Journal of Biomechanics xxx (2018) xxx–xxx

Contents lists available at ScienceDirect

Journal of Biomechanics
journal homepage: www.elsevier.com/locate/jbiomech
www.JBiomech.com

Joint moments’ contributions to vertically accelerate the center of mass

during stair ambulation in the elderly: An induced acceleration approach
Vera Moniz-Pereira a,⇑, Thomas M. Kepple b, Silvia Cabral a, Filipa João a, António P. Veloso a
a
Universidade de Lisboa, Faculdade de Motricidade Humana, CIPER, LBMF, P-1499-002 Lisboa, Portugal
b
C-Motion, Inc., Germantown, MD, USA

a r t i c l e i n f o a b s t r a c t

Article history: Falls are a serious problem faced by the elderly. Older adults report mostly to fall while performing loco-
Accepted 31 July 2018 motor activities, especially the ones requiring stair negotiation. During these tasks, older adults, when
Available online xxxx compared with young adults, seem to redistribute their lower limb joint moments. This may indicate that
older adults use a different strategy to accelerate the body upward during these tasks. The purposes of
Keywords: this study were to quantify the contributions of each lower limb joint moment to vertically accelerate
Stair ascent the center of mass during stair ascent and descent, in a sample of community-dwelling older adults,
Stair descent
and to verify if those contributions were correlated with age and functional fitness level. A joint moment
Joint kinetics
Induced acceleration analysis
induced acceleration analysis was performed in 29 older adults while ascending and descending stairs at
Older adults their preferred speed. Agreeing with previous studies, during both tasks, the ankle plantarflexor and the
knee extensor joint moments were the main contributors to support the body. Although having a smaller
contribution to vertically accelerate the body, during stair descent, the hip joint moment contribution
was related with the balance score. Further, older adults, when compared with the results reported pre-
viously for young adults, seem to use more their knee extensor moment than the ankle plantarflexor
moment to support the body when the COM downward velocity is increasing. By contributing for a better
understanding of stair negotiation in community dwelling older adults, this study may help to support
the design of interventions aiming at fall prevention and/or mobility enhancement within this
population.
Ó 2018 Elsevier Ltd. All rights reserved.

1. Introduction
Approximately 30% of older adults living in the community fall
each year and the impact of such events on elderly’s quality of life
and on health care costs is well documented (World Health
Organization, 2007). Gait and balance problems are among the
strongest and potentially modifiable risk factors for falling
(Ambrose et al., 2013; Deandrea et al., 2010). Further, falls are
reported to occur mostly while walking and dealing with steps
and stairs (Lord et al., 2007; Timsina et al., 2017).
The biomechanical and motor control requirements necessary
to successfully negotiate with stairs are high, not only increasing
fall risk but, more importantly, the risk of fall related injuries
(Jacobs, 2016). In particular, stair walking requires a higher
demand at the knee and ankle joints, when compared with level
⇑ Corresponding author at: Laboratório de Biomecânica e Morfologia Funcional,
Faculdade de Motricidade Humana, Estrada da Costa, 1499-002 Cruz Quebrada –
Dafundo, Portugal.
E-mail address: veramps@fmh.ulisboa.pt (V. Moniz-Pereira).
walking (Jacobs, 2016). This could be especially critical for older
adults who, compared with younger adults, redistribute their joint
moments during stair ambulation (Karamanidis and Arampatzis,
2011, 2009; Novak and Brouwer, 2011; Reeves et al., 2009, 2008).
While ascending stairs, older adults apply lower knee extensor
and ankle plantarflexion joint moments, and higher hip extensor
joint moments, when compared with younger adults
(Karamanidis and Arampatzis, 2009; Reeves et al., 2009). This prox-
imal shift in movement strategy has been mostly attributed to
changes in neuro-muscular system occurring with aging (Jacobs,
2016; Karamanidis and Arampatzis, 2009). While descending stairs,
older adults seem also to apply lower ankle plantarflexor joint
moments, higher hip extensor joint moments and the same or
higher knee extensor joint moments, when compared with young
subjects (Karamanidis and Arampatzis, 2011; Novak and Brouwer,
2011; Reeves et al., 2008). By applying a lower leading limb plan-
tarflexor joint moment during early stance and a higher trailing
limb knee extensor joint moment during mid stance, older adults
seem to use more their trailing limb to prepare the double support
phase than younger adults (Karamanidis and Arampatzis, 2011).

https://doi.org/10.1016/j.jbiomech.2018.07.040
0021-9290/Ó 2018 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Moniz-Pereira, V., et al. Joint moments’ contributions to vertically accelerate the center of mass during stair ambulation
in the elderly: An induced acceleration approach. J. Biomech. (2018), https://doi.org/10.1016/j.jbiomech.2018.07.040
2 V. Moniz-Pereira et al. / Journal of Biomechanics xxx (2018) xxx–xxx
This joint moment redistribution that occurs with ageing may
be the cause why older adults appear to have more difficult in con-
trolling fast vertical displacements of the center of mass (COM)
during stair ambulation (Jacobs, 2016).
While the mentioned studies provided a great insight about the
age effect on the generation of lower limb joint moments during
these tasks, they do not allow a direct measure between the joint
moments and the COM acceleration. Induced acceleration analysis
(IAA) is based on the dynamic coupling principles that each joint
moment applied to a body will accelerate all joints of the body
(Zajac and Gordon, 1989) and allows the direct quantification of
the contribution of each lower limb joint moment to the COM
acceleration. To the best of our knowledge, IAA was only used to
analyze lower limb function during stair ambulation in healthy
adults (Harper et al., 2018; Lin et al., 2015).
Therefore, this study aims to quantify, using IAA, the lower limb
joint moments’ contributions to vertically accelerate the COM dur-
ing stair ascent and descent in a sample of community dwelling
older adults. Further, because age may not be the only factor con-
tributing to this joint moment redistribution (Hortobágyi et al.,
2016) and because, in the field, professionals use simpler func-
tional fitness measures to determine fall risk level and to monitor
the effects of exercise interventions on gait and balance perfor-
mance, we also aimed to verify if those contributions are correlated
with age and commonly used functional fitness tests. It was
hypothesized that joint moment contributions would be related
with both age and functional fitness level in the tested sample.
2. Methods
2.1. Study design
A cross-sectional study was conducted.
2.2. Participants
Sample size was determined based on Cohen’s work (Cohen,
1988), considering a two tailed test, 5% level of significance, 80%
power and an effect size of 0.5. A minimal number of 28 partici-
pants was obtained. Nevertheless, to account for withdraws,
non-attenders and to increase precision 33 participants from the
Biomechanics of Locomotion in the Elderly Project (PTDC/
DES/72946/2006) (Moniz-Pereira et al., 2012) were invited to
participate.
Participants were considered eligible if they were above 60
years, living in the community independently, and able to ascend
and descend a flight of stairs without using the handrail. The exclu-
sion criteria included any reported neurologic or orthopedic condi-
tion, as well as any uncorrected visual problem, that would affect
their locomotion pattern.
From the 33 invited participants, data from 4 different partici-
pants were excluded from each task due to problems in data acqui-
sition or modeling. Thus, in order to maximize sample size and
power (Knudson, 2017), two similar groups of 29 participants were
included for analysis. All the participants were informed about the
aims and study protocols, agreed to participate and signed
the informed consent. The Faculty Ethics Committee approved
the study protocol.
2.3. Procedures
On their first session, participants answered a health and falls
questionnaire and performed different functional fitness tests, fol-
lowing the procedures described elsewhere (Moniz-Pereira et al.,
2012). In the health questionnaire participants were asked about
Please cite this article in press as: Moniz-Pereira, V., et al. Joint moments’ contributions to vertically accelerate the center of mass during stair ambulation
in the elderly: An induced acceleration approach. J. Biomech. (2018), https://doi.org/10.1016/j.jbiomech.2018.07.040
their demographic data, general health, medication intake (and
associated diseases) and fall history. This questionnaire was used
to verify the eligibility of the participants according to the previously
mentioned criteria. The functional fitness tests performed included
the 8 foot Up-and-Go (UG) test, a measure of motor agility from
Senior Fitness Test battery (Rikli and Jones, 1999), and items 4 (step
up and over), 5 (tandem walk), 6 (stand on one leg) and 7 (stand on
foam eyes closed) from Fullerton Advanced Balance Scale (Rose
et al., 2006). In the UG test, a higher functional fitness level is present
if the subject takes less time to complete the test. The sum of the bal-
ance tests scores represents a total score and a better result is
obtained with a higher score (in a maximum of 16 points). The
results of the mentioned functional fitness tests have been previ-
ously associated with fall risk (Hernandez and Rose, 2008; Rose
et al., 2002).
During the second visit, anthropometric measures (body mass,
stature and trochanteric height) and motion capture tests were
performed. The participants were barefoot and wore tight black
shorts and t-shirts. Thirty passive markers and four marker clusters
were placed based on the calibrated anatomical system technique
(Cappozzo et al., 1995) by the same researcher. Specifically, six
markers were placed on the trunk, one on top of each acromion,
one on the C7 spinous process and three on the sternum area
(placed so that soft tissue artifact and collinearity was avoided).
At the pelvis, two markers were placed on each posterior superior
iliac spines and two along each iliac crest. A virtual marker was
created in each anterior superior iliac spine using a digitizing poin-
ter. Markers were also placed on the lateral and medial femur epi-
condyles, the lateral and medial ankle malleoli and on the top of
the first and fifth metatarsal heads. Each foot had also one marker
on the heel, other laterally on the middle of the foot and a third one
between the two metatarsal heads. Finally, the mentioned marker
clusters were attached to both thighs and shanks.
Kinematic and kinetic data were collected at 200 Hz using 8
infrared cameras (Oqus 300, Qualisys AB, Sweden) synchronized
in time and space with two force plates (9281B and 9283U014,
Kistler, Switzerland).
For the stairs trials, a wooden staircase with three steps was
built. Each step was 15 cm high and 27 cm deep, except for the last
step which depth was 80 cm. The first force platform was imbedded
on the floor in front of the staircase while the second was covered
by the first step. This step was securely fixed to the second force
platform and was built ensuring the rigidity of the structure. Each
force platform was independent of the surrounding wooden pieces.
A similar wooden structure has been used previously (Alcock et al.,
2014) and the staircase mounting error associated with this struc-
ture type was shown to be negligible (Chesters et al., 2014).
For the ascend task, participants stood still in front of the stair-
case and stepped on the floor in front of the staircase prior to
climbing up the stairs and stopped on top of the staircase. For
the descend task, participants stood still on top of the staircase
and continued for one more step after stepping on the floor. They
walked at their comfortable pace and to use a step over step pat-
tern during both ascending and descending tasks. Practice trials
were performed before collecting 10 trials (five with each limb)
from each task.
2.4. Data processing
Three trials of each task, in which the right limb would step on
the first step of the staircase, were selected to be analyzed, in order
to maximize sample size.
Kinematic and kinetic data were filtered using a fourth order
Butterworth low pass filter at 10 Hz.
The biomechanical model built for each participant had 7 seg-
ments (feet, shanks, thighs and a head-arms-trunk segment). The
 V. Moniz-Pereira et al. / Journal of Biomechanics xxx (2018) xxx–xxx
local coordinate system of each segment were defined in accor-
dance with Robertson et al. (2014). The hip joint center was com-
puted using the pelvis markers, through a regression equation
proposed by Bell et al. (1990), the knee joint center was the mid-
point of the epicondyles and ankle joint center the mid-point of
the malleoli (Robertson et al., 2014). Global optimization (Lu and
O’Connor, 1999) was used to estimate the position and orientation
of the segments in which three rotations were allowed at the hip,
one at the knee (flexion/extension) and two at the ankle (dorsi/
plantar flexion, and external/internal rotation), while also restrain-
ing all joints’ translations. Segment masses were determined
according to Dempster (Dempster, 1955), whereas the remaining
inertial parameters were computed based on Hanavan (1964).
Basic stride events were computed based on ground reaction
force data in order to obtain spatial-temporal parameters. Lower
limb joint moments were determined through a standard inverse
dynamics approach, expressed relatively to the proximal segment
and normalized to subjects’ body mass.
IAA was performed (Kepple et al., 1997) to determine the con-
tribution of each lower limb joint moment to the vertical acceler-
ation of the COM of the body. Briefly, these contributions are
directly quantified by inputting each joint moment one at a time
in Newton’s equation of motion, setting all the other terms to zero
and solving it in order to obtain its contribution to body’s COM
acceleration. During this process, the foot was fixed to the floor
during foot flat and allowed to rotate about the center of pressure
for the rest of the time (Kepple et al., 2002). The accuracy of the
model was measured for each subject by computing the absolute
differences between the COM acceleration derived from the force
platform and the one induced through the model. The mean error
was 4.8% of the total range of accelerations and less than 8% for all
the subjects in both tasks. The IAA was performed during two
phases of the stride: when the COM acceleration was above zero
(i.e. when the vertical ground reaction force (VGRF) was higher
than body weight) and when the COM acceleration was below zero
(i.e. when VGRF was lower than body weight). Because joint
moments and induced accelerations’ peaks may not represent the
entire phase under analysis and because during these phases it is
not always possible to have a clear curve peak, lower limb joint
angular impulses and change in COM velocity were computed by
performing the time integral of the respective curves, during the
mentioned phases.
Data processing was performed in Visual 3D (Professional
Version v5.02.27, C-Motion, Inc, USA) and waveform plotting in
Microsoft Excel (Microsoft Office Professional Plus 2016, Microsoft
Corporation, USA).
2.5. Statistical analysis
Statistical analysis was performed in IBM SPSS Statistics (ver-
sion 23, IBM, USA).
In order to verify the correlations between the above mentioned
biomechanical variables (i.e. spatial-temporal parameters, joint
angular impulses and changes in COM velocity) with age and func-
tional fitness levels, Spearman correlation coefficient was calcu-
lated and, when significant (p < 0.05), a minimum of 0.50 was
considered to be an acceptable correlation. As joint moments and
their contribution to the COM acceleration may also be related
with speed, correlations with cadence were also computed.
3. Results
Participants of both groups had a mean age of 72 years
(63–81 years), about 60% were women, took approximately
5.0 s (3.4–7.5 s) to complete the UG test and had an average
balance score of 15 points (11–16 points) (Table 1).
Please cite this article in press as: Moniz-Pereira, V., et al. Joint moments’ contributions to vertically accelerate the center of mass during stair ambulation
in the elderly: An induced acceleration approach. J. Biomech. (2018), https://doi.org/10.1016/j.jbiomech.2018.07.040
3
Table 1
Participants characteristics.



Stair ascent X ± r (Xmid) Stair descent X ± r (Xmid)
Age (y) 71.6 ± 4.3 (72.0) 71.7 ± 3.8 (72.0)
Body mass (kg) 69.7 ± 12.8 (69.0) 71.4 ± 13.6 (70.0)
Body height (m) 1.61 ± 0.10 (1.57) 1.61 ± 0.10 (1.59)
Trochanteric height (m) 0.88 ± 0.05 (0.86) 0.88 ± 0.05 (0.86)
BMI (kg/m2) 26.7 ± 3.2 (26.7) 27.4 ± 3.3 (26.8)
UG (sec) 4.91 ± 0.92 (4.97) 4.95 ± 1.03 (4.94)
TB (points - max 16) 14.7 ± 1.5 (15.0) 14.7 ± 1.4 (15.0)
n (%) n (%)
Gender (female) 18 (62) 17 (59)
UG = 8 foot up and go test; TB = total balance score;
During stair ascent, the phase when the COM vertical accelera-
tion is above zero, and the slope of the COM vertical velocity is pos-
itive (VGRF > W), starts just before the foot is placed on the first
step and lasts for about the duration of the double support phase,
for some subjects, or until the end of early stance of the single sup-
port phase, for others (thus including the weight acceptance and
the beginning of the pull up phase - Fig. 1). The highest joint rota-
tional impulse was produced at the ankle of the trailing limb (left),
followed by the leading limb (right) knee, ankle and hip joints
(Table 2). In agreement, the IAA results showed that the highest
contribution to the change in the COM vertical velocity was the
trailing limb (left) ankle joint plantarflexor moment, followed by
the leading limb (right) knee, ankle and hip extensor joint
moments (Fig. 1, Table 2). No acceptable correlations were found
between both joint angular impulses and joint moments’ contribu-
tions and age, functional fitness or cadence.
The phase when the COM vertical acceleration is below zero,
and the slope of the COM vertical velocity is negative
(VGRF < W), was longer, including most of the single support phase
until the next foot placement (thus including the remaining of the
pull up and the forward continuance phase). The highest joint
rotational impulses were produced at the support (right) limb by
the ankle, knee and hip joints (Fig. 1, Table 3). Though the ankle
plantarflexor and knee extensor joint rotational impulses were
similar, the knee joint contribution to the upward COM accelera-
tion was about 1.5 times the ankles’ joint contribution. Correla-
tions were only acceptable between the duration of the phase
and right ankle plantarflexor impulse with cadence, showing that
the higher the cadence is, the briefer this phase is and the lower
is the ankle plantaflexor impulse applied (Table 3).
During stair descent, the phase when the COM vertical acceler-
ation is below zero, and the slope of the COM vertical velocity is
negative (VGRF < W), starts approximately during early swing
and ends at the beginning of the double support phase or, in some
cases, during late swing (thus including the forward continuance
and the controlled lowering phases). During this phase, the support
of the body was mainly performed by the support limb (right)
ankle plantarflexor and knee extensor joint moments (Fig. 1,
Table 4). Interestingly, although sagittal plane hip joint moment’s
contribution to the change in COM vertical velocity was lower than
5% (
0.12 ± 0.20 m/s;
4.1 ± 5.9%), significant negative correla-
tions were found between both the absolute and normalized con-
tribution and the balance score (Rs =
0.50; Rs =
0.51; p < 0.01),
meaning that the subjects with a lower balance level had a higher
contribution of the hip joint moment to support the body.
The downward velocity is then halted (COM vertical accelera-
tion is above zero, VGRF > W) during double support phase until
the early swing (the weight acceptance phase – Fig. 1) firstly by
the plantarflexor joint moment produced at the ankle of the lead-
ing (left) limb, followed by the knee extensor joint moment of the
same limb and the trailing (right) limb ankle and knee extensor
4 V. Moniz-Pereira et al. / Journal of Biomechanics xxx (2018) xxx–xxx

Fig. 1. Mean (all subjects) of center of mass (COM) vertical velocity, lower limb joint moments in the sagittal plane and lower limb joint moments’ contributions to COM
vertical acceleration. Positive values represent upward velocity, plantarflexor joint moment, extensor joint moment and upward acceleration. LHIP, RHIP, GRAV, LKNEE,
RKNEE, LANKLE and RANKLE stands for left hip, right hip, gravity, left knee, right knee, left ankle and right ankle.

joint moments (Table 5). A higher ankle plantarflexor joint
moment contribution during this phase was shown to be signifi-
cantly correlated with a higher cadence. No other acceptable corre-
lations were found.
4. Discussion
To the extent of our knowledge, this is the first study to quan-
tify, using IAA, the contributions of lower limb joint moments to
vertically accelerate the COM during stair ambulation in a sample
of older adults living in the community. We further aimed to verify
if those contributions were correlated with age and typical func-
tional fitness tests used in the field to determine fall risk level
and to monitor the effects of exercise interventions on gait and bal-
ance performance. By contributing for a better understanding of
stair ambulation movement patterns in community dwelling older
adults, this study may help to support the design of interventions
aiming at fall prevention and/or mobility enhancement within this
population.
The results showed that in stair ascent, during the weight
acceptance and the beginning of the pull up phase, the support
of the body was mainly produced at the trailing limb plantarflexor
joint moment, followed by the leading limb knee, hip and ankle
extensor and plantarflexor joint moments. Moreover, the knee
extensor, ankle plantarflexor and hip extensor joint moments pro-
duced in the support limb, supported the body during the rest of
the stance. These results are in agreement with the results reported
for healthy adults in both inverse dynamics (McFadyen and
Winter, 1988) and induced acceleration studies (Harper et al.,
2018; Lin et al., 2015) which show that the hip and knee extensors
have a higher contribution to support the body at early to mid
stance, while the ankle plantarflexors contribution is higher at late
stance. Contrary to our hypothesis, no acceptable correlations were
found between joint moments (and respective contributions to
vertically accelerate the COM) and both age and functional fitness

Please cite this article in press as: Moniz-Pereira, V., et al. Joint moments’ contributions to vertically accelerate the center of mass during stair ambulation
in the elderly: An induced acceleration approach. J. Biomech. (2018), https://doi.org/10.1016/j.jbiomech.2018.07.040
 V. Moniz-Pereira et al. / Journal of Biomechanics xxx (2018) xxx–xxx 5

Table 2
Mean ± sd (Median) for: spatial-temporal parameters, lower limb joint rotational impulses (sagittal plane) and respective contributions (5%) when the COM acceleration was
above zero (VGRF > W) during stair ascent. Correlations (Spearman’s rank correlation coefficients) with age, functional fitness. VGRF > W duration was normalized to induced
acceleration interval duration and joint moments’ contributions were normalized to total change in velocity.


Stair ascent (n = 29) X ± r (Xmid) AGE UG TB CAD

Spatial-temporal parameters
VGRF > W Duration (s) 0.31 ± 0.07 (0.31) 0.32 0.03 0.15
0.36
Normalized VGRF > W duration (%) 41.9 ± 8.4 (40.4) 0.13 0.07 0.20 0.14
Cadence (steps/min) 79.8 ± 10.8 (80.0)
0.05 0.15 0.07
Joint rotational impulses (VGRF > W)
Right hip (Nms/kg) 0.08 ± 0.04 (0.07) 0.46* 0.39*
0.31 0.31
Right knee (Nms/kg) 0.14 ± 0.07 (0.12) 0.23 0.08 0.03
0.13
Left ankle (Nms/kg) 0.24 ± 0.04 (0.24) 0.02
0.32 0.45*
0.37*
Right ankle (Nms/kg) 0.09 ± 0.05 (0.08) 0.14
0.02 0.09 0.13
Joint moments’ contributions (VGRF > W)
Gravity contribution (m/s) 0.41 ± 0.25 (0.36) 0.13
0.18
0.22
0.37*
Normalized gravity contribution (%) 11.4 ± 5.5 (10.8)
0.01
0.25
0.27
0.29
Right hip contribution (m/s) 0.21 ± 0.10 (0.20) 0.39* 0.41*
0.20 0.15
Normalized right hip contribution (%) 6.0 ± 2.5 (6.1) 0.29 0.48y
0.34 0.31
Right knee contribution (m/s) 0.74 ± 0.37 (0.56) 0.21 0.11 0.07
0.04
Normalized right knee contribution (%) 19.8 ± 6.6 (18.3) 0.16 0.14
0.04 0.09
Left ankle contribution (m/s) 1.21 ± 0.86 (1.61) 0.11
0.01
0.11
0.17
Normalized left ankle contribution (%) 35.5 ± 25.5 (46.3)
0.08
0.02
0.20
0.06
Right ankle contribution (m/s) 0.22 ± 0.14 (0.19) 0.09 0.02 0.10 0.20
Normalized right ankle contribution (%) 6.1 ± 3.6 (6.3) 0.05 0.08 0.06 0.40*

VGRF = Vertical Ground Reaction Force; W = Weight; UG = 8 foot up and go test; TB = total balance score; CAD = cadence.
*
p < 0.05.
y
p < 0.01.

Table 3
Mean ± sd (Median) for: spatial-temporal parameters, lower limb joint rotational impulses (sagittal plane) and respective contributions (5%) when the COM acceleration was
below zero (VGRF < W) during stair ascent. Correlations (Spearman’s rank correlation coefficients) with age, functional fitness. VGRF < W duration was normalized to induced
acceleration interval duration and joint moments’ contributions were normalized to total change in velocity.


Stair ascent (n = 29) X ± r (Xmid) AGE UG TB CAD

Spatial-temporal parameters
VGRF < W Duration (s) 0.43 ± 0.09 (0.44)
0.01
0.23
0.04
0.62y
Normalized VGRF < W duration (%) 58.1 ± 8.4 (59.6)
0.13
0.07
0.20
0.14
Cadence (steps/min) 79.8 ± 10.8 (80.0)
0.05 0.15 0.07
Joint rotational impulses (VGRF < W)
Right hip (Nms/kg) 0.09 ± 0.07 (0.10) 0.04 0.10
0.13
0.09
Right knee (Nms/kg) 0.21 ± 0.10 (0.18) 0.21
0.25
0.10
0.37
Right ankle (Nms/kg) 0.23 ± 0.06 (0.22)
0.08
0.42* 0.08
0.51y
Joint moments’ contributions (VGRF < W)
Gravity contribution (m/s) 0.86 ± 0.39 (0.77)
0.27
0.08 0.01
0.48y
Normalized gravity contribution (%) 23.0 ± 7.6 (21.8)
0.38* 0.13 0.02 0.05
Right hip contribution (m/s) 0.36 ± 0.31 (0.38) 0.03 0.19
0.10
0.09
Normalized right hip contribution (%) 9.6 ± 8.0 (10.9)
0.01 0.32
0.12 0.16
Right knee contribution (m/s) 1.53 ± 0.68 (1.44) 0.20
0.24
0.14
0.45*
Normalized right knee contribution (%) 40.5 ± 12.1 (40.1) 0.29
0.24
0.06
0.16
Right ankle contribution (m/s) 0.96 ± 0.21 (1.01)
0.10
0.27 0.12
0.45*
Normalized right ankle contribution (%) 27.1 ± 7.7 (26.3)
0.09 0.14 0.04 0.25

VGRF = Vertical Ground Reaction Force; W = Weight; UG = 8 foot up and go test; TB = total balance score; CAD = cadence.
*
p < 0.05.
y
p < 0.01.

test results. These findings contrast with recent level and incline
gait studies showing that both age and physical capacity mediate
the redistribution of joint moments (Hortobágyi et al., 2016;
Kuhman et al., 2018). However, besides focusing on different loco-
motor tasks, the mentioned studies used other measures of func-
tional fitness and/or considered different levels of statistical
significance in order to interpret the results. Also, the lack of
strength in our correlations, despite of guaranteeing the minimal
sample size and power, may be due to the high functional fitness
level of the participants, who scored above the average of previous
population reports (Marques et al., 2014; Moniz-Pereira et al.,
2013).
During the forward continuance and controlled lowering phases
in stair descent, the support limb knee extensors, followed by the
ankle plantarflexors, were the mainly contributors for the support.
The braking of this downward velocity was then performed by the
ankle plantarflexors followed by knee extensor joint moments
from both limbs, but especially, from the leading limb. These
results are also in accordance with previous inverse dynamics
(McFadyen and Winter, 1988) and induced acceleration (Lin
et al., 2015) studies in healthy adults. However, Lin et al. (2015)
showed a higher contribution of ankle plantarflexors, namely
soleus, when compared with knee extensors, to support the body
throughout all stance. These differences may be due to the differ-
ences in the stair cycle analyzed or due to the age difference
between samples, as previous results (Karamanidis and
Arampatzis, 2011) showed that older adults prepare the double
support phase more with their knee extensors than their ankle

Please cite this article in press as: Moniz-Pereira, V., et al. Joint moments’ contributions to vertically accelerate the center of mass during stair ambulation
in the elderly: An induced acceleration approach. J. Biomech. (2018), https://doi.org/10.1016/j.jbiomech.2018.07.040
6 V. Moniz-Pereira et al. / Journal of Biomechanics xxx (2018) xxx–xxx

Table 4
Mean ± sd (Median) for: spatial-temporal parameters, lower limb joint rotational impulses (sagittal plane) and respective contributions (5%) when the COM acceleration was
below zero (VGRF < W) during stair descent. Correlations (Spearman’s rank correlation coefficients) with age, functional fitness. VGRF < W duration was normalized to induced
acceleration interval duration and joint moments’ contributions were normalized to total change in velocity.


Stair descent (n = 29) X ± r (Xmid) AGE UG TB CAD

Spatial-temporal parameters
VGRF < W duration (s) 0.37 ± 0.10 (0.36) 0.22
0.03 0.12
0.54y
Normalized VGRF < W duration (%) 56.4 ± 8.6 (57.0) 0.15
0.02 0.06
0.30
Cadence (steps/min) 93.2 ± 13.6 (94.5)
0.19 0.02 0.02
Joint rotational impulses (VGRF < W)
Right hip (Nms/kg)
0.05 ± 0.07 (
0.05) 0.13 0.35
0.46*
0.31
Right knee (Nms/kg) 0.26 ± 0.11 (0.25) 0.09
0.05
0.06
0.45*
Right ankle (Nms/kg) 0.29 ± 0.08 (0.30) 0.27
0.01
0.14
0.45*
Joint moments’ contributions (VGRF < W)
Gravity contribution (m/s) 0.43 ± 0.22 (0.39) 0.07
0.25 0.28
0.58y
Normalized gravity contribution (%) 13.2 ± 4.1 (13.1)
0.08
0.25 0.31
0.17
Right knee contribution (m/s) 1.55 ± 0.71 (1.39) 0.03
0.21 0.19
0.40*
Normalized right knee contribution (%) 48.4 ± 13.7 (46.2)
0.07
0.23 0.28
0.07
Right ankle contribution (m/s) 1.41 ± 0.54 (1.43) 0.13 0.01
0.06
0.21
Normalized right ankle contribution (%) 44.6 ± 11.9 (46.4)
0.06 0.12
0.18 0.30

VGRF = Vertical Ground Reaction Force; W = Weight; UG = 8 foot up and go test; TB = total balance score; CAD = cadence.
*
p < 0.05.
y
p < 0.01.

Table 5
Mean ± sd (Median) for: spatial-temporal parameters, lower limb joint rotational impulses (sagittal plane) and respective contributions (5%) when the COM acceleration was
above zero (VGRF > W) during stair descent. Correlations (Spearman’s rank correlation coefficients) with age, functional fitness. VGRF > W duration was normalized to induced
acceleration interval duration and joint moments’ contributions were normalized to total change in velocity.


Stair descent (n = 29) X ± r (Xmid) AGE UG TB CAD

Spatial-temporal parameters
VGRF > W duration (s) 0.28 ± 0.06 (0.27) 0.07
0.06 0.14
0.41*
Normalized VGRF > W duration (%) 43.6 ± 8.6 (43.0)
0.15 0.02
0.06 0.30
Cadence (steps/min) 93.2 ± 13.6 (94.5)
0.19 0.02 0.02
Joint rotational impulses (VGRF > W)
Left knee (Nms/kg) 0.07 ± 0.05 (0.07) 0.08 <0.01
0.04
0.04
Right knee (Nms/kg) 0.11 ± 0.07 (0.09) 0.21
0.01
0.06
0.33
Left ankle (Nms/kg) 0.22 ± 0.04 (0.22)
0.21
0.10 0.30 0.31
Right ankle (Nms/kg) 0.11 ± 0.07 (0.10) 0.19 0.01
0.03
0.38*
Joint moments’ contributions (VGRF > W)
Gravity contribution (m/s) 0.34 ± 0.15 (0.36)
0.13
0.29 0.12 0.04
Normalized gravity contribution (%) 10.6 ± 4.6 (10.8)
0.15
0.27 0.02 0.29
Left knee contribution (m/s) 0.57 ± 0.40 (0.54)
0.05
0.09 <
0.01 0.11
Normalized left knee contribution (%) 18.2 ± 12.9 (17.7)
0.12
0.15
0.02 0.17
Right knee contribution (m/s) 0.28 ± 0.22 (0.20) 0.27 0.06
0.06
0.35
Normalized right knee contribution (%) 8.1 ± 5.1 (6.8) 0.29 0.13
0.12
0.32
Left ankle contribution (m/s) 1.15 ± 0.55 (1.21)
0.21
0.04 0.08 0.35
Normalized left ankle contribution (%) 36.7 ± 18.7 (35.7)
0.22
0.03
0.03 0.56y
Right ankle contribution (m/s) 0.54 ± 0.34 (0.46)
0.15 0.11
0.03
0.30
Normalized right ankle contribution (%) 15.8 ± 7.8 (15.1) 0.10 0.13
0.09
0.27

VGRF = Vertical Ground Reaction Force; W = Weight; UG = 8 foot up and go test; TB = total balance score; CAD = cadence.
*
p < 0.05.
y
p < 0.01.

plantarflexors while stepping down. Also according with the previ-
ous mentioned studies (Lin et al., 2015; McFadyen and Winter,
1988), the hip musculature had a small (<5%) contribution to sup-
port the body in stair descent. However, this contribution was cor-
related with the balance score. The participants who had poorer
balance used more the hip extensors to brake the downwards
COM velocity during the weight acceptance phase.
A limitation of this study is that the analysis focused on data
from the ground and the first step of a 3-step staircase, which
may not accurately represent a steady state movement pattern
(Cluff and Robertson, 2011; Whatling and Holt, 2010). Neverthe-
less, the fact that our joint moments’ results are consistent with
previous studies (Novak and Brouwer, 2011; Reeves et al., 2009,
2008) give us confidence about our findings.
In summary, this study directly quantified the lower limb joint
moments’ contributions to vertically accelerate the COM during
stair ambulation and verified the correlations of those contribu-
tions with age and functional fitness in a group of community
dwelling older adults. During both stair ascent and descent tasks,
ankle plantarflexor and knee extensor joint moments were the
main contributors to support the body. During stair ascent, no
acceptable correlations were found between both joint angular
impulses and induced acceleration contributions and age or func-
tional fitness. Although the hip joint extensor moment showed a
small contribution to vertically accelerate the body, during stair
descent, this contribution was negatively related with balance
level. Also, comparing with previous reports in healthy adults
(Lin et al., 2015), older adults seem to use more their knee extensor
moment than the ankle plantarflexor moment to brake the COM
downward velocity during this task. Future studies, including older
adults with a poorer functional fitness and considering joint
moments’ contributions to anterior-posterior and mediolateral

Please cite this article in press as: Moniz-Pereira, V., et al. Joint moments’ contributions to vertically accelerate the center of mass during stair ambulation
in the elderly: An induced acceleration approach. J. Biomech. (2018), https://doi.org/10.1016/j.jbiomech.2018.07.040
 V. Moniz-Pereira et al. / Journal of Biomechanics xxx (2018) xxx–xxx
COM accelerations, will allow a better understanding of the influ-
ence of functional fitness parameters on balance and locomotion.
Conflicts of interest
Thomas M. Kepple is the Chief Science Officer at C-Motion (the
company that commercializes the motion analysis software Visual
3D). His role at C-Motion had no impact on the current results.
There are no other known conflicts of interest associated with this
publication.
Acknowledgements
The authors are grateful to all the subjects who volunteered to par-
ticipate in this study.
This work was supported by the Portuguese Foundation for
Science and Technology (project number: PTDC/DES/72946/2006;
and PhD Grant number: SFRH / BD / 36670 / 2007). The funding
source of the study had no involvement in the study design, data
collection, analysis and interpretation, or in the preparation and
submission of this manuscript.
References
Alcock, L., O’Brien, T.D., Vanicek, N., 2014. Biomechanical demands differentiate
transitioning vs. continuous stair ascent gait in older women. Clin. Biomech. 29,
111–118. https://doi.org/10.1016/j.clinbiomech.2013.10.007.
Ambrose, A.F., Paul, G., Hausdorff, J.M., 2013. Risk factors for falls among older
adults: a review of the literature. Maturitas 75, 51–61. https://doi.org/10.1016/
j.maturitas.2013.02.009.
Bell, A.L., Pedersen, D.R., Brand, R.A., 1990. A comparison of the accuracy of several
hip center location prediction methods. J. Biomech. 23, 617–621.
Cappozzo, A., Catani, F., Croce, U. Della, Leardini, A., 1995. Position and orientation
in space of bones during movement: anatomical frame definition and
determination. Clin. Biomech. 10, 171–178.
Chesters, T., Alcock, L., O’Brien, T.D., Vanicek, N., Dobson, C.A., 2014. Frequency
domain characteristics of kinetic data from force platforms mounted in
instrumented stairways. Gait Post. 39, S118–S119. https://doi.org/10.1016/
j.gaitpost.2014.04.164.
Cluff, T., Robertson, D.G.E., 2011. Kinetic analysis of stair descent: Part 1. Forwards
step-over-step descent. Gait Post. 33, 423–428. https://doi.org/10.1016/
j.gaitpost.2010.12.016.
Cohen, J., 1988. Statistical Power Analysis for the Behavioral Sciences. Psychology
Press - Taylor and Francis Group.
Deandrea, S., Lucenteforte, E., Bravi, F., Foschi, R., La Vecchia, C., Negri, E., 2010. Risk
factors for falls in community-dwelling older people: a systematic review and
meta-analysis. Epidemiology 21, 658–668. https://doi.org/10.1097/
EDE.0b013e3181e89905.
Dempster, W.T., 1955. Space Requirements of the Seated Operator (WADC Technical
report). Wright-Patterson Air Force Base, Ohio.
Hanavan, E.P., 1964. A Mathematical Model of the Human Body. AMRL-TR-64-102.
AMRL. TR. 1–149.
Harper, N., Wilken, J., Neptune, R., 2018. Muscle function and coordination of stair
ascent. J. Biomech. Eng. 140, 1–11. https://doi.org/10.1115/1.4037791.
Hernandez, D., Rose, D.J., 2008. Predicting which older adults will or will not fall
using the Fullerton Advanced Balance scale. Arch. Phys. Med. Rehabil. 89, 2309–
2315. https://doi.org/10.1016/j.apmr.2008.05.020.
Hortobágyi, T., Rider, P., Gruber, A.H., DeVita, P., 2016. Age and muscle strength
mediate the age-related biomechanical plasticity of gait. Eur. J. Appl. Physiol.
116, 805–814. https://doi.org/10.1007/s00421-015-3312-8.
Jacobs, J.V., 2016. A review of stairway falls and stair negotiation: Lessons learned
and future needs to reduce injury. Gait Post. 49, 159–167. https://doi.org/
10.1016/j.gaitpost.2016.06.030.
Please cite this article in press as: Moniz-Pereira, V., et al. Joint moments’ contributions to vertically accelerate the center of mass during stair ambulation
in the elderly: An induced acceleration approach. J. Biomech. (2018), https://doi.org/10.1016/j.jbiomech.2018.07.040
7
Karamanidis, K., Arampatzis, A., 2011. Altered control strategy between leading and
trailing leg increases knee adduction moment in the elderly while descending
stairs. J. Biomech. 44, 706–711. https://doi.org/10.1016/j.jbiomech.2010.10.040.
Karamanidis, K., Arampatzis, A., 2009. Evidence of mechanical load redistribution at
the knee joint in the elderly when ascending stairs and ramps. Ann. Biomed.
Eng. 37, 467–476. https://doi.org/10.1007/s10439-008-9624-7.
Kepple, T.M., Siegel, K.L., Stanhope, S.J., 2002. Comparison of two foot - floor
interfaces in induced acceleration analysis. Gait Posture 16 (Suppl.), S67–S68.
Kepple, T.M., Siegel, K.L., Stanhope, S.J., 1997. Relative contributions of the lower
extremity joint moments to forward progression and support during gait. Gait
Post. 6, 1–8. https://doi.org/10.1016/S0966-6362(96)01094-6.
Knudson, D., 2017. Confidence crisis of results in biomechanics research. Sport.
Biomech. 16, 425–433. https://doi.org/10.1080/14763141.2016.1246603.
Kuhman, D., Willson, J., Mizelle, J.C., DeVita, P., 2018. The relationships between
physical capacity and biomechanical plasticity in old adults during level and
incline walking. J. Biomech. 69, 90–96. https://doi.org/10.1016/j.
jbiomech.2018.01.006.
Lin, Y., Fok, L.A., Schache, A.G., Pandy, M.G., 2015. Muscle coordination of support,
progression and balance during stair ambulation. J. Biomech. 48, 340–347.
https://doi.org/10.1016/j.jbiomech.2014.11.019.
Lord, S.R., Sherrington, C., Menz, H.B., Close, J., 2007. Falls in Older People Risks
Factors and Strategies for Prevention. Cambridge University Press.
Lu, T.W., O’Connor, J.J., 1999. Bone position estimation from skin marker co-
ordinates using global optimisation with joint constraints. J. Biomech. 32, 129–
134.
Marques, E.A., Baptista, F., Santos, R., Vale, S., Santos, D.A., Silva, A.M., Mota, J.,
Sardinha, L.B., 2014. Normative functional fitness standards and trends of
Portuguese older adults: cross-cultural comparisons. J. Aging Phys. Act. 22, 126–
137. https://doi.org/10.1123/japa.2012-0203.
McFadyen, B., Winter, D., 1988. An integrated biomechanical analysis of normal
stair ascent and descent. J. Biomech. 21, 733–744.
Moniz-Pereira, V., Carnide, F., Machado, M., Andre, H., Veloso, A.P., 2012. Falls in
Portuguese older people: procedures and preliminary results of the study
Biomechanics of Locomotion in the Elderly. Acta Reum. Port. 37, 324–332.
Moniz-Pereira, V., Carnide, F., Ramalho, F., André, H., Machado, M., Santos-Rocha, R.,
Veloso, A.P., 2013. Using a multifactorial approach to determine fall risk profiles
in portuguese older adults. Acta Reum. Port. 38, 263–272.
Novak, A.C., Brouwer, B., 2011. Sagittal and frontal lower limb joint moments during
stair ascent and descent in young and older adults. Gait Post. 33, 54–60. https://
doi.org/10.1016/j.gaitpost.2010.09.024.
Reeves, N.D., Spanjaard, M., Mohagheghi, A.A., Baltzopoulos, V., Maganaris, C.N.,
2009. Older adults employ alternative strategies to operate within their
maximum capabilities when ascending stairs. J. Electromyogr. Kinesiol. 19,
e57–e68. https://doi.org/10.1016/j.jelekin.2007.09.009.
Reeves, N.D., Spanjaard, M., Mohagheghi, A.A., Baltzopoulos, V., Maganaris, C.N.,
2008. The demands of stair descent relative to maximum capacities in elderly
and young adults. J. Electromyogr. Kinesiol. 18, 218–227. https://doi.org/
10.1016/j.jelekin.2007.06.003.
Rikli, Roberta E., Jones, C.J., 1999. Development and validation of a functional fitness
test for community-residing older adults. J. Aging Phys. Act. 7, 129–161.
Robertson, D.G.E., Caldwell, G.E., Hamill, J., Kamen, G., Whittlesey, S.N., 2014.
Research Methods in Biomechanics. Human Kinetics.
Rose, D., Jones, C., Lucchese, N., 2002. Predicting the probability of falls in
community-residing older adults using the 8-foot up-and-go: a new measure
of functional mobility. J. Aging Phys. Act. 10, 466–475.
Rose, D.J., Lucchese, N., Wiersma, L.D., 2006. Development of a multidimensional
balance scale for use with functionally independent older adults. Arch. Phys.
Med. Rehabil. 87, 1478–1485. https://doi.org/10.1016/j.apmr.2006.07.263.
Timsina, L.R., Willetts, J.L., Brennan, M.J., Marucci-Wellman, H., Lombardi, D.A.,
Courtney, T.K., Verma, S.K., 2017. Circumstances of fall-related injuries by age
and gender among community-dwelling adults in the United States. PLoS One
12, e0176561. https://doi.org/10.1371/journal.pone.0176561.
Whatling, G.M., Holt, C.A., 2010. Does the choice of stair gait cycle affect resulting
knee joint kinematics and moments? Proc. Inst. Mech. Eng. Part H J. Eng. Med.
224, 1085–1093. https://doi.org/10.1243/09544119JEIM724.
World Health Organization. (2007). WHO Global Report on Falls Prevention in Older
Age. World Health Organization. Retrieved from http://www.who.int/ageing/
publications/Falls_prevention7March.pdf.
Zajac, F.E., Gordon, M.E., 1989. Determining muscle’s force and action in multi-
articular movement. Sport Sci. Rev Exerc. https://doi.org/10.1249/00003677-
198900170-00009.