Attentional and neuromotor deficits in ADHD

Attentional andneuromotor deficits inADHD

J Steger PhD, Laboratory for Biomechanics, Eidgenössiche

Technische Hochschule (ETH) Zurich;

K Imhof PhD;

E Coutts lic phil;

R Gundelfinger MD;

H-Ch Steinhausen MD PhD;

D Brandeis*PhD, Department of Child and Adolescent

Psychiatry, University of Zurich, Switzerland.

*Correspondence to last author at Department of Child and

Adolescent Psychiatry, University of Zurich,

Neumünsterallee 9/ Postfach, CH-8032 Zürich, Switzerland.

E-mail: [email protected]

In order to classify attention-deficit–hyperactivity disorder(ADHD) in 11-year-old children, the role of specificattentional and motor deficits was examined. Participantscomprised 22 children with ADHD (19 male, 3 female;median age 11 years, range 8.8 to 13.5 years) and 20 controlchildren (17 male, 3 female; median age 10.6 years, range 8.2to 12.6 years). Neuromotor assessment indicated that whileboth groups needed more time to complete finger compared tohand movements, this increase was more pronounced inchildren with ADHD. Reaction-time testing with continuous-force recording identified both motor and attentional deficitsin children with ADHD. Longer intervals between force onsetand force peak, and higher rate of responses with multipleforce peaks (particularly in the bilateral condition) revealedspecific deficits in the speed and quality of their motor output.Increase in errors and variability of force onsets indicatedattentional deficits. Prediction analysis indicated that force-onset variability contributed significantly to groupclassification which was 85.7% correct. Neither neuromotorassessment nor specific motor deficits contributedsignificantly to classification, indicating that pure motor-speed measures play a minor role in characterizing ADHD inthis age range.

Attention-deficit–hyperactivity disorder (ADHD) is charac-

terized by developmentally inappropriate attention, overac-

tivity, and impulsivity, but several lines of evidence suggest

that children with ADHD also have motor deficits. Some

research suggests that the overlap between ADHD and motor

deficits is close to 50% and therefore high enough to warrant a

separate diagnostic category (deficit in attention, motor con-

trol, and perception DAMP; Gillberg 1995) while other

research suggests that motor processing is typically unaffected

in ADHD (Leung and Connolly 1998). Such large discrepan-

cies are likely to result from addressing different populations,

and from differences in the assessment and conceptualization

of motor deficits, ADHD, and their combination.

Motor deficits in ADHD have been assessed in a wide

range of tests. In the neuromotor assessment battery of

Denckla and Rudel (1978), the speed of pure motor func-

tions is strongly represented. The time needed for repetitive

simple movements or movement sequences in the different

tasks, is measured for finger, hand, and foot. Denckla and

Rudel (1978), Carte and colleagues (1996), and Schuerholz

and coworkers (1997) all used these tasks to examine chil-

dren with ADHD with a mean age below 10 years. Denckla

and Rudel (1978) reported that slowing of gross movements

(i.e. heel–toe taps) differentiated the ADHD and control

groups most accurately while Carte and colleagues (1996)

found that hand and foot movements were significantly

slower in children with ADHD. Both Denckla and Rudel

(1978) and Carte and colleagues (1996) showed that neuro-

motor parameters contributed significantly to group dis-

crimination between children with ADHD and control

children. Schuerholz and colleagues (1997) reported that

even though children with ADHD showed no overall slow-

ing of finger movements, they had a larger time discrepancy

between simple and patterned movements, and performed

fewer tasks within the age-appropriate speed limits than

children with Tourette syndrome. More subtle motor

deficits which affect movement quality rather than speed are

also found in ADHD. These include motor overflow, i.e.

involuntary associated movements such as mirror movements

(Denckla and Rudel 1978). Motor overflow also appears to cor-

relate with inattentiveness in normally developing school-age

children (Waber et al. 1985, Lazarus and Todor 1991).

More complex tests of fine motor skills based on handwrit-

ing, manual dexterity, ball skills, and balance have also

indicated deficits in children with ADHD. However these

tests tend to measure both pure motor as well as visuomotor

or higher cognitive functions, similar to neuropsychological

frontal-lobe tests such as copying a complex design or maze

tracing. Compared to control children, children with ADHD

were found to have impaired handwriting skills (McMahon

and Greenberg 1977, Barkley 1990), poorer fine visuomotor

ability (Whitmont and Clark 1996), and poorer performance

in visuomotor frontal-lobe tests (review in Barkley et al.

1992). In addition, different types of motor deficits appear

to characterize different ADHD subgroups (Piek et al. 1999).

Children of the inattentive subtype of ADHD had significantly

poorer manual dexterity skills, whereas children of the com-

bined subtype experienced greater difficulty with balance.

Direct experimental manipulation of premotor processing

via low or irregular event rate (Chee et al. 1989, van der Meere

et al. 1992, Földényi et al. 2000) or via stimulus–response

incompatibility (van der Meere et al. 1989) have also revealed

172 Developmental Medicine & Child Neurology 2001, 43: 172–179

some specific premotor deficits in ADHD. However, these

premotor deficits are derived from manipulations affecting

higher-order regulation, and are less consistent than the

attentional and inhibitory deficits (Oosterlaan and Sergeant

1998) revealed by increases of response-time variability and

errors of omission and commission, i.e. misses and false alarms

(Földényi et al. 2000). The fact that neither choice reaction

time nor movement time are increased in children with ADHD

(Leung and Connolly 1998) suggests that there is little evi-

dence for specific motor-speed deficits in simple motor tests.

In summary, the majority of studies with neuromotor

standard assessments indicate that the quality and speed of

simple movement sequences discriminate between those

with ADHD and control children. However this conclusion

receives little support from a study using simple motor labo-

ratory tests that measure motor sequencing speed on a trial-

by-trial basis (Leung and Connolly 1998). Continuous-force

recordings rather than discrete button presses would be

required to assess the quality of motor output in such labora-

tory tests. Other types of motor coordination have not been

examined on a trial-by-trial basis in ADHD children.

Bilateral responding (i.e. synchronization of unilateral

responses) reflects a basic form of motor coordination that

results in slower reaction times (around 5%) than unilateral

response (Jeeves and Dixon 1970, Ohtsuki 1981, Anson and

Bird 1993). This bilateral reaction-time deficit is present even if

separate blocks of uni- and bilateral trials permit optimal atten-

tional strategies and response preparation (Kawabe 1989).

This may be due to additional motor coordination rather than

attentional or cognitive coordination. In addition, the reduc-

tion of this bilateral timing deficit in patients with split-brains

(Gazzaniga and Sperry 1966, Wyke 1971) indicates that it is

mediated by the corpus callosum. As some studies indicate

that certain callosal regions are reduced in children with

ADHD (Semrud-Clikeman et al. 1994, Castellanos et al. 1996,

but see Overmeyer et al. 2000), impaired bilateral coordina-

tion may be expected in this group.

The aims of the present study were: (1) to determine spe-

cific motor processing deficits in children with ADHD using

neuromotor assessment and continuous-force recordings

in a reaction-time test involving bilateral coordination, and

(2) to clarify the contribution of attentional and motor

measures to the prediction of an ADHD diagnosis.

MethodPARTICIPANTS

Forty-two right-handed children (IQ>80; age range 8.2 to

13.5 years) were selected for this study. The 22 children with

ADHD (mean age 10.9 years, range 8.8 to 13.5 years) and 20

sex- and age-matched control children (mean age 10.6 years

range 8.2 to 12.6 years) were grouped on the basis of DSM-

III-R (American Psychiatric Association 1987) diagnoses

derived from a structured interview (Schaffer et al. 1993) car-

ried out by trained undergraduate psychology students. The

children with ADHD were recruited from our patients (n=11),

from parents’ organizations (n=5), from an epidemiological

field study (n=4; Steinhausen et al. 1998), and from nearby

schools through teachers (n=1), or school psychologists

(n=1). They all met the ADHD diagnosis and had no comor-

bid disorders except for oppositional defiant disorder (11

children) and/or conduct disorder (two children), and/or

one isolated phobia (five children). Control children were

recruited from regular schools through teachers or laboratory

staff members or from the same field study. No DSM-III-R diag-

noses were allowed except for a single isolated phobia (two

children). Control children’s T-scores (mean 50, SD 10) on the

Attention Problem Scale of the Child Behavior Check List

(CBCL; Achenbach 1991a) had to be lower than 55, i.e. less

than half a standard deviation above the normative value. The

sample description (for both the full group and subgroup

selected for additional analyses) is given in Table I.

At initial assessment the children’s parents and teachers

reported on behavior problems at home (CBCL) and at school

(Teacher’s Report Form, TRF; Achenbach 1991b). In the first

laboratory session all children underwent neuromotor assess-

ment and several neuromotor tests. An abbreviated version of

the revised Wechsler Intelligence Scale for Children (WISC)

with German normative values (Bründler and Schallberger

1988) was used to determine IQ. In a second laboratory ses-

sion a reaction-time test involving bilateral coordination was

performed in addition to other neuropsychological tests. All

children were free of medication at the time of testing; those

children with ADHD taking methylphenidate (n=7) had been

off medication for at least 24 hours.

Attentional and Neuromotor Deficits in ADHD J Steger et al. 173

Table I: Description of sample

Full group SubgroupADHD Control p ADHD Control pn=22 n=20 n=14 n=14

Age 10.93 10.55 >0.3 10.54 10.78 >0.5

Sex (F/M) 3/19 3/17 0/14 0/14

IQ 98.67 105.56 <0.05 101.43 104.10 >0.3

CBCL attention problems 66.13 43.10 <0.0001 64.31 43.11 <0.0001

(T-scores)

Comorbidity (nr children)

Multiple comorbidities 4 – – 2 – –

Single comorbidity 10 2 – 8 2 –

ODD 11 – – 7 – –

CD 2 – – 1 – –

Phobias 5 2 – 4 2 –

CBCL, Child Behavior Checklist; ODD, Oppositional defiant disorder; CD, Conduct disorder.

NEUROMOTOR ASSESSMENT

Each child was examined by trained child psychiatrists using

a neuromotor assessment battery (Largo and Caflisch 1999)

based on Denckla’s (1974) six tasks of repeated movements.

The children were instructed to execute each of the follow-

ing tasks rapidly but precisely while the time needed to com-

plete the movements was measured: (1) Finger repetition

(simple movement), the child touches index finger to

thumb (20 repetitions). (2) Finger sequention (patterned

movement), the child touches each finger successively to

thumb, moving from the index finger to the little finger each

time (5 repetitions). (3) Hand pat (simple movement), the

child pats the thigh with the palm of the hand, with the heel

of the palm resting on the thigh (20 repetitions). (4) Hand

pronation–supination (patterned movement), the child

pats the thigh with the back and palm of the hand alternately

(10 repetitions). (5) Toe taps (simple movement), the child

taps toe with the heel resting on the floor (20 repetitions).

(6) Heel–toe (patterned movement), the child alternates

touching heel and toe to the floor in a rocking motion (10

repetitions).

REACTION-TIME TEST

Stimuli and testFigure 1 illustrates how pictures of an aeroplane or a man

(both subtending 0.95˚ horizontally and vertically) were pre-

sented on a computer monitor every 1.65 seconds for 180

seconds in black on a light grey background. Vertical lines

above and below the center of the picture served as fixation

marks. These pictures appeared with equal probability on

the left, right, or both sides simultaneously (1.33˚ from fixa-

tion). On target trials (p=0.67), the aeroplane which required

a response was presented. On non-target trials (p=0.33), the

man was presented at the same screen positions as the aero-

plane (left, right, bilateral), and the child had to refrain from

responding. The six test conditions (target or non-target, each

in left unilateral, right unilateral, and bilateral conditions)

were presented in randomized order. Children were seated

1.2 metres from the monitor. They responded as quickly as

possible with the hand corresponding to the target position

(left, right, or simultaneously with both hands), using oppos-

ing pressure of thumb and index finger (precision grip). After

a short practise run, three experimental blocks, each contain-

ing 60 targets and 30 non-targets, followed.

Force recording and processingResponse force between thumb and index finger was mea-

sured by force transducers (strain gauges) and continuously

recorded for each hand at 256 Hz. The force pulses produced

by the responses were analyzed as shown in Figure 1.

The onsets of the force pulses (force-onset latencies)

were automatically marked as the time points at which the

force signals exceeded the noise level. In addition, the

time between force onset and force peak, and the peak force

were computed. To be coded as a correct trial, the force-onset

latency had to be within a time window of 200 and 800 ms and


Figure 1: Selected stimuli (not drawn to scale) and force-pulse analysis in reaction-time test. Pulses in force curves for left(bottom) and right (top) hand reflect correct responses for this arbitrary stimulus sequence. Inset: black line indicates,correct; dashed small lines indicate incorrect; and dashed multipeak lines indicate imperfect right-hand responses to rightunilateral target. (a), parameters force-onset latency, (b) time from force onset to force peak, (c) peak force.

Left hand

Right hand

0 ms 200 ms 800 ms

10 (N)

Force (N)

Time

0 s 1.65 s 3.3 s 4.95 s

Right handLeft hand

peak force had to be higher than 0.5 N (grey box in Fig. 1).

In addition to the traditional error types (omission and

commission errors, side errors, unilateral responses in the

bilateral condition, and bilateral responses in unilateral con-

ditions), several imperfect but correct responses were coded

and analysed. These were: (1) trials with force-onset laten-

cies shorter than 200 ms or longer than 800 ms, (2) trials with

mirror movements (correct but additional response pulses

below 0.5 N with the incorrect hand, see Fig. 1), (3) trials

with multiple force peaks (see Fig. 1), and (4) trials with

poor synchronization of bilateral responses (defined as

force-onset latency differences of more than 60 ms ). These

imperfect responses were not included in the total error

count in order to maintain compatibility with previous error

analyses. All errors were transformed into percentages and

time measures were log transformed (base 10).

DATA ANALYSIS

Group and condition effects were first tested separately for

the parameters of the neuromotor assessment, and the reac-

tion-time test. Multivariate analysis of variance for repeated

measures covaried for IQ (MANCOVA/GLM) were used for

this purpose. Equivalent F values for Wilk’s Lambda are

reported. These multivariate analyses were followed by uni-

variate analyses of covariance (ANCOVA) for specific tasks or

test conditions. Due to the fact that sex and IQ were suspect-

ed to correlate with an ADHD diagnosis, all analyses were

also performed on subgroups matched for sex and IQ

(instead of covarying for IQ). These subgroup results are

only reported if there were discrepancies with the full

group analyses. As no significant differences were obtained

between left- and right-sided responses, their mean was

taken for all analyses. For the neuromotor assessment, the

time (in seconds) needed to complete the movements in the

six tasks was analyzed with the factors group (ADHD versus

control), extremity (finger versus hand versus leg), move-

ment mode (simple versus patterned), and with age as an

additional covariate. For the reaction-time test, the time and

error measures listed above were analyzed with the factors

group, bilaterality (mean unilateral versus bilateral), and tar-

get/non-target (targets versus non-targets, only for error

parameters). These analyses were performed both with and

without age as an additional covariate.

For the group prediction (ADHD or control group, full

groups only), the best group discriminator from the neuro-

motor analysis, and the best group discriminator from the

reaction-time analyses were selected. These two measures

and the IQ were entered together in a logistic regression.

Due to several measures in the reaction-time test indicating

high significance group differences (p<0.001) logistic regres-

sions were computed for each of those measures alone to

determine the one with the best predictive value. Specific

settings of all these maximum-likelihood predictions includ-

ed: forward and backward stepwise method, entry p=0.05,

removal p=0.1; classification cut off p=0.05; and 20 maxi-

mum iterations. In case of discrepancies between forward

and backward solutions, the window between entry and

removal criterion was reduced (entry p=0.05, removal

p=0.06). Subanalyses with single predictors were run to clari-

fy the effects. Significance values of the Wald statistic are based

on the χ2 distribution.

ResultsTable I shows that the mean IQ was lower for children with

ADHD than for control children whereas the analysis of the

subgroups (n=14 for each group, males only) showed no

significant IQ differences. According to the CBCL, children

with ADHD had significantly higher attention-problem scores

than control children. No significant age differences between

the ADHD and control children were found.

NEUROMOTOR ASSESSMENT

The multivariate analysis of the neuromotor measures indi-

cated that whether children with ADHD needed more time to

complete movements than control children depended on the

extremity (group ×extremity, F[2,37]=3.77, p<0.05). Post-

hoc analysis of this interaction indicated that the additional

time needed to complete finger compared to hand move-

ments was increased in the ADHD group (group ×extremity,

F[1,38]=7.27, p<0.01), but none of the three extremities

revealed slowing when tested individually except for a trend

for slower finger movements (F[1,38]=3.22, p=0.08).

Univariate analysis also showed no significant group differ-

ences (see Table II) and only a trend for slower sequential fin-

ger movements in the ADHD group (F[1,38]=2.89, p=0.1).

A main effect of extremity (F[2,37]=4.96, p<0.05) indi-

cated that finger and foot movements were slower than

hand movements. Patterned movements were slower than

simple movements (8.41 versus 5.84 seconds, F[1,38]=10.44,

p<0.01). In addition, the covariate age significantly affected

the time to complete movements, F[1,38]=8.00, p<0.01).

The negative correlations between age and time to complete

finger (r=–0.260, p=0.048), hand (r=–0.436, p=0.002), and

foot (r=–0.318, p=0.02) movements indicated that younger

children were slower with all extremities. No significant

resutls were obtained for the covariate IQ.

REACTION-TIME TEST

The multivariate analysis of the values shown in Table III

indicated that in children with ADHD, force onsets were

more variable (force onset, SD 110.0 versus 84.6 ms ;

F[1,39]=12.51, p<0.001) but not significantly later than in

control children. In addition, the time from force onset to

force peak was prolonged in ADHD compared to control

children (225.2 versus 192.9 ms F[1,39]=4.42, p<0.05).

Bilateral responding was significantly slower than unilateral

responding (bilateral reaction-time deficit), but only in the

subgroup analysis (443.3 versus 422.4 ms , F[1,26]=12.74, p<0.001). No significant group differences were found for the


Table II: Neuromotor task measures (unilateral), mean (SD)

ADHD (n=22) Controls (n=20)

Finger

Repetition 5.85 (1.28) 5.68 (1.10)

Sequention 10.37 (2.47) 8.87 (2.06)

Hand

Pat 4.84 (1.13) 5.07 (1.09)

Pronation–supination 7.00 (1.30) 6.88 (1.29)

Foot

Toe taps 7.20 (2.19) 6.39 (1.25)

Heel–toe 9.23 (3.17) 8.10 (2.25)

bilateral reaction-time deficit. Children with ADHD made

more errors than control children (13.8% versus 4.7%;

F[1,39]=15.39, p<0.001); this was most pronounced for

bilateral non-targets (F[1,39]=4.89, p<0.05). They also

made more commission errors than control children

(F[1,39]=9.76, p<0.005) and responded more often with

multiple force peaks (F[1,39]=18.36, p<0.001), particularly

in the bilateral condition (F[1,39]=5.69, p<0.05). No signifi-

cant group differences were found for omission errors or

responses outside the latency window (200 to 800 ms).

Compared to control children, those with ADHD also made

fewer synchronized bilateral responses in the bilateral

condition (F[1,39]=20.21,p<0.001) and made more bilateral

responses in the unilateral conditions (F[1,39]=15.36,

p<0.001). No significant group differences were found for side

errors, mirror movements, unilateral responses in the bilateral

condition, or peak force.

Significant effects of the covariate IQ indicated that children

with lower IQ made fewer synchronized bilateral responses

(F[1,39]=4.13, p<0.05) but they also made fewer erroneous

bilateral responses in the unilateral conditions (F[1,39]=7.52,

p<0.01). Introducing age as an additional covariate revealed

no significant age effects on reaction-time test parameters

except for two trends (force onset latency and variability

p<0.1). Group and condition effects remained unchanged.

GROUP PREDICTION

From the neuromotor assessment measures, only the time

difference between finger and hand movements (group ×extremity, neuromotor interaction) discriminated between

the groups. In the reaction-time test, individual logistic regres-

sion analysis of several highly discriminating measures (Table

IV) indicated that response variability was the best predictor.

IQ was included because of the significant group differences.

Predictions based on these measures from neuromotor

assessment and the reaction-time test are shown in Table V.

The final equation contained overall response variability and

IQ, but not the neuromotor interaction measure, as signifi-

cant predictors. This equation resulted in the correct classifi-

cation of 85.71% of all participants. Twenty-one of 22

children with ADHD (95.45%) and 15 of 20 control children

(75%) were correctly classified. The initial backward solution

retained both the IQ and the neuromotor interaction at a

non-significant level, but more similar entry- and removal-

criteria produced consistent forward and backward solutions

identical to the initial forward approach. When entering each

predictor separately, response variability resulted in 80.95%,

IQ 64.29%, and neuromotor interaction measure 69.5% cor-

rect classification.

DiscussionNEUROMOTOR ASSESSMENT

Neuromotor assessment indicated that 11-year-old children

with ADHD had no clear neuromotor-speed deficits except

that they were slower than control children when perform-

ing finger movements compared to hand movements. This

interaction suggests that children with ADHD have specific

problems with motor tasks that require fine motor skills.

Finding only a trend towards slowing across all finger move-

ments, is consistent with the results of a recent report using

the same tasks and analyses controlling for age and IQ

(Schuerholz et al. 1997). Their finding of more pronounced

slowing with patterned compared to repetitive finger move-

ments was not obtained in our data. However, the lack of

deficits for hand or foot movements alone contrasts with pre-

vious studies reporting differences for both hand and foot

movements (Denckla and Rudel 1978, Carte et al. 1996).

Considering that several standard findings (significant slow-

ing for younger children, patterned movements, and fin-

ger/foot compared to hand movements) support the validity


Table III: Reaction time test measures, mean (SD)

ADHD (n=22) Control (n=20)Unilateral Bilateral Unilateral Bilateral

Force onset (ms) 421.7 (70.8)c 445.2 (85.5)a 415.2 (72.6) 439.9 (68.2)

Variability of force onsets (ms) 110.3 (26.7)c 109.3 (27.1)a 84.2 (13.0) 85.5 (15.4)

Time from onset to peak (ms) 228.8 (52.3)a 221.6 (49.0)a 197.7 (39.2) 188.1 (42.5)

Total errors (%) 15.5 (9.0)a 12.0 (10.0)a 6.8 ( 3.0) 2.7 (2.3)

Errors (%)

Target 13.0 (11.7)a 6.6 (15.3)d 5.3 (3.3) 0.8 (1.1)

Non-target 18.0 (11.3)c 17.4 (13.1)d 8.2 (4.7) 4.7 (4.2)

Omissions 4.5 (10.1)c 3.8 (10.4)d 0.8 (1.6) 0.3 (0.8)

Commissions 12.1 (7.7)b 5.5 (6.9)c 6.5 (3.7) 1.7 (3.3)

Multiple peaks 6.9 (6.1)a 19.0 (9.2)c 2.2 (2.1) 8.8 (5.9)

Window 3.2 (3.5)a 2.5 (3.3)c 2.0 (3.9) 0.6 (1.2)

Side 1.0 (2.0)a – 0.6 (0.9) –

Bi–uni 7.4 (5.3)d – 3.9 (3.4) –

Mirror movements 3.8 (3.6)d – 2.9 (2.1) –

Uni–bi – 2.8 (5.2)dd – 0.5 (1.0)

Synchronicity – 10.7 (5.0)d – 5.5 (3.5)

Individual ANCOVA results with IQ covaried: a p<0.05; b p<0.01; c p<0.005; d p<0.001.

Multiple peaks, responses with multiple force peaks; Side, left hand response in right-side condition (vice

versa); Window, responses outside the latency window; Bi–uni, bilateral responses in unilateral conditions;

Uni–bi, unilateral responses in bilateral condition; Synchronicity, synchronicity of bilateral responses.

Multiple peaks, window, mirror movements, and synchronity are imperfect but correct responses.

k

of our neuromotor assessment, it appears likely that group

characteristics are responsible for these discrepancies. One

explanation for these discrepancies might be the older mean

age of the children in our groups. Gillberg and colleagues

(1983) and Gillberg and Gillberg (1989) showed that neuro-

motor deficits observed in children with ADHD at the age of 6

or 7 years often disappeared at age 10 and 13 years even

though behavioural problems persisted. Whereas the studies

of Denckla and Rudel (1978) and Carte and coworkers (1996)

included children younger than 8 years, the children partici-

pating in the present study were all older than 8 years, i.e. at

an age where neuromotor deficits in ADHD are less promi-

nent. In addition, our ADHD group may have contained chil-

dren with less severe ADHD than the sample of Carte (1996),

which consisted only of children with severe ADHD who had

been on stimulant treatment. Such a correlation between

motor deficits and severity was confirmed by Whitmont and

Clark (1996). Whereas additional analyses confined to chil-

dren younger than 10 or to children with severe ADHD failed

to reveal neuromotor deficits for finger or foot movements

alone, the small samples (n<10) available for these analyses

do not allow firm conclusions.

REACTION-TIME TEST

The force-onset latencies from the reaction-time test also

indicated no general slowing in ADHD compared to control

children, which is consistent with other studies using simple

motor tests (Barkley et al. 1992, Leung and Connolly 1998)

and high event rates (van der Meere et al. 1992). Additional

motor coordination demands during bilateral responding

resulted in a bilateral reaction-time deficit, i.e. in slower

force onsets than with unilateral responding in the IQ-

matched subgroups. This effect was similar in size to that

observed in adults (Jeeves and Dixon 1970, Anson and Bird


Table IV: Evaluation of best predictor of reaction-time test for diagnostic status

Method Variables Parameter estimates Standard error Adjusted odds ratio p Overall classification %

Enter Variability 0.1046 8.0232 1.1103 0.005 85.71

IQ –0.1019 3.8771 0.9032 0.049

Enter Total errors 55.0279 7.7888 7.912E+23 0.005 83.33

IQ –0.1526 3.9819 0.8585 0.046

Enter Synchronicity 45.1713 8.3961 4.146E+19 0.004 80.95

IQ –0.1822 5.9702 0.8334 0.015

Enter Multiple peaks 14.6179 9.3773 2.230E+06 0.002 78.57

IQ –0.0634 1.7476 0.9386 0.186

Enter Bi–uni 37.0397 9.0808 1.219E+16 0.003 73.81

IQ –0.1755 7.4096 0.8390 0.007

Multiple peaks, responses with multiple peaks; Bi–uni, bilateral responses in unilateral conditions; Synchronicity, synchronicity of bilateral

responses.

Table V: Predicting diagnostic status through logistic regression

Method Entry/removal Variables Parameter Standard error Adjusted odds p Overallprobability estimates ratio classification %

Forward 0.05/0.1 Variability 0.1046 8.0232 1.1103 0.005 85.71

Movement – – – 0.0919

time

IQ –0.1019 3.8771 0.9032 0.05

Backward 0.05/0.1 Variability 0.0917 6.1202 1.0960 0.0134 88.10

Movement 0.4620 2.5943 1.5872 0.1072

time

IQ –0.0951 3.0749 0.9093 0.0795

Forward 0.05/0.06 Variability 0.1046 8.0232 1.1103 0.005 85.71

Movement – – – 0.0919

time

IQ –0.1019 3.8771 0.9032 0.05

Backward 0.05/0.06 Variability 0.1046 8.0232 1.1103 0.005 85.71

Movement – – – 0.0919

time

IQ –0.1019 3.8771 0.9032 0.05

Enter 0.05/0.1 Variability 0.0972 8.7194 1.1021 0.0031 80.95

Enter 0.05/0.1 Movement 0.6229 7.0179 1.8644 0.0081 69.05

time

Enter 0.05/0.1 IQ –0.0957 5.2194 0.9087 0.0223 64.29

Variability, overall variability of reaction time test; Movement time, increased time to complete finger movements compared to hand

movements (group×extremity, neuromotor variables).

1993), which suggests that the underlying processes were

relatively mature in our 11-year-old children. Whether the

bilateral deficit in adults is also eliminated by covarying IQ

remains to be examined. The groups did not differ in the size

of their bilateral deficits. This contrasts with the reduced

bilateral deficit of patients with split-brains and suggests that

those callosal functions supporting bilateral motor coordi-

nation are unaffected in children with ADHD.

Measuring continuous-force output in this test provided

direct measures of motor function that are unavailable with

typical reaction-time testing. Due to the fact that peak force

was comparable between ADHD and control children, the

prolonged time from force onset to force peak in children

with ADHD must reflect a pure motor-speed deficit. Another

type of specific motor deficit was revealed by the increased

rate of multiple response peaks in children with ADHD

which was most pronounced under bilateral coordination

demands. Deterioration of the quality of motor output,

therefore, increased with demands. This may contribute to

the high rates of motor deficits in children with ADHD which

are obtained if qualitative aspects of motor performance in

more complex skills are also scored (Gillberg 1995, Kadesjo

and Gillberg 1998).

Mirror movements are another aspect of motor function

that can only be measured with continuous-force recordings

in reaction-time studies. The mirror movements were not

significantly increased in the group with ADHD, which con-

trasts with a neuromotor assessment study in younger chil-

dren (Denckla and Rudel 1978) and parallels the lack of

significant neuromotor timing differences in this age range.

Future studies with younger children, who are known to

exhibit more mirror movements, could clarify the relation-

ship between mirror movements derived from quantifica-

tion of continuous-force output and from visual ratings

during neuromotor assessments.

Children with ADHD made more errors than control chil-

dren and showed increased variability of force-onset latencies

during correct responding. These findings are consistent

with previous studies (Mitchell et al. 1990, Schachar 1991,

Oosterlaan and Sergeant 1998, Rubia et al. 1998, Földényi et

al. 2000) and are thought to reflect attentional deficits in chil-

dren with ADHD. In particular, high response variability has

been associated with non-optimal regulation of attentional or

energetic states (Oosterlaan and Sergeant 1998). Similarly, the

fewer synchronized bilateral responses of children with

ADHD may be due to the increased variability of the participat-

ing unilateral responses. As neither the latency nor the vari-

ability of bilateral responses were specifically increased in

the group with ADHD, contributions from a specific transcal-

losal coordination deficit appear less likely.

Children with ADHD made slightly more unilateral instead

of bilateral responses than control children, but they made

many more erroneous bilateral responses on unilateral trials,

and particularly on unilateral non-targets (commission errors).

This suggests a bias towards bilateral responding throughout

the test and could reflect attentional response–control deficits

in children with ADHD. Commission errors were more fre-

quent in ADHD than in control children, which is consistent

with findings from other tests using similarly high response

rates (50%) either throughout or after cues (Michael et al.

1981, van Leeuwen et al. 1998). The lack of significant group

differences in omission errors is also consistent with previous

studies; omission errors are typically prominent only in vigi-

lance tests with low-response probability.

Taken together, continuous-force measurements during

the reaction-time test revealed both specific motor and atten-

tional deficits in children with ADHD but no clear age effects

and no general motor slowing. This finding is consistent with

recent event-related potential findings during the same test,

which indicate that both premotor and attentional brain

activity is reduced in boys with ADHD despite normal laten-

cies (Steger et al. 2000). The neuromotor assessment also

indicated no general motor slowing in children with ADHD,

but instead resulted in clear age effects. Both modes of test-

ing thus indicated that children with ADHD have no general

motor slowing but specific deficits which differ from those

characterizing younger children.

Correlating neuromotor with reaction-time parameters

provided further insight into the determinants of movement

speed during the neuromotor assessment. Finger movement

time correlated with force-onset latency during correct

responding (r=0.387, p=0.011) but not with the latency

between force onset and peak or with error rate in the reac-

tion-time test. This suggests that the time needed to com-

plete finger movements primarily varies with force-onset

latencies and is relatively unaffected by movement time and

errors of movement or sequencing in this child population.

GROUP PREDICTION

Prediction analysis indicated that force-onset variability and IQ

contributed to 85.7% correct group classification. Force-onset

variability provided the main contribution to group discrimi-

nation and IQ a minor contribution. Neuromotor interaction

(finger versus hand) not only led to poorer classification than

force-onset variability when tested separately, but was

removed from the final equation in the iterative procedures.

Similarly, the motor parameter in the reaction-time test led to

a lower classification than latency variability and errors did.

This contrasts with the results from neuromotor studies of

younger children and with significant group differences for

individual neuromotor measures. In the study by Carte and

colleagues (1996), neuromotor variables provided one of the

highest independent contributions to the 82% correct group

classification (using discriminant analysis of eight age-adjust-

ed measures), similar to the 80.6% correct classification

obtained by Denckla and Rudel (1978).

In conclusion, even in simple motor reaction-time tests,

attentional measures seem to discriminate better than motor

measures between 11-year-old children with ADHD and con-

trol children. Strong support for the conclusion that motor

deficits are less pronounced in this age range comes from the

fact that subtle deficits, which are unlikely to allow visual scor-

ing, could be measured reliably through continuous-force

recording. Further studies have to prove the clinical feasibility

of this new method of assessing neuromotor function.

Accepted for publication 8th June 2000.

AcknowledgmentsWe wish to thank the child psychiatrists of the clinic assessing theneuromotor state of all children and H Haas from the Laboratory forBiomechanics (ETH Zurich) for the construction and technical supportof the force measurement device. This work was supported by theSwiss National Science Foundation (31-43790.95 and 32-43144.95).


ReferencesAchenbach TM. (1991a) Manual for the Child Behavior Check-

List/4-18 and 1991 Profile. Burlington, VT: Department ofPsychiatry, University of Vermont.

— (1991b) Manual for the Youth Self-Report and 1991Profile.Burlington, VT: Department of Psychiatry, University of Vermont.

American Psychiatric Association. (1987) Diagnostic and StatisticalManual of Mental Disorders. 3rd revised edn. Washington DC:American Psychiatric Association.

Anson JG, Bird YN. (1993) Neuromotor programming: bilateral andunilateral effects on simple reaction time. Human MovementScience 12: 37–50.

Barkley RA. (1990) Attention Deficit Hyperactivity Disorder: ahandbook for diagnosis and treatment. New York: GuildfordPress.

— Grodzinsky G, DuPaul GJ. (1992) Frontal lobe functions inattention-deficit disorder with and without hyperactivity: a reviewand research report. Journal of Abnormal Child Psychology 20:163–88.

Bründler M, Schallberger U. (1988) HAWIK-R Ergänzungsband mitNormentabellen und Testanweisungen für die deutschsprachigeSchweiz. Bern, Switzerland: Huber. (In German).

Carte ET, Nigg JT, Hinshaw SP. (1996) Neuropsychological functioning,motor speed, and language processing in boys with and withoutADHD. Journal of Abnormal Child Psychology 24: 481–98.

Castellanos FX, Giedd JN, Marsh WL, Hamburger SD, Vaituzis AC,Dickstein DP, Sarfatti SE, Vauss YC, Snell JW, Lange N et al. (1996)Quantitative brain magnetic resonance imaging in attention-deficit–hyperactivity disorder. Archives of General Psychiatry53: 607–16.

Chee P, Logan G, Schachar R, Lindsay P, Wachsmuth R. (1989) Effectsof event rate and display time on sustained attention inhyperactive, normal, and control children. Journal of AbnormalChild Psychology 17: 371–91.

Denckla MB. (1974) Development of motor co-ordination innormal children. Developmental Medicine & Child Neurology16: 729–41.

— Rudel RG. (1978) Anomalies of motor development inhyperactive boys. Annals of Neurology 3: 231–3.

Földényi M, Imhof K, Steinhausen H-C. (2000) Zur klinischenValidität der computerunterstützten Testbatterie (TAP) beiKindern mit Aufmerksamkeits- / Hyperaktivitätsstörungen.Zeitschrift für Neuropsychologie. 11: 154–167.

Gazzaniga MS, Sperry RW. (1966) Simultaneous doublediscrimination response following brain bisection.Psychonomic Science 4: 261–2.

Gillberg C. (1995) Clinical Child Neuropsychiatry. New York:Cambridge University Press.

— Carlstrom G, Rasmussen P. (1983) Hyperkinetic disorders inseven-year-old children with perceptual, motor and attentionaldeficits. Journal of Child Psychology and Psychiatry24: 233–46.

Gillberg IC, Gillberg C. (1989) Children with preschool minorneurodevelopmental disorders. IV: Behaviour and schoolachievement at age 13. Developmental Medicine & ChildNeurology 31: 3–13.

Jeeves MA, Dixon NF. (1970) Hemisphere differences in responserates to visual stimuli. Psychonomic Science 20: 249–51.

Kadesjo B, Gillberg C. (1998) Attention deficits and clumsiness inSwedish 7-year-old children. Developmental Medicine & ChildNeurology 40: 796–804.

Kawabe S. (1989) Effects of force output and preparatory set onpremotor time of simultaneous bilateral response. Perceptualand Motor Skills 68: 619–25.

Largo RM, Caflisch JA. (1999) Neuromotor development fromkindergarten age to adolescence. In: AF Kalverboer and AGramsbergen, editors. Handbook on Brain and Behavior inHuman Development. (Forthcoming).

Lazarus JC, Todor JI. (1991) The role of attention in the regulationof associated movement in children. Developmental Medicine& Child Neurology 33: 32–9.

Leung PW, Connolly KJ. (1998) Do hyperactive children have motororganization and/or execution deficits? Developmental Medicine& Child Neurology 40: 600–7.

McMahon SA, Greenberg LM. (1977) Serial neurologic examinationof hyperactive children. Pediatrics 59: 584–7.

Michael RL, Klorman R, Salzman LF, Borgstedt AD, Dainer KB.(1981) Normalizing effects of methylphenidate on hyperactivechildren’s vigilance performance and evoked potentials.Psychophysiology 18: 665–77.

Mitchell S, Bradley VA, Welch JL, Britton PG. (1990) Coma arousalprocedure: a therapeutic intervention in the treatment of headinjury. Brain Injury 4: 273–9.

Ohtsuki T. (1981) Increase in simple reaction time of kneeextension induced by simultaneous bilateral performance.Perceptual and Motor Skills 53: 27–30.

Oosterlaan J, Sergeant JA. (1998) Response inhibition and responsere-engagement in attention-deficit–hyperactivity disorder,disruptive, anxious and normal children. Behavioral BrainResearch 94: 33–43.

Overmeyer S, Simmons A, Santosh J, Andrew C, Williams SC, TaylorA, Chen W, Taylor E. (2000) Corpus callosum may be similar inchildren with ADHD and siblings of children with ADHD.Developmental Medicine & Child Neurology 42: 8–13.

Piek JP, Pitcher TM, Hay DA. (1999) Motor coordination andkinaesthesis in boys with attention-deficit–hyperactivity disorder.Developmental Medicine & Child Neurology 41: 159–65.

Rubia K, Oosterlaan J, Sergeant JA, Brandeis D, v. Leeuwen T. (1998)Inhibitory dysfunction in hyperactive boys. Behavioral BrainResearch 94: 25–32.

Schachar R. (1991) Childhood hyperactivity. Journal of ChildPsychology and Psychiatry 32: 155–91.

Schaffer DM, Fisher PS-S, Cohen P, Piacentini J, Davies M, ConnersCK, Regier D. (1993) The diagnostic interview schedule forchildren, revised version (DISC-R) I: Preparation, field testing,interrater reliability, and acceptability. Journal of the AmericanAcademy of Child and Adolescent Psychiatry 32: 643–50.

Schuerholz LJ, Cutting L, Mazzocco MM, Singer HS, Denckla MB.(1997) Neuromotor functioning in children with Tourettesyndrome with and without attention-deficit–hyperactivitydisorder. Journal of Child Neurology 12: 438–42.

Semrud-Clikeman M, Filipek PA, Biederman J, Steingard R, KennedyD, Renshaw P, Bekken K. (1994) Attention-deficit–hyperactivitydisorder: magnetic resonance imaging morphometric analysis ofthe corpus callosum. Journal of the American Academy of Child& Adolescent Psychiatry 33: 875–81.

Steger J, Imhof K, Steinhausen H-C, Brandeis D. (2000) Brainmapping of bilateral interactions in attention-deficit–hyperactivity disorder and control boys. ClinicalNeurophysiology 111: 1141–56

Steinhausen HC, Metzke CW, Meier M, Kannenberg R. (1998)Prevalence of child and adolescent psychiatric disorders: the Zurichepidemiological study. Acta Psychiatrica Scandinavica 98: 262–71.

van der Meere J, van Baal M, Sergeant J. (1989) The additive factormethod: a differential diagnostic tool in hyperactivity and learningdisability. Journal of Abnormal Child Psychology 17: 409–22.

— Vreeling HJ, Sergeant J. (1992) A motor presetting study inhyperactive, learning disabled and control children. Journal ofChild Psychology and Psychiatry 33: 1347–54.

van Leeuwen TH, Steinhausen HC, Overtoom CCE, Pascual-MarquiRD, van’t Klooster B, Rothenberger A, Sergeant JA, Brandeis D.(1998) The continuous performance test revisited withneuroelectric mapping: impaired orienting in children withattention deficits. Behavioral Brain Research 94: 97–110.

Waber DP, Mann MB, Merola J, Bauermeister M, Cohen C, Ferber R,Wolff PH. (1985) Motor overflow and attentional processes innormal school-age children. Developmental Medicine & ChildNeurology 27: 491–7.

Whitmont S, Clark C. (1996) Kinaesthetic acuity and fine motorskills in children with attention-deficit–hyperactivity disorder: apreliminary report. Developmental Medicine & Child Neurology38: 1091–8.

Wyke M. (1971) The effects of brain lesions on the performance ofbilateral arm movements. Neuropsychologia 9: 33–44.


Documents

Attentional and neuromotor deficits in ADHD