Anticipatory postural adjustments during load catching by standing subjects

Abstract

Objectives: (1) To study differences in the generation of anticipatory postural adjustments (APAs) in arm and trunk/leg muscles prior to catching a load released either by the subject him-/herself or by the experimenter. (2) To study the importance of different mechanical characteristics of the load at impact for the generation of APAs prior to load catching.

Methods: Standing subjects were asked to catch loads dropped onto the left hand from different heights either by the experimenter or by the subject's right hand. The load mass and release height were manipulated to keep either the mass or the momentum of the load at impact constant. APAs were quantified with integral electromyographic indices.

Results: APAs were observed in leg, trunk and arm muscles prior to load impact for both self- and experimenter-release trials. Kinetic energy showed higher correlations with the magnitude of APA than momentum, but only in experimenter-release trials.

Conclusions: Subjects can generate APAs in both arm and trunk/leg muscles in the absence of an explicit voluntary action. The relative importance of kinetic energy and momentum for defining the magnitude of APAs can reflect the difference in the sources of information used to prepare for the forthcoming perturbation during self- and experimenter-released load catch.

Introduction

Anticipatory postural adjustments (APAs) represent changes in the muscle activity that occur prior to an expected mechanical perturbation. Most commonly, APAs have been studied in experiments when subjects' own actions perturbed the posture of the body or of a limb (Cordo and Nashner, 1982, Friedli et al., 1984, Crenna et al., 1987, Massion, 1992). Perturbations associated with such actions can be due to two major factors: (1) reaction forces due to the coupling of body segments; and (2) quick changes in the mass distribution of the body induced by changes in the body geometry or load manipulations (Massion, 1992, Bouisset and Zattara, 1987).

APAs associated with perturbations of the first type have been extensively studied in the leg and trunk muscles prior to arm movements. These studies have shown scaling of the magnitude and temporal onset of the APAs with changes in the forthcoming dynamic requirements of the task, e.g. during arm movements at different velocities (Horak et al., 1984) and during arm movements against different inertial loads (Horak et al., 1984, Aruin and Latash, 1995a). The second group of perturbations, caused by abrupt changes in the body or segment mass distribution, have been studied using unloading and loading procedures (Paulignan et al., 1989, Lacquaniti and Maioli, 1989a, Aruin and Latash, 1995b, Aruin and Latash, 1996, Bennis et al., 1996). The magnitude of APAs has been shown to scale with changes in the magnitude of a forthcoming perturbation during unloading (Aruin and Latash, 1996) and loading (e.g. in catching, Lacquaniti and Maioli, 1989a; also see Santello and McDonagh, 1998).

APAs have generally been viewed as postural adjustments to self-induced perturbations, while the possibility of feedforward postural preparation to externally induced perturbations has not been widely accepted. This view comes from a number of experiments involving loading and unloading manipulations examining APAs occurring during self- and experimenter-triggered perturbations. When loading or unloading was performed or triggered by the subject, APAs were observed in both arm and leg/trunk muscles, while these adjustments were not seen when a similar perturbation was triggered by an experimenter, even when the timing of the perturbation was predictable (Dufossé et al., 1985, Paulignan et al., 1989, Struppler et al., 1993, Aruin and Latash, 1995b, Scholz and Latash, 1998). In particular, in a study of APAs associated with dropping loads, Aruin and Latash (1996) did not see APAs prior to experimenter-triggered unloadings even when the subjects were instructed to watch the experimenter triggering the unloadings. Struppler et al. (1993) showed APAs in the biceps brachii prior to taps delivered to the wrist by the subject's other hand; such APAs were absent when the taps were delivered by another person, even if the timing of the taps was well predictable. In contrast, there are several experiments suggesting that postural adjustments occur prior to catching a load released by another person. Lacquaniti and Maioli, 1989a, Lacquaniti and Maioli, 1989b described APAs prior to catching a load released by an experimenter, but only when visual information was not restricted. When Lacquaniti and Maioli (1989b) substituted visual information with an auditory cue on the release time, the subjects were unable to show APAs, even when the release height and the mass of the load were known in advance.

These studies suggest that when adequate information for an accurate prediction of the impact time is available, APAs can be generated, even if the perturbation is not self-triggered. However, APAs in trunk/leg muscles have never been compared between self- and experimenter-triggered perturbations within the same study. Therefore, we plan to test whether catching a load released by an experimenter is accompanied by APAs in both arm and trunk/leg muscles, and if so, to compare these APAs with those observed during catching the same load released by the subject him-/herself.

By their very nature, APAs are generated prior to an expected perturbation so that their magnitude should be defined based on a predicted rather than an actual magnitude of the perturbation. We hypothesize that the relations between APAs prior to load catching and different mechanical characteristics of the load may be task-dependent and, in particular, depend on whether the perturbation is induced by the subject or the experimenter. In these two situations, different time constraints may be imposed on the APA preparation such that different physical variables may be used in predicting the magnitude of a forthcoming perturbation. For example, during load manipulations, the magnitude (and/or timing) of APAs in some muscles may be related to a predicted mass redistribution after the load is dropped or caught, or it may be related to predicted transient forces acting on body segments at impact. Lacquaniti and Maioli (1989a) have reported that during catching, the magnitude of APAs in arm muscles correlated linearly with the momentum of the load at impact. This finding looks reasonable from a mechanical view since the time integral of forces exerted by the load on the catching hand is defined by the load's momentum. However, if a subject has limited time to prepare for load impact (as during load falls from relatively low heights triggered by an experimenter), the information on load velocity during the fall may be too late to be used for the APA generation. Then, one may expect the subject to switch to alternative sources of information that are available in advance, e.g. the mass of the load (m) and its distance from the catching hand (h) prior to the release. This information may be sufficient for a relatively direct assessment of the kinetic energy of the load at impact (from the law of conservation of energy, kinetic energy at impact equals potential energy prior to load release and can be calculated as the product mgh, where g is the gravity constant). Indeed, Bennis et al. (1996) reported a high correlation between a kinematic index of APAs and load kinetic energy.

Within the present study, we used a load catching paradigm similar to that of Lacquaniti and Maioli, co-varying mass and height to create conditions in which there were different momentums as well as the same momentum at the time of load impact on the subject's hand. Two major issues have been in the center of the study: (1) the role of the subject's action in the generation of APAs; and (2) the role of different mechanical characteristics of a falling object in the generation of APAs in the arm and leg/trunk muscles. To get insight into these issues, we focused on the following specific experimental questions. Will APAs in the arm and in the trunk/leg muscles be seen during catching a load released by an experimenter? Will the relation between the mechanical characteristics of the load at impact and APAs be the same for subject- and experimenter-release conditions? Will momentum or kinetic energy be a better predictor of APA magnitude? Will the magnitude of APAs and the magnitude of muscle responses after the load impact show similar relations to mechanical characteristics of the load at impact?