Postural responses to anterior and posterior perturbations applied to the upper trunk of standing human subjects

Abstract

This study concerned the effects of brisk perturbations applied to the shoulders of standing subjects to displace them either forwards or backwards, our aim being to characterise the responses to these disturbances. Subjects stood on a force platform, and acceleration was measured at the level of C7, the sacrum and both tibial tuberosities. Surface EMG was measured from soleus (SOL), tibialis anterior (TA), the hamstrings (HS), quadriceps (QUAD), rectus abdominis (RA) and lumbar paraspinal (PS) muscles. Trials were recorded for each of four conditions: subjects' eyes open (reference) or closed and on a firm (reference) or compliant surface. Observations were also made of voluntary postural reactions to a tap over the deltoid. Anterior perturbations (mean C7 acceleration 251.7 mg) evoked activity within the dorsal muscles (SOL, HS, PS) with a similar latency to voluntary responses to shoulder tapping. Responses to posterior perturbations (mean C7 acceleration -240.4 mg) were more complex beginning, on average, at shorter latency than voluntary activity (median TA 78.0 ms). There was activation of TA, QUAD and SOL associated with initial forward acceleration of the lower legs. The EMG responses consisted of an initial phasic discharge followed by a more prolonged one. These responses differ from the pattern of automatic postural responses that follow displacements at the level of the ankles, and it is unlikely that proprioceptive afferents excited by ankle movement had a role in the initial responses. Vision and surface properties had only minor effects. Perturbations of the upper trunk evoke stereotyped compensatory postural responses for each direction of perturbation. For posterior perturbations, EMG onset occurs earlier than for voluntary responses.

Fig. 1

Recording arrangements. Subjects were asked to correct for postural disturbances (anterior or posterior perturbations) while standing upright on the force platform. EMG was recorded over the left side from the soleus (SOL: A) and tibialis anterior (TA: B), hamstrings (HS: C), quadriceps (QUAD: D), lumbar paraspinal (PS: E) and rectus abdominis muscles (RA: F) with accelerometers positioned over C7, the sacrum (shown in black) and the tibial tuberosities (shown in grey). Perturbations were applied at the level of the shoulders

Fig. 2

Realignment of raw data and averaging. Data shown from a single subject following posterior perturbations on a rigid surface during eyes open, triggered by the experimenter pressing a footswitch. The left column shows the five individual trials superimposed for C7 accelerations, unrectified EMG from tibialis anterior and CoP displacements (a). The right column shows each individual trial realigned to the onset of C7 accelerations and with EMG signals rectified (b). Averaged data for the realigned traces are shown in grey and were measured for data analysis. Note that the consistency of the footswitch trigger was good but that it systematically underestimated the true onset of the perturbation as shown by the C7 accelerometer

Fig. 3

Anterior and posterior perturbations. Grand means of accelerometry, EMG and CoP recordings following anterior perturbations (a; left column) and posterior perturbations (b: right column) on a rigid surface during eyes open. Anterior perturbations showed an initial period with phasic bursts in soleus (SOL), hamstrings (HS) and paraspinal (PS) and prolonged contractions in SOL and HS. Posterior perturbations showed an initial phasic burst in quadriceps (QUAD) associated with cocontraction in TA and SOL, followed by prolonged activation in TA and QUAD. EMG responses were generally small for the HS, rectus abdominis (RA) and PS muscles. Upward deflections in accelerometry traces reflect anterior displacement in the anterior–posterior (AP) plane and rightward displacement in the mediolateral (ML) plane; asterisk denotes significant changes in rectified EMG compared to baseline levels

Fig. 4

Rectified EMG levels during anterior and posterior perturbations. Muscle groups which demonstrated significant changes in mean rectified EMG levels from baseline are shown for anterior (left column) and posterior (right column) perturbations. Anterior perturbations showed significant increases from baseline for soleus (SOL), hamstrings (HS) and paraspinal (PS) muscles. The greatest modulation in amplitude for posterior perturbations was observed for tibialis anterior (TA), whereas a similar but less marked effect was seen for quadriceps (QUAD). SOL showed an initial increase in EMG amplitude from baseline to 0.2 s with a subsequent decrease over the 0.2- to 1-s time interval, followed by a later increase from 1 to 2.5 s

Fig. 5

Voluntary postural reactions. EMG and CoP recordings following active posterior and anterior lean (n = 8) triggered by a tap to the shoulder. For comparison, responses to posterior and anterior perturbations are shown in grey. For posterior lean, SOL is active initially (the later TA activity acts to restore the initial posture), while the opposite applies to anterior lean. Initial cocontraction was not evident for either direction of voluntary movement. Note similar initial displacements for the CoP (indicated by the black arrows)