Research in Developmental Disabilities 34 (2013) 2917–2926

Contents lists available at SciVerse ScienceDirect

Research in Developmental Disabilities

Understanding macrographia in children with autism

spectrum disorders

Beth P. Johnson a,*, James G. Phillips a, Nicole Papadopoulos a,b, Joanne Fielding a,Bruce Tonge a, Nicole J. Rinehart a,b

a School of Psychology and Psychiatry, Monash University, Wellington Road, Clayton, VIC, Australiab School of Psychology, Deakin University, 221 Burwood Highway, Burwood, VIC, Australia

A R T I C L E I N F O

Article history:

Received 10 April 2013

Received in revised form 2 June 2013

Accepted 3 June 2013

Available online

Keywords:

Autism spectrum disorders

Fine motor

Manual dexterity

Handwriting

A B S T R A C T

It has been consistently reported that children with autism spectrum disorders (ASD)

show considerable handwriting difficulties, specifically relating to accurate and consistent

letter formation, and maintaining appropriate letter size. The aim of this study was to

investigate the underlying factors that contribute to these difficulties, specifically relating

to motor control. We examined the integrity of fundamental handwriting movements and

contributions of neuromotor noise in 26 children with ASD aged 8–13 years (IQ> 75), and

17 typically developing controls. Children wrote a series of four cursive letter l’s using a

graphics tablet and stylus. Children with ASD had significantly larger stroke height and

width, more variable movement trajectory, and higher movement velocities. The absolute

level of neuromotor noise in the velocity profiles, as measured by power spectral density

analysis, was significantly higher in children with ASD; relatively higher neuromotor noise

was found in bands>3 Hz. Our findings suggest that significant instability of fundamental

handwriting movements, in combination with atypical biomechanical strategies,

contribute to larger and less consistent handwriting in children with ASD.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Autism spectrum disorders (ASD) refers to a group of neurodevelopmental disorders that share a triad of impairments incommunication, social interactions, and restricted and stereotyped patterns of interests and behaviors. In addition to their corediagnostic symptoms, it is now recognized that there are also pervasive motor abnormalities associated with ASD (Fournier,Hass, Naik, Lodha, & Cauraugh, 2010). The need for further investigation of fine motor and handwriting impairments, inparticular, is highlighted by the high rate of clinical referrals for assistance in this domain. A study of academic performance inboys with Asperger’s disorder (AD) found that 58% of children those aged between 6 and 11 years had received occupationaltherapy to address fine motor issues such as poor handwriting and scissor manipulation (Church, Alisanski, & Amanullah, 2000).Cartmill, Rodger, and Ziviani (2009) also reported 40% of occupational therapists’ caseloads were comprised of children withASD, and 86% of those cases with twenty or more consecutive referrals were for assistance with handwriting or fine motor skills.Despite the clear clinical and academic importance of evidence-based handwriting therapies for children with ASD, few studieshave attempted to characterize the nature of handwriting difficulties in these disorders.

Handwriting is a complex skill, dependent on smooth integration of higher-order linguistic planning working memory,proprioceptive and visuomotor control (Cartmill et al., 2009; Kushki, Chau, & Anagnostou, 2011). A review of the

* Corresponding author. Tel.: +61 3 9905 4040.

E-mail address: beth.johnson@monash.edu (B.P. Johnson).

0891-4222/$ – see front matter � 2013 Elsevier Ltd. All rights reserved.

http://dx.doi.org/10.1016/j.ridd.2013.06.003

B.P. Johnson et al. / Research in Developmental Disabilities 34 (2013) 2917–29262918

handwriting literature in ASD by Kushki et al. (2011) identified consistent deficits in overall legibility and letter formation. Ofthe small number of studies that have examined handwriting in children and adolescents with ASD, qualitative measureshave identified problems with legibility (Church et al., 2000; Fuentes, Mostofsky, & Bastian, 2009) and poor letter formation(Cartmill et al., 2009; Fuentes et al., 2009). Difficulty controlling handwriting size is also commonly reported in children withboth high functioning autism (HFA) and AD (Cartmill et al., 2009; Myles et al., 2003). Furthermore, Beversdorf et al. (2001)found that adults with ASD exhibited macrographia (atypically large handwriting). The overarching aim of this study was totherefore move beyond descriptions of handwriting quality and understand the process of producing handwritingmovements in children with ASD, specifically focusing on motor control.

All handwriting is comprised of a base unit, a stroke, which is a simple up or down movement; consecutive up and downstrokes form a loop, e.g. the cursive letter l (Plamondon & Guerfali, 1998). Strokes and loops are chained together to formmore complex letters and whole words (Plamondon & Guerfali, 1998). The motor plan for simple handwriting strokes can beor upscaled or downscaled dependent on the context, for example the letter l or e (Phillips, Ogeil, & Best, 2009). If the internalrepresentation (motor plan) for basic handwriting units is poorly formed or poorly controlled, this has downstreamconsequences for consistent formation of letters and planning of whole words (Phillips, Ogeil, & Best, 2009; Plamondon &Guerfali, 1998). The approach of using a series of chained loops, such as ‘l’s or ‘e’s to understand the control of handwritingmovements has previously been used in children with ADHD (Langmaid, Papadopoulos, Johnson, Phillips, & Rinehart, 2012;Tucha, Mecklinger, Walitza, & Lange, 2006), Huntington’s disease (Phillips, Bradshaw, Chiu, & Bradshaw, et al., 1994),Parkinson’s disease (Caligiuri, Teulings, Filoteo, Song, & Lohr, 2006), and alcohol dependence (Phillips, Ogeil, & Muller, 2009).Studies assessing the control of simple fine motor movements, such as ‘l’s or small circles, have found that limiting externalcues, such as writing guides, is effective in eliciting and assessing the integrity of the internal representation of the motorplan (Langmaid et al., 2012; Martin, Phillips, Iansek, & Bradshaw, 1994).

Following on from findings of inconsistent letter formation in children with ASD (Cartmill et al., 2009; Church et al., 2000;Fuentes et al., 2009), the first specific aim of this study was to therefore assess the integrity of the internal representation ofsimple handwriting movements in children with ASD, using a cursive l’s task. It was hypothesized that children with ASDwould show more unstable motor plans for simple handwriting movements (strokes) than typically developing children.Previous studies of children with poor handwriting (van Galen, Portier, Smits-Engelsman, & Schomaker, 1993) and ADHD(Langmaid et al., 2012) have indicated that this would manifest as overall movement size and trajectory being larger andmore variable. A recent kinematic evaluation of large circle drawing in children aged 4–8 years with ASD revealed that timingconsistency during drawing continuous circles did not differ between ASD and controls, therefore we did not anticipate anygreater variability in movement duration (Fleury, Kushki, Tanel, Anagnostou, & Chau, 2013). However, we predicted greatermovement velocity (van Galen et al., 1993) and reduced index of ballisticity (i.e. few number of cycles of acceleration anddeceleration during a movement; Langmaid et al., 2012).

The second aim of this study was to investigate whether inconsistent letter formation may also relate to inherently greaternoise within the neuromotor system (Smits-Engelsman & Van Galen, 1997; van Galen et al., 1993). Previous studies haveestimated neuromotor noise using power spectral density analysis, in which the velocity profile of an entire stroke is decomposedinto its frequencies using Fast Fourier Analysis (Smits-Engelsman & Van Galen, 1997; van Galen et al., 1993). As the duration ofeach complete l is approximately 1 second, low frequencies (1–2 Hz) relate to the periodicity of the task itself, representing theadductive and abductive finger and wrist movements required to write consecutive letter l’s (van Galen, van Doorn, & Schomaker,1990). Superimposed upon this signal are frequencies associated with visual and proprioceptive feedback corrections, tremor,and neuromotor recruitment. Handwriting movements require ongoing visual feedback to maintain an accurate movementtrajectory, resulting in rapid corrections to movement trajectory that can seen within the 3 to 4 Hz range (Miall, Weir, Wolpert, &Stein, 1993; Smits-Engelsman & Van Galen, 1997). Uncontrolled movements such as neuromotor tremor and reflex-basedcontractions, occur at frequencies greater than 5 Hz (van Galen et al., 1990). In tasks that are highly automated, or well learned,the relative contribution of low frequency (1–4 Hz) decreases, and there is a shift toward frequencies >5 Hz (van Galen et al.,1993). Poor handwriting in typically developing children is associated with greater energy within the 4–7 Hz band of the velocityprofile (van Galen et al., 1993). We therefore predicted higher absolute spectra values in the ASD group relative to typicallydeveloping controls across all frequencies. Further, we predicted children with ASD would show relatively higher frequencies inbands>4 Hz compared to typically developing children. In light of reports of atypical visuomotor control as a contributing factorto handwriting deficits in ASD (Cartmill et al., 2009), we also predicted higher relative energy within the 3–4 Hz band.

The final aim of this study was to explore whether age, IQ measures, motor skills and the spectral band associated withvisual correction of movement (3–4 Hz) were associated with kinematic features of handwriting movements in ASD. Earlierinvestigations of handwriting quality in ASD found that motor skills as measured by Physical and Neurological Examinationfor Subtle (Motor) Signs (PANESS; Fuentes et al., 2009), and perceptual reasoning index (Fuentes, Mostofsky, & Bastian, 2010)were significantly predictive of letter formation.

2. Method

2.1. Participants

This study was approved by Monash University and Southern Health Human Research Ethics Committees. Parents ofparticipants provided informed consent prior to the commencement of the study, and written assent was provided by the

Table 1

Participant demographics and scores.

ASD TD

Mean SD Mean SD

Age (months) 136.77 19.68 125.71 18.25

VCI/VIQ 103.08 16.65 104.06 11.26

PRI/PIQ 106.08 14.09 103.71 12.75

Full scale IQ 100.69 14.18 106.06 12.88

M-ABC 2Manual dexterity 7.28* 2.75 9.82 4.36

Aiming & catching 8.04* 3.28 13.18 4.61

Balance 8.84* 3.22 11.65 4.20

Total M-ABC-2 score 7.64* 3.47 11.88 4.46

ASD: autism spectrum disorders; TD: typically developing; PIQ: performance IQ; VIQ: verbal IQ; M-ABC-2: movement assessment battery for children 2nd

edition; SD: standard deviation.

* ASD vs. TD, p< .05

B.P. Johnson et al. / Research in Developmental Disabilities 34 (2013) 2917–2926 2919

participants in accordance with the Declaration of Helsinki. Forty three boys aged 8–13 years participated in the study: 26with ASD (16 with HFA, 10 with AD) and 17 typically developing (TD) children (see Table 1 for a summary of participantcharacteristics).

Boys with ASD were recruited from private pediatricians in Melbourne, Victoria and the Autism Victoria database. Reportsfrom pediatricians were reviewed to ensure that all children were diagnosed according to Diagnostic and Statistical Manualof Mental Disorders – 4th edition, revised (DSM-IV-TR; American Psychiatric Association, 2000) criteria for autistic disorder.Further diagnostic information was gathered using the Social Responsiveness Scale (SRS; Constantino & Gruber, 2007),Developmental Behavior Checklist–Parent Version (DBC-P; Einfeld & Tonge, 2002), structured parent interviews, direct childobservations and information from teachers and other therapists involved in the assessment process. The DBC-P has goodpsychometric properties, includes five subscales (disruptive/antisocial, self-absorbed, communication disturbance, anxiety,social relating), and provides an autism screening algorithm (autism-related items are weighted and collated to calculate anoverall risk index; Brereton, Tonge, Mackinnon, & Einfeld, 2002; Witwer & Lecavalier, 2007). Participants were excluded ifthey presented with any co-morbid neurological (e.g. tuberous sclerosis), genetic (e.g., Fragile� syndrome) or psychiatricdiagnosis (e.g. Tourette’s syndrome). No children in the ASD groups were reported to have been taking any type ofmedication.

Typically developing boys were recruited from public schools and community-wide organizations. The DBC-P (Einfeld &Tonge, 2002), and the SRS (Constantino, 2002) were used to exclude the presence of autism, AD or other previously listedpsychiatric diagnosis in typically developing boys; no participants were excluded from the study.

All clinical children completed the Wechsler Intelligence Scale for Children – 4th edition (WISC-IV; Wechsler, 2003),while TD children were administered either the WISC-IV (8 TD boys) or Wechsler Abbreviated Scales of Intelligence (9 TDboys; WASI; Wechsler, 1999). All children had a full-scale IQ (FSIQ), VCI/verbal IQ (VIQ) and PRI/performance IQ (PIQ) greaterthan 75. Analysis of variance (ANOVA) revealed that HFA, AD and TD groups did not differ on FSIQ [F(2, 48) = 2.75, p = .07], VIQ[F(2, 48) = 1.60, p = .20] or PIQ [F(2,48) = 0.13, p = .88] (see Table 1 for participant characteristics).

Motor skills of all children were assessed using the Movement Assessment Battery for Children – 2nd edition (M-ABC-2;Henderson, Sugden, & Barnett, 2007), which has previously been used to assess motor performance in children with ASD(Green et al., 2009; Papadopoulos, McGinley, Tonge, Bradshaw, & Bradshaw, et al., 2012). The M-ABC-2 consists of eightitems grouped in three sections: manual dexterity, ball skills and balance, with age dependent items used for each section.No TD participants were classified in the ‘‘definite motor impairment range’’ as defined by the M-ABC-2 guidelines.

2.2. Apparatus and procedure

Participants were asked to write four, cursive letter l’s at their preferred size. Participants practiced writing five sets offour cursive letter l’s at their preferred size using a pen and non-lined paper, to ensure they understood the task. Participantsthen performed the same task on a Wacom UD-1212-R digitizing tablet with non-inking stylus, controlled by a laptopcomputer. The digitizing tablet had an active surface area of 30.48 cm� 30.48 cm, a resolution of 0.001 cm and a samplingfrequency of 100 Hz. Experiments were designed and conducted using MovAlyzeR (Version 5.8.1) software by NeuroscriptL.L.C.

2.3. Data analysis

MovAlyzeR software (Version 5.8.1) was used to filter, segment each cursive letter l’s into up and down strokes, andextract kinematic features. Variables were calculated and averaged for each stroke across trials.

Handwriting variables were grouped according to (1) stroke metrics, (2) dynamics (3) consistency and (4) power spectraldensity analysis.

B.P. Johnson et al. / Research in Developmental Disabilities 34 (2013) 2917–29262920

(1) S

Tab

Mea

St

Ve

Ho

St

St

Du

Pe

Re

In

St

CO

CO

CO

CO

ASD

*

troke metrics included stoke height and width (centimeters), and trajectory (slant; measured as the direction frombeginning point to endpoint of a stroke, angles measured in degrees from vertical).

(2) S

troke dynamics were stroke duration (seconds), peak velocity (cm/s), asymmetry ratio and ballisticity. The asymmetryratio, calculated as the proportion spent in acceleration relative to the total movement duration, was calculated todetermine the shape of the velocity profile; efficient movements have a symmetrical shape, with equal time spent inaccelerative and decelerative phases. An index of ballisticity was as an indicator of smoothness of the movementendpoint (pen on tablet), calculated by dividing the number of zero crossings in the acceleration function by the numberof zero crossings in the velocity function, was used as a measure of handwriting dysfluency (Phillips, Ogeil, & Muller,2009).

(3) T

he consistency of stroke height, width, trajectory and duration were calculated as a coefficient of variability (COV;division of the mean by its standard deviation).

(4) F

or the power spectral density analysis, the velocity function was decomposed into contributing movement frequencies(1–10 Hz) using Fourier analysis (Phillips, Ogeil, & Muller, 2009; van Galen et al., 1993). Absolute energy within eachfrequency and across frequencies was estimated using the absolute power spectra. However the relative proportion ofthe velocity spectrum at each frequency was expressed as a percentage of total power across the 1–10 Hz spectrum wasalso compared between groups, as differences in movement velocity can contribute to higher absolute spectra values(van Galen et al., 1993; van Galen et al., 1990). Relative power spectra were calculated by dividing the power scores ineach band by the summed power of the whole spectrum.

2.3.1. Statistical analysis

A series of one-way ANOVAs were used to compare IQ, MABC-2 scores and handwriting variables across HFA, AD and TDgroups. Bonferroni corrections were applied across each group of variables: (1) stroke metrics (height, width, slant), (2)dynamics (duration, velocity, ballisticity) and (3) consistency (COV of height, width, slant and duration). Normality wasdetermined using Levene’s test of Normality. For the power spectral density analysis, mixed model ANOVAs were used tocompare absolute and relative energy across 1–10 Hz bands between each group.

There was no statistical difference between HFA or AD on IQ, MABC-2 or any handwriting variables (vertical size,horizontal size, stroke trajectory, duration, peak velocity, relative time to peak velocity, index of ballisticity, COV of verticalsize, COV of horizontal size, COV of slant, COV of duration and power density spectra) therefore data across both groups werecollapsed into a single ASD group. Handwriting variables were therefore compared between the ASD and TD group using t

tests, with Bonferroni corrections applied across each group of variables.Correlational analyses were conducted within the ASD group with kinematic stroke measures as the dependent variables;

age, PRI/PIQ, VCI/VIQ, MABC-2 scores, and total SRS score as the independent variables.

3. Results

3.1. Movement assessment battery for children – 2nd edition

There was a significant difference between ASD and TD groups on all components of the MABC-2, with the ASD groupperforming more poorly on all measures: manual dexterity (t(41) = 2.32, p = .025), aiming and catching (t(41) = 4.22,p< .001), balance (t(41) = 2.40, p = .019) and total MABC-2 score (t(41) = 3.46, p = .001).

le 2

ns (SD) of stroke metrics and velocity profile.

ASD TD

Mean SD Mean SD

roke metrics

rtical size (cm) 5.65* 5.96 1.79 1.13

rizontal size (cm) 1.48* 1.12 0.81 0.67

roke trajectory (degrees) 20.22* 11.24 17.35 9.66

roke dynamics

ration (seconds) 0.45 0.25 0.42 0.15

ak velocity (cm/second) 26.35* 30.11 8.19 5.65

lative time to peak velocity 0.49 0.04 0.48 0.03

dex of ballisticity 2.38 1.61 2.20 1.07

roke consistency

V vertical size 4.77 2.85 4.99 2.21

V horizontal size 1.10 0.70 1.31 0.85

V trajectory 0.13* 0.14 0.23 0.12

V duration 5.00 2.10 5.83 6.09

: autism spectrum disorder; COV: coefficient of variability; TD: typically developing; SD: standard deviation.

HFA vs. TD, p< .05.

B.P. Johnson et al. / Research in Developmental Disabilities 34 (2013) 2917–2926 2921

3.2. Stroke metrics

Table 2 summarizes handwriting measures for ASD and TD groups.There was a statistically significant difference between groups in vertical stroke size between (t(41) = 4.22, p< .001,

d = 0.83; Table 2) and horizontal stroke size (t(41) = 2.47, p< .018, d = 0.70), with both vertical and horizontal stroke sizebeing significantly larger in the ASD group (Figs. 1 and 2). Handwriting trajectory (slant) was not significantly differentbetween groups.

3.3. Stroke dynamics

Despite significantly increased movement size, there was no statistically significant difference between groups inmovement duration. Peak velocity was significantly greater in the ASD group (t(41) = 4.22, p< .006, d = 0.77).

There were no statistically significant differences between groups with respect to relative time to peak velocity or theindex of ballisticity, indicating comparable movement efficiency and smoothness to the TD group.

3.4. Stroke consistency

There was no significant difference in the COV of either vertical stroke size, horizontal stroke size or stroke durationbetween groups. However children with ASD had significantly lower COV of stroke trajectory (t(41) = 0.29, p< .001, d = 0.75),indicating greater variability of handwriting trajectory.

3.5. Power spectral density analysis

Mixed model ANOVA revealed that there was a significant interaction between absolute power density spectra anddiagnosis (F(1.06, 43.39) = 4.97, p = .03). Mauchly’s test of sphericity had been violated (chi-square= 2256.42, p� .05),[(Fig._1)TD$FIG]

Fig. 1. Vertical stroke size in children with high-functioning autism, Asperger’s disorder and typically developing children.[(Fig._2)TD$FIG]

Fig. 2. An example of macrographic handwriting in a child with ASD.

[(Fig._3)TD$FIG]

Fig. 3. Power spectral density functions, expressing both (a) absolute and (b) relative power for the vertical velocity profiles of children with autism

spectrum disorders (black circles) and typically developing children (white circles). Star designates significant difference between groups (p< .05).

B.P. Johnson et al. / Research in Developmental Disabilities 34 (2013) 2917–29262922

therefore the corrected degrees of freedom using Greenhouse–Geisser estimates of sphericity (epsilon = 0.50) are reported.Post hoc tests revealed absolute energy was higher in the ASD group across all spectra (Fig. 4).

There was a significant interaction between relative power density spectra and diagnosis (F(1.38, 133.73) = 6.54, p = .007).Mauchly’s test of sphericity had been violated (chi-square = 1508.12, p� .05), therefore the corrected degrees of freedomusing Greenhouse–Geisser estimates of sphericity (epsilon = 0.50) are reported. Post hoc tests revealed that relative powerscores were higher in the ASD group across all spectra, except for at 1–2 Hz, where the TD group had relatively higher power(Fig. 3).

3.6. Correlational analyses

Spearman’s correlations were performed to investigate the relationship between characteristic handwriting features inASD (vertical stroke size, COV of stroke trajectory and absolute spectra values) and motor (MABC-2) and IQ measures.Vertical stroke size and stroke velocity were significantly negatively correlated with manual dexterity scores (Fig. 4; strokesize: r =�.39, p = .027) and relative energy within the 3–4 Hz band (r = .70, p =<.001). COV of stroke trajectory wassignificantly positively correlated with aiming and catching scores; higher COV scores are indicative of less variability (Fig. 4;r =�.42, p = .018), but not relative energy within the 3–4 Hz band. There were no significant correlations between strokevariables and IQ measures.

[(Fig._4)TD$FIG]

Fig. 4. (a) A correlation between stroke size and manual dexterity score on the Movement Assessment Battery for Children – 2nd edition in children with

autism spectrum disorder. (b) A correlation between coefficient of variability of stroke trajectory and aiming and catching score on the Movement

Assessment Battery for Children – 2nd edition in children autism spectrum disorder.

B.P. Johnson et al. / Research in Developmental Disabilities 34 (2013) 2917–2926 2923

4. Discussion

The overall aim of this study was to investigate factors that contribute to difficulties producing consistent letter formationin children with ASD, specifically, investigating the integrity of fundamental handwriting movements. We found that simplehandwriting movements were considerably larger and movement trajectory was more variable in children with ASD. This

B.P. Johnson et al. / Research in Developmental Disabilities 34 (2013) 2917–29262924

finding now extends previous reports of macrographia in adults with ASD (Beversdorf et al., 2001) to children with ASD.Furthermore, handwriting size was associated with manual dexterity scores on the M-ABC-2, indicating that macrographicwriting movements in ASD are linked to fine motor coordination deficits, but not age or IQ. Similarly, greater variability of thetrajectory of handwriting movements was associated with aiming and catching, which may be indicative of fundamentaleye-hand coordination difficulties in children with ASD. These findings are consistent with previous qualitative assessmentsof handwriting in children with ASD aged 8–13 (Fuentes et al., 2009), in which it was found that motor skills measured by thePANESS were predictive of letter formation. That simple handwriting movements in ASD were markedly larger and morevariable may reflect impaired internal regulation, or instability, of fundamental handwriting motor plans.

Large handwriting is common in early school years as children begin to learn to write, and becomes smaller and lessvariable as children become more proficient at writing. Further investigation is required to establish whether macrographiain ASD is attributable to disruption of motor control networks, or simply delayed maturation of handwriting skills. Childhoodconditions that have reported abnormally large handwriting are ADHD and children with cerebellar lesions (Frings et al.,2010), while in adults, macrographia has been associated with cerebellar lesions (Haymaker, 1956; Holmes, 1917),Parkinson’s disease (Caligiuri et al., 2006; Slavin et al., 1999) and Huntington’s disease (Phillips et al., 1994). Each of thesedisorders affects part of the network that underpins motor planning and refinement of motor error, namely the striatum,prefrontal cortex or cerebellum. Children with ASD have been found to have reduced connectivity across the prefrontal/basalganglia/cerebellar motor network (Langen et al., 2009; Mostofsky et al., 2009; Sivaswamy et al., 2010). Striatal abnormalitiesin children with ASD have been associated with poorer motor performance on the PANESS (Qiu, Adler, Crocetti, Miller, &Mostofsky, 2010). Furthermore, children with ASD show reduced cerebellar activation during a sequential fine motor task(Mostofsky et al., 2009); this may have implications for refining movement trajectories and establishing automated motorplans for handwriting movements as well as efficient coordination of both proximal (shoulder, arm) and distal joints (fingers,wrist).

A primary aim of this study was to examine how children with ASD produce handwriting movements, in order tounderstand how this gives rise to qualitative handwriting impairments, such as more variable handwriting trajectories. Ourfindings that children with ASD had higher movement velocities, and significantly higher relative power in spectra >4 Hz,are consistent with that of Smits-Engelsman and Van Galen (1997), who found TD children with poor handwriting showedrelatively lower power bands (0–3 Hz), and higher power in 4–7 and 8–10 Hz bands. Relative time to peak velocity andmovement ballisticity did not differ between groups, indicating smooth overall velocity profiles, however both relative andabsolute noise was observed in 4–10 Hz spectral bands. Greater values within the 4–7 and 8–10 Hz range represents noiseassociated with neuromotor tremor and recruitment (Phillips, Ogeil, & Muller, 2009; Smits-Engelsman & Van Galen, 1997;van Galen et al., 1990). That overall movement velocity was not different to TD children, but that there was greater energy at4–10 Hz bands may indicate that children with ASD require higher energy input to achieve the same movement smoothness.The relatively higher energy within the neuromotor system of children with ASD, and impaired ability to control size andtrajectory of movement, may reflect both a fundamental difficulty in developing strategies to inhibit neuromotor noise andrefine movement error, in particular with respect to coordination of the hands/wrist (van Galen et al., 1993). Higher energyand inefficient motor control may also be a contributing factor to slower handwriting in children with ASD (Cartmill et al.,2009; Kushki et al., 2011).

The ASD group was characterized by higher movement velocities and increased energy across 3–10 Hz spectra.Continuous volitional movements require ongoing visual feedback; delay of the visuomotor system in providing updatedcorrections to movement trajectories in response to visual feedback is approximately 300 ms, or 3–4 Hz (Miall et al., 1993).Higher power within this frequency is proposed to represent rapid, periodic motor adjustments in response to updatedvisuomotor commands. Children with ASD showed higher relatively power within the 3–4 Hz band, which was significantlypositively correlated with vertical stroke size. While this may represent atypical visuomotor integration in children withASD, energy within this band could also comprise additional periodic signal from additional recruitment of effectors requiredto achieve such large movement trajectories, such as the elbow or wrist.

Our study revealed that both HFA and AD share impaired internal regulation of simple handwriting movements, asevidenced by atypically large movements. Previous studies that have directly compared upper limb kinematics between HFAand AD have revealed subtle differences between the two groups (Papadopoulos, McGinley, Tonge, Bradshaw, & Saunders,et al., 2012; Rinehart et al., 2006). However it is important to note the tasks used by Rinehart et al. (2006) and Papadopoulos,McGinley, Tonge, Bradshaw, and Saunders, et al. (2012) provided greater external and contextual cues, whereas the presentstudy limited external cues in order to best elicit the internal representation of the motor plan. The absence of cues to guidebehavior has been found to result in higher rates of aberrant responses, such as disinhibition and increased perseverativebehaviors in children with HFA and AD (Rinehart et al., 2006; Turner, 1999). Impairments in internal regulation of simplehandwriting motor plans may potentially be aided by the provision of external cues. The use of visual guides in behavioraland educational instructional instruction has been a mainstay for improving overall cognitive and behavioral outcomes forchildren with ASD (Ozonoff & Cathcart, 1998; Quill, 1997).

Macrographic handwriting in children with ASD, along with its associated increase in neuromotor noise, may alsorepresent a preference for using proximal over distal musculature, or greater postural instability. Studies in typicallydeveloping children have revealed that greater postural and proximal limb instability, but not distal musculature instability,are associated with more inaccurate and extraneous movements, and thicker lines while drawing (Miyahara, Piek, & Barrett,2008). Postural training, in the context of writing, may help to improve handwriting control and stability, and reduce

B.P. Johnson et al. / Research in Developmental Disabilities 34 (2013) 2917–2926 2925

neuromotor noise. Future research is required to confirm whether training to improve postural and limb stability areeffective in constraining handwriting size and reducing neuromuscular effort.

4.1. Future directions

In conclusion, children with ASD show poor integrity and control of fundamental handwriting movements, whichencompasses control of handwriting size, trajectory, speed and neuromotor noise. The macrographic movements observedwithin the ASD group were striking. That previous studies have reported more variability in handwriting size and shape, butnot size, suggests that visual cues, such as writing guides, may help to contain macrographic handwriting in children withASD. In light of recommendations that visual guides may assist motor learning in children with ASD, further research shouldinvestigate how the presence of visual guides modulates the control of handwriting movements in this population.

References

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders(4th ed., text rev). http://dx.doi.org/10.1176/appi.-books.9780890423349

Beversdorf, B., Anderson, J. M., Manning, S. E., Anderson, S. L., Nordgren, R. E., Felopulos, G. J., et al. (2001). Brief report: Macrographia in high-functioning adultswith autism spectrum disorder. Journal of Autism and Developmental Disorders, 31(1), 97–101. http://dx.doi.org/10.1023/A:1005622031943

Brereton, A., Tonge, B., Mackinnon, A., & Einfeld, S. (2002). Screening young people for autism with the developmental behaviour checklist. Journal of AmericanAcademy of Child & Adolescent Psychiatry, 41(11), 1369–1375.

Caligiuri, M., Teulings, H., Filoteo, J., Song, D., & Lohr, J. (2006). Quantitative measurement of handwriting in the assessment of drug-induced parkinsonism. HumanMovement Science, 25, 510–522. http://dx.doi.org/10.1016/j.humov.2006.02.004

Cartmill, L., Rodger, S., & Ziviani, J. (2009). Handwriting of eight-year-old children with autism spectrum disorder: An exploration. Journal of Occupational Therapy,Schools and Early Intervention, 2(2), 103–118. http://dx.doi.org/10.1080/19411240903146426

Church, C., Alisanski, S., & Amanullah, S. (2000). The social, behavioral, and academic experiences of children with Asperger syndrome. Focus on Autism and OtherDevelopmental Disabilities, 15(1), 12–20. http://dx.doi.org/10.1177/108835760001500102

Constantino, J. N., & Gruber, C. P. (2007). Social responsiveness scale (SRS) manual (vol. 2). Los Angeles, CA: Western Psychological Services.Constantino, J. (2002). The social responsiveness scale Los Angeles: Western Psychological Services.Einfeld, S., & Tonge, B. (2002). Manual for the developmental behaviour checklist (2nd ed.). Clayton, Australia: Monash University Centre for Developmental

Psychology and Psychiatry, University of New South Wales.Fleury, A., Kushki, A., Tanel, N., Anagnostou, E., & Chau, T. (2013). Statistical persistence and timing characteristics of repetitive circle drawing in children with ASD.

Developmental Neurorehabilitation, 1–10. http://dx.doi.org/10.3109/17518423.2012.758184 [Ahead of print]Fournier, K., Hass, C., Naik, S., Lodha, N., & Cauraugh, J. (2010). Motor coordination in autism spectrum disorders: A synthesis and meta-analysis. Journal of Autism &

Developmental Disorders, 40(10), 1227–1240. http://dx.doi.org/10.1007/s10803-010-0981-3Frings, M., Gaertner, K., Buderath, P., Christiansen, H., Gerwig, M., Hein-Kropp, C., et al. (2010). Megalographia in children with cerebellar lesions and in children

with attention-deficit/hyperactivity disorder. Cerebellum, 9(3), 429–432. http://dx.doi.org/10.1007/s12311-010-0180-yFuentes, C., Mostofsky, S., & Bastian, A. (2009). Children with autism show specific handwriting impairments. Neurology, 73, 1532–1537. http://dx.doi.org/

10.1212/WNL.0b013e3181c0d48cFuentes, C., Mostofsky, S., & Bastian, A. (2010). Perceptual reasoning predicts handwriting impairments in adolescents with autism. Neurology, 75, 1825–1829.

http://dx.doi.org/10.1212/WNL.0b013e3181fd633dGreen, D., Charman, T., Pickles, T., Chandler, S., Loucas, T., Siminoff, E., et al. (2009). Impairment in movement skills of children with autistic spectrum disorders.

Developmental Medicine & Child Neurology, 51(4), 311–316. http://dx.doi.org/10.1111/j. 1469-8749.2008.03242.xHaymaker, W. (1956). Bing’s local diagnosis in neurological diseases. St. Louis, MO C. V. Mosby.Henderson, S. E., Sugden, D. A., & Barnett, A. L. (2007). Movement assessment battery for children (2nd ed.). London, England: Pearson Assessment.Holmes, G. (1917). The symptoms of acute cerebellar injuries due to gunshot injuries. Brain, 40, 461–535, doi:10.1093/brain/40.4.461.Kushki, A., Chau, T., & Anagnostou, E. (2011). Handwriting difficulties in children with autism spectrum disorders: A scoping review. Journal of Autism &

Developmental Disorders, 41(12), 1706–1716. http://dx.doi.org/10.1007/s10803-011-1206-0Langen, M., Schnack, H. N. H., Lahuis, B., de Jonge, M., van Engeland, H., & Durston, S. (2009). Changes in the developmental trajectories of striatum in autism.

Biological Psychiatry, 66, 327–333. http://dx.doi.org/10.1016/j.biopsych.2009.03.017Langmaid, R., Papadopoulos, N., Johnson, B., Phillips, J., & Rinehart, N. (2012). Handwriting in Children with ADHD. Journal of Attention Disorders http://dx.doi.org/

10.1177/1087054711434154 [Ahead of print]Martin, K., Phillips, J., Iansek, R., & Bradshaw, J. (1994). Inaccuracy and instability of sequential movements in Parkinson’s disease. Experimental Brain Research, 102,

131–140. http://dx.doi.org/10.1007/BF00232445Miall, R. C., Weir, D. J., Wolpert, D. M., & Stein, J. F. (1993). Is the cerebellum a Smith predictor? Journal of Motor Behavior, 25, 203–216. http://dx.doi.org/10.1080/

00222895.1993.9942050Miyahara, M., Piek, J. P., & Barrett, N. C. (2008). Effect of postural instability on drawing errors in children: A synchronized kinematic analysis of hand drawing and body

motion., 27(5), 705–713. http://dx.doi.org/10.1016/j.humov.2008.03.001Mostofsky, S., Powell, S., Simmonds, D., Goldberg, M., Caffo, B., & Pekar, J. (2009). Decreased connectivity and cerebellar activity in autism during motor task

performance. Brain, 132, 2413–2425. http://dx.doi.org/10.1093/brain/awp088Myles, B., Huggins, A., Rome-Lake, M., Hagiwara, T., Barnhill, G., & Griswold, D. (2003). Written language profile of children and youth with Asperger syndrome:

From research to practice. Education and Training in Developmental Disabilities, 38(4), 362–369.Ozonoff, S., & Cathcart, K. (1998). Effectiveness of a home program intervention for young children with autism. Journal of Autism & Developmental Disorders, 28(1),

25–32.Papadopoulos, N., McGinley, J., Tonge, B., Bradshaw, J., Bradshaw, J. L., Saunders, K., et al. (2012). Motor proficiency and emtional/behavioural disturbance in

autism and Asperger’s disorder: Another piece in the neurological puzzle? Research in autism spectrum disorders, 6(1), 286–292. http://dx.doi.org/10.1016/j.rasd.2011.05.010

Papadopoulos, N., McGinley, J., Tonge, B., Bradshaw, J., Saunders, K., Murphy, A., et al. (2012). An investigation of upper limb motor function in high functioningautism and Asperger’s disorder using a repetitive Fitts’ aiming task. Autism, 16(6), 627–640. http://dx.doi.org/10.1177/1362361311418692

Phillips, J. G., Bradshaw, J. L., Chiu, E., & Bradshaw, J. A. (1994). Characteristics of handwriting of patients with Huntington’s disease. Movement Disorders, 9(5), 521–530. http://dx.doi.org/10.1002/mds.870090504

Phillips, J., Ogeil, R., & Best, C. (2009). Motor constancy and the upsizing of handwriting. Human Movement Science, 28(5), 578–587. http://dx.doi.org/10.1016/j.humov.2009.07.004

Phillips, J., Ogeil, R., & Muller, R. (2009). Alcohol consumption and handwriting: A kinematic analysis. Human Movement Science, 28, 619–632. http://dx.doi.org/10.1016/j.humov.2009.01.006

B.P. Johnson et al. / Research in Developmental Disabilities 34 (2013) 2917–29262926

Plamondon, R., & Guerfali, W. (1998). The generation of handwriting with delta-lognormal synergies. Biological Cybernetics, 78(2), 119–132. http://dx.doi.org/10.1007/s004220050419

Quill, K. (1997). Instructional considerations for young children with autism: The rationale for visually cued Instruction. Journal of Autism & DevelopmentalDisorders, 27(6), 697–714. http://dx.doi.org/10.1023/A:1025806900162

Qiu, A., Adler, M., Crocetti, D., Miller, M. I., & Mostofsky, S. H. (2010). Basal ganglia shapes predict social, communication, and motor dysfunctions in boys withautism spectrum disorder. Journal of the American Academy of Child & Adolescent Psychiatry, 49(6), 539–551. http://dx.doi.org/10.1016/j.jaac.2010.02.012

Rinehart, N. J., Bellgrove, M. A., Tonge, B. J., Brereton, A., Howells-Rankin, D., & Bradshaw, J. L. (2006). An examination of movement kinematics in young peoplewith high-functioning autism and Asperger’s disorder: Further evidence for a motor planning deficit. Journal of Autism and Developmental Disorders, 36(6),757–767. http://dx.doi.org/10.1007/s10803-006-0118-x

Sivaswamy, L., Kumar, A., Rajan, D., Behen, M., Muzik, O., Chugani, D., et al. (2010). A diffusion tensor imaging study of the cerebellar pathways in children withautism spectrum disorder. Journal of Child Neurology, 25(10), 1223–1231. http://dx.doi.org/10.1177/0883073809358765

Slavin, M., Phillips, J., Bradshaw, J., Hall, K., & Presnell, I. (1999). Consistency of handwriting movements in dementia of the Alzheimer’s type: A comparison withHuntington’s and Parkinson’s diseases. Journal of the International Neuropsychological Society, 5(1), 20–25.

Smits-Engelsman, B., & Van Galen, G. (1997). Dysgraphia in children: Lasting psychomotor deficiency or transient developmental delay? Journal of ExperimentalChild Psychology, 67, 164–184. http://dx.doi.org/10.1006/jecp.1997.2400

Tucha, O., Mecklinger, L., Walitza, S., & Lange, K. (2006). Attention and movement execution during handwriting. Human Movement Science, 25, 536–552. http://dx.doi.org/10.1016/j.humov.2006.06.002

Turner, M. A. (1999). Generating novel ideas: Fluency performance in high-functioning and learning disabled individuals with autism. Journal of Child Psychologyand Psychiatry, 40(2), 189–201.

Van Galen, G. P., Portier, S. J., Smits-Engelsman, B. C. M., & Schomaker, L. R. B. (1993). Neuromotor noise and deviant movement strategies as an explanatory groundfor poor handwriting in children. Acta Psychologica, 82(1–3), 161–178. http://dx.doi.org/10.1016/0001-6918(93)90010-O

Van Galen, G. P., Van Doorn, R. R. A., & Schomaker, L. R. B. (1990). Effects of motor programming on spectral density function of finger and wrist movements. Journalof Experimental Psychology: Human Perception and Performance, 16, 755–765. http://dx.doi.org/10.1037/0096-1523.16.4.755

Wechsler, D. (1999). Wechsler Abbreviated Scale of Intelligence (WASI). San Antonio, TX: Harcourt Assessment.Wechsler, D. (2003). Wechsler Intelligence Scale for Children (4th ed.). San Antonio, TX: Harcourt Assessment.Witwer, A., & Lecavalier, L. (2007). Autism screening tools: An evaluation of the Social Communication Questionnaire and the Developmental Behaviour Checklist

– Autism Screening Algorithm. Journal of Intellectual and Developmental Disability, 32(3), 179–187. http://dx.doi.org/10.1080/13668250701604776