Content uploaded by Amar Gandavadi
Author content
All content in this area was uploaded by Amar Gandavadi on Mar 03, 2015
Content may be subject to copyright.
People are increasingly spending more time
sitting – for work, for travel and for relax-
ation. However, it is important that the
seat is comfortable, maintains a healthy
posture and facilitates task performance. The pur-
pose of this article is to describe a study investigat-
ing the effects of two different seated postures upon
the muscular activation and the ability to perform a
simple upper limb skill.
BIOMECHANICS
The seated posture changes the demands placed
on the musculoskeletal system, since there is a ten-
dency to gradually adopt a position in which the
pelvis is rotated backward to compensate for mus-
cular tightness in the Hamstring muscles (Pheasant,
1996). Such rotation at the pelvis is accompa-
nied by flexion of the lumbar spine, which may
be close to the end of the available range. In this
position the muscles relax and the body weight is
supported by passive structures such as the spinal
ligaments. Therefore, slump sitting has been noted
to be associated with higher intradiscal pressures
(Nachemson, 1975).
International Journal of Therapy and Rehabilitation, November 2005, Vol 12, No 11 485
Effect of two seating
positions on upper limb
function in normal subjects
The need for good sitting posture is further rein-
forced, since prolonged sitting has been shown to
predispose individuals to the development of low
back pain (Magora, 1972; Wilder and Pope, 1996).
Electromyographic studies suggest that sitting in a
slouched or reclined posture relaxes the trunk mus-
cles and requires minimal muscle activity to support
the body weight. However, this finding is ambigu-
ous, since the reduction in muscular activity may
not be healthier in terms of minimizing the onset of
low back pain.
To sit in an anterior pelvic tilt position, the mus-
cles must be activated to achieve and maintain the
position. Further, habitual slump sitting could result
in muscle fibre imbalance with slow oxidative fibres
changing to fast twitch fibres (Cram and Vinitzky,
1995). Such fibre changes may result in loss of abil-
ity to maintain an upright seated posture for longer
than a short period before fatigue intervenes and the
person then slumps into a posteriorly tilted posture
(Garlick, 1998).
Of recent years, there have been attempts to
design seating that maintains the pelvis in a rela-
tively neutral position (i.e. neither posteriorly nor
anteriorly tilted). To achieve this position, the hips
Many upper limb functions are performed in a sitting position. However, if seating is inadequate and
poorly designed, back pain and reduced upper limb control may result. This study investigates pelvic
posture and performance in an upper limb task.
In total, 15 normal healthy volunteers (aged 18–30 years) were seated in posterior and anterior pelvic
tilt positions and performed a simple upper limb task. The parameters measured were electromyography
of the lumbar paravertebral muscles, time taken to complete the task and the task error rate. The data
were analysed by repeated measures analysis of variance (ANOVA) and post-hoc t tests. The results
indicate that, when seated in an anterior pelvic tilt position, the error rate decreased (P=0.013) and the
electrical activity increased (P=0.008). The time taken to complete the task was not significantly different.
Since the error rate decreased when seated in the anterior pelvic tilt position, it might be concluded
that this posture facilitated task skill. It is suggested that the increase in electrical activity is related to
the increased load on the lumbar muscles in maintaining the posture. This study’s generalizability is
limited, with small subject numbers and the use of a non-functional task. It is, however, a beginning in
addressing the interrelationship between the seated posture and skilled upper limb performance.
Key words: upper limb function, seating, pelvic posture, back pain
Gandavadi A, Ramsay J, James G (2005) Effect of two seating positions on upper limb function in normal subjects. Int
J Ther Rehabil 12(11): 485–490
Clinical
Amar Gandavadi
is PhD student and
Jill Ramsay and Gill
James are Lecturers in
Physiotherapy, School
of Health Sciences,
Physiotherapy, University
of Birmingham,
Edgbaston, Birmingham
B15 2TT, UK.
Correspondence to:
Amar Gandavadi
Amar Gandavadi, Jill Ramsay, Gill James
ered a suitable non-invasive method for assessing
electrical activity in muscles. A wire game was used
for quantifying the upper limb function. This is a
simple game that needs no special skill to play, and
is an easy way of measuring errors in the activity.
Ethics
The research was approved by the ethics committee
of the School of Health Sciences before commence-
ment of the research. All subjects were volunteers
recruited from an advertisement placed on a notice-
board in the university. Subjects provided informed
consent and could withdraw at any time without
explanation.
Subject selection
Participants were included in the study if they had
no low back or upper limb pain at the time of the
experiment. Participants were excluded if they had
any spinal deformities, acute or persistent back
pain or shoulder or upper limb injury, or lacked full
range upper limb movement, because the presence
of these conditions could influence the outcome of
the study.
Materials
Materials used were as follows:
■ Adjustable-height treatment couch.
■ The Bambach Saddle Seat (Figure 1).
■ The NeuroTrac dual channel electromyograph
(Verity Medical Ltd., www.veritymedical.co.uk)
■ A wire game (Figure 2): This consists of a base
with two upright arms perpendicular to the base
International Journal of Therapy and Rehabilitation, November 2005, Vol 12, No 11 487
486 International Journal of Therapy and Rehabilitation, November 2005, Vol 12, No 11
Clinical
must be positioned at an angle of less than 90º.
Mandel (1976, 1981) argued that a forward sloping
seat surface decreases the hip flexion requirements.
Various seats have been designed adopting these
principles – most notably, the ‘kneeling’ chair.
Evaluation of such seats has been largely confined
to examining user comfort (Drury and Flancher,
1985). Such investigations showed that users of the
kneeling chair complained of increased discomfort
in the knees, thereby limiting its usefulness.
Recently, a saddle-type seat (manufactured by
Bambach; www.bambach.co.uk) has been mar-
keted. This seat maintains the pelvis in an anteriorly
tilted posture but does not bear body weight on the
knees, and should therefore be investigated since
theoretically, it may be more user friendly and there-
fore be used more widely.
SEATING AND UPPER LIMB SKILL
One aspect of seating that has been relatively
neglected is the effect upon upper limb function.
Reissner (1972) argues that sitting in an anterior
pelvic tilt position provides pelvic stability, which
is necessary to allow for dynamic spinal move-
ments and for upper limb stability (Nwaobi, 1987).
However, he fails to explain how pelvic stability
affects upper limb function. Logically, sitting in an
anterior pelvic tilt position promotes ease of main-
tenance of the natural lumbar curve and may mini-
mize discal pressures (Keegan, 1953; Mandal, 1981;
Bridger et al, 1992).
Some work has been carried out investigating chil-
dren with disabilities such as cerebral palsy. Myhr
(1994), for example, showed greater ability to use
the upper limb when seated in an anterior pelvic tilt
position, thereby resulting in improved task qual-
ity and diversity. Biomechanically, more upper limb
effort is required to perform movements in the pos-
terior pelvic tilted position owing to the gravitational
pull, thus rendering arm and hand function non-
optimal (Ariyaratnam et al, 2000). However, there
appears to have been little work investigating the
effect of the seating posture upon skill performance
in a population of people without such disabilities.
METHODOLOGY
Research design
A same-subject experimental design was selected,
since the focus of interest is comparing the two dif-
ferent pelvic positions while performing the same
experimental task. Surface electromyography
(EMG) was used to record the electrical activity
of the paraspinal muscles. EMG has been used in
similar studies (Finsen et al, 1998; Callaghan and
McGill, 2001) and within its constraints, is consid-
Figure 1. The Bambach Saddle Seat (reproduced with
permission from Bambach).
and a sinuous metal wire connecting the arms. A
jockey is inserted into the metal wire. The jockey
consists of a hand piece with a small metal ring
in front of it and with a wire attached to it. The
metal ring of the jockey is inserted into the sinu-
ous metal wire.
■ Stopwatch.
■ Tape measure.
Pilot study
A pilot study was carried out with one subject in
order to test the methodology and to assess reliabil-
ity in measuring the parameters. The experimenter
and two physiotherapists measured all the param-
eters. No differences were observed in the results.
Task
The task involved passing a metal ring into a sinuous
wire, tracing its path to the end. A light illuminated
each time the metal ring contacted the wire. In order
to control for possible learning effects, the subject
could not see the light. All the trials were therefore
carried out under the same conditions. The wire game
was placed on a height-adjustable couch at distance
of 1.5 forearm lengths from the trunk midline of
the subject and at elbow height of the seated subject
when the subject’s hand was placed on the thigh.
The subjects were seated for 30 mins before the
task was carried out, in order to acclimatize them to
the sitting position and to standardize the condition
of the subjects before the experiment. Various time
periods have been used in similar studies – from
very short (e.g. mins) to prolonged (Callaghan and
McGill, 2001). A total of 30 mins acclimatization
was used because it was long enough to allow tis-
sues to adapt, but not so long as for fatigue to occur.
Conditions and parameters measured
Figure 3 shows the subject seated on the Bambach
Saddle Seat with the pelvis anteriorly tilted. The
back was unsupported and the feet were fully sup-
ported on the floor by adjusting the height of the
International Journal of Therapy and Rehabilitation, November 2005, Vol 12, No 11 487
486 International Journal of Therapy and Rehabilitation, November 2005, Vol 12, No 11
Clinical
Bambach chair. The hip was kept at an obtuse angle
from the horizontal, allowing the pelvis to fall in the
anterior tilted position.
Figure 4 shows the subject seated on a couch with
the back slumped and unsupported in a posterior
pelvic tilt position. The feet were fully supported on
the floor by adjusting the height of the couch. The
order of seated positions was counterbalanced to
control for any possible order effects.
The parameters measured were:
■EMG activity of the lumbar paravertebral mus-
cles – mean of peak amplitudes of three trials.
The peak amplitudes are sudden surges in ampli-
tude, which were critical to the task being per-
formed. The mean of these peak amplitudes was
used as it uses all the data and improved the reli-
ability of the data. Two electrodes are placed in
the back at the L1/L2 spinal level and two elec-
trodes are placed at the L4/L5 spinal level (Cram
and Vinitzky, 1995).
■Time taken for task completion. The mean of the
three trials was used.
Figure 2. The wire game.
Figure 3. Subject seated on
the Bambach Saddle Seat
in an anterior pelvic tilt
position (reproduced with
permission from Bambach).
Figure 4. Subject
seated on a couch in
a posterior pelvic tilt
position (reproduced with
permission from Bambach).
488 International Journal of Therapy and Rehabilitation, November 2005, Vol 12, No 11
Clinical
■Task error rate. This was measured as the number
of times the light illuminated during the task. The
mean of the three trials was used.
Data analysis
All these hypotheses are two-tailed. Two-way, same-
subject repeated measures analyses of variance
(ANOVA) were performed to test the hypothesis that
the seated posture makes a difference to the param-
eters stated above. Post-hoc tests (the t test) were
calculated to examine significant differences where
appropriate. Throughout, a 0.05 level of significance
was used for rejection of the null hypotheses.
Procedure
■The subjects were seated for 30 mins and given a
magazine or newspaper to read.
■The height of the couch or seat was adjusted so
the feet were flat on the floor, with the thighs
fully supported.
■After 30 mins, three trials of the upper limb task
were carried out.
■The seated condition was then adjusted and a
further three trials of the upper limb task was
carried out.
■The time for completion of the task was measured
using a stop clock and the numbers of errors were
noted in accordance to the number of light glows.
Simultaneously, EMG measurements were taken.
RESULTS
A total of 15 subjects aged 18–30 years who ful-
filled the inclusion criteria were recruited to take
part in the experiment. Figure 5 shows the mean and
standard error for all three parameters. The error rate
and task time had both fallen when the subjects were
seated in the anterior pelvic tilt position (Bambach
seat), while the mean EMG measurement increased
in that condition.
The ANOVA results were significant (F=216.105,
P<0.000). Post hoc t tests were significant for errors
(t=2.863, P=0.013) and for EMG (t=–3.106,
P=0.008). The time taken to complete the task was
not significant. It is clear that both the error rate and
the EMG measurements (P=0.008) were signifi-
cantly affected by the seating posture. While the error
rate decreased when using the Bambach seat, the
mean peak EMG reading increased. In both cases,
therefore, the null hypothesis can be rejected and the
experimental hypothesis accepted.
DISCUSSION
The study shows that there was a statistically signifi-
cant difference in the subject’s upper limb function
relative to this task in the anterior pelvic tilted posi-
tion (i.e. seated on the Bambach chair). This sug-
gests that the principles of pelvic tilt and dynamic
seating enhance natural spinal curves, good seating
and upper limb function (Keegan, 1953; Mandal,
1981, Bridger et al, 1992; McClenaghan et al, 1992).
The fall in the error rate was highly significant
(P=0.013), with 12 participants improving their per-
formance in the anterior pelvic tilt position. Reissner
(1972) reported that an anteriorly tilted pelvis pro-
vides pelvic stability and is necessary to allow for
dynamic spinal movements. Such improved stabil-
ity should enable subjects to access the wire game
more easily, because movement occurs at the hip
without the characteristic slouching so often seen
Figure 5. (a) Bar chart for
errors made in the two
seating positions, showing
mean ±2 standard error
(SE) of mean; (b) error bar
chart for electromyography
readings in the two seating
positions, showing mean
±2 SE of mean; (c) error
bar chart for time taken
for activity, showing mean
±2 SE of mean.
7
22
6
20
5
18
4
16
3
14
2
12
10
40
30
20
10
Errors: Couch
EMG: Couch
Activity time Activity time
Errors: Bambach Saddle Seat
EMG: Bambach
a.
c.
b.
Number of errors made (mean ± 2E)
Electromyography readings (mean ± 2E) (µA)
Time taken (mean ± 2E) (s)
in other positions and facilitates the pelvic stabil-
ity necessary for good upper limb movement for
desktop tasks. Conversely, poor performance by the
subjects sitting in the posterior tilted position was
noted – a finding which apparently concurs with the
study reported by Ariyaratnam et al (2000).
An increase in EMG activity was noted when
subjects were seated in the Bambach seat.
Biomechanically, in the anterior pelvic tilted posi-
tion, the line of gravity falls anterior to the axis of
rotation at the ischial tuberosities, gravity acts to
flex the head and trunk so as to pull it down. In
order to maintain a stable posture, the back and
neck muscles generate an equal torque in the oppo-
site direction (Norkin and Levangie, 1983). There is
a reflex activation of the postural muscles as well as
the order in which these muscles are recruited for
postural alignment and stability. This supports the
reason for increased EMG activity of the back mus-
cles reported in this study.
When asked to comment on the two postures, par-
ticipants indicated that they believed that their upper
limb performance was enhanced by their improved
posture. However, three of the participants reported
that they were more comfortable sitting in the pos-
terior pelvic tilt position and perceived difficulties
in performing the task in the anterior pelvic tilt posi-
tion. The reason for this might be a muscle fibre
imbalance associated with the habitual slumped
position – instead of slow oxidative fibres of postural
or antigravity muscles, the fast twitch muscle fibres
can be activated (Cram and Vinitzky, 1995).
This study should be regarded as a pilot study,
owing to its small subject numbers and the narrow
age band from which subjects were recruited. It
would be useful to replicate the study with a larger
sample drawn from a wider age range. Additionally,
it would be useful to study subjects with either low
back pain or neck or upper limb dysfunctions. The
task is unfunctional and therefore adopting a similar
methodology but using a functional work-related task
should be attempted. However, such a study would
require a method of evaluating skilled performance.
CONCLUSIONS
Within these limitations, however, it is possible to
draw several positive conclusions. It appears that the
ability to carry out upper limb tasks may be influ-
enced by the seated posture – specifically, the angle
of pelvic tilt. Interestingly, it may be that the altered
posture facilitates skilled performance, although
this aspect needs further investigation.
The increase in EMG activity, however, indicates
caution since subjects may need time to adapt to
sitting with the pelvis tilted anteriorly. Even allow-
ing for such training time, this position may not suit
everyone and in those cases, it is possible to specu-
late that performance will be adversely affected.
However, for those people whom this sitting posture
suits, it could logically create healthier working pos-
tures and may thereby decrease morbidity in those
individuals with work-related dysfunctions.
The authors gratefully acknowledge the assistance of Bambach for
the loan of a Bambach Saddle Seat. They also thank all the sub-
jects who participated in the study. This study was carried out in
part-fulfillment of the regulations for the degree of MSc in Health
Sciences at the University of Birmingham by the first author.
Conflict of interest: none.
Ariyaratnam A, Chia C, Martin N (2000) The Effect Pelvis-
Positioning Angle has on Upper Extremity Function in
Children with Cerebral Palsy. Best Practice Guidelines. Curtin
University, Curtin
Bridger R, Orkin D, Henneberg M (1992) A quantitative investiga-
tion of lumbar and pelvic postures in standing and sitting: inter-
relationship with body position and hip muscle length. Int J Ind
Ergon 9: 235–44
Callaghan JP, McGill SM (2001) Low back joint loading and kin-
ematics during standing and unsupported sitting. Ergonomics
44(3): 280–94
Cram JR, Vinitzky I (1995) Effects of chair design on back muscle
fatigue. J Occup Rehabil 5(2): 101–13
Drury CG, Flancher M (1985) Evaluation of a forward-sloping
chair. Appl Ergon 16: 41–7
Finsen L, Christensen H, Bakke M (1998) Musculoskeletal disor-
ders among dentists and variation in dental work. Appl Ergon
20(2): 119–25
Garlick D (1998) ‘The Garlick report’, direction. A Journal on the
Alexander Technique 1(8): 8, 118–19, 302
Keegan JJ (1953) Alterations of the lumbar curve relative to pos-
ture and seating. J Bone Joint Surg 35A: 589–603
Magora A (1972) Investigation of the relation between low back
pain and occupation. 3. Physical requirements: sitting, standing
and occupation. Ind Med Surg 41(12): 5–9
Mandel AC (1976) Work chair with tilted seat. Ergonomics 19:
157–64
Mandal AC (1981) The seated man (horno sedens): the seated
work position, theory and practice. Appl Ergon 12(1): 19–26
McClenaghan BA, Thombs L, Milner M (1992) Effects of seat
surface inclination on postural stability and function of the
upper extremities of children with cerebral palsy. Dev Med
Child Neurol 34: 40–8
Myhr U (1994) Influence of different seat and backrest inclinations
on the spontaneous positioning of the upper extremities in non-
disabled children. Physiother Theory Pract 10: 191–200
Nachemson A (1975) Towards a better understanding of low-back
pain: a review of the mechanics of the lumbar disc. Rheumatol
Rehabil 14: 129–43
Norkin C, Levangie P (1983) Posture. In: Norkin C, Levangie P,
eds. Joint Structure and Function. Davis, Philadelphia: 367–94
Nwaobi OM (1987) Seating orientations and upper extremity func-
tion in children with cerebral palsy. Phys Ther 67(8): 1209–12
Pheasant S (1996) Bodyspace: Anthropometry, Ergonomics and the
Design of Work. 2nd edn. Taylor and Francis, London: 60–75
Reissner F (1972) The dynamic sitting posture. Quintessenz 4: 73–80
Wilder DG, Pope MH (1996) Epidemiological and aetiological
aspects of low back pain in vibration environments – an update.
Clin Biomech 11(2): 61–73
■When subjects were seated in an anterior pelvic tilt position, upper limb
task performance was significantly better than when they were seated in a
posterior pelvic tilt position.
■There was increased electrical activity in the lumbar paravertebral muscles in
the anterior pelvic tilt position.
■The decreased error rate in the anterior pelvic tilt position (using the
Bambach Saddle Seat) appears indicative of enhanced upper limb function.
■Sitting in an anterior pelvic tilt position should promote healthier working
postures and may minimize discal pressure.
Key points
490 International Journal of Therapy and Rehabilitation, November 2005, Vol 12, No 11
Clinical
iJtR