Clinical Biomechanics 18 (2003) 745750 www.elsevier.

com/locate/clinbiomech

Biomechanical analysis of sit-to-stand movement in normal and obese subjects

F. Sibella
a

a,*

, M. Galli a, M. Romei a, A. Montesano b, M. Crivellini

Department of Bioengineering, Politecnico di Milano, p.zza Leonardo da Vinci 32, 20133 Milano, Italy b Ospedale S. Giuseppe, Piancavallo, Verbania, Italy Received 16 January 2003; accepted 17 June 2003

Abstract Objective. Main purpose of this study was to develop a biomechanical model for the analysis of sit-to-stand movement in normal and obese subjects. Design. A biomechanical model describing sit-to-stand was developed using kinetic and kinematic experimental data. Trunk exion, feet movement, knee and hip joint torques were assumed as sensible indexes to discriminate between normal and obese subjects. Background. Sit-to-stand is a functional task that may become dicult for certain patients. The analysis of its execution provides useful biomechanical information on the motor ability of selected subjects. Methods. Sit-to-stand was recorded using an optoelectronic system and a force platform in 40 obese patients and 10 normal subjects. A biomechanical model was developed using inverse dynamics equations. Results. Kinematic and kinetic indexes evidenced dierences in motion strategy between normal and obese subjects. Obese subjects rise from the chair limiting trunk exion (mean value: 73.1) and moving their feet backwards from initial position (mean deviation: 50 mm). Normal subjects, instead, show a higher trunk exion (mean value: 49.2, a lower angular value between trunk and the horizontal means increased exion) and xed feet position (mean deviation: 5 mm). As for kinetics, obese patients show knee joint torque higher than hip torque (maximum knee torque: 0.75 Nm/kg; maximum hip torque: 0.59 Nm/kg), while normal subjects show opposite behaviour (maximum knee torque: 0.38 Nm/kg; maximum hip torque: 0.98 Nm/kg). Relevance We found dierences in motion strategy between normal and obese subjects performing sit-to-stand movement, which may be used to plan and evaluate rehabilitative treatments. 2003 Elsevier Ltd. All rights reserved.
Keywords: Obesity; Sit-to-stand; Biomechanics; Optoelectronic system; Force platform; Rehabilitation

1. Introduction Obesity is one of the commonest pathologies in industrialised countries. Many clinical studies (Tagliaferri et al., 1998; Brozek et al., 1993) investigated its consequences, but, from a biomechanical point of view, research is still poor and no biomechanical models of specic movements were found in a literature review. Sit-to-stand movement (STS) is an important functional task that may become dicult for certain patients. In particular, rising from a chair requires adequate torques
*

Corresponding author. E-mail address: sibella@biomed.polimi.it (F. Sibella).

to be developed about each body joint. The accurate analysis of the requirements for STS execution will provide useful biomechanical information on the motor ability of selected patients. In fact, STS has been a topic for many studies (Coglin et al., 1994; Doorenbosch, 1994; Kathleen, 1991), but it has never been analysed in obese patients using a biomechanical model. The main aim of this study was to develop a biomechanical model for a quantitative description of STS motion strategy in normal and obese subjects. In fact, because of evident dierences in mass distribution between these two populations, correspondent dierences in STS motion strategy are expected both for kinematics and kinetics aspects.

0268-0033/$ - see front matter 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0268-0033(03)00144-X

746

F. Sibella et al. / Clinical Biomechanics 18 (2003) 745750

2. Methods Ten volunteers subjects [seven men and three women; mean age (SD): 26.5 years (2.5)] were enrolled as control group (CG). Forty obese subjects [all women mean age (SD): 48.5 years (13.5)], recovered at the Auxologico hospital, (Piancavallo, Verbania, Italy), were selected for the study. The hospital ethical committee gave its approval for the study. Inclusion criteria for normal subjects were: no musculoskeletal pathologies, body mass index (BMI; calculated as weight [kg]/height2 [m2 ]) <25 kg/m2 , no previous skeletal fractures, no low back pain. Mean BMI (SD) of CG group was: 23 kg/m2 (2.2). Inclusion criteria for obese subjects were: BMI > 30 kg/m2 , able to stand from an armless chair, patients scheduled for a rehabilitative treatment. Mean BMI (SD) of the obese group was: 37.9 kg/m2 (4.9). Rehabilitative treatment included: low-fat diet, gymnastic sessions for 2 h twice a week, a 5 km walk each day, massages and stretching lessons. The treatment was 2 week long. All subjects were enrolled before the treatment. A motion optoelectronic measurement system (E L I T E , B.T.S. S.p.A., Milano, Italy), 6 TVC, acquisition frequency set at 50 Hz, provided the three-dimensional coordinates of reective passive markers; a force platform (A M T I , Newton, MA, USA), acquisition frequency set at 500 Hz provided the ground reaction forces. 2.1. Experimental setup Sixteen reective passive markers (diameter: 15 mm) were positioned on bony landmarks as follows: over the spinosus process of C7, over the sacrum midway between the posterior superior iliac spines and, bilaterally, over the acromion, the anterior superior iliac spine, the great trochanter, the femoral condyle, the lateral malleolus, the fth metatarsal head and the heel. An oce chair, armless and backless, was adjusted vertically for each subject to obtain the same knee exion angle (xed at 90). A reference marker was placed on the chair. Each subject was asked to stand up at self-selected speed with arms crossed over the trunk and with the feet selfpositioned over the force platform (no xed distance between the feet was imposed) for 10 acquisition trials. Interval time between each trial was xed at 30 s. 2.2. Kinematic analysis STS strategy was rst analysed taking into account trunk movement: the vertical coordinates of shoulder markers (Y coordinate) were extracted and trunk angle a dened as the angle between the horizontal line and the segment representing the trunk was calculated (Fig. 1). Asymmetry in feet movement was investigated: the

Fig. 1. Biomechanical model of STS movement (four segments).

horizontal coordinates X of markers positioned over both lateral malleola were analysed and the dierence between left and right side anteriorposterior absolute displacement (diff jX left X rightj mm) was calculated. Since marker diameter is 15 mm, a dierence less than 7.5 mm was considered negligible and the movement symmetric. A comparison between the rst and last trial was made for each subject to nd dierences (if existing) due to repeated movements. 2.3. Biomechanical model STS movement was assumed to be a symmetric movement in frontal and horizontal plane as found in literature (Ludin et al., 1995) and this hypothesis was veried through the symmetry analysis of feet movement. A simplied biomechanical model of the human body was developed to investigate dynamic interactions at knee and hip joint considering only the sagittal plane. The model is based on a 4-segment representation of the body. Each segment is modelled as a rigid non-extensible one-dimensional segment and it is linked with the next through an ideal joint (Fig. 1). The rst segment represents the head, trunk and upper limbs (HAT), the second represents the thighs (THIGH), the third represents the legs (LEG) and the fourth represents the feet (FOOT). The developed model was implemented using Matlab. It was used rst to calculate torques at hip Mhip , knee Mknee and ankle Mankle joints for normal subjects, then it was rened to correctly explain the movement strategy of obese population. The fat mass of obese subjects was

F. Sibella et al. / Clinical Biomechanics 18 (2003) 745750

747

modelled through a hemisphere positioned on the abdomen. The hemisphere was supposed to be homogeneous and fat density value was set at 0.99 g/cm3 . An ideal BMI value was xed at 23 kg/m2 (this value coincides with the mean value of CG); from this value an ideal weight for each obese subject was calculated as follows: 23 ideal weight [kg]/(eective height2 ) [m2 ] and the dierence between the eective weight of each obese patient and her ideal weight represented the fat mass that had to be modelled through the hemisphere. Since all our obese patients are women, the fat mass was distributed 40% on the abdomen and 60% on the thighs (pear shape distribution); therefore only 40% of the calculated fat mass was considered to calculate the hemisphere dimension. From mass and density values, the radius R of the hemisphere for each obese patient was calculated as follows: R (mass (g)/density [g/cm3 ])1=3 and from R value, the position of the fat centre of mass Gs with respect to hip joint was extracted. Then, inertial moments with respect to the axial centre of mass, which is perpendicular to the sagittal plane, were evaluated to write the equilibrium equations of the model. Anthropometric parameters came from the measures obtained by Winter (1979). Each segment length was calculated using the distance between the sagittal projections of the correspondent markers coordinates. The origin of the absolute reference system was positioned on the left malleolus, horizontal and vertical components of HATsegment centre of mass acceleration were calculated using the following equations (Fig. 1): XG1 L3 cos c L2 cos b d1 cos a YG1 L3 sin c L2 sin b d1 sin a By deriving G aG x X 1 1 _2 _ 2 L3 sin cc L2 cos bb L3 cos cc d1 cos aa _ 2 d1 sin aa L2 sin bb G aG y Y 1 1 L3 sin cc _ 2 L2 cos bb L3 cos cc _ 2 d1 cos aa d1 sin aa _2 L2 sin bb 4 3 1 2

Fourth segmentFOOT (Fig. 2): Mankle D G4 KP4 D P KFy 0 8

where Mankle unknown moment; P4 weight vector mfoot g, D G4 G4 centre of mass position with respect to D, D P P ground reaction position with respect to D, Fy vertical component of ground reaction force, inertial vector not signicant. Third segmentLEG (Fig. 2): Mknee Mankle D G3 KP3 D C KRLEGx RLEGy 0 9

where Mknee unknown knee moment, Mankle ankle moment, P3 weight vector of the third segment mleg g, D G3 G3 centre of mass position with respect to D, D C C position with respect to D, RLEGx and RLEGy horizontal and vertical component of the reaction force on C , inertial vector not signicant.

where Li ith-segment length; d1 distance between G1 and B (Fig. 1). For each segment, dynamic equilibrium equations along the horizontal and vertical axes and torques equilibrium equation were written F x m ax F y m ay _ c Mc C 5 6 7
Fig. 2. Foot, shank and HAT model.

Only the equations used to obtain joint moments are reported in the following:

748

F. Sibella et al. / Clinical Biomechanics 18 (2003) 745750

First segmentHAT (Fig. 2):

0 Mhip B G1 KP1 B G1 KF1x F1y JG1 a 10

calculated in the rst and last trial of the experimental session for each group. Signicance level was set at P 0:05. 3. Results 3.1. Symmetry analysis Dierences between left and right side anteriorposterior displacement of malleolus markers were calculated for each analysed subject of both groups. For all subjects we found a dierence <7 mm [mean di (SD): 5.3 mm (1.4)]. 3.2. Kinematic results Control group of healthy subjects. 90% of the analysed CG subjects use a rising strategy characterised by a high degree of trunk exion and a feet movement near to zero mm (see Table 1 for numerical data). In Fig. 3 the Y coordinate of shoulder marker is plotted vs movement cycle to give a qualitative idea of the movement strategy. Comparison between the rst and the tenth trial shows a slight decrease of trunk exion (that means an increased angle value), but again no feet movement in 100% of the analysed cases. No dierences between men and women were found during this analysis. Obese subjects. 100% of the analysed subjects limit their trunk exion. This strategy (In Fig. 4 the Y coordinate of shoulder marker is plotted) results dierent from that adopted by CG subjects (Fig. 3). A feet movement backwards from the initial position is always to observe (Fig. 4). The comparison between rst and last trial does not evidence any change in motion strategy during the experimental session. All data relative to kinematic analysis of both groups are summarised in Table 1. t-Test performed on CG vs obese group for all kinematic parameters conrmed statistical signicance of the results P < 0:05. 3.3. Kinetic results The analysis was mainly focused on the torques generated at lower limb joints. Particular attention was given to knee and hip joint moments.

where Mhip unknown hip moment, P1 weight vector mHAT g mtrunk mhead mupperlimbs g, B G1 centre of mass G1 position with respect to B, F1x and F1y horizontal and vertical component of the inertial vector m1 aG1 , JG1 centre of mass inertial moment with respect to the perpendicular from the sagittal plane. First segment obeseHAT modied adding the virtual belly Mhip B G1 KP1 B G1 KF1x F1y JG1 a B Gs KPGs B Gs KF1sx F1sy 0 JGs a 11

where (B Gs ) centre of mass Gs position (of the virtual belly) with respect to B, PGs weight vector of the fat mass mfat mass g, F1sx and F1sy inertial vector horizontal and vertical component ms aGs [fat mass], JGs centre of mass inertial moment with respect to the perpendicular from the sagittal plane [fat mass]. All calculated joint moments were normalised dividing them by height and weight of each subject to permit inter-subject comparisons and they were analysed vs movement cycle. Movement cycle was dened using a angle according to Millington et al. (1992). In detail, motion cycle begins when a decreases of 0.5 in 20 ms and it ends when the variations of a value remain within 0.5 and the analysed subject is in standing position. Standing position was evaluated considering the vertical coordinate of shoulder markers. Knee and hip joint moments were plotted in graphs and their maximum values were extracted for the rst and last trial of each subject. Two more indexes representing fatigue M % and work P % of each joint were calculated (Nigg and Herzog, 1995) R Mi 2 dt motioncycle Mi % R 100 12 P3 2 k 1 Mk dt motioncycle R 2 P dt motioncycle i Pi % R 100 13 P3 2 k 1 Pk dt motioncycle where Pi Mi wi , Pi muscular power in the ith joint, Mi moment at the ith joint, wi angular velocity for the ith joint. 2.4. Statistical analysis Statistical analysis of the results was performed using Students t-test between CG and obese group parameters and considering separately the rst and last trial of STS. The same test was also used to compare the parameters

Table 1 Kinematic results: a angle, X coordinate deviation of malleola markers: mean (SD) Mean a () Control group (1st trial) Control group (10th trial) Obese group (1st trial) Obese group (10th trial) 49.2 56.1 73.1 75.3 (0.5) (0.1) (1.2) (2.0) X coord. deviation of the malleola (mm) 5 (1.7) 5 (1.5) 50 (2.3) 47.2 (1.6)

F. Sibella et al. / Clinical Biomechanics 18 (2003) 745750

749

Control group of healthy subjects. The curve representing hip and knee joint moments (Mhip and Mknee ) is shown in Fig. 5 (upper part) for a single subject taken as example. 100% of the analysed subjects show a high value of Mhip maximum that leads to the minimisation of Mknee maximum for the rst trial. The parameters calculated for last trial, instead, show a decreased Mhip maximum and consequently an increased Mknee maximum. The same results can be observed from the fatigue and work indexes (M % and P %) calculated for both joints. Numerical comparisons can be seen in Table 2. Obese subjects. A qualitative curve of hip and knee moment is plotted in Fig. 5 (bottom part) for a single subject taken as example. Two dierences arise from the comparison between CG and obese group graphs: rst, in obese patients the higher peak corresponds to knee and not to hip joint moment as in the CG; second, for obese group the comparison between the rst and the last trial shows no changes in motion strategy. M % and P % conrm kinetic results: if compared to CG, obese group shows higher values of both indexes at the knee and, consequently, lower values at the hip (for numerical data see Table 2). t-Test performed on CG vs obese group for all kinetic parameters conrmed statistical signicance of the results P < 0:05.
Fig. 3. CG rising strategy: Y shoulder coordinate (left) and X malleolus coordinate (left).

Fig. 4. Obese group rising strategy: Y shoulder coordinate (left) and X malleolus coordinate (left).

Fig. 5. CG and obese group rising strategy: hip (light grey) and knee (black) joint moments.

750

F. Sibella et al. / Clinical Biomechanics 18 (2003) 745750

Table 2 Kinetic results for control and obese group: mean (SD) Mean Mhip maximum (Nm/kg) Control group (1 trial) Control group (10th trial) Obese group (1st trial) Obese group (10th trial) 0.98 0.88 0.59 0.57 (0.04) (0.02) (0.04) (0.06) Mean Mknee maximum (Nm/kg) 0.38 0.45 0.75 0.77 (0.08) (0.03) (0.08) (0.08) M % hip 76.6 70.3 52.5 50.3 M % knee 11.7 17.5 35.2 36.4 P % hip 62.5 57.5 33.8 33.5 P % knee 24.7 36.7 64.6 65.1

4. Discussion 4.1. Kinematic results Kinematic analysis permitted to evaluate the dierences in STS motion strategy between normal (CG) and obese population. In particular, normal people rise from the chair by exing the trunk forward and keeping the feet in their initial position; on the contrary, obese subjects rise from the chair by limiting the forward trunk exion and moving the feet backwards from the initial position. In the CG, from the rst to the last trial a small decrease in forward trunk exion is to notice, even though the exion degree remains always sensibly higher than that of the obese group. Vice versa, no changes in forward trunk exion values are visible for the obese group. 4.2. Kinetic results Kinetic analysis conrms and explains kinematic results. The developed model, implemented to calculate torques at the dierent joints, permitted to explain the STS dierences between normal and obese subjects. To high degrees of trunk exion correspond a high hip joint moment that, in general, leads to a high low back loading and a minimisation of knee joint torque. Obese subjects tend to minimise trunk exion: this kinematic strategy brings to a minimisation of hip joint torque (and therefore a minimisation of low back loading); but it maximises the moment at knee joint. No evident changes occurred from the rst to the last STS trial conrming that the described strategy is the only possible strategy for obese people, who tend not to load their low back to stand up. We veried that dierences found in motion strategy cannot be due to age dierences between the analysed groups by a comparison between a small group (10 subjects) of obese people aged between 23 and 28 years and our CG (mean age 26.5 years).

5. Conclusions We suggested a biomechanical model for the comprehension of STS movement in normal and obese people. In particular a model of the belly was introduced to underline the role of the fat mass in the impairment of obese people during STS movement. STS movement was chosen because of its high repeatability in normal subjects and because it is one of the commonest movements in daily activities. Our ndings may be useful to evaluate the eect of rehabilitative treatments, which are expected to modify STS strategy. The proposed model is very simple and it can be further developed. Nevertheless it was able to explain the dierences between normal and obese subjects during STS motion, both from a kinematic and from a kinetic point of view.

References
Brozek, A. et al., 1993. Densitometric analysis of body composition. Annuals of the New York Academy of Sciences 110, 113 126. Coglin, S. et al., 1994. Transfer strategy used to rise from a chair in normal and low-back pain subjects. Clinical Biomechanics 9, 85 92. Doorenbosch, C., 1994. Biomechanics and muscular activity during sit-to-stand transfer. Clinical Biomechanics 9, 235244. Kathleen, M., 1991. Rising from a chair. Physiotherapy 77, 15 19. Lundin, T.M. et al., 1995. On the assumption of bilateral lower extremity joint moment symmetry during STS task. Technical note. Journal of Biomechanics 28, 109112. Millington, P.J. et al., 1992. Biomechanical analysis of the sit-to-stand motion. Archives of Physical Medicine and Rehabilitation 73, 609 617. Nigg, M., Herzog, W., 1995. Biomechanics. John Wiley & Sons, NY. Tagliaferri, M. et al., 1998. Obesit ae dispendio energetico. Aggiornamento Medico 22, 4855. Winter, D., 1979. Biomechanics of Human Movement. John Wiley & Sons, NY.