Presented at NACOB 98: North American Congress on Biomechanics Canadian Society for Biomechanics - American Society of Biomechanics University of Waterloo Waterloo, Ontario, Canada August 14-18, 1998

---

ANALYSIS OF SIT-TO-STAND IN AN ELDERLY POPULATION USING
A TELESCOPIC INVERTED PENDULUM MODEL

E. Papa and A. Cappozzo
Dipartimento di Scienze Biomediche, Sezione di Fisiologia e Bioingegneria dell'Uomo,
Università degli Studi di Sassari, 07100 Sassari, Italy

INTRODUCTION

The quantitative assessment of motor ability in disabled, or at risk of disablement (e.g. elderly), subjects is presently one of the leading aims of clinical biomechanics. For field applicability, measurements must be carried out using a least perceivable to the subject and "essential" experimental apparatus. Since data thus obtained do not necessarily lend themselves to straight interpretation, they may be fed to a model of the musculo-skeletal system embodying its invariant aspects as well as those of the motor task investigated. Such a model is, therefore, expected to provide richer, physiology-related, and, thus, easier to interpret, information than found at its input. This may help discriminating different motor strategies. In this framework, the present study investigated the sit-to-stand (STS) motor task in elderly and young individuals using experimental and analytical methods consistent with the requirements and aims illustrated above. STS was selected since it is a demanding task in terms of lower limb strength, joint ranges of motion and balance control. Further, being an activity of daily living, it is not affected by motivation and learning.

REVIEW AND THEORY

Various studies analyzed STS in general (Pai et al., 1990; Schultz et al., 1992) and in elderly populations in particular (Vander Linden et al., 1994). However, the applicability in a clinical context of the analysis has been overlooked. In this study a telescopic inverted pendulum (TIP) model was used (figure). This is composed of a telescopic massless link, hinged at a point effectively approximating the base of support, and joining the latter point with the center of mass (CM) of the portion of the body in motion. The link varies its length by effect of a force (linear) actuator (LA) and its orientation by effect of two couple (rotational) actuators. One acts in the frontal (FA) and the other in the sagittal (SA) plane. These actuators may be seen as muscle equivalent effectors. The generalized coordinates of the TIP model are the Cartesian coordinates of the CM relative to an inertial frame, and its parameters are the relevant mass and the location of the stationary hinge. For the analysis of STS, two TIP models were used in temporal sequence. Prior to the loss of contact between subject and seat (seat off, SO) only the head-arms-trunk system (HAT) moves and therefore is accounted for in the model. Following the SO the whole body (WB) is considered. The CM trajectory may be estimated by double integration of the inertial components of the external reaction force (RF).

PROCEDURES

Two sample healthy populations of 24 elderly (8 males, 16 females; 58 ÷ 80 years; 55 ÷ 106 kg; 1.46 ÷ 1.71 m) and 16 young adults (7 male, 9 females; 22 ÷ 34 years; 48 ÷ 84 kg; 1.58 ÷ 1.78 m) were investigated after informed consent was obtained. The experimental apparatus consisted of: (1) a modular seat without backrest; (2) a six-component force plate, positioned under both subjectÕs feet and seat; (3) rulers and scale for anthropometric and initial posture data collection. Force plate signals were acquired (100 samples/s) while subjects stood up from sitting (chair height at 80% of knee height, arms folded across the chest, vertical trunk, free feet position). The motor task was executed five times for each of two self-determined speeds: natural (NS), and as high as possible (HS). The initial conditions for RF double integration were derived from a stick model of the subject. The CM trajectory, the hinge location and the relevant mass were fed to the appropriate TIP model. The three model actuators were characterized through the force/couple they generate (normalized relative to body mass and body mass times body height, respectively) and their linear/angular velocity versus time. Given the sagittal symmetry of STS in the subjects analyzed, the FA carried no useful information. The following parameters were selected: initial ankle dorsiflexion angle (q); duration of the motor task (T); SO occurrence time; maximal value of the SA propulsive couple prior to SO (C1) and immediately after SO (C2); maximal value of the LA propulsive force prior to SO (F1) and after SO (F2); occurrence of F2 (Tf) in percent of T; maximal value of the SA angular velocity prior to SO (W1) and immediately after SO (W2); maximal value of the LA velocity prior to SO (V1) and after SO (V2). In addition, the maximal value of the A-P (Rx) and vertical (Ry) components of the RF were calculated. Descriptive statistics and t-test assuming unequal variances were applied to the above listed parameters.

RESULTS

The table reports sample mean value (mv) and standard deviation (sd) of the parameters for each population and speed. The young and elderly populations showed significant differences (p<0.05) for q and, at both speeds, for the parameters C2, F2, W1, W2, Tf and Ry. All other parameters in the table were able to discriminate the two populations at HS only.

DISCUSSION

While NS exhibited no significant difference, HS was 21% higher than NS in the elders and 33% in the young subjects. Prior to SO, elders flexed the trunk more rapidly (W1) than the young adults did. Immediately after SO and in order to accomplish stability, they exhibited a larger C2 (ankle dorsiflexion). This is related to their tendency to start the motor task with the ankles less dorsiflexed (feet more anterior), mostly because of a limited range of motion. The young subjects exhibited a larger (F2) and anticipated (Tf) action of LA. The actuator parameters show variations with speed that indicate that the physical functional reserve decreases with age, and seem to more effectively discriminate the two populations at HS, indicating that both NS and HS should be tested.

In conclusion, for motor ability assessment, it seems that the motor task strategy can be identified using the TIP model, but not through an analysis of the RF, and as effectively as it could be done with a multifactorial analysis.

REFERENCES

(1) Pai Y.C. et al. Med. Sci. Sports Exerc., 22, 378-384, 1990.

(2) Schultz A.B. et al. J. Biomech., 25, 1383-1391, 1992.

(3) Vander Linden D.W. et al. Arch. Phys. Med. Rehabil., 75, 653-660, 1994.

ACKNOWLEDGMENTS

Supported by Istituto Superiore di Sanitˆ, Italy. Experiments were carried out in collaboration with Prof. M. Marchetti, his team and the local administration of the city of Genzano.