BRAIN AND LANGUAGE 9, 315-323 (1980)

Limb Apraxia and Motor Performance

K. YORK HAALAND, BRUCE E. PORCH, AND

HAROLD D. DELANEY
Alhrrquerqtte Vefercrm Administrrrtion Mediccr/ Center rend
Unit,crsity ofNrn> Mexico

This study was designed to determine if motor deficits in limb apraxia are task
specific. Non-brain-damaged patients and apraxic and nonapraxic patients with
left hemisphere damage performed language and limb apraxia tests and six motor
tasks with the left hand. Contrary to previous data, no significant group differ-
ences occurred on a finger tapping task. Although task complexity or sequencing
requirements affected group differences, greatest apraxic impairment was noted
on a task of precise steadiness, but only when its response inhibition requirements
were increased. This pattern of deficits appears to be associated with greater
involvement of the premotor area in the apraxic group.

INTRODUCTION
Liepmann (1913) initially suggested patients with ideomotor limb
apraxia demonstrated a general impairment of skilled motor movement.
Heilman (1975) supported this contention by demonstrating finger tapping
deficits in patients with, but not without, ideomotor limb apraxia. How-
ever, Pieczuro and Vignolo (1967) found no significant differences be-
tween apraxic and nonapraxic groups on a manual dexterity task. Heil-
man suggested this discrepancy was associated with the differences
between the tasks used in the two studies, but other data (Haaland, Clee-
land, & Carr, 1977) in left and right hemisphere-damaged patients would
indicate that the manual dexterity task should be the more sensitive of
the two tasks. Heilman also did not consider the possibility that the two
studies differed in their criteria of limb apraxia and in the composition
of their apraxic and nonapraxic groups.
The present investigation was designed to determine in a single study
whether motor deficits in apraxic and nonapraxic patients are influenced
This research was supported by the Veterans Administration. David Flaherty provided
technical support and James Allender provided several suggestions for analysis. Address
reprint requests to K. York Haaland, Psychology Service, VA Medical Center, 2100 Ridge-
crest Drive S.E., Albuquerque, NM 87108.

0093-934X/80/0203 15-09$02.00/O
CopyrIght % 1980 by Academic Pre\\. Inc.
All right\ of reproductwn in any form reerved.
316 HAALAND, PORCH, AND DELANEY

by the motor task administered. Six motor tasks and a limb apraxia
battery were administered to non-brain-damaged controls and to apraxic
and nonapraxic patients with left hemisphere damage. In order to ensure
that the apraxic group’s deficits were specifically gestural rather than
more general, language tests were administered and used as a basis for
matching the apraxic and nonapraxic groups.

SUBJECTS AND METHODS

Thirty-three right-handed patients were given a limb apraxia battery, language test, and six
motor tests after informed consent had been obtained. Twenty-two of these patients were
stroke patients with damage restricted to the left hemisphere according to brain or CT scan,
EEG, or neurologic evaluation while the other 11 were an age-matched group of non-brain-
damaged individuals.
The stroke group was subdivided into 11 patients who demonstrated ideomotor limb
apraxia and 11 patients who did not. In each group 10 strokes were thrombo-embolic in
origin and one was hemorrhagic. The ideomotor limb apraxia battery consisted of 15
demonstrations partially derived from Kaplan’ and Brown (1972) including 5 imitations of
symbolic gestures (for example, salute), 5 imitations of gestures miming object use without
object present (for example, brush teeth), and 5 imitations of meaningless gestures (for
example, point to ear with index finger). All administrations of the limb apraxia battery were
videotaped to allow more careful qualitative scoring; delays were not considered errors, but
distortion of the fine (hand position) or gross part of the movement (arm position or
movement) as well as error in final target or orientation of the arm and hand relative to the
target were considered inaccuracies. Scoring was done independently by two judges and the
correlations between judges’ scores were .78 for the control group, .84 for the apraxic group,
and .99 for the nonapraxic group.
The patients had to make errors on at least 5 out of 15 imitations to be considered apraxic.
This criterion was specified before the study began and was conservative compared to
Heilman’s criterion for limb apraxia (personal communication, 1977). Like Pieczuro and
Vignolo (1967), all of our patients classified as apraxic were worse than any control patient
while none of the nonapraxic patients performed below the worst control patient.
The stroke and non-brain-damaged control groups were matched for age and education
(see Table l), and no significant group differences were seen using two analyses of variance.
All patients demonstrated right hemiparesis, with the criterion for right hemiparesis being
that the left hand had to surpass the right hand on a grip strength measure by at least 1 SD.
Between-group comparisons indicated that the apraxic and nonapraxic groups did not differ
in grip strength for their right hand (r(20) = .1943, p > .05). Mean grip strength (kg) in the
apraxic group was 8 and 31 for the right and left hands, in the nonapraxic group was 10 and
36, and in the control group was 46 and 41.
Scores on subtests I and IV-XII of the Porch Index of Communicative Ability (PICA)
(Porch, 1967), which do not require pantomimic gestures or writing (subtests I, IV-XII),
were used as a measure of language skill for matching the brain-damaged groups. These
subtests were chosen rather than the standard PICA overall score since the overall scores
are affected by gestural skills which the apraxic group lack, and this would have resulted in
poorly equated groups as far as aphasic involvement was concerned. The subtests used
measure auditory comprehension, reading, visual matching, and four verbal skills of naming,
description, sentence completion, and repetition. On the basis of this composite language
score two patients in the apraxic group were discarded so language score could be matched
between the apraxic and nonapraxic groups (t(20) = 1.6559, p > .05). Other descriptive data

’ E. Kaplan, 1977, personal communication.

 LIMB APRAXIA AND MOTOR PERFORMANCE 317

TABLE 1
DESCRIPTIVE DATA

Control group Nonapraxic group Apraxic group

Measure Mean .SD Mean SD Mean SD

Age (years) 60.1 11.0 58.5 9.4 63.4 10.5

Education (years) 13.6 2.7 10.3 3.4 11.5 4.1
Limb apraxia (Z) 89.1 9.1 82.4 6.1 53.0 14.5
PICA (nongestural) (%) 78.5 12.9 65.2 23.5
PICA (verbal) (!Z) 56.5 13.5 40.4 18.5

and the final mean percentiles of the composite language measure based upon comparison
with a large random group of left hemisphere aphasic patients are presented in Table 1.
Auditory comprehension was also group matched using two sections of the PICA, one
requiring pointing to object by function (for example, point to the one used for cleaning
teeth) and another requiring pointing to object by name (for example, point to the tooth-
brush); a t test was computed to assess the group match and indicated the groups were not
significantly different (t(20) = .1864, p > .05). Similar auditory comprehension skills de-
crease the possibility that group performance deficiencies on the apraxia battery and other
motor tasks could be due to decreased comprehension. Even though not used as matching
variables, there were no significant differences between the apraxic and nonapraxic groups
on the two reading tasks (r(20) = I .2805, p > .05) or the two visual matching tasks (t(20) =
.9299, p > .05) of the PICA. However, the apraxic group’s verbal skills were significantly
worse than the nonapraxic group using the four verbal subtests of the PICA which require
verbal naming, description, sentence completion, and repetition (t(20) = 2.3262, p < .05).
Six motor tasks developed as part of the Wisconsin Motor Battery (Matthews & Klove,
1964) were administered. They varied in complexity and precision from grip strength to a
measure of fine finger dexterity (grooved pegboard). All motor tasks except the grip strength
task were administered only to the left arm, the arm ipsilateral to lesion.
(I) Grip strength. The maximum score on a hand dynamometer on one of two trials on
each hand.
(2) Finger tapping speed. Patients were required to tap as rapidly as possible on the key of
a counter. The mean of five IO-set trials was recorded for the left hand: this task was also
used by Heilman (1975).
(3) Static steadiness. The patient placed an electric stylus for 15 set in each of nine holes
ranging in diameter from 12 to 3 mm. The total number of side contacts and duration of
contact per error were recorded for the left hand.
(4) Vertical groove steadiness. Using the same response measures (side contacts and
duration of contact per error), this task required the patient to run the stylus up and down a
vertical groove (5 mm wide and 25 cm long) for two trials with the left hand. This task adds
movement in one dimension to a steadiness task.
(5) Maze coordination. This test also used number of side contacts and duration of side
contacts made with the stylus as the response measures, but required the subject to trace
through a maze (8 mm wide) with both horizontal and vertical movement.
(6) Grooved pegboard. This task, which was similar to that used by Pieczuro and Vignolo
(1967). sampled manipulative dexterity using a pegboard containing 25 holes in a 5 x 5 array
with randomly positioned slots. The pegs have an edge along one side that must be rotated to
match the slot before they can be inserted. The patients performed the task in a specified
right-to-left order, and time to complete the task was recorded.
To allow for direct comparison among all six tasks, data from all groups were converted to
standardized 7 scores with a mean of 50 and .SD of 10. These conversions were based on a
normative population of 50 subjects who were over 35 years of age.
318 HAALAND, PORCH, AND DELANEY

RESULTS
Results on the motor tasks administered are graphed in Fig. 1. Since
standard T scores were not available for the duration of contact per error
measure only number of side contacts was graphed for the static steadi-
ness, vertical groove, and maze coordination tasks.
Multiple comparisons with the control group using the Dunnett proce-
dure (Dunnett, 1955) for each task indicated there were significant differ-
ences between the control and apraxic groups on the static steadiness,
maze coordination, and grooved pegboard tasks (p < .05). There were,
however, no significant differences between these two groups on mea-
sures of finger tapping or vertical groove steadiness. These same compari-
sons between the control and nonapraxic group were not significant (p >
.OS>for the finger tapping task even when, like Heilman, the mean of three
tapping trials was used. Significant differences were found between the
apraxic and nonapraxic groups only on the static steadiness task while the
apraxic patients performed significantly worse than control patients on
the static steadiness, maze coordination, and grooved pegboard tasks.
Since the static steadiness task was most sensitive to the motor differ-
ences between the apraxic and nonapraxic group, this task, as well as the
other steadiness tasks of vertical groove and maze coordinations, were
further examined by analyzing group differences in the duration of contact
per error measure. This measure should theoretically measure sensitivity

q CONTROL
q NON-APRAXICS
q APRAXICS

TAP
L
STATIC YERTlCAL MAZE PEGS

FIG. 1. Mean left-hand performance across all tasks for control group and groups with
and without ideomotor limb apraxia.
 LIMB APRAXIA AND MOTOR PERFORMANCE 319

to feedback since every time a patient touched the side, tactual input
changed and a click was heard. Thus, duration of contact per error is a
measure of the system’s efficiency of response to that feedback. These
data are summarized in Table 2. Analyses of variance were computed for
each task separately with group effects reaching significance for static
steadiness (p < .005) and maze coordination (p < .025). The pattern of
results using this measure of the steadiness tests was exactly the same as
that reported for number of errors. A direct comparison between apraxic
and nonapraxic patients was significant (p < .05) only for the static
steadiness task while comparisons between the control and apraxic pa-
tients were significant for the static steadiness and maze coordination
tasks (p < .025), again using the Dunnett procedure. This suggests that
not only did the apraxic group have more difficulty maintaining steadiness
on these tasks but also once the error was made, it took them longer to
correct their errors.
Since the static steadiness task was most sensitive to motor differences
between the apraxic and nonapraxic groups, the impact of precision
requirements on this task was directly examined. The static steadiness
task requires the patient to hold the stylus in each of nine holes which
progressively decrease in diameter, thereby increasing the precision re-
quirements of the task. The specific task requirements do not change
qualitatively, but as the holes decrease in diameter the patient is required
to even further inhibit hand and arm movement to minimize error.
Changes in performance across these holes would indicate the effect of
increasing precision requirements. Thus, two separate one-way analyses
of variance were computed for the d(ffewncrs in errors and the diffeferemcz
in mean duration of contact per error between the first four holes (diame-
ter range = 12-5 mm) and the last four holes (diameter range = 4-3 mm)
of the steadiness test. Significant (p < .Ol) group differences were found
only for the difference in errors so separate t tests were computed to
examine these group effects on the first four and the last four holes. No
significant group effects were demonstrated on the first four holes, but
significant apraxic versus nonapraxic (p < .Ol) and apraxic versus control
(p < .Ol) differences were noted on the last four holes. These data suggest

TABLE 2
DURATION OF CONTACT PER ERROR

Control Nonapraxic Apraxic

group group group
Test Mean SD Mean SD Mean SD

Static steadiness (set) 0.08 0.04 0.09 0.04 0.13 0.04

Vertical groove steadiness (set) 0.09 0.04 0.12 0.07 0.17 0.14
Maze coordination (set) 0.08 0.05 0.10 0.08 0.17 0.09
320 HAALAND, PORCH, AND DELANEY

task precision requirements which increase response inhibition require-

ments are particularly important variables differentially affecting motor
performance in these apraxic and nonapraxic patients.

DISCUSSION
Relationship to Other Studies of Limb Apraxia
These data support Liepmann’s (1913) and Heilman’s (1975) contention
that patients with ideomotor limb apraxia also demonstrate other motor
deficits. It also demonstrates that when overall language skills are the
same between two groups, limb apraxia and other motor deficits can still
be differentially demonstrated. However, contrary to Heilman’s report,
we found no differences in finger tapping between apraxic and nonapraxic
patients. While the nonapraxic group did not differ from the control
group, the apraxic group’s performance was significantly impaired when
compared to the control group on the static steadiness, maze coordina-
tion, and grooved pegboard tasks. This pattern of results was unchanged
even if the two patients initially discarded to achieve the language match
were included, so the difference between the two studies is more likely
associated with differences in limb apraxia criteria rather than Heilman’s
failure to match on language impairment. Heilman may have used stiffer
criteria of limb apraxia so his apraxic group would have been so severely
apraxic they would have been impaired even on the very simple finger
tapping task. He did not provide other behavioral measures or specify
limb apraxia criteria, and with only 11 apraxic patients in the present
study it was not possible to distinguish levels of apraxia to directly
examine this hypothesis.
Like Pieczuro and Vignolo (1967) our apraxic, but not our nonapraxic,
patients performed worse than any control patient on the limb apraxia
battery. This suggests our patient population should be similar to theirs,
but, in contrast to our results, they found ideomotor apraxia was indepen-
dent of fine finger dexterity, even though the dexterity measures used in
the two studies were similar. While the apraxic patients in the Pieczuro
and Vignolo study had lesions primarily of the left posterior retrorolandic
areas (inferred from language, visual field, and motor data), both our
apraxic and nonapraxic patients had greater frontal lobe involvement.
This inference is based upon the fact that both groups were right
hemiparetic. More importantly, it is likely our apraxic group’s frontal
lesions were different from those of our nonapraxic group since their
pattern of language impairment was different. These differences in pattern
were characterized by similar auditory comprehension and reading skills
but impaired verbal skills in the group with limb apraxia. Speech was
nonfluent in many cases similar to what has been described in patients
with Broca’s aphasia and damage to Broca’s area (Geschwind, 1965)
which includes the left premotor area (Luria, 1966). According to Luria
 LIMB APRAXIA AND MOTOR PERFORMANCE 321

(1966) damage to the premotor area leads to disruption of movements

requiring sequencing of unique responses with preservation of move-
ments requiring sequencing of the same response. On this premise, finger
tapping which requires sequencing of the same response should not be
disrupted by premotor involvement, but more complex tasks (maze coor-
dination and grooved pegboard) which require sequencing of unique re-
sponses would be affected.
Sensory versus Motor Contribution
Although sensory and motor contributions to response cannot be
methodically separated with these results, the duration of contact per
error measure of the steadiness tasks has a substantially greater sensory
(auditory, proprioceptive) component than the error measure and can best
be considered a measure of response to direct feedback. The pattern of
results for these two measures was the same on all the steadiness tasks
suggesting sensory and motor factors were of similar influence. However,
when the effect of task precision was directly examined on the static
steadiness task, group differences occurred on the error measure but not
on the measure of response to direct feedback. This result could be
associated with the apraxic group’s greater impairment of response inhibi-
tion. This finding is in accord with our emphasis upon frontal and possible
premotor, rather than postcentral involvement. Wyke (1966) takes a simi-
lar stand in her study of postural arm drift. She infers deficits after parietal
damage are associated with impairment of sensory-motor interaction
while the deficits of the frontal patients are more specifically motor and
associated with impairment in the balanced control of flexors and exten-
sors. The emphasis upon motor rather than sensory factors in our apraxic
patients is in contrast to Trevarthen’s discussion of hemispheric asym-
metry of motor control (1978) and Flower’s discussion of hand preference
(1975) which focus upon sensory factors. However, proponents of open
loop control of motor movement deemphasize the import of sensory
factors (Keele, 1968) and suggest the importance of this variable may be
heavily influenced by task requirements (Glencross, 1977).

Theoretical Considerations
Task complexity which may be related to difficulty changing limb
position (Kimura, 1977) differentiated apraxic versus control but not
nonapraxic versus control patients. This was evidenced by the apraxic
group’s poorer performance on the more complex maze coordination and
grooved pegboard tasks. As the number of serial sensory and motor require-
ments increases, the possibility of task disruption increases since the
response is dependent upon a larger number of sensory-motor interac-
tions. Using different tasks Luria (1966) has emphasized that disintegra-
tion of complex serial movements is the most permanent sign of premotor
322 HAALAND, PORCH, AND DELANEY

lesions while simple isolated movements are preserved. The effect of task
complexity has also been emphasized by Haaland et al. (1977).
However, despite the apparent importance of task complexity, these
data emphasize the even greater importance of precision requirements in
directly differentiating patients with and without limb apraxia. This
finding relates to the clinically observed clumsiness of apraxics described
by Geschwind (1963, 1965, 1975) and associated with lesion or disconnec-
tion of areas in the left hemisphere which are crucial to motor program-
ming. The area most frequently associated with this disruption has been
the premotor area which is more likely to be involved in our apraxic group
since verbal deficits were more common in that group than in the nona-
praxic group.
Accuracy on the static steadiness task is dependent upon inhibition of
hand movement, and the degree of inhibition required increases as the
hole diameter decreases (and precision requirements increase). It is only
when inhibition requirements are high (with the four holes of smallest
diameter) that the apraxic patients have greater difficulty inhibiting re-
sponses than the nonapraxic patients. The relationship of this deficit to
response inhibition is emphasized by the fact that these same patients
have no increased difficulty responding to feedback as hole diameter
decreases (i.e., duration of side contact does not significantly change as
hole diameter decreases). The apraxic group’s increasing deficits on the
smaller diameter holes of the static steadiness task appear to be as-
sociated with decreased ability to inhibit responses. This characteristic
has also been described in patients with premotor damage (Luria, 1966).
These patients have difficulty aborting a movement after it is begun and
frequently will follow a command but not discontinue the response after it
is complete (e.g., maintaining a raised hand). This same mechanism may
be operating at a different level in the static steadiness task.

Conclusion
These results provide quantitative experimental support of task-specific
motor deficits in patients with limb apraxia. More specifically, they em-
phasize the relatively greater importance of task precision and response
inhibition requirements over task complexity in differentiating the motor
deficits of patients with and without ideomotor or limb-kinetic apraxia.

REFERENCES
Brown, J. 1972. Aphasia, apruxiu, and agnosia. Springfield, IL: Charles C Thomas.
DeAjuriagnerra, J., & Tissot, R. 1969. The apraxias. In P. .I. Vincken & G. W. Bruyn (Eds.),
Handbook of clinicul neurology. Amsterdam: American Elsevier. Vol. 4.
DeRenzi, E., Pieczuro, A., & Vignolo, L. 1966. Oral apraxia and aphasia. Cortex, 2,50-73.
Dunnett, C. W. 1955. A multiple comparison procedure for comparing several treatments
with a control. Journal of the Americrrn Statistical Association, 50, 1096-1121.
 LIMB APRAXIA AND MOTOR PERFORMANCE 323

Flowers, K. 1975. Handedness and controlled movement. British Journul ofPsychology, 66,
39-52.
Geschwind, N. 1963. Sympathetic dyspraxia. Trrrn.yactions of the Americcrn Neurologicrrl
Associnrion, 88, 219-220.
Geschwind, N. 1965. Disconnexion syndromes in animals and man. Bruin, 88, 585-644.
Geschwind, N. 1975. The apraxias: Neural mechanisms of disorders of learned movement.
Americtrn Scientkf. 63, 188-195.
Glencross, D. J. 1977. Control of skilled movements. PsychoLogicul Bullefin. 84, 14-29.
Goodglass, H., & Kaplan, E. 1%3. Disturbance of gesture and pantomine in aphasia. Bruin,
86, 703-720.
Haaland, K. Y., Cleeland, C. S., & Carr, D. 1977. Motor performance after unilateral
hemisphere damage in patients with tumor. Archives of Neurology. 34, 556-559.
Htcaen. H. 1969. Aphasic, apraxic, and agnosic syndromes in right and left hemisphere
lesions. In P. J. Vincken & G. W. Bruyn (Eds.), Handbook o.f clinicrrl neurology.
Amsterdam: American Elsevier. Vol. 4.
Heilman, K. 1975. A tapping test in apraxia. Cortr.r, 11, 259-263.
Keele, S. W. 1968. Movement control in skilled motor performances. P.svcho/ogiccr/ &l/r-
fin, 70, 387-403.
Kimura, D. 1977. Acquisition of a motor skill after left-hemisphere damage. Brain. 100,
527-542.
Kimura, D., & Archibald, Y. 1974. Motor functions of the left hemisphere. Bruin, 97,
337-350.
Liepmann, H. 1913. Motor aphasia, anarthria, and apraxia. Trmsacrions qf rhr 17th Interm-
tionrrl Congwss of Medicine, Sec. XI. Pt. II, 97-106.
Luria, A. 1966. Higher corticul funcriom. New York: Basic Books.
Matthews. C. G., & Klove, H. 1964. Instruction munurrljtir the adult neuropsychology tat
huftery. Madison: Univ. of Wisconsin Medical School.
Pieczuro, A., & Vignolo, L. 1967. Studio sperimentale sull’ aprassis ideomotoria. Sistemcr
NerIwu. 19, 131-143.
Porch, B. 1967. Porch Index qf Communicutit,e Ability: Administration rrnd interpretorion.
Palo Alto, CA: Consulting Psychologist Press. Vol. 2.
Semmes, J. 1968. Hemispheric specialization: A possible clue to mechanism. Nertro-
psychologiu, 6, 1 l-26.
Tower, S. S. 1940. Pyramidal lesions in the monkey. Brain, 63, 36-90.
Trevarthen. C. 1978. Manipulative strategies of baboons and origins of cerebral asymmetry.
In M. Kinsbourne (Ed.), Asvmnzetri~rrl~irn~tion of the hrrrin. Cambridge: Cambridge
Univ. Press.
Vaughan, H. G., & Costa, L. 1962. Performance of patients with lateralized cerebral lesions.
II. Sensory and motor tests. Journul of Nervous und Mental Disetrse, 34, 237-243.
Wyke, M. 1966. Postural arm drift associated with brain lesions in man. Archives of
Neurology, 15, 329-334.
Wyke, M. 1967. Effects of brain lesions on rapidity of arm movement. Neurology. 17,
1113-l 120.
Wyke, M. 1968. The effects of brain lesions in the performance of an arm-hand precision
task. Neurop~24.(.hOIOKi(l, 6, 125-134.
Wyke, M. 1971. The effects of brain lesions on the learning performance of a bimanual
coordination task. Cortex. 7, 59-72.