Accelerat ing t he world's research.

Influence of laptop computer design

and working position on physical
exposure variables
Mats Hagberg, Hélène Moffet, Lena Karlqvist

Clinical Biomechanics

Cite this paper Downloaded from Academia.edu 

Get the citation in MLA, APA, or Chicago styles

Related papers Download a PDF Pack of t he best relat ed papers 

T he impact of comput er display height and desk design on muscle act ivit y during informat io…
Robin Burgess-Limerick, Clare Pollock

Charact erizat ion of post ure and comfort in lapt op users in non-desk set t ings
Judit h Gold, Jeffrey B Driban

Kinemat ic and elect romyographic differences bet ween mouse and t ouchpad use on lapt op comput ers
Carmela Cont e
 Clinical Biomechanics 17 (2002) 368–375
www.elsevier.com/locate/clinbiomech

Influence of laptop computer design and working position

on physical exposure variables
a,*
H. Moffet , M. Hagberg b, E. Hansson-Risberg c, L. Karlqvist c

a
Department of Rehabilitation, Faculty of Medicine, Laval University and Center for Interdisciplinary Research in Rehabilitation
and Social Integration, 525, Blvd Hamel, B-77, Que., Canada G1M 2S8
b
Department of Occupational and Environmental Medicine, Sahlgrenska Academy and Hospital, Goteborg University, St Sigfridsgatan 85, 412 66
Goteborg, Sweden
c
National Institute for Working Life, 112 79 Stockholm, Sweden
Received 7 April 1998; accepted 2 April 2002

Abstract
Objective. To evaluate the impact of two laptop designs (with or without palm rest) and two work situations (on desk or lap) on
neck and upper limb posture, muscle activity and productivity.
Design and methods. Eight healthy subjects performed a standardized typing task of 15 min duration. During the last 5 min of
each test, the neck, upper arm and trunk postures were captured by a three-dimensional video system, wrist motion was measured by
a biaxial electrogoniometer and muscle activity of four neck and upper limb muscles was recorded.
Results. Only minor differences in postures, wrist positions and productivity were observed when comparing the two laptop
designs in the same situation. Larger differences were found when comparing the two situations (desk or lap). In the desk situation,
the subjects bent their heads forward less, had less backward trunk inclination and wrist extension, but more elevation of the upper
arm. Higher electromyographic (EMG) levels in the trapezius and deltoid muscles and lower EMG levels in the wrist extensors were
also found in the desk situation.
Conclusions. Our findings do not favor one particular laptop design because only small differences in physical exposure were
found. However, the workstation set up influenced the physical exposure variables, and was pinpointed as the main determinant to
be considered when doing laptop work even-though no ideal situation was found. Greater physical (muscular and articular) con-
straints seem to be imposed to the shoulder region in the desk situation whereas the head-neck and wrist segments appear to be more
stressed in the lap situation.

Relevance
Laptop computers are often used although the physical exposure in laptop work and the impact of different laptop designs have
not been systematically assessed. A better understanding of these factors may help formulate some recommendations for laptop
users. Ó 2002 Elsevier Science Ltd. All rights reserved.

Keywords: Computer; Neck and upper limb posture; Motion analysis; Muscle activity

1. Introduction physical and technical characteristics of laptops, much

Portable computers (laptops) are widely used by
professionals who need to travel and work in different
places with a computer system. This is not surprising
considering that they offer high technology performance
in a compact, light, self-sufficient (battery provided) and
portable device. In parallel to the improvement in the
*
Corresponding author. Department of Rehabilitation, Faculty of
Medicine, Laval University, Quebec City, Quebec, Canada G1K 7P4.
E-mail address: helene.moffet@rea.ulaval.ca (H. Moffet).
0268-0033/02/$ - see front matter Ó 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 2 1 - 9 2 9 0 ( 0 2 ) 0 0 0 6 2 - 3
effort has been invested to achieve ergonomically satis-
factory laptop designs. No study has, however, seriously
tested laptops under practical conditions.
Laptop work conditions differ considerably from that
of traditional visual display terminal (VDT). In general,
laptops are not used in an office environment with ad-
justable furniture. In addition, the adjustability of the
different laptop components is limited. For instance, the
screen and keyboard are not detachable and therefore
cannot be adjusted independently as recommended for
traditional VDT work [1–3] except for screen inclination.
 H. Moffet et al. / Clinical Biomechanics 17 (2002) 368–375
The inherent adjustability restrictions of laptops and
design characteristics, in addition with the variable yet
fixed of typical laptop workstation are critical factors
which may negatively affect task performance, comfort
and working postures. Considering that, in traditional
VDT operators, several significant relationships were
discovered between the designs of workstations, pos-
tures or muscle loads on one hand and the incidence of
complaints or medical findings on the other [2,4–7], it is
important to determine the physical exposure in laptop
work and to determine the impacts of different laptop
designs.
The purpose of this study was to compare neck and
upper limb postures, muscle activity and productivity in
two typical work situations: on desk and on lap with two
different laptop designs (with and without palm rest) in a
group of inexperienced users.
2. Subjects
Eight healthy subjects, four men and four women,
aged between 21 and 31 years (median: 26.5 years) with
a height ranging from 164 to 182 cm (median, women:
169 cm; men ¼ 179 cm) and habitual users (median ¼
69%; range ¼13–100% of their working time) of a tra-
ditional computer system (past experience: median ¼ 4
years, range ¼1.5–5 years) participated voluntarily in
one laboratory session. Half of the subjects (two men
and two women) were IBM compatible users while the
other four subjects were Macintosh users. None of them
were familiar with laptops (portable computers) nor
were they professional secretaries. The sample of subjects
was restricted to non-experienced laptop users to ensure
the same baseline experience with both laptop designs.
This choice was considered the best alternative in the
context of the present study even though it has some
implications for the generalization of the results in other
populations. This study was approved by the Ethical
Commitee of the Karolinska Institute.
3. Methods
3.1. Work tasks
The subjects performed a standardized typewriting
test with two different laptops (laptop A: the Apple
Powerbook 140 with a built-in palm rest and a keyboard
positioned close to the screen; laptop B: the Notebook-
Altima SLC 33, an IBM compatible portable without
palm rest and a keyboard positioned close to the front
part of the laptop base) in two situations (D: on desk;
L: on lap). The sequence of the four typewriting tests
(2 laptops  2 situations) was systematically alternated
among subjects. Before each test, the subjects were
369
asked to adjust at their own convenience some compo-
nents of the workstation (see below). No familiarization
period with laptops was provided prior to the mea-
surements considering that stable posture, electromyo-
graphic (EMG) and productivity measures can be
obtained after the first 10 min of recording in non-
experienced laptop users [8]. Each test lasted 15 min [8]
and consisted of typing a new written text with a com-
parable degree of difficulty at a free work pace without
correcting any keying mistakes. A 1 min pause was in-
troduced after the first 5 min of typing to allow the
subjects to readjust, when needed, the workstation and
then the subjects were asked to type continuously for
the last 10 min without modifying the workstation set-
ting.
3.2. Workstation dimensions
Two typical work situations with laptops were sim-
ulated: the desk and lap situations. Standardized
workstations with unadjustable desk and chair were
used considering that laptops are commonly used in
places where adjustable furniture is not available. The
desk and seat heights were determined according to the
European and Swedish Standards for fixed office tables
and chairs [2,9]. The chair had no wheels, a seat 46 cm
high and a backrest slightly tilted backwards (about
100°). The desk (desk situation only) was 73 cm high
and measured 120 cm per 70 cm. Only three workstation
components were adjusted by the subjects before and
after the first 5 min of each test: (1) position of the
laptop on desk or on lap, (2) inclination of the screen
and (3) position of the chair on the floor. In both the
desk and lap situations, a document holder was used
and it was placed just beside the screen of the laptops.
The two laptops had similar characteristics when con-
sidering the screen (20 cm  14 cm), keyboard (27 cm 
9 cm) and key (1.2 cm  1.4 cm) sizes, key depression
force (about 0.30 N) and the screen readability. Main
differences in laptop design were the keyboard position
and the presence or absence of a palm support. Laptop
A offered a buit-in palm rest at the front of the laptop
base and consequently the keyboard was positioned
close to the screen whereas laptop B had no palm rest
and a keyboard positioned close to the front of the
laptop base. An overview of the workstations and lap-
tops is shown in Fig. 1.
3.3. Experimental procedure
A short health interview and a physical examination
of the neck and upper limbs including specific tests to
detect carpal tunnel syndrome (Tinel and Phanel’s tests)
were carried out by a physiotherapist (HM). All subjects
were free of symptoms before tests.
370 H. Moffet et al. / Clinical Biomechanics 17 (2002) 368–375
Fig. 1. Illustration of the standardized workstation used during tests in
the desk (upper part of the figure) and lap (lower part of the figure)
situations. Postures were measured with the MacReflex video system
using five reflectives markers placed on the subject’s right side (see
methods for details) and three reference markers placed on a board.
The defined angles were calculated as following: (A) head bending:
angle between the lines joining the reference markers 2 and 3 (fixed
vertical line) and the markers 4 and 5 (head vertical axis), (B) upper
arm elevation: 90° minus the angle between the lines joining the ref-
erence markers 1 and 2 (fixed horizontal line) and the markers 6 and 7
(upper arm longitudinal axis), (C) trunk inclination: 90° minus the
angle between the lines joining the reference markers 1 and 2 (fixed
horizontal line) and the markers 8 and 6 (trunk longitudinal axis), (D)
trunk–upper arm angle: angle between the lines joining the markers 6
and 7 (upper arm longitudinal axis) and the markers 6 and 8 (trunk
longitudinal axis).
During tests, muscle activity (EMG) from four mus-
cles of the subject’s right side was picked up by surface
electrodes (bipolar silver–silver chloride, Medicotest A/
S, Olstykke, Denmark; contact diameter: 6 mm). The
muscles tested and the electrode locations were the fol-
lowing: (1) cervical portion of the descending part of the
trapezius and cervical erector spinae muscle (one elec-
trode at the level of spinous process of C3 and the other
30 mm lower around C6 about 1 cm lateral to the
spinous processes), (2) lateral portion of the descending
part of the trapezius muscle (the two electrodes were
placed along the line joining the spinous process of C7
and the acromion, lateral to the mid-portion of the
distance between these two anatomical landmarks), (3)
anterior portion of the deltoid (parallel to the muscle
fibers and in the mid-portion of the belly) and, (4) wrist
extensors (distal electrode at 33% of the distance be-
tween the lateral epicondyle and the ulnar styloid and,
over the bellies of the extensor digitorum communis,
extensor carpi radialis brevis and longus). The inter-
electrode distance was 20 mm except for the cervical
extensors (30 mm). The EMG signals were preampli-
fied about 200 times by a preamplifier placed close to
the electrodes and thereafter sent to the data acquisi-
tion unit (DAU-200) of the Physiometer (PHY-400; 10
channels A/D converter; P R E M E D , Oslo, Norway) which
amplified (the gain settings are automatically selected
and vary dynamically according to the magnitude of the
input signal) and sampled the EMG inputs at a rate of
1600 Hz. Then, the root mean square (RMS) values
were calculated (bandwidth 15–800 Hz) and sent 10
times per second to the serial interface of the IBM-PC.
The quality of the EMG signals was checked continu-
ously on two oscilloscopes during the experiment.
The right wrist motions (flexion-extension and radial-
ulnar deviations) were measured by a biaxial flexible
electrogoniometer (Penny & Giles M110, Blackwood,
UK). Before each test, the zero degree position in both
planes of motion were set by placing the wrist in neutral
position (forearm and hand in line) against an hori-
zontal surface. During tests, the angle signals were
sampled at a frequency of 10 Hz by the DAU-200 of the
Physiometer, simultaneously to the EMG signals, and
recorded by an IBM-PC.
The head, trunk and upper arm positions were re-
corded by a three-dimensional (3D) video system, the
MacReflex measurement system (Qualisys Inc., Glas-
tonbury, Connecticut, USA) during the last 5 min of
each test. Two cameras coupled with video processors
detected the eight reflective markers placed on the sub-
ject’s right side (five markers) and on a reference board
(three markers), and calculated the position of each
marker (3D coordinates) at the pre-determined sampling
frequency of 5 Hz. Body landmarks for marker place-
ment on the subjects were: (1) when the subject is in erect
position, over the ear along the vertical line crossing the
masto€ıd process (2) masto€ıd process, (3) lateral part of
the acromion, (4) lateral epicondyle of the elbow, (5)
greater trochanter. From these coordinate values and by
means of a specialized software (Wingz for MacReflex),
the following angles between body segments were cal-
 H. Moffet et al. / Clinical Biomechanics 17 (2002) 368–375
culated: (1) head bending, (2) upper arm elevation, (3)
trunk inclination and (4) trunk–upper arm angle (angle
definitions in Fig. 1). In addition, a video-tape of the
subjects was taken during tests to get an overview of the
posture and motions of the subjects and to determine
whether the palm rest or desk was used to support the
upper limbs.
Immediately after each test, measurements of the
workstation setting were made and the subjects scored
the intensity of perceived discomfort from different body
regions using a category rating scale with ratio proper-
ties (CR10 scale) described by Borg [10].
Productivity was determined from the typed texts by
calculating the number of characters written (quantity
of work) and the absolute and relative number of keying
mistakes (quality of work) done by the subjects.
4. Data analysis and statistics
Only the data acquired during the last 5 min of each
test was used for the analyses. This decision was made in
light of a previous study showing that stable posture,
muscle load and productivity measurements are ob-
tained after the first 10 min of work [8]. The amplitude
probability distribution function (APDF) was computed
for the EMG and angle (wrist motions) recordings and
the low (P ¼ 0:1), median (P ¼ 0:5) and peak (P ¼ 0:9)
values were determined [11–13]. Only the results from
the median muscle loads (in lV) are reported in this
study because similar results were found when low,
median and peak muscle loads were compared between
tests. For wrist motions, median (P ¼ 0:5) angles and
range of motion (RoM) in both planes are reported.
Range of motion was calculated by substracting low
angle from peak angle values. Thereby, in the remaining
text, the variable ‘‘range of motion or RoM’’ refers to
the range of motion that was used most (80%) of the
recording time. Mean head, upper arm and trunk angles
(n ¼ 1500 frames; 5 Hz during 5 min) were calculated
and used as the representative variables for posture
analyses.
For each test, median values and range (n ¼ 8 sub-
jects) were calculated for the different variables. Com-
parisons of the performances between tests were made
by establishing, (1) the absolute difference between sit-
uations when working either with laptop A or laptop B,
and (2) the absolute difference between laptops when
working either on desk or on lap. Thereafter, the mean
difference (n ¼ 8 subjects) and its 95% confidence in-
terval were computed. The muscle load levels were not
compared between the two laptops because small dif-
ferences in baseline EMG levels between laptops were
observed when the subjects rested their hands over the
keyboards. This was probably due to a difference in
electromagnetic fields generated by laptops.
371
5. Results
5.1. Workstation settings
5.1.1. Comparison of the two laptops in each situation
In the desk situation, the distances from the eye to the
screen (or document holder) and from the desk edge to
the laptop base (or screen) were greater when the sub-
jects worked with laptop B than with laptop A (range of
mean differences: 3–9 cm). The distance from the desk
edge to the middle row of the keyboard was, however,
the same with both laptops. In the lap situation, as in the
desk situation, the distances from the eye to the screen
or to the document holder were greater with laptop B
(mean difference: 5 cm). Laptop A was placed on lap
closer to the abdomen (about 2 cm) than laptop B. In
contrast to the desk situation, the middle row of the
keyboard of the two laptops was not placed at the same
distance in the lap situation. The middle row of laptop A
was placed about 6 cm further from the abdomen than
that of the laptop B.
5.1.2. Comparison of the same laptop in both situations
For both laptops, the distances from the eye to the
screen (or document holder), although not significant,
tended to be slightly greater in the lap situation than in
the desk situation (mean differences: 1–6 cm). A greater
inclination of the screen was observed for both laptops
in the lap situation (mean difference for laptop A and B:
10° and 14°).
5.2. Head, upper arm and trunk postures
The characteristic posture (Fig. 1 and Table 1) of the
subjects during tests was the following: the head bent
forward, the trunk slightly inclined backward, the upper
arm always abducted in reference to the trunk line and
elevated in front of the trunk in the desk situation and
slightly elevated behind the trunk in the lap situation.
Only minor differences in posture of the head, upper
arm and trunk were observed when the two laptop
performances were compared in the same situation.
Larger differences were found when comparing the desk
and lap situations. In the desk situation, in comparison
with the lap situation, the subjects had lower amplitudes
of forward head bending and backward trunk inclina-
tion and a higher amplitude of upper arm elevation
(Table 1).
5.3. Wrist motions
The wrist was positioned in extension and ulnar de-
viation during tests, as indicated by the median angle
values in Table 2. For both laptops, the wrist was sig-
nificantly more extended (about 8°) in the lap situation
372 H. Moffet et al. / Clinical Biomechanics 17 (2002) 368–375

Table 1
Head, upper arm and trunk postures
Median value (range) Mean of the differences [95% CI]
A B Desk-Lap A-B

Desk Lap Desk Lap A B Desk Lap

Head 17 (11–29) 23 (21–39) 17 (10–28) 25 (19–40) 9 [13 to 5] 10 [15 to 5] 0 [2 to 1] 1 [3 to 1]
bending ð°Þ
Upper arm 9 (3–22)a 1 (19 to 7)a 5 (1–28)a 2 (12 to 4)a 12 [2–23] 11 [1–22] 3 [2 to 7] 1 [1 to 4]
elevation ð°Þ
Trunk 2 (8 to 7)b 6 (14 to 3)b 3 (10 to 6)b 8 (15 to 2)b 4 [1 to 9] 7 [2–12] 1 [2 to 3] 3 [1 to 7]
inclination
ð°Þ
Trunk–upper 7 (2–19) 6 (1–17) 6 (2–24) 11 (4–24) 1 [5 to 8] 4 [12 to 3] 0 [4 to 4] 5 [7 to 4]
arm angle ð°Þ
Abbreviations: CI: confidence interval; A: laptop A; B: laptop B; desk: typing with laptop on desk; lap: typing with laptop on lap.
a
Positive value means an elevation of the upper arm in front of the trunk (combined motion of flexion and abduction); negative value means an
elevation of the upper arm in posterior of the trunk (combined motion of extension and abduction).
b
Positive value means forward trunk inclination or flexion; negative value means backward trunk inclination or extension.

Table 2
Wrist motions
Median (range) Mean of the differences [95% CI]
A B Desk-Lap A–B

Desk Lap Desk Lap A B Desk Lap

Range of motion in 15 (13–22) 14 (12–21) 17 (12–21) 14 (11–19) 1 [2 to 4] 2 [1 to 4] 0 [2 to 2] 0 [2 to 2]
flexion-extension ð°Þ
Range of motion in 10 (6–15) 10 (8–12) 10 (7–15) 10 (7–12) 0 [1 to 1] 1 [1 to 3] 1 [2 to 0] 0 [2 to 2]
ulnar-radial deviation
ð°Þ
Median angle in 17 (8–30)a 28 (17–33)a 24 (8–36)a 32 (21–40)a 8 [14 to 1] 8 [16 to 0] 4 [12 to 4] 4 [11 to 3]
flexion-extension
plane ð°Þ
Median angle in 7 (14 to 6)b 9 (24 to 6)b 10 (28 to 6)b 5 (19 to 8)b 5 (0–10) 6 [13 to 2] 4 [3 to 10] 7 [11 to 3]
radial-ulnar deviation
plane ð°Þ
Abbreviations: CI: confidence interval; A: laptop A; B: laptop B; desk: typing with laptop on desk; lap: typing with laptop on lap.
a
Positive value represents extension.
b
Negative value represents ulnar deviation; positive value represents radial deviation.

compared to the desk situation. No differences in me- 5.5. Productivity

dian angles were, however, found between laptops
except for the degree of ulnar deviation in the lap situ-
ation. A larger degree of ulnar deviation was observed
when using laptop A on lap. The RoM was about 15° in
the flexion-extension plane with 10° of radial-ulnar de-
viation. No differences were found in the RoM of the
wrist between situations or laptops.
5.4. Muscle load
The performance of the subjects was characterized
by higher muscle load levels in the trapezius and del-
toid muscles when using both laptops with lower mus-
cle load levels in the wrist extensors (laptop A only) in
the desk situation in comparison with the lap situa-
tion (Fig. 2). No differences were found in the muscle
load levels of the cervical extensors between test situa-
tions.
On the desk, the subjects wrote more characters with
laptop A than with laptop B (Table 3). The magnitude
of the difference in productivity between laptops was,
however, very small since it represented less than 4% of
the overall productivity. Moreover it varied largely
within the group of subjects as showed by the broad
confidence interval. The number (absolute and relative)
of errors was comparable between laptops in each situ-
ation and between situations for each laptop.
5.6. Perceived discomfort
Less than half of the subjects reported discom-
fort during tests. The intensity of discomfort was weak
(score ¼ 2), very weak (score ¼ 1) or extremely weak
(score ¼ 0:5) in five subjects while three subjects scored
their discomfort at four (between moderate and strong)
 H. Moffet et al. / Clinical Biomechanics 17 (2002) 368–375 373

6.1. Comparison of postures between laptops

The position of the keyboard and screen are major

Fig. 2. Bar graph illustrating the mean of the differences in muscle load
levels (lV) between desk and lap situations and 95% confidence in-
tervals for the differences in four neck and upper arm muscles when
typing on two different laptop computers; laptop A (shaded bars) and
laptop B (white bars). Positive differences indicate higher muscle load
in the desk situation whereas negative differences indicate higher
muscle load in the lap situation.
in the neck–shoulder or lower back regions after the
test with laptop B on the lap. No clear increase in the
prevalence or intensity of discomfort was found from
the first to the last test. More subjects tended to report
discomfort (with an intensity of 2–4) in the neck–
shoulder regions after the tests on lap (with both lap-
tops) than after the tests on desk.
6. Discussion
The purposes of this study were to determine the
impact of laptop designs on neck and upper limb pos-
ture, muscle activity and productivity, and to compare
the use of laptops in two typical work situations, on an
ordinary desk and on lap. The results showed that there
were no marked differences in productivity and only
small differences in physical exposures in a typing task
between the two laptop designs. The workstation set up
(desk and lap), however, had great influence on the
physical exposures.
determinants of the main postural elements of VDT
operators, according to Grandjean [2]. In the present
study, the subjects adopted comparable head, upper
limb and trunk postures when they used the two differ-
ent laptop designs on the desk. This finding could lar-
gely be explained by the fact that the subjects positioned
both laptops on the desk to insure the same keyboard to
desk edge distance. In addition, the screens of both
laptops were adjusted to a similar degree of inclination
and were at a similar vertical distance (height) from the
floor. Only the eye to screen and eye to document holder
distance slightly differed between the two laptop designs.
It is doubtful that a difference of less than 5 cm in the
eye to screen distance was sufficient to induce detectable
changes in posture.
It was particularly interesting to note that the subjects
adjusted intuitively the keyboard (middle row) at a
constant distance to the desk edge, about 20 cm, irre-
spective of laptop design. This finding supports the as-
sumption that the keyboard is a major determinant for
the placement of laptops on desk. Similar middle row-
desk edge distances were previously reported for tradi-
tional keyboard users [1,14]. On the lap, the keyboard of
laptop B was placed on average, 6 cm closer to the ab-
domen than that of laptop A. To compensate for the
keyboard placed closer to the abdomen, the subjects
adopted a slightly more abducted (5°) and extended
position of the upper arm when typing with laptop B
without changing the head and trunk postures. Proba-
bly, as in the desk situation, the keyboard distance
was an important factor in determining the position of
the laptops on the lap. In this latter situation, how-
ever, the dimension of the thighs and the need for a
stable and comfortable placement of the laptop when
typing, most likely played an important role in position
choice of the laptop in addition to other factors such as
the anthropometric characteristics and postural habits
of the subjects as well as their workstation setting
preference.

Table 3
Productivity
Median (range) Mean of the differences [95% CI]
A B Desk-lap A–B
Desk Lap Desk Lap A B Desk Lap
Number of 597 (476–806) 580 (498–795) 583 (467–738) 580 (506–811) 23 [4 to 49] 1 [38 to 40] 22 [1 to 43] 1 [22 to 23]
characters
Number of 9 (2–24) 11 (8–22) 15 (10–20) 11 (3–34) 2 [5 to 2] 0 [8 to 7] 4 [10 to 1] 3 [10 to 4]
errors
Percent of 2 (0–4) 2 (1–4) 2 (2–3) 2 (1–6) 0 [1 to 0] 0 [1 to 1] 1 [2 to 0] 0 [2 to 1]
errorsa
Abbreviations: CI: confidence interval; A: laptop A; B: laptop B; desk: typing with laptop on desk; lap: typing with laptop on lap.
a
Number of errors in percent of the number of written characters.
374 H. Moffet et al. / Clinical Biomechanics 17 (2002) 368–375
Similar amplitudes of wrist extension and a slightly
smaller magnitude of lateral deviation was found in this
study as compared to a recent laboratory study in 25
typists working on a standard computer workstation
[15]. However, no difference was found in the wrist
motions when typing on the desk with laptop A or B.
This suggested that the presence or absence of a built-in
palm rest and the keyboard positioning had no signifi-
cant impact on the wrist posture adopted when typing.
From the videotape recordings, we observed that five
out of eight subjects used the palm rest of laptop A
during tests, at least for short pauses. Interestingly, the
same five subjects also used support when typing with
laptop B. They rested their forearm–wrist segments on
the desktop or on the edge of the laptop B. This habit
obviously minimized the impact of the built-in palm rest
on wrist posture. Different profiles of use of the built-in
palm rest might, however, be found in more experienced
subjects. On the lap, a larger degree of ulnar deviation at
the wrist was observed when using laptop A as com-
pared with laptop B. This could in part be related to
the fact that half of the subjects rested their palms on
the built-in palm rest of laptop A, whereas, none of the
subjects used support when typing with laptop B.
Finally, one can assume that the design of laptop A is
more advantageous than that of laptop B because it
contributes indirectly to reduce the visual distance (eye
to screen distance). Indeed, the subjects tended to place
the keyboards of both laptops at a similar distance
which implied that the screen of laptop A was placed
closer to the subject’s eyes than that of laptop B, espe-
cially in the lap situation. Although this may be bene-
ficial, the difference in the visual distance between the
two laptops was rather small, less than 5 cm, and re-
mained within the limits of the recommended range. In
the present study, the eye to screen center distance
varied from 48 to 59 cm across the four tests while the
acceptable range for the viewing distance is reported to
be from 45 to 70 cm [7,16]. Considering that the screen
dimension of laptops are smaller than that of ordinary
computers, it is understandable that the screen to eye
distance values lie in the lower limit of this reference
range. This result is consistent with the work of Saito
et al. [17] who demonstrated that the viewing distance
was markedly shorter (8 cm) with laptop computers as
compared to desktop computers.
6.2. Comparison of posture and muscle load between
situations
The posture of VDT operators is, in part, determined
by the physical dimensions of the workstation [16]. In
the present study, the physical dimensions of the two
tested workstations varied considerably. This obviously
implied more changes in posture of the users than the
effects of laptop designs. Major changes in the vertical
and sagittal position of the keyboard and screen were
observed between the desk and lap situations. In the lap
situation, to adjust for the lowering of the keyboard and
the screen, the subjects had larger amplitudes of forward
head bending, backward trunk inclination and wrist
extension and had reduced amplitudes of upper arm
elevation. The largest postural differences (about 10°)
were observed in the head, upper arm and wrist seg-
ments. A larger amplitude of wrist extension has also
been observed when the keyboard was lower than the
elbow height during desktop computer work [18]. The
increased forward head bending in persons working on a
laptop computer as compared with a desktop computer
has been previously associated to the lower vertical
position of the laptop screen [17,19]. The screen height
was also found to be a significant factor affecting the
neck posture and extensor muscle load in VDT work
[20]. In our study, although the adjustment of the screen
in a more horizontal position could have help minimize
forward head bending and trunk inclination angles in
the lap situation, mean differences of 10° and 5° in the
head and trunk postures respectively, remained between
the two situations. A greater inclination of the screen in
the lap situation as compared to the desk situation was
expected since the degree of inclination of the screen was
previously found to be negatively correlated to its height
[3,14].
Changes in posture between test situations occurred
in parallel with major changes in muscle activity in three
out of the four muscles tested. Higher muscle load levels
in the trapezius and deltoid muscles and lower mus-
cle load level in the wrist extensors were found in the
desk situation as compared to the lap situation. These
differences in muscle load cannot be attributable to dif-
ferent levels of productivity since the number of charac-
ters written and the number of keying errors were
similar in both situations. On the other hand, differences
in posture, especially the higher elevation of the upper
arm and the reduced amplitude of wrist extension on
desk may be related to these differences in muscle load
levels. In contrast with previous work, however, the
results of this study are inconclusive with regards to the
relationship of the flexed neck position and increased
load of the cervical neck extensors [17,20–22].
It is likely that the subjects adjusted the workstation
and laptop components to minimize physical strain by
favoring more neutral postures. In this context and
considering that more extreme postures have been as-
sociated with increased complaints [2,5], it is likely that
more stress (muscle load and articular constraints) was
imposed on the shoulder-upper arm regions in the desk
situation whereas more constraints were imposed to the
head-neck and wrist segments in the lap situation. Such
an interpretation is in agreement with other studies that
showed complaints in the shoulder-neck regions and
EMG levels in the trapezius muscles increased as the
 H. Moffet et al. / Clinical Biomechanics 17 (2002) 368–375
vertical distance of the keyboard from the floor in-
creased, while complaints in the wrist region were
more frequent with a lower positioning of the keyboard
[4,23].
7. Conclusion
These results highlight the importance of the choice
of workstation (desk or lap) rather than laptop design.
Greater physical (muscular and articular) constraints
seem to be imposed to the shoulder region in the desk
situation whereas the head-neck and wrist segments ap-
pear to be more stressed in the lap situation. These re-
sults need to be confirmed in a larger number of subjects
with different level of experience while performing var-
ious durations and types of task such as editing.
Acknowledgements
The authors gratefully acknowledge financial support
from Laval University, Quebec, Canada and the Swed-
ish National Institute of Occupational Health, Stock-
holm, Sweden. The authors thank Ms. Louise Bertrand
who made the drawings.
References
[1] Grandjean E, Hunting W, Nishiyama K. Preferred VDT work-
station settings, body posture and physical impairments. App
Ergon 1984;15:99–104.
[2] Grandjean E. Fitting the task to the man. In: A Textbook of
Occupational Ergonomics. 4th ed. London: Taylor and Francis;
1988.
[3] Miller W, Suther TW. Display station anthropometrics: preferred
height and angle settings of CRT and keyboard. Hum Factors
1983;25:401–8.
[4] Bergqvist U, Wolgast E, Nilsson B, Voss M. Musculoskeletal
disorders among visual display terminal workers––individual,
ergonomic and work organizational factors. Ergonomics 1995;38:
763–76.
[5] H€unting W, L€ aubli T, Grandjean E. Postural and visual loads at
VDT workplaces. I. constrained postures. Ergonomics 1981;24:
917–31.
[6] Sauter SL, Schleiffer LM, Knutson SJ. Work posture, workstation
design, and musculoskeletal discomfort in a VDT data entry task.
Hum Factors 1991;33:151–67.
375
[7] Stammerjohn LW, Smith MJ, Cohen BF. Evaluation of work
station design factors in VDT operations. Hum Factors
1981;23:401–12.
[8] Moffet H, Hagberg M. Variation in physical exposure measures
during keyboard work on a portable computer. A methodological
study to determine optimal test duration. In Second International
Scientific Conference on Prevention of Work-related Musculo-
skeletal Disorders (PREMUS 95), Montreal, 1995, Abstract no
1072.
[9] Chaffin DB, Andersson GBJ. In: Occupational Biomechanics. 2nd
ed. New York: John Wiley; 1991.
[10] Borg G. Psychophysical basis of perceived exertion. Med Sci Sport
Exer 1982;14:377–81.
[11] Ericson B-E, Hagberg M. EMG signal level versus external
force: a methodological study on computer aided analysis. In:
Biomechanics VI-V. Baltimore: University Park Press; 1978.
p. 251–5.
[12] Hagberg M. The amplitude distribution of surface EMG in static
and intermittent static muscular performance. Eur J App Physiol
1979;40:265–72.
[13] Jonsson B. Kinesiology, with special reference to electromyo-
graphic kinesiology. In: Cobb WA, Van Duijn H, editors.
Contemporary Clinical Neurophysiology. Amsterdam: EEG;
1978. p. 417–28 (suppl 34).
[14] Grandjean E, Hunting W, Pidermann M. VDT workstation
design: preferred settings and their effects. Hum Factors
1983;25:161–75.
[15] Serina ER, Tal R, Rempel D. Wrist and forearm postures and
motions during typing. Ergonomics 1999;42:938–51.
[16] Green RA, Briggs CA, Wrigley TV. Factors related to working
posture and its assessment among keyboard operators. App Ergon
1991;22:29–35.
[17] Saito S, Miyao M, Kondo T, Sakakibara H, Toyoshima H.
Ergonlomic evaluation of working posture of VDT operation
using personal computer with flat panel display. Ind Health
1997;35:264–70.
[18] Simoneau GG, Marklin RW. Effect of computer keyboard slope
and height on wrist extension angle. Hum Factors 2001;43:
287–98.
[19] Straker L, Jones KJ, Miller J. A comparison of the postures
assumed when using laptop computers and desktop computers.
App Ergon 1997;28:263–8.
[20] Villanueva MBG, Jonai H, Sotoyama M, Hisanaga N, Takeuchi
Y, Saito S, et al. Sitting posture and neck and shoulder muscle
activities at different screen height settings of the visual display
terminal. Ind Health 1997;35:330–6.
[21] Sch€uldt K, Ekholm J, Harms-Ringdahl K, Nemeth G, Arborelius
U. Effects of changes in sitting work posture on static neck and
shoulder activity. Ergonomics 1986;29:1525–37.
[22] Hamilton N. Source document position as it affects head position
and neck muscle tension. Ergonomics 1996;39:593–610.
[23] Bendix T, Jessen F. Wrist support during typing––a controlled,
electromyographic study. App Ergon 1986;17:162–8.