Persistent Handwriting Difficulties in Children With ADHD After Treatment With Stimulant Medication

Abstract

Objective: Children with ADHD often present with handwriting difficulties. However, the extent to which motor and attention skills influence performance in this group has not yet been explored. The objective of this study was to examine the factors associated with change in handwriting performance. Method: This study examines the factors associated with change in handwriting performance of 49 children newly diagnosed with ADHD (mean age = 8.4 [SD=1.3] years) prior to and 3 months following use of a stimulant medication. Results: Handwriting legibility and speed improved significantly at follow-up evaluation. However, most of the children with legibility difficulties at baseline continued to demonstrate difficulties when evaluated 3 months after initiation of medication. Change in handwriting legibility was best determined by improvements in visual-motor integration skills (β = 0.07-0.10; p < .001), while the change in speed did not appear to be consistently related to a single factor. Conclusion: Handwriting difficulties are common in children with ADHD, and medication alone is not sufficient to resolve these challenges.

Keywords

attention deficit hyperactivity disorder handwriting motor skill disorder motor performance methylphenidate stimulant medication

Introduction

Motor difficulties are frequently reported in children with ADHD, and indeed the coexisting developmental coordination disorder (DCD) occurs in approximately 50% (Kaplan, Wilson, Dewey, & Crawford, 1998; Miyahara, Mobs, & Doll-Tepper, 2001; Pitcher, Piek, & Hay, 2003; Watemberg, Waiserberg, Zuk, & Lerman-Sagie, 2007). Use of stimulant medication is the current standard of practice for treating individuals with ADHD, and its positive impact on disruptive behavior has been widely demonstrated (Santosh, 2007). In fact, the impact of the medication often extends beyond behavioral improvement to include improved motor skills for a subset of children (Bart, Podoly, & Bar-Haim, 2010; Brossard-Racine, Shevell, Snider, Bélanger, & Majnemer, 2012; Flapper, Houwen, & Schoemaker, 2006).

A high prevalence of handwriting difficulties has been reported in children with ADHD (Brossard-Racine, Majnemer, Shevell, & Snider, 2008). There is a common belief that stimulant medication also improves handwriting performance; however, the level of evidence supporting this is weak as the majority of studies that evaluated handwriting production used qualitative observations of handwriting. While some reported a significant improvement (Lerer, Lerer, & Artner, 1977; Tucha & Lange, 2001; Whalen, Henker, & Finck, 1981), others did not observe a notable change (Lufi & Gai, 2007) between handwriting quality of children with or without medication intervention. To our knowledge, only two studies have used standardized evaluations of handwriting in children with ADHD. Rosenblum, Epsztein, and Josman (2008) did not observe a change in handwriting quality, whereas Flapper et al. (2006) did. These discrepancies can be explained in part by the difference in the handwriting evaluation tools used and also by the fact that Flapper et al. (2006) specifically recruited children with a dual diagnosis of ADHD-DCD. Preliminary evidence has identified a link between motor capacity, visual-motor integration skills, and handwriting legibility in medication-naïve children newly diagnosed with ADHD (Brossard-Racine, Majnemer, Shevell, Snider, & Bélanger, 2011). However, the positive influence of stimulant medication on motor and handwriting skills has not been investigated. Flapper et al. (2006) reported that methylphenidate improved fine motor skills and handwriting quality in children with ADHD-DCD. Nevertheless, the relationship between motor and handwriting skills was not reported. The effect of medication on fine motor skills in children with ADHD-DCD might have partially explained the change in handwriting quality. The impact of stimulants on motor skills of children with ADHD but without DCD appears to be of lower magnitude when compared with children with the dual condition (Brossard-Racine, Shevell, et al., 2012); therefore interaction between attention, motor, and handwriting skills might be of a different order than in children with simple ADHD. The extent to which motor and/or behavioral difficulties underlie performance in everyday activities, such as handwriting, has not been adequately studied in children with ADHD, with and without motor difficulties.

Therefore, the primary objective of this study was to identify the determinants of handwriting performance in a community-based sample of children newly diagnosed with ADHD. We conducted a pre/post cohort study to evaluate whether the use of stimulant medication alone, as typically implemented in a community setting, was associated with handwriting performance changes and evaluated potential factors associated with improvement in handwriting performance 3 months following daily use of stimulant medication.

Method

Participants

Children newly diagnosed with ADHD, for whom a long-acting stimulant medication had been prescribed by their evaluating physician, were recruited. ADHD diagnosis was made following the physician’s comprehensive investigation with the child, his family, and his schoolteacher as recommended by the Canadian Attention Deficit Hyperactivity Disorder Resource Alliance’s (CADDRA; 2011) practice guidelines. Diagnosis and subtype were confirmed by applying Diagnostic and Statistical Manual of Mental Disorders (4th ed.; DSM-IV; American Psychiatric Association, 1994) criteria. French-speaking or English-speaking children between 6 and 11 years of age were included. Exclusion criteria were the following: (a) intelligence quotient (IQ) < 80, (b) congenital malformation affecting the dominant upper limb, (c) other neurological or developmental disorders (e.g., Tourette syndrome, autism, cerebral palsy), (d) use of an assistive device to write independently (e.g., hand-splint, prosthesis), (e) another medication that could affect motor skills or behavior (e.g., psychostimulant or antiepileptic), and (f) previous or current rehabilitation services to improve motor skills.

Assessment Measures

Handwriting

The Evaluation Tool of Children’s Handwriting (ETCH; Amundson, 1995) is a criterion-referenced standardized evaluation of handwriting legibility and speed. Both a cursive (ETCH-C) and a manuscript (ETCH-M) version have been developed and can be used in children from Grade 1 and over. The child is asked to complete handwriting tasks (described below) in the manner that the child would typically perform these tasks (i.e., without specific instruction on the quality and speed of execution). Tasks are representative of the writing tasks the children experience in the typical school environment and include alphabet writing from memory in lower/uppercase letters, numeral writing from memory, near-point copying, far-point copying, dictation, and sentence composition. Once the evaluation is completed, word, letter, and/or numeral legibility are scored using specific criteria in the manual. Legibility scores for each task are averaged, and a total percentage (0%-100%) of word, letter, and numeral legibility is derived. Higher scores imply better legibility. The time to complete each task is recorded in seconds and then transformed into speed scores (letters/minute). The interrater reliability is acceptable (Pearson coefficient average r = .84) for the manuscript version and cursive version (r = .88; Amundson, 1995). In the ETCH-M, test–retest reliabilities were r = .63-.77 (Diekema, Deitz, & Amundson, 1998) with total scores being more consistent than individual tasks. For the cursive version, test–retest reliability intraclass correlations (ICC) for the total scores were .24 for numeral, .65 for words, and .61 for letters (Duff & Goyen, 2010). Cutoff values (normal/abnormal) were not established by the developer; however, preliminary values were recently published for the ETCH-M (Brossard-Racine, Mazer, Julien, & Majnemer, 2012) and ETCH-C (Koziatek & Powell, 2002). Furthermore, cutoff values derived from typically developing Canadian first graders (Feder, Majnemer, Bourbonnais, Blayney, & Morin, 2007) using −1.5 standard deviation below the mean have been reported. ETCH cutoff values used in this study appear in Table 1.

Table 1.

ETCH Cutoff Values Used to Discriminate Children With Handwriting Difficulties From Children Without Difficulties.

Studies	Grade	Legibility components	Cutoff values (%)
Feder, Majnemer, Bourbonnais, Blayney, and Morin (2007)	1	ETCH-M Total Word	0.45^a
Feder, Majnemer, Bourbonnais, Blayney, and Morin (2007)		ETCH-M Total Letter	0.64^a
Brossard-Racine, Mazer, et al. (2012)	2-4	ETCH-M Total Word	0.75
Brossard-Racine, Mazer, et al. (2012)		ETCH-M Total Letter	0.76
Koziatek and Powell (2002)	4	ETCH-C Total Word	0.75
		ETCH-C Total Letter	0.82

Note: ETCH-C = Evaluation Tool of Children’s Handwriting-Cursive; ETCH-M = Evaluation Tool of Children’s Handwriting-Manuscript.

Determined by the mean minus 1.5 standard deviation.

Behavior

The Parent version of the Conners’ Global Index (CGI-P; Conners, 2001) is a norm-referenced 10-item questionnaire that identifies children with behavioral difficulties. Both hyperactive and inattentive symptoms are captured by this measure. Questions are classified into (a) the restless-impulsivity index and (b) the emotional lability index. When summed together, these subscales generate the CGI-P total score. Items are scored on a 4-point scale and then transformed into a T-score (standard score) that can be compared with normative data for age and gender before transformation into a percentile. Higher scores indicate more severe ADHD symptoms. A T-score ≥61 is a marker of significant problems, while scores between 56 and 60 are considered slightly atypical (borderline).

Motor skills

The Movement Assessment Battery for Children (M-ABC; Henderson & Sugden, 1992) is a norm-referenced standardized evaluation used to discriminate children with or without motor difficulties between 4 and 12 years of age. Eight tasks are grouped under three subscales: (a) manual dexterity index, (b) ball skills index, and (c) balance index. Together, the sum of these indices yields the total impairment score. A higher score means poorer performance. Standard scores can be transformed into percentiles. Scores between the 5th and 15th percentiles indicate risk for motor difficulties (borderline performance), whereas scores <5th percentile indicate definitive motor problems.

The Developmental Test of Visual Motor Integration (VMI; Beery, Buktenica, & Beery, 2005) is a standardized evaluation that discriminates between children with and without impaired visual-motor integration skills. A maximum of 30 shapes can be copied and scored according to specific criteria. Impairment was determined by standard scores < 85.

Procedures

At diagnosis of ADHD, a hospital- or community-based pediatrician or neurologist introduced the study to the family. Introduction of the medication and individual’s dosage titration respected CADDRA’s guidelines and was conducted as part of the patient follow-up with his or her physician as typically conducted in the community. Parents who did not wish to comply to use of the prescribed medication for their child were not included. Only medication-naïve children were enrolled in the study. Recruitment took place at the two pediatric hospitals of the region (ADHD clinic, neurology outpatient clinic) and in several community-based pediatric clinics, between 2007 and 2011.

If no IQ report was available, the Wechsler Abbreviated Scale of Intelligence (WASI; Wechsler, 1999) was administered by a psychologist. An occupational therapist (OT) then performed the baseline assessment (T0) which included the ETCH, the M-ABC, and the VMI. Parents completed the CGI-P and a short demographic questionnaire. Three months after initiation of stimulant medication, the baseline assessments were repeated (T2). No other intervention had been administered between the two evaluations. Evaluators were blinded to the specific study hypotheses and to medication status. In addition, all evaluations were independently scored by another OT, trained for scoring consistency and blinded to the child’s identity and timing of testing (pre/post). Parental informed consent and child’s assent were obtained prior to the first evaluation. This study had ethical approval from the hospital’s Research Ethics Board and scientific approval from the Montreal Children’s Hospital-Research Institute.

As a secondary hypothesis, we added another testing point for handwriting to capture the immediate effect of the stimulant medication on handwriting performance. The ETCH was readministered within the 1st week after initiation of medication (T1). For the last two assessments (T1 and T2), we verified that the child had taken their medication that day and that the evaluation was conducted when the child was still under the effect of the stimulant (2-8 hr post-ingestion).

Data Analysis

Descriptive statistics were used to characterize the sample and performance on all measures. Assumption of normality was first verified. Wilcoxon signed rank tests were conducted to identify differences in handwriting between two evaluation periods (T0/T1/T2).

Multivariate analyses were performed to obtain adjusted estimates of motor and behavior effects on handwriting legibility and speed at T2. Mixed effect models with random intercept for participants were used to determine whether change in motor and behavior scores was related with change in handwriting outcomes between T0 and T2. For each outcome, models were selected based on the Akaike Information Criterion and validated using fivefold cross-validation. All multivariate models contained a restricted cubic spline (Harrell, 2001) for age to correctly capture the relationship between the outcome and age. Visits with missing data for at least one variable in the model were omitted. Diagnostic plots of the residuals were produced for final multivariate models, and important outliers were deleted. Analyses were performed using SAS 9.0 and R 2.13.

Results

Group Characteristics

Sixty-eight children were referred to the project. Of those, 13 children were subsequently excluded because (a) 1 had received rehabilitation services for fine motor difficulties, (b) 2 were outside the age range targeted, (c) 5 had already started using the stimulant medication by the time they were contacted by the research coordinator, (d) 2 families did not intend to use the medication, and (e) 3 lived too far away to accommodate assessment. Therefore, a total of 55 children were enrolled in the study. Of those, six children completed baseline assessment but not the T2 assessment because either the family decided to stop medication during the 3-month period (1) or not to use the medication as planned (4) or to switch to a short-acting methylphenidate (1). These 6 children were only included in the linear mixed-effect models of change. The majority of the children were in Grade 2 (23/55), followed by Grade 4 (12/55), then Grade 1 (8/55), Grade 3 (7/55), Grade 5 (4/55), and Grade 6 (1/55). Total family income was greater than CAD 60,000 for 80.0% of the sample. Spoken language at school was French for more than half of the sample (32/55), English in 18/55, and 5 children attended an officially bilingual school (French–English). According to the ADHD classification of the DSM-IV, 4 children were typically hyperactive, 27 were predominantly inattentive, and 24 fit within the combined subtype. A total of 49 children completed T0 and T2 assessments (10 female; mean age 8.4 [1.3] years). For 29 children of the 49 followed, the long-acting stimulant medication taken was Concerta (methylphenidate chlorhydrate, long-acting); 16 children took Biphentin (methylphenidate chlorhydrate, long-acting), while 5 were taking Adderal-XR (mixed amphetamine salts, long-acting). The 6 children who did not complete all of the study were not significantly different in age or on baseline assessment results than the rest of the sample. Mean time between T0 and T2 was 132.9 (SD=51.4) days, and T2 was performed on an average of 105.5 (SD=21.2) days (3.5 months) after initiation of medication.

Detailed description of the findings on the behavior and motor subscales before and after medication have been reported elsewhere (Brossard-Racine, Shevell, et al., 2012). Overall, significant improvements were observed on all three CGI-P subscales between T0 and T2 (mean difference = 10.67 (SD=9.91), p < .001). CGI-P Total index ranged from 41 to 90 (M = 67.3 (SD=11.9)) at baseline and ranged from 40 to 80 (M = 55.6 (SD=9.3)) after treatment. Twenty-five children (51.0%) continued to score in the clinical range for behavioral difficulties (CGI-P ≥ 56) at T2. Motor impairment was highly prevalent at baseline when 73.5% (36/49) of the sample obtained an M-ABC total impairment score under the 15th percentile. At T2, M-ABC total impairment score improved significantly, t(48) = 5.62, p < .001, mean difference = 3.97 (SD=4.94); however, motor impairment persisted in 55.1% (27/49) of the sample. At all instances of evaluation, manual dexterity was the most impaired motor index and remained impaired in 59.1% (29/49) of the sample at T2. Change in visual-motor integration did not reach significance.

Handwriting Performance

Forty-three children were tested with the ETCH-M and 12 with the ETCH-C. Most of the participants were right handed (48/55). Handwriting legibility and speed was extremely variable between children at all points of testing. Baseline handwriting skills in children newly diagnosed with ADHD, prior to the use of medication, has been previously reported (Brossard-Racine et al., 2011).

At baseline, total word legibility could not be obtained for two participants due to the extremely poor produce of the children. Overall, word legibility difficulties were present in 44.7% (21/47) of the sample and letter legibility difficulties in 65.3% (32/49), while at T2, difficulties were now present in 38.8% (19/49) of the sample in terms of word legibility and in 42.9% (21/49) of the sample for letter legibility based on the evaluation’s cutoffs. Legibility difficulties persisted in most (i.e., presented under cutoff value at both evaluations), with 14 of 21 with word legibility problems (66.7%) and 18 of 32 with letter legibility difficulties (56.3%) demonstrating ongoing performance challenges. Chi-square analyses to determine whether word and letter legibility difficulties were more frequent in children with persisting motor difficulties (M-ABC total impairment ≤15th percentile at T2) or without motor difficulties were not significant. Means and standard deviation of the handwriting outcomes at T0 and T2 are presented in Table 2 with tests of differences.

Table 2.

Handwriting Performance (M ± SD) at Baseline and After Treatment and Results of the Wilcoxon Signed Ranks Tests (n = 49).

	T0	T2	Z	p value
Legibility (%)
Total word	67.4 ± 28.0	74.0 ± 23.5	−2.24	.025*
Total letter	69.6 ± 17.3	74.3 ± 15.3	−2.46	.014**
Total numeral	78.9 ± 18.3	82.0 ± 16.3	−1.55	.120
Speed
Alphabet lowercase (s)	126.1 ± 91.8	108.0 ± 85.0	−2.63	0.009**
Alphabet uppercase (s)	125.3 ± 70.4	110.5 ± 51.9	−1.63	.104
Near-point copying (letters/min)	27.5 ± 13.0	35.4 ± 16.5	−3.09	.002**
Far-point copying (letters/min)	25.9 ± 10.6	31.2 ± 14.9	−2.57	.010**
Sentence composition (letters/min)	37.8 ± 18.6	43.6 ± 19.7	−2.125	.034*

Note: T0 = baseline evaluation; T2 = after 3 months of treatment with stimulant medication; s = seconds; min = minute.

p ≤ .05. **p ≤ .01.

As the study was already underway when we made the addendum to the protocol to explore the direct effect of the medication on the handwriting performance, only 29/55 children completed the evaluation at T1, which was conducted on average 5.6 (SD=4.9) days after initiation of medication. Demographic information, motor, and behavior scores at baseline were not significantly different in children who were asked to participate in T1 when compared with those who did not participate. No significant difference in any of the handwriting legibility scores was found between either T0 and T1 or between T1 and T2. Three children who performed below the letter legibility cutoff and one who performed below the word legibility cutoff at baseline improved enough to move up to the level without handwriting difficulty at T1. These children remained in the latter group at T2. Regarding speed, the only significant difference found was for the lowercase alphabet writing speed from memory tasks between T0 and T1 (mean T0 = 130.95, mean T1 = 98.73, Z = −2.33, p = .02). Overall, group mean comparisons between T0 and T2 indicated that children wrote significantly faster (shorter time) in all tasks at T2 except in the uppercase alphabet written from memory tasks, which did not reach significance.

A minimal change of 10% in total word legibility and 6% in total letter legibility has been reported to correspond to a minimal clinically important difference (MCID; i.e., a change visually detected by expert clinicians; Brossard-Racine, Mazer, et al., 2012). At T2, for word legibility, in terms of proportion of the sample, 36.2% (17/47) of participants improved by at least 10%, 17.0% (8/47) got worse by at least 10%, and 46.8% changed by less than 10% (22/47). For letter legibility, 40.8% (20/49) increased by 6%, 42.9% (21/49) did not change enough to correspond to a MCID, and 16.3% (8/49) of the sample became worse by at least 6%. Only 20.4% (10/49) of the sample improved sufficiently to be clinically detectable by expert clinicians in both word and letter legibility. Children who increased at least MCID criterion in word legibility had significantly worse word, χ²(2, 47) = 11.34, p = .003, and letter legibility, χ²(2, 49) = 12.36, p = .002, at baseline when compared with the other two groups. Also, children who improved enough to move above the legibility cutoff had significantly greater percentage change in word, χ²(3, 47) = 23.60, p < .001, and letter legibility, χ²(3, 49) = 18.19, p < .001.

Factors Associated With Change in Handwriting Performance

Mixed effect models were used to look at the effect of change scores in the motor and behavioral assessments as well as change scores in handwriting (dependent variable). The standardized coefficients with their p values for the motor and behavior variables selected in the final models are presented in Table 3. Consistently, change in VMI score was associated with change in handwriting legibility. Change in speed of writing was associated with motor change in different subscales to include change in M-ABC total impairment index, balance index, and VMI score. No significant difference in the amount of change was found between ADHD subtypes and between children with and without persisting motor difficulties.

Table 3.

Multivariate Mixed Models of Motor and Behavioral Skills Associated With Change in Handwriting Performance Before and After Use of Medication.

Dependent variables (outcomes)
Independent variables^a	n ^a	Standardized β coefficient	p value (coefficient)
Word legibility ^b	55
VMI		0.10	<.001
Letter legibility ^b	55
VMI		0.06	<.001
Numeral legibility^c	55
VMI		0.07	<.001
Alphabet lower-case ^d	53
CGI-P Total		12.60	.026
M-ABC ball skills index		−3.36	.583
Alphabet upper-case ^b	54
CGI-P restless impulsivity		4.91	.381
M-ABC balance index		17.69	.006
VMI		15.42	.018
Near-point copying speed ^e	48
CGI-P Total		−1.96	.259
M-ABC Total impairment		−3.49	.043
Far-point copying speed ^e	48
CGI-P Total		−0.41	.795
Sentence composition speed ^f	50
CGI-P Total		1.36	.527
M-ABC Total impairment		−4.16	.053
VMI		−3.50	.090

Note: CGI-P = Connor Global Index–Parental version; M-ABC = Movement ABC; VMI = The Developmental Test of Visual Motor Integration.

Number of subjects with at least one visit used in the analysis.

Adjusted for a cubic spline with three knots for age.

Adjusted for gender and a cubic spline with three knots for age.

Adjusted for language spoken at school and a cubic spline with three knots for age.

Adjusted for time since baseline and a cubic spline with three knots for age.

Adjusted for time since baseline, language spoken at home, income and a cubic spline with three knots for age.

Factors Associated With Handwriting Performance After Treatment With Medication

Significant multivariate models for the group as a whole and for children with or without persisting motor difficulties are presented in Table 4. Consistently, handwriting legibility was best modulated by performance on the VMI. No significant model was noted for speed in any of the tasks evaluated at T2.

Table 4.

Multiple Linear Regression Models of Variables Associated With Handwriting Legibility Following Pharmacologic Treatment in Children With ADHD (N = 49).

Dependent variables
Independent variables^a	Adj R²	SMSE	β¹	p value
Word legibility	0.44	20.1		<.001
M-ABC bi			0.04	.15
VMI			0.12	<.001
Letter legibility	0.52	13.9		<.001
M-ABC bi			0.04	.04
VMI			0.08	<.001
Numeral legibility	0.34	15.8		<.001
VMI			0.07	<.001

Note: adj R² = adjusted R²; SMSE = square root of mean square error obtained with cross-validation; β¹ = standardized beta coefficient.

Coefficients are adjusted for age (there were no other important explanatory variables or confounder); M-ABC bi = Movement ABC balance skills index; VMI = the Developmental Test of Visual Motor Integration.

Discussion

With this study, we found that children with ADHD, with or without motor difficulties, significantly improved in handwriting legibility 3 months after initiation and treatment with stimulant medication. This corroborates preliminary studies that either used qualitative measures of Latin-based handwriting (Lerer et al., 1977; Tucha & Lange, 2001; Whalen et al., 1981) or quantitative measure of handwriting but in children with ADHD-DCD (Flapper et al., 2006). This is the first study that compares amount of change in handwriting performance to a predetermined MCID in handwriting as established by expert clinicians to better appreciate the clinical relevance of the “statistically significant change” after intervention. We found that about a third of the sample improved by at least a MCID in word and letter legibility, and these were more likely to be the children with the most apparent handwriting difficulties at baseline. These findings are suggesting that the medication has a beneficial effect primarily in children with ADHD who presented more severe handwriting difficulties at diagnosis. However, although mean performance of the sample significantly improved, more than half of the children who demonstrated handwriting difficulties at baseline continued to demonstrate challenges over time. Unfortunately, due to the pre/post setting of the study, it is difficulty to affirm that the modest change observed is entirely attributed to the effects of the medication rather than maturation or typical practice over the 3-month period. Moreover, the additional exploratory period of testing (T1) that was introduced within the 1st week of medication initiation did not reveal a substantive change in speed or legibility in the subset that was evaluated. Such a dramatic immediate change would have been more easily attributed to the medication. Handwriting remediation that does not include repeated handwriting practice sessions have been shown to be ineffective (Hoy, Egan, & Feder, 2011). Therefore, it is unlikely that children with legibility difficulties will have these specific problems resolved with medication alone, in the absence of handwriting remediation intervention. However, use of the medication in reducing the presence of coexisting inattentive and hyperactive behaviors could enhance the success of handwriting intervention.

An association between inattentive and hyperactive symptoms and handwriting legibility has been reported in preterm children (Feder et al., 2005). Such a relationship was not apparent in our sample. The lack of obvious association could be due to the skewed distributions of behavioral scores obtained in this population.

Significant improvements in all three behavioral subscales were found. Nevertheless, half of the sample continued to score above the clinical cutoff on the CGI-P total index indicating persistence of attentional difficulties. However, these results do not suggest that the medication was ineffective, rather, it illustrates that some ongoing behavioral difficulties were perceived by the parents to persist months after treatment with the medication. It should be noted that in this study, the parental version of the CGI was the only measure of problematic behavior which may have introduced a potential bias as the parents may have preferentially rated their child when the medication effect had worn off (e.g., after school).

Motor skill difficulties were highly prevalent in this sample and also improved significantly following pharmacologic treatment. Although the change in visual-motor integration skills was not statistically significant, an improvement in this component was associated with improvement in letter, word, and numeral legibility. This is the first study that identifies a change in a motor component as a determinant of concomitant improvements in handwriting performance. Fine motor difficulties were highly prevalent in our sample. Such difficulties are often reported in children with ADHD or ADHD-DCD (Bart et al., 2010; Brossard-Racine, Shevell, et al., 2012; Miyahara et al., 2001; Pitcher et al., 2003), and in children overall with handwriting difficulties (Smith-Engelsman, Niemeijer, & van Galen, 2001).

After 3 months of medication treatment, visual-motor integration was the most important determinant of handwriting legibility in children with and without persisting motor difficulties, consistent with the results obtained in medication-naïve children (Brossard-Racine et al., 2011). Although VMI scores did not improve in all children, those with improvements were more likely to have enhancement in legibility. Visual-motor integration has also been identified as a good predictor of legibility in premature children (Feder et al., 2005), in children with other minor neurological dysfunction (Van Hoorn, Maathuis, Peters, & Hadders-Algra, 2010), and in children with motor deficits (Maeland, 1992; Malloy-Miller, Polatajko, & Anstett, 1995; Volman, van Schendel, & Jongmans, 2006).

Handwriting difficulties were not more frequent in children with persisting motor difficulties as compared with those without motor problems. Although performance on the balance index also contributed to the variance of the models for legibility, the association was weak and therefore was not an important determinant of performance. Associations between balance skills and speed of writing have been anecdotally reported in medication-naive children with ADHD (Brossard-Racine et al., 2011). However, further evidences are required to validate whether this could be explained by diminished postural stability (proximally) that could concomitantly affect both balance and writing skills.

The ETCH is the only latin-based standardized handwriting evaluation that includes such a broad variety of handwriting tasks that are similar to those that the children will be exposed to in the classroom. Judging handwriting based on the use of an evaluation with a strong ecological validity increases our ability to have a representative sample of the child’s performance in his everyday life. Due to sample size limitation, which restricts multiple comparisons, we did not analyze change for each of the subtasks but rather used the total scores. However, as these tasks involve different cognitive and sensori-motor components, it would be interesting for future studies to explore and compare differences in change in performance between the various subtasks.

Speed of writing significantly increased (shorter time) in all tasks at T2 except for the uppercase alphabet writing and was not significant between T0 and T1 except for the lowercase alphabet writing. An improvement in speed was not associated with greater legibility. Placebo-control studies have not reported a significant change in speed with medication (Flapper et al., 2006; Rosenblum et al., 2008), and studies of handwriting interventions have reported the necessity for intensive practice to increase speed appreciably (Hoy et al., 2011). Children who wrote faster (shorter speed time at T2) did not differ significantly in terms of motor and behavioral skills than those who were slower in their writing speed. However, a faster speed does not necessary translate to better handwriting since both writing too fast or too slow can negatively affect the overall performance. Therefore, interpretation of change in speed in this sample remains preliminary, and further studies are required.

Limitations

These results require further validation as our sample was not compared with an age-matched control group, which would have isolated the possible maturation effect. Nevertheless, age-appropriate standardized tools were used to evaluate behavior, handwriting, and motor skills, and the results were compared with published normative data obtained in age-matched peers and to clinically established cutoff. The design of this study also limited our ability to ascertain the timing of effect of behavioral changes on performance. However, the families recruited in this study had made the decision with the treating physician to initiate treatment with medication. Therefore, it would have been difficult ethically to randomize children recommended for pharmacologic treatment to a control group (either no medication or suboptimal dosage).

Although the sample is relatively small, to our knowledge we are the first to report these findings in a cohort of newly diagnosed children. Only 29/55 of the children completed T1, and this sample size may have been too small to enable the reliable detection of immediate significant changes related to the medication, and results have therefore to be interpreted carefully. Changes in medication dosage and possible medication side effects were not recorded due to limited access to medical records across systems. However, the parents verbally confirmed treatment compliance, and dosage was believed to be within the optimal range for most children at T2 as individual medication titration followed CADDRA’s recommendations (CADDRA, 2011). Finally, multiple comparison tests were conducted. If we chose to use the Bonferroni correction, the level of significance would be reduced to p < .006, and none of the differences would have remained significant according to this correction. The need to validate our results in a larger sample should therefore be emphasized.

Finally, although the psychometric properties of the ETCH are acceptable overall, it is important to consider that the cursive version of the ETCH has only moderate interrater reliability for the total word and letter legibility scores. This could have influenced our ability to detect change in the 12 children evaluated with this version and might have affected the overall mean score.

Conclusion

Handwriting difficulties are highly present in our sample of children with ADHD and were not limited to children with comorbid motor difficulties. After 3 months of daily use of stimulant medication, handwriting legibility and speed significantly improved statistically but the level of change in legibility was not clinically detectable in most. There is a common presumption that stimulant medication dramatically improves handwriting skills and that these changes are clinically meaningful, although evidence to support this is limited. In our study, only a subset of children with ADHD improved substantially in their handwriting legibility 3 months after initiation of stimulant medication. Interestingly, the children who were more likely to improve were those presenting with greater handwriting problems, suggesting that the medication has a more substantial performance in this particular subset. Nevertheless, handwriting continues to be a problem in most children with ADHD and therefore requires rehabilitation intervention to optimize legibility. To our knowledge, this study is novel in highlighting existing relationships between motor skills and performance in an essential everyday school activity in children newly diagnosed with ADHD. Nevertheless, this study focused only on a subset of possible explanatory variables for handwriting, and it should be viewed as a first step toward understanding the mechanisms underlying handwriting capacity in this population.

Footnotes

Acknowledgements

We thank the evaluators and the research assistants for their assistance in this study and Victoria Surtees for manuscript correction. Special thanks to the children and parents who participated in this study.

Authors’ Note

Preliminary results of this study were presented at the 2011 DCD-9 conference in Lausanne.

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: During completion of this study, Marie Brossard-Racine received doctoral support from the Richard and Edith Strauss Foundation, the Foundation of Stars/Montreal Children’s Hospital Studentships, and the Fond de recherche en santé du Québec (FRSQ). The project was collectively funded by the Canadian Association of Occupational Therapy Foundation, the Montreal Children’s Hospital-Research Institute (MCH-RI), and Janssen-Ortho Inc. The research laboratory benefited from infrastructure provided by the MCH-RI and Centre de Recherche Interdisciplinaire en Réadaptation (CRIR), both supported by the FRSQ.

Author Biographies

Marie Brossard-Racine, OT, PHD, is a postdoctoral fellow at the Advanced Pediatric Brain Imaging Research Laboratory in the Departments of Diagnostic Imaging and Radiology/Fetal and Transitional Medicine, Children’s National Medical Center. At the time of this study, she was a PhD candidate at the School of Physical and Occupational Therapy, McGill University.

Michael Shevell, MD, CM, is a chair of pediatrics, McGill University, and a physician-in-chief at the Montreal Children’s Hospital. He is professor in the Departments of Neurology and Neurosurgery, Pediatrics. He is a project lead on the Cerebral Palsy Demonstration Project, a national interdisciplinary program of research that is part of NeuroDevNet, a Network of Centres of Excellence.

Laurie Snider, OT, PhD, is an associate professor, School of Physical and Occupational Therapy, McGill University and Research Associated, Montreal Children’s Hospital, McGill University Health Center.

Stacey Ageranioti Bélanger, MD, CM, is a director of the ADHD clinic, Centre Hospitalier Universitaire-Sainte-Justine.

Marilyse Julien, MSc is a statistical consultant, Centre Hospitalier Universitaire Sainte-Justine.

Annette Majnemer, OT, PhD, is a director and associate dean, School of Physical and Occupational Therapy, McGill University, and has cross-appointments in the Departments of Neurology and Neurosurgery and Pediatrics. She is a research associate at the Montreal Children’s Hospital and is coeditor of Physical & Occupational Therapy in Pediatrics and associate editor of Developmental Medicine and Child Neurology.

References

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders. Washington, DC: Author.

Amundson

S. J.

(1995). Evaluation Tool of Children’s Handwriting: ETCH examiner’s manual. Homer, AK: O.T. KIDS.

Bart

Podoly

Bar-Haim

(2010). A preliminary study on the effect of methylphenidate on motor performance in children with comorbid DCD and ADHD. Research in Developmental Disabilities, 31, 1443-1447.

Beery

K. E.

Buktenica

N. A.

Beery

N. A.

(2005). Beery-Buktenica Developmental Test of Visual-Motor Integration (5th ed.). Minneapolis, MN: NCS Pearson.

Brossard-Racine

Majnemer

Shevell

Snider

(2008). Handwriting performance in children with attention deficit hyperactivity disorder (ADHD). Journal of Child Neurology, 23, 399-406.

Brossard-Racine

Majnemer

Shevell

Snider

Bélanger

A. S.

(2011). Handwriting capacity in children newly diagnosed with attention deficit hyperactivity disorder. Research in Developmental Disabilities, 32, 2927-2934.

Brossard-Racine

Mazer

Julien

Majnemer

(2012). Validating the use of the Evaluation Tool of Children’s Handwriting-Manuscript (ETCH-M) to identify handwriting difficulties and detect change in school-aged children. American Journal of Occupational Therapy, 66, 414-421.

Brossard-Racine

Shevell

Snider

Bélanger

A. S.

Majnemer

(2012). The motor skills in children newly diagnosed with Attention Deficit Hyperactivity Disorder prior to and following treatment with stimulant medication. Research in Developmental Disabilities, 33, 2080-2087.

Canadian Attention Deficit Hyperactivity Disorder Resource Alliance. (2011). Canadian Attention Deficit Hyperactivity Disorder Resource Alliance (CADDRA): Canadian ADHD practice guidelines (3rd ed.). Toronto, Ontario, Canada: Author.

10.

Conners

C. K.

(2001). Conner’s Rating Scales-Revised: Technical manual. North Tonawanda, NY: Multi-Health Systems.

11.

Diekema

S. M.

Deitz

Amundson

S. J.

(1998). Test-retest reliability of the Evaluation Tool of Children’s Handwriting-manuscript. American Journal of Occupational Therapy, 52, 248-255.

12.

Duff

Goyen

T. A.

(2010). Reliability and validity of the Evaluation Tool of Children’s Handwriting-Cursive (ETCH-C) using the general scoring criteria [Validation Studies]. American Journal of Occupational Therapy, 64, 37-46.

13.

Feder

K. P.

Majnemer

Bourbonnais

Blayney

Morin

(2007). Handwriting performance on the ETCH-M of students in a grade one regular education program. Physical & Occupational Therapy in Pediatrics, 27, 43-62.

14.

Feder

K. P.

Majnemer

Bourbonnais

Platt

Blayney

Synnes

(2005). Handwriting performance in preterm children compared with term peers at age 6 to 7 years. Developmental Medicine & Child Neurology, 47, 163-170.

15.

Flapper

B. C. T.

Houwen

Schoemaker

M. M.

(2006). Fine motor skills and effects of methylphenidate in children with attention deficit-hyperactivity disorder and developmental coordination disorder. Developmental Medicine & Child Neurology, 48, 165-169.

16.

Harrell

F. E.

(2001). Regression modeling strategies. New York, NY: Springer.

17.

Henderson

S. E.

Sugden

D. A.

(1992). Movement Assessment Battery for Children: Manual. London, England: The Psychological Corporation.

18.

Hoy

M. M.

Egan

M. Y.

Feder

K. P.

(2011). A systematic review of interventions to improve handwriting. Canadian Journal of Occupational Therapy, 78, 13-25.

19.

Kaplan

B. J.

Wilson

B. N.

Dewey

Crawford

S. G.

(1998). DCD may not be a discrete disorder. Human Movement Science, 17, 471-790.

20.

Koziatek

S. M.

Powell

N. J.

(2002). A validity study of the Evaluation Tool of Children’s Handwriting-Cursive. American Journal of Occupational Therapy, 56, 446-453.

21.

Lerer

R. J.

Lerer

M. P.

Artner

(1977). The effects of methylphenidate on the handwriting of children with minimal brain dysfunction. Journal of Pediatrics, 91, 127-132.

22.

Lufi

Gai

(2007). The effect of methylphenidate and placebo on eye-hand coordination functioning and handwriting of children with attention deficit hyperactivity disorder. Neurocase, 13, 334-341.

23.

Maeland

A. F.

(1992). Handwriting and perceptual motor skills in clumsy, dysgraphic, and normal children. Perceptual & Motor Skills, 75, 1207-1217.

24.

Malloy-Miller

Polatajko

H. J.

Anstett

(1995). Handwriting error patterns of children with mild motor difficulties. Canadian Journal of Occupational Therapy, 62, 258-267.

25.

Miyahara

Mobs

Doll-Tepper

(2001). Severity of hyperactivity and the comorbidity of hyperactivity with clumsiness in three sample sources: School, support group and hospital. Child: Care, Health and Development, 27, 413-424.

26.

Pitcher

T. M.

Piek

J. P.

Hay

D. A.

(2003). Fine and gross motor ability in males with ADHD. Developmental Medicine & Child Neurology, 45, 525-535.

27.

Rosenblum

Epsztein

Josman

(2008). Handwriting performance of children with attention deficit hyperactive disorders: A pilot study. Physical & Occupational Therapy in Pediatrics, 28, 219-234.

28.

Santosh

P. J.

(2007). Pharmacotherapy in ADHD. In Taylor

(Ed.), People with hyperactivity: Understanding and managing their problems (pp. 118-133). London, England: Mac Keith Press.

29.

Smith-Engelsman

B. C. M.

Niemeijer

A. S.

van Galen

G. P.

(2001). Fine motor deficiencies in children diagnoses as DCD based on poor grapho-motor ability. Human Movement Science, 20, 161-182.

30.

Tucha

Lange

K. W.

(2001). Effects of methylphenidate on kinematic aspects of handwriting in hyperactive boys. Journal of Abnormal Child Psychology, 29, 351-356.

31.

Van Hoorn

J. F.

Maathuis

C. G.

Peters

L. H.

Hadders-Algra

(2010). Handwriting, visuomotor integration, and neurological condition at school age. Developmental Medicine & Child Neurology, 52, 941-947. doi:10.1111/j.1469-8749.2010.03715.x

32.

Volman

M. J. M.

van Schendel

Jongmans

M. J.

(2006). Handwriting difficulties in primary school children: A search for underlying mechanisms. American Journal of Occupational Therapy, 60, 451-460.

33.

Watemberg

Waiserberg

Zuk

Lerman-Sagie

(2007). Developmental coordination disorder in children with attention-deficit-hyperactivity disorder and physical therapy intervention. Developmental Medicine & Child Neurology, 49, 920-925.

34.

Wechsler

(1999). Wechsler Abbreviated Scale of Intelligence. San Antonio, TX: Psychological Corporation.

35.

Whalen

C. K.

Henker

Finck

(1981). Medication effects in the classroom: Three naturalistic indicators. Journal of Abnormal Child Psychology, 9, 419-433.