

Note. ERN = error-related negativity; CRN = correct-related negativity.

p < .05. **p < .01.

The N100 component

We then determined whether differences existed prior to error commission in the first attention stage between the two participant groups (Figures 4 and 5). A significant main effect of Group was found, F(2, 25) = 3.902, p < .05, η² = .238, with a greater N100 for the ADHD group than for the ADHD + RD group (Table 4, and Figures 4 and 5).

Figure 4.

Stimulus-locked N100 and N400 waveforms for erroneous and correct responses in participants with ADHD and ADHD + RD.

Figure 5.

Topographical maps of the N100 component for the ADHD and ADHD + RD groups.

The N400 component

Greater N400 was found for the ADHD group compared with the ADHD + RD group, main effect of group: F(3, 24) = 4.002, p < .05, η² = .33. Main effect of response was also observed, F(3, 24) = 3.745, p < .05, η² = .319, with greater N400 amplitudes prior to errors compared with correct responses (main effect of response; see Table 4 and Figures 4 and 6).

Figure 6.

Topographical maps of the N400 component for the ADHD and ADHD + RD groups.

Correlations Between Reading and Behavioral Measures

To examine the relationship between the words/pseudowords reading, and between reading ability and comprehension, Pearson correlations were performed.

Correlations between words/pseudowords read per minute and oral and reading comprehension

Lower error rates on the words/pseudowords task and faster reading speed were associated with higher comprehension scores.

Correlations between executive functions and reading measures

Significant Pearson correlations were obtained in the entire research population between several speed of processing and reading measures. Slower words/pseudowords reading was associated with slower speed of processing, lower memory scores, and poorer shifting ability. Inaccurate reading was also associated with poor shifting abilities and with lower fluency abilities (see Table 5 for these correlations).

Table 5.

Correlations Between Behavioral Reading and Executive Measures and Between ERP Components and Behavioral Measures in the Entire Sample (N = 28).

Measure	Variable 1	Variable 2	Measure used (as Variable 2)	Correlation (r)
Behavioral—single words/pseudowords and text reading	Oral reading speed	Reading accuracy	1-min words and pseudowords reading task	−.466, −.424
	Reading comprehension	Reading speed	1-min pseudowords reading task	−.468, −.488*
	Reading comprehension	Reading speed	Oral reading speed (real words reading)	.411*
Behavioral—speed of processing and reading	Object naming time	Reading accuracy	1-min pseudowords reading task	−.55**
	Object naming time	Reading accuracy	Real words reading (from lexical decision task)	−.512**
	RAS naming time	Reading speed	1-min pseudowords and words reading tasks	.48, .489*
Behavioral—memory and reading	Working memory (standard score digit span)	Reading speed	1-min pseudowords reading task	.519**
		Reading speed	Oral reading task (real words reading)	.513**
		Reading accuracy	1-min pseudowords and words reading tasks	−.442, −.406
Behavioral—shifting and reading	Shifting	Reading speed	1-min words reading task	.402*
Behavioral—shifting and reading	Shifting	Reading accuracy	Words reading (from lexical decision task)	.418*
Behavioral—fluency and reading	Fluency (semantic)	Reading accuracy	1-min pseudowords reading task	−.533**
Electrophysiology	ERN	Speed of processing	Coding	−.403*
		Speed of processing	RAS naming	−.45*
		Working memory	Digit span	−.384*
		Reading	Reading rate for words	−.385*
			Reading rate for pseudowords	−.388*
			Error rate in oral reading	.417*
	N400	Speed of processing	Symbol search	−.561**

Note. ERP = Event-Related Potentials; RAS = rapid alternating stimulus; ERN = error-related negativity. Description of the correlations between the behavioral measures and between the ERP measures (“Variable 1” column) and behavioral measures (“Variable 2” and “Measure used” columns). Correlation scores and p value are presented in the right column.

p < .05. **p < .01. ***p < .001.

Correlations Between Event-Related Potentials and Behavioral Measures

ERN and executive functions

Significant Pearson correlations in the entire research population between speed of processing parameters and ERN suggested that slower speed of processing was associated with a smaller ERN amplitude. Also, increased memory ability was associated with a larger ERN amplitude.

ERN and reading measures

Greater ERN amplitudes were associated with faster reading rates and fewer reading errors (see Table 5 for the correlation results).

N400 and executive functions

Faster speed of processing was significantly associated with greater N400 amplitude.

Discussion

The present study aimed to find an objective biomarker that will help to clearly differentiate IQ- and age-matched children with ADHD + RD from their peers with ADHD. Due to both populations’ deficits in reading and executive functions, we were looking for a nonbehavioral biomarker that reflects these abilities in the ADHD and ADHD + RD populations. These two groups generally have similar levels of performance in the classroom setting, a finding we validated in the present study, making it difficult to differentiate between them. Although we did not find any significant differences between the groups behaviorally, the ADHD + RD group showed a trend of greater severity of deficits. The ADHD + RD group struggled in the phonological and orthographical routes, read less words/pseudowords per minute, and made more reading errors, whereas the ADHD group read relatively more accurately, though still slowly. Automatic and fluent reading relies on automatic and fluent word recognition (Frith, 1985). Based on the relatively better reading scores of the ADHD group compared with the ADHD + RD group, we can postulate that ADHD participants experience better phonological ability (see Table 2). Still, while ADHD participants could read words accurately, they could not do this automatically (i.e., fast enough), for single words and for oral reading, enough to be considered normal readers. Oral reading was not only correlated with accuracy in single-word reading but was also correlated with decoding ability (e.g., pseudowords reading), corroborating findings by Fuchs, Fuchs, and Hosp (2001). The present study found that decoding ability was impaired in the ADHD group. Although oral reading and comprehension in both groups were impaired, accurate word reading in the ADHD group was less affected.

Going back to our initial question as to etiologies for these two disabilities, it seems that both groups share common behavioral characteristics, but our examination of the underlying mechanisms revealed differences between the two. Based on the standardized tests that were administered in the present study, we might ask instead how the ADHD group was able to “compensate” better than the ADHD + RD group, for the same difficulties they share.

Automaticity in reading is related to speed of processing abilities, meaning the ability to perceive information, process it, and provide an output (Horowitz-Kraus & Breznitz, 2011b). The correlations found between reading and speed of processing measures support this relatedness. Both groups demonstrated a severe deficit in speed of processing (as demonstrated by below normal scores in these measures, see Table 2), though the ADHD + RD group was significantly worse. The slow speed of processing together with concomitantly impaired memory abilities may create a “bottleneck” for phonological information entering the phonological loop (Breznitz & Share, 1992). We hypothesize that the heavy load of phonological information weighing on impaired working memory abilities and slow speed of processing (as manifested by low standard scores in the “digit-span” and “decoding” and “symbol search” subtests) result in erroneous and slow reading. The observed correlations suggest that reading difficulties worsen as memory and speed of processing are increasingly impaired. Then, when children with ADHD + RD are required to shift from automatized recognition of words to decoding (when encountering an unfamiliar word), shifting ability (which is impaired in this group) may also contribute to reading severity. It is important to note, though, that the relations between speed of processing and executive functions/the components of executive functions are yet to be discovered (Travis, 1998). Travis (1998) suggested that speed of processing is the most fundamental component underlying cognitive development (see also Bunce & Macready, 2005), and that the role of shaping development is reserved to executive functions maturation. Although several studies pointed at a possible relations between these two factors (Nelson, Yoash-Gantz, Pickett, & Campbell, 2009; Prins et al., 2005), there is not a clear conclusion as to the existence of causality or to the direction of it. Further study should look at this point in depth. Executive functions and reading ability can be demonstrated physiologically through activation of the error-detection mechanism while reading. In previous studies, we demonstrated that deficits in working memory and in speed of processing have a negative effect on the activation of the error-detection mechanism, resulting in decreased ERN amplitudes (Horowitz-Kraus & Breznitz, 2009, 2011a). We also found that reading difficulties caused a reduction in ERN amplitudes in children and adults with dyslexia (Horowitz-Kraus, 2011). These findings are consistent with the decreased ERN and the smaller differences between ERN and CRN amplitudes that were found in the ADHD + RD group compared with children with ADHD. It is possible that, because the ERN is part of the learning circuitry in the brain (Falkenstein et al., 1991), children with ADHD + RD experience difficulties in learning from their reading mistakes and therefore have more reading errors than the ADHD group. Also, and consistent with our hypothesis, the ADHD + RD group exhibited more severe impairments in reading and, physiologically, they showed decreased N400 amplitudes when performing reading errors. As the N400 component reflects the level of semantic incongruity (unsuitable semantic meaning) and might be related to the mental lexicon (see Horowitz-Kraus & Breznitz, 2008), it is possible that the impairment in error monitoring while reading affects one’s ability to learn from one’s reading mistakes, as we previously suggested, in relation to adults and children with dyslexia (Horowitz-Kraus, 2011).

Furthermore, the differences between the ADHD and the ADHD + RD groups are already present at the perception stage, even before the involvement of the mental lexicon. Our previous observation of decreased N100 amplitudes in readers with dyslexia (also reported by Bonte & Blomert, 2004; Brunswick & Rippon, 1994) suggests that poor reading ability might be accompanied by less efficient visual perception, which was supported in this study by decreased overall N100 amplitudes in the ADHD + RD group. This leads us to assume that the impaired cascade of information processing in children with ADHD + RD starts at the very first stage of perception. It might be possible that children with ADHD + RD perceive a stimulus inefficiently and then process the stimulus slower than the children in the ADHD group. They have relatively lower working memory capacity as well (below normal score in a digit-span test, see Table 2), so it becomes a more significant task to ascertain the semantic meaning of the presented words/pseudowords. Decreased N400 activation implies a lower mental lexicon capacity (and may be the absence of words stored in the lexicon, as we found in the significantly impaired orthographic ability), as evidenced by lessened activation of the error-detection mechanism in cases of reading error in the ADHD + RD group. There is evidence that a lower activation of the error-detection mechanism occurs when the desired or the actual responses are missing (Nieuwenhuis et al., 2001). This evidence might explain the decreased ERN in children with ADHD + RD. The ADHD + RD children showed lower activation of their error-detection mechanism, and they exhibited a smaller distinction between ERN and CRN amplitudes, which leads us to postulate that they might not be able to learn from their mistakes in an efficient manner; therefore, they continue making reading errors, even more so than the ADHD group.

Conclusion

This study suggests that although children with ADHD + RD and ADHD might seem to share common reading and executive disabilities, their observable similarities in behavioral symptoms represent only the tip of the iceberg. When we separated the different abilities into their subabilities (e.g., reading into phonology, orthography, and fluency; and executive functions into speed of processing, memory, fluency, and error-detection system), we found differences in these underlying mechanisms for basic reading and executive functions between the two groups. We therefore suggest that the ADHD + RD group is, in reality, more impaired than the ADHD group. The activation of the error-detection mechanism also pointed at differences between the groups and was found to reflect reading and executive disabilities. This particular mechanism is clearly demonstrated in ERN and CRN amplitudes; therefore, we suggest these endpoints as possible biomarkers to objectively differentiate between ADHD + RD and ADHD children. These results enable clinicians to treat both groups differently, according to their impairments in each domain (i.e., in reading and executive functions), and to evaluate the success of the chosen intervention by examining the error-detection mechanism activation (see Horowitz-Kraus & Breznitz, 2009).

These conclusions should be taken into account with the present study’s limitations. First, we worked with a relatively small sample size. Also, the inclusion of only two females in the ADHD + RD group, which was not an equal number to males, is a drawback. Finally, because the focus of the present study was to look at the differences between the ADHD + RD and the ADHD groups, we did not include a group of children only experiencing reading difficulties or even a nonclinical group of children. Therefore, we are unable to conclude whether either group showed significant impairment relative to a normative group for any measure. We did, however, use many normalized tests in this study, to be able to describe the performance of each group in relation to generalized normative data (using standard scores, see Table 2). Still, the present study’s results are capable of concluding that the results of children with ADHD + RD are different from those with only ADHD. A future study should enroll an additional group of children with only RD, as well as a nonclinical group of children, to better understand if ADHD + RD involves a separate etiology than RD and how ADHD + RD and ADHD differ from a nonclinical group. Third, although we suggested the ERN and CRN as possible biomarkers, the electroencephalogram used in the present study examines only cortical activation over the scalp. Therefore, we cannot conclude with certainty that the brain regions known to reflect executive functions, reading processing (e.g., orthographic reading: fusiform gyrus, phonological processing: angular gyrus) or semantic processing (inferior frontal gyrus; see van der Mark et al., 2011), are activated to a lower extent in children with ADHD + RD compared with ADHD. Studies using functional magnetic resonance imaging that can provide this spatial information may be able to validate our results and confirm that the error-detection mechanism activation is an objective measure that can differentiate the ADHD + RD population from the ADHD group. These studies could potentially support the previously mentioned “double dissociation” theory by showing whether the ADHD + RD demonstrates unique brain activation patterns that are absent in the ADHD group, or vice versa.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the Edmond J. Safra Research Foundation (Grant 22436).

Author Biography

Tzipi Horowitz-Kraus, PhD, is the program director of the newly formed Reading and Literacy Discovery Center at Cincinnati Children’s Hospital Medical Center. She received her BS (2002) in biology and her MS in neurobiology from the Department of Neurobiochemistry at the Tel-Aviv University, Israel (2004). Her MA in the clinical program for diagnosis and treatment of learning disabilities (Summa Cum Laude, 2007) and her PhD (2009) were awarded by the University of Haifa in Israel from the Edmond J. Safra Brain Center for the Study of Learning Disabilities. Her PhD focused on characterizing event-related potentials in children and adults with dyslexia and learning disabilities. After completion of her PhD, she conducted postdoctoral research at University of Haifa that determined behavioral and electroencephalographic measures following intervention programs for children with learning disabilities as well as studies aimed at developing objective measures to assess the effectiveness of interventions for reading difficulties. In 2011, She was awarded a Fulbright Scholarship and moved to the USA to join the Pediatric Neuroimaging Research Consortium Cincinnati Children’s Research Foundation to continue postdoctoral research using advanced neuroimaging tools, including functional magnetic resonance imaging (fMRI), diffusion tensor imaging (DTI), and electroencephalogram (EEG), to better understand the role of executive functions in reading disabilities in children to develop more effective interventions.

References

August

G. J.

Garfinkel

B. D.

(1990). Comorbidity of ADHD and reading disability among clinic-referred children. Journal of Abnormal Child Psychology, 18, 29-45.

Biederman

Monuteaux

M. C.

Doyle

A. E.

Seidman

L. J.

Wilens

T. E.

Ferrero

Faraone

S. V.

(2004). Impact of executive function deficits and attention-deficit/hyperactivity disorder (ADHD) on academic outcomes in children. Journal of Consulting and Clinical Psychology, 72, 757-766.

Bonte

M. L.

Blomert

(2004). Developmental dyslexia: ERP correlates of anomalous phonological processing during spoken word recognition. Brain Research Cognitive Brain Research, 21, 360-376.

Breznitz

(1989). Teacher evaluation of pupils’ reading performance: Criteria for assessment. Reading Improvement, 26, 174-180.

Breznitz

(1992). Effects of accelerated reading rate on memory for text. Journal of Educational Psychology, 84, 193-199.

Brunswick

Rippon

(1994). Auditory event-related potentials, dichotic listening performance and handedness as indices of lateralisation in dyslexic and normal readers. International Journal of Psychophysiology, 18, 265-275.

Bunce

Macready

(2005). Processing speed, executive function, and age differences in remembering and knowing. Quarterly Journal of Experimental Psychology, 58A, 155-168.

Castellanos

F. X.

Lee

P. P.

Sharp

Jeffries

N. O.

Greenstein

D. K.

Clasen

L. S.

Rapoport

J. L.

(2002). Developmental trajectories of brain volume abnormalities in children and adolescents with attention-deficit/hyperactivity disorder. Journal of the American Medical Association, 288, 1740-1748.

Clarke

A. R.

Barry

R. J.

McCarthy

Selikowitz

(2002). EEG analysis of children with attention-deficit/hyperactivity disorder and comorbid reading disabilities. Journal of Learning Disabilities, 35, 276-285.

10.

CogniFit Personal Coach. (2008). Training program and database. Yokneam, Israel: CogniFit.

11.

Conners

C. K.

(1989). Conners’ rating scales manual. North Tonawanda, NY: Multi-Health System.

12.

de Jong

C. G. W.

Oosterlaan

Sergeant

J. A.

(2006). The role of double dissociation studies in the search for candidate endophenotypes for the comorbidity of attention deficit hyperactivity disorder and reading disability. International Journal of Disability, Development and Education, 53, 177-193.

13.

Denckla

M. B.

Rudel

R. G.

(1976). Rapid automatized naming (RAN): Dyslexia differentiated from other learning disabilities. Neuropsychologia, 14, 471-479.

14.

Falkenstein

Hohnsbein

Hoormann

Blanke

(1991). Effects of crossmodal divided attention on late ERP components. II. Error processing in choice reaction tasks. Electroencephalography and Clinical Neurophysiology, 78, 447-455.

15.

Faraone

S. V.

Sergeant

Gillberg

Biederman

(2003). The worldwide prevalence of ADHD: Is it an American condition? World Psychiatry, 2, 104-113.

16.

Frith

(1985). Beneath the surface of developmental dyslexia. In Patterson

K. E.

Marshall

J. C.

Coltheart

(Eds.), Surface dyslexia (pp. 301-330). London, England: Erlbaum.

17.

Frost

(2001). Unpublished word list in Hebrew. Jerusalem, Israel: Hebrew University. Retrieved from http://atar.mscc.huji.ac.il/~frost

18.

Fuchs

L. S.

Fuchs

Hosp

K. M.

(2001). Oral reading fluency as an indicator of reading competence: A theoretical, empirical and historical analysis. Scientific Studies of Reading, 5, 239-256.

19.

Gooch

Snowling

Hulme

(2011). Time perception, phonological skills and executive function in children with dyslexia and/or ADHD symptoms. Journal of Child Psychology and Psychiatry, 52, 195-203.

20.

Gratton

Coles

M. G. H.

Donchin

(1983). A new method for off-line removal of ocular artifact. EEG Clinical Neurophysiology, 55, 468-484.

21.

Greven

C. U.

Rijsdijk

F. V.

Asherson

Plomin

(2012). A longitudinal twin study on the association between ADHD symptoms and reading. Journal of Child Psychology and Psychiatry, 53, 234-242.

22.

Horowitz-Kraus

(2011). Does development affect the error-related negativity of dyslexic and skilled readers in reading? An ERP study. Developmental Neuropsychology, 36, 914-932.

23.

Horowitz-Kraus

Breznitz

(2008). An error-detection mechanism in reading among dyslexic and regular readers: An ERP study. Clinical Neurophysiology, 119, 2238-2246.

24.

Horowitz-Kraus

Breznitz

(2009). Can the error detection mechanism benefit from training the working memory? A comparison between dyslexics and controls: An ERP study. PLoS ONE, 4, e7141.

25.

Horowitz-Kraus

Breznitz

(2011a). Error detection mechanism at words and sentences: a comparison between readers with dyslexia and skilled readers. International Journal of Developmental Disabilities and Education, 58, 33-45.

26.

Horowitz-Kraus

Breznitz

(2011b). Reaction time and accuracy in erroneous vs correct responses among dyslexic and regular readers: From letters to sentences. Dyslexia, 17, 72-84.

27.

Horowitz-Kraus

Breznitz

(2013). Compensated dyslexics have a more efficient error detection system than non-compensated dyslexics. Journal of Child Neurology, 28, 1266-1276. doi:10.1177/0883073812460917

28.

Jacobson

L. A.

Ryan

Martin

R. B.

Ewen

Mostofsky

S. H.

Denckla

M. B.

Mahone

E. M.

(2011). Working memory influences processing speed and reading fluency in ADHD. Child Neuropsychology, 17, 209-224.

29.

Jasper

H. H.

(1958). The ten-twenty electrode system of the International Federation of Societies for Electroencephalography: Appendix to report of the committee on methods of clinical examination in electroencephalography. EEG Clinical Neurophysiology, 10, 371-375.

30.

Johnson

(2002). Behaviour problems in children and adolescents with learning disabilities. Internet Journal of Mental Health. Retrieved from http://www.ispub.com/ostia/index.php?xmlFilePath=journals/ijmh/vol1n2/learning.xml

31.

Katz

L. J.

Brown

F. C.

Roth

R. M.

Beers

S. R.

(2011). Processing speed and working memory performance in those with both ADHD and a reading disorder compared with those with ADHD alone. Archives of Clinical Neuropsychology, 26, 425-433.

32.

Kutas

Federmeier

K. D.

(2000). Electrophysiology reveals semantic memory use in language comprehension. Trends in Cognitive Sciences, 412, 463-470.

33.

Kutas

Hillyard

S. A.

(1980). Reading senseless sentences: Brain potentials reflect semantic incongruity. Science, 207, 203-205.

34.

Mangina

C. A.

Beuzeron-Mangina

(2009). Similarities and differences between learning abilities, “Pure” learning disabilities, “Pure” ADHD and comorbid ADHD with learning disabilities. International Journal of Psychophysiology, 73, 170-177.

35.

Mannuzza

Klein

R. G.

Bessler

Malloy

LaPadula

(1993). Adult outcome of hyperactive boys. Educational achievement, occupational rank, and psychiatric status. Archives of General Psychiatry, 50, 565-576.

36.

Marzocchi

G. M.

Oosterlaan

Zuddas

Cavolina

Geurts

Redigolo

Sergeant

J. A.

(2008). Contrasting deficits on executive functions between ADHD and reading disabled children. Journal of Child Psychology and Psychiatry, 49, 543-552.

37.

Nelson

L. A.

Yoash-Gantz

R. E.

Pickett

T. C.

Campbell

T. A.

(2009). Relationship between processing speed and executive functioning performance among OEF/OIF veterans: Implications for postdeployment rehabilitation. Journal of Head Trauma Rehabilitation, 24, 32-40.

38.

Nicolson

R. I.

Fawcett

A. J.

Dean

(1995). Time estimation deficits in developmental dyslexia: Evidence of cerebellar involvement. Biological Sciences/The Royal Society, 259, 43-47.

39.

Nieuwenhuis

Ridderinkhof

K. R.

Blom

Band

G. P. H.

Kok

(2001). Error-related brain potentials are differentially related to awareness of response errors: Evidence from an antisaccade task. Psychophysiology, 38, 752-760.

40.

Pennington

B. P.

Groisser

Welsh

M. C.

(1993). Contrasting cognitive deficits in attention deficit hyperactivity disorder versus reading disability. Developmental Psychology, 29, 511-523.

41.

Prins

N. D.

van Dijk

den Heijer

Vermeer

S. E.

Jolles

Koudstaal.

P. J.

Breteler

M. M. B.

(2005). Cerebral small-vessel disease and decline in information processing speed, executive function and memory. Brain, 128, 2034-2041.

42.

Purvis

K. L.

Tannock

(2000). Phonological processing, not inhibitory control, differentiates ADHD and reading disability. Journal of the American Academy of Child and Adolescent Psychiatry, 39, 485-494.

43.

Raven

J. C.

(1969). Guide to the standard progressive matrices. London, England: H. K. Lewis.

44.

Rucklidge

J. J.

Tannock

(2002). Neuropsychological profiles of adolescents with ADHD: Effects of reading difficulties and gender. Journal of Child Psychology and Psychiatry, 43, 988-1003.

45.

Semrud-Clikeman

Walkowiak

Wilkinson

Minne

E. P.

(2010). Direct and indirect measures of social perception, behavior, and emotional functioning in children with Asperger’s disorder, nonverbal learning disability, or ADHD. Journal of Abnormal Child Psychology, 38, 509-519.

46.

Sesma

H. W.

Mahone

E. M.

Levine

Eason

S. H.

Cutting

L. E.

(2009). The contribution of executive skills to reading comprehension. Child Neuropsychology, 15, 232-246.

47.

Sexton

C. C.

Gelhorn

Bell

Classi

(2012). The co-occurrence of reading disorder and ADHD: Epidemiology, treatment, psychosocial impact, and economic burden. Journal of Learning Disabilities, 45, 538-564. doi:10.1177/0022219411407772

48.

Shany

Lachman

Shalem

Bahat

Zeiger

(2006). “Aleph-Taph” a test for the diagnosis of reading and writing disabilities, based on national Israeli norms. Tel Aviv, Israel: Yesod.

49.

Shatil

Nevo

Breznitz

(2008). Elull learning disabilities diagnostic test—The reading sub-test. Haifa, Israel: University of Haifa.

50.

Skounti

Philalithis

Galanakis

(2007). Variations in prevalence of attention deficit hyperactivity disorder worldwide. European Journal of Pediatrics, 166, 117-123.

51.

Trani

D. M.

Casini

M. P.

Capuzzo

Gentile

Bianco

Menghini

Vicari

(2011). Executive and intellectual functions in attention-deficit/hyperactivity disorder with and without comorbidity. Brain & Development, 33, 452-469.

52.

Travis

(1998). Cortical and cognitive development in 4th, 8th and 12th grade students: The contribution of speed of processing and executive functioning to cognitive development. Biological Psychology, 48, 37-56.

53.

Vakil

Blachstein

Wertman-Elad

Greenstein

(2012). Verbal learning and memory as measured by the Rey-Auditory Verbal Learning Test: ADHD with and without learning disabilities. Child Neuropsychology, 18, 449-466. doi:10.1080/09297049.2011.613816

54.

van der Mark

Klaver

Bucher

Maurer

Schulz

Brem

Brandeis

(2011). The left occipitotemporal system in reading: Disruption of focal fMRI connectivity to left inferior frontal and inferior parietal language areas in children with dyslexia. NeuroImage, 54, 2426-2436.

55.

Van de Voorde

Roeyers

Wiersema

J. R.

(2010). Error monitoring in children with ADHD or reading disorder: An event-related potential study. Biological Psychology, 84, 176-185.

56.

Wechsler

(1991). Wechsler intelligence scale for children (3rd ed.). San Antonio, TX: The Psychological Corporation.

57.

Willcutt

E. G.

Doyle

A. E.

Nigg

J. T.

Faraone

S. V.

Pennington

B. F.

(2005). Validity of the executive function theory of attention-deficit/hyperactivity disorder: A meta-analytic review. Biological Psychiatry, 57, 1336-1346.

58.

Willcutt

E. G.

Pennington

B. F.

Boada

Ogline

J. S.

Tunick

R. A.

Chhabildas

N. A.

Olson

R. K.

(2001). A comparison of the cognitive deficits in reading disability and attention-deficit/hyperactivity disorder. Journal of Abnormal Psychology, 110, 157-172.

59.

Wolf

Bally

Morris

(1986). Automaticity, retrieval processes, and reading: A longitudinal study in average and impaired readers. Child Development, 57, 988-1000.