effects of limiting anterior displacement of the center of foot pressure on anticipatory postural...

Journal of Electromyography and Kinesiology xxx (2013) xxx–xxx

Contents lists available at ScienceDirect

Journal of Electromyography and Kinesiology

journal homepage: www.elsevier .com/locate / je lek in

Effects of limiting anterior displacement of the center of foot pressureon anticipatory postural control during bilateral shoulder flexion

1050-6411/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.jelekin.2013.07.015

⇑ Corresponding author. Tel.: +81 76 265 2225; fax: +81 76 234 4219.E-mail address: [email protected] (K. Fujiwara).

Please cite this article in press as: Fujiwara K, Yaguchi C. Effects of limiting anterior displacement of the center of foot pressure on anticipatory pcontrol during bilateral shoulder flexion. J Electromyogr Kinesiol (2013), http://dx.doi.org/10.1016/j.jelekin.2013.07.015

Katsuo Fujiwara a,⇑, Chie Yaguchi b

a Department of Human Movement and Health, Graduate School of Medical Science, Kanazawa University, 13-1 Takara-machi, Kanazawa 920-8640, Japanb Department of Physical Therapy, Faculty of Human Science, Hokkaido Bunkyo University, 5-196-1, Kogane-chuo, Eniwa 061-1449, Japan

a r t i c l e i n f o

Article history:Received 28 August 2012Received in revised form 19 July 2013Accepted 28 July 2013Available online xxxx

Keywords:Anticipatory postural controlArm movementCenter of foot pressureElectromyogramStrategy

a b s t r a c t

In bilateral shoulder flexion with the arms moving from the sides of the body to the horizontal level whilestanding, no preceding activation of the triceps surae (TS) with respect to focal muscles has been found.Considering that preceding activation would offer a useful indicator of anticipatory postural control, itwas attempted to induce preceding activation by limiting the anterior displacement range of the centerof foot pressure in the anteroposterior direction (CoPap). Subjects were 13 healthy young adults. The 50%anterior range of CoPap displacement caused by shoulder flexion was calculated, and the floor inclined bythe subject’s weight when CoPap extended beyond that range. Subjects were instructed not to incline thefloor during shoulder flexion. Under the limitation condition, the ankle and knee joints plantarflexed andextended at 1.1�, respectively, with no hip movement; that is, the whole body inclined backward by piv-oting at the ankle. This limitation resulted in preceding muscle activation of TS as well as erector spinaeand biceps femoris, and no significant differences in onset time were seen between these muscles. Theseresults demonstrated that by limiting CoPap anterior displacement, preceding activation of TS could beinduced with backward inclination of the whole body.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

With rapid movements of the arms while standing, the posturalmuscles of the legs and trunk are automatically activated beforethe focal muscles of the arms, to moderate any disturbances of pos-ture and equilibrium caused by the arm movement (Belen’kiı̆ et al.,1967). Cordo and Nashner (1982) reported that the timing of pos-tural muscle activation changes according to internal and externalconditions. In addition, they proposed that the preceding muscleactivation is clearly observed in those postural muscles playingthe most important roles in balance maintenance. Marked preced-ing activation is found in the biceps femoris (BF), and the onsettime was found to be earlier under conditions of rapid (Lee et al.,1987) and self-paced shoulder flexion (Fujiwara et al., 2011b),anterior initial position of the center of gravity (COG) (Fujiwaraet al., 2003) and postural movement patterns in which the legsand trunk or trunk alone were inclined backward (Fujiwara et al.,2007).

However, when shoulder flexion has been performed in previ-ous studies with the arm moving from the side of the body tothe horizontal level, preceding activation of the triceps surae (TS)has not been observed (Bouisset and Zattara, 1981). This

presumably results from a lack of task condition in which the pre-ceding activation of TS is necessary for postural control.

The present study was designed to provide useful backgroundinformation in order to subsequently select tasks in which preced-ing activation of TS is recognized. For postural control during tran-sient floor translation, a postural movement pattern focused on thehip joint (hip strategy) is adopted with a narrow base of support inthe anteroposterior direction, while with a wide support base, themovement pattern focused on the ankle joint (ankle strategy)(Horak and Nashner, 1986). The muscles around the hip joint andthe lower leg muscles are mainly activated in these hip and anklestrategies, respectively. These findings indicate that in transientfloor translation with a wide support base, torque around the anklestarts to change earliest, so the focus of postural control is put onthe lower leg muscles, while with a narrow support base, the focusis switched to the muscles around the hip joint. However, no stud-ies have examined activation patterns of postural muscles duringbilateral shoulder flexion on a stable platform when subjects tryto keep the center of foot pressure in the anteroposterior direction(CoPap) within a restrictive range of the support base.

We considered that with bilateral shoulder flexion from thebody side position, preceding activation of TS might be inducedby narrowing only the anterior range of the support base, whilekeeping a wide posterior range. In this case, to limit anterior dis-placement of CoPap, a strategy in which the center of body mass

ostural

http://dx.doi.org/10.1016/j.jelekin.2013.07.015

mailto:[email protected]


http://www.sciencedirect.com/science/journal/10506411

http://www.elsevier.com/locate/jelekin


2 K. Fujiwara, C. Yaguchi / Journal of Electromyography and Kinesiology xxx (2013) xxx–xxx

moves backward with the postural control focused on the ankle isadopted, and any preceding activation of TS would be clearlyapparent. In the present study, the floor was inclined when CoPapextended beyond a determined anterior range. Based on the feed-back information of floor inclination, the subjects were required tocontrol their posture to avoid inclining the floor.

A standing posture with the upper limbs at the side of the bodyis fundamental and arm movements from this position are fre-quently performed during daily life. In the elderly, decreased mus-cle strength of the lower legs leads to deterioration in balancefunction (Horak et al., 1989). Fujiwara et al. (2011a) reported thatduring shoulder flexion from hands at the side of the body, thesoleus (Sol) activated slightly earlier following muscle training ofTS, but clear preceding activation was not observed. If precedingactivation of TS is able to be induced for young adults during shoul-der flexion from this hand position, this task condition could beeffectively applied for the elderly (Woollacott and Manchester,1993).

We therefore investigated the onset timing of TS and posturalmovement patterns by narrowing the anterior range of the supportbase, first in young adults. Our working hypothesis was that duringbilateral shoulder flexion with limitation of CoPap anterior dis-placement, the preceding activation of TS would be induced, withbackward inclination of the whole body pivoting at the ankles.

2. Methods

2.1. Subjects

Subjects comprised 7 men and 6 women. Mean values (stan-dard deviation (SD)) for age, height, weight, foot length and lowerleg length were 24.8 (5.0) years, 165.5 (8.1) cm, 59.3 (8.1) kg, 24.5(1.5) cm and 40.5 (2.3) cm, respectively. No subjects reported anyhistory of neurological or orthopedic impairment. In accordancewith the Declaration of Helsinki, all subjects provided informedconsent after receiving an explanation of the experimental proto-col, which was approved by our institutional ethics committee.

2.2. Apparatus

To measure CoPap, a force platform (FPA34; Electro-design,Noda, Japan) was used. CoPap position was shown as a percentageof the distance from the heel in relation to foot length (%FL). CoPapelectronic signals were sent simultaneously to three devices: acomputer (PC9801BX; NEC, Tokyo, Japan) to determine CoPapposition; another computer (Dimension E521; Dell Japan, Kawasa-ki, Japan) for analysis; and an oscilloscope (DS6612; Iwatsu, Tokyo,Japan) to monitor the results. The onset time of postural muscles isinfluenced by CoPap position just before shoulder flexion (Fujiwaraet al., 2003). To control initial CoPap positions, the first computer,which received CoPap data via an analog-to-digital (A/D) converter(PIO9045; I/O-Data, Kanazawa, Japan) with a 20-Hz sampling rateand 12-bit resolution, generated a buzzing sound when CoPap waslocated within a range of ±1 cm of the quiet standing posture (QSPrange). Since the SD for CoPap fluctuation during QSP for 60 s wasapproximately 0.5 cm for young healthy subjects (Goshima, 1986),the QSP range corresponds to ±2 SDs of the fluctuation.

The 50% range of most anterior displacement of the CoPapcaused by shoulder flexion was calculated. To limit the anteriordisplacement range of CoPap, we used a handmade inclinationboard on the force platform that underwent forward-inclinationby 10� due to the subject’s weight when CoPap extended beyondthat range (Fig. 1).

Arm acceleration was recorded using a miniature unidirectionalaccelerometer (AS-5GB; Kyowa, Tokyo, Japan) placed on the dorsal

Please cite this article in press as: Fujiwara K, Yaguchi C. Effects of limiting ancontrol during bilateral shoulder flexion. J Electromyogr Kinesiol (2013), http:

surface of the right wrist. To detect the beginning of arm move-ment, the axis of sensitivity before shoulder flexion was alongthe anteroposterior direction.

To measure arm, trunk and leg motions in the sagittal planeduring shoulder flexion, a position sensor system (C5949; Ham-amatsu Photonics, Hamamatsu, Japan) was used. The sensor headwas placed 4.0 m away from the right side of the subject. Light-emitting diode (LED) targets were placed over the following land-marks on the right side: the wrist; spinous process of C7; midpointof the greater trochanter (GT); lateral epicondyle (LE); lateral mal-leolus (LM); and the fifth metatarsal head (MH) (Fig. 1). The x- andy-coordinates of LED targets were recorded at 0.3-mm resolution.

A fixation point was presented at the center of an eye-trek face-mounted display (FMD011F; Olympus, Tokyo, Japan). Warning (S1)and response (S2) signals were auditory stimuli delivered via ear-phones with frequency, intensity, duration and inter-stimulusinterval of 2000 Hz, 35 dB above the auditory threshold, 100 msand 2 s, respectively.

Surface electrodes (30-mm diameter, 13-mm diameter capturearea, P-00-S; Ambu, Ballerup, Denmark) were used in bipolar der-ivation to record electromyographic (EMG) activity of the follow-ing muscles: the anterior deltoid (AD) as a focal muscle forshoulder flexion; rectus abdominis (RA) at the level of the navel,erector spinae (ES) at the level of the iliac crest, rectus femoris(RF) at the midpoint between the anterior inferior iliac spine andupper border of the patella, long head of BF at the midpoint be-tween the ischial tuberosity and head of the fibula, tibialis anterior(TA), medial head of gastrocnemius (GcM) and Sol as postural mus-cles (Fig. 1). Electrode locations for AD, TA, GcM and Sol were themidportion of the muscle belly. Electrodes were placed on the rightside of the body with an inter-electrode (center-to-center) distanceof about 3 cm. A ground electrode was placed over the right LM.These electrodes were fixed after shaving and cleaning the skinwith alcohol. Inter-electrode impedance, as measured by animpedance tester, was reduced to below 5 kX. EMG signals fromelectrodes were amplified (�4000) and band-pass filtered(5–500 Hz) using an analog amplifier (Biotop-6R12; NEC-Sanei, To-kyo, Japan; common mode rejection ratio, 86 dB; input impedance,>10 MX).

All electrical signals, including CoPap, arm acceleration, x- andy-coordinates of each LED, S1 and S2 signals and each EMG, weresent to the computer for analysis via A/D converters (ADJ-98;Canopus, Kobe, Japan) with a 1000-Hz sampling rate and 16-bitresolution.

2.3. Procedure

All measurements were performed on the force platform whilestanding barefoot with feet 10 cm apart and parallel, elbowsextended, and hands positioned on the thigh anterior to GT(Fig. 1). Subjects were instructed to gaze at the fixation point dur-ing all measurements.

First, after the experimenter told the subject to maintain a sta-bilized QSP and confirmed the stabilization, CoPap fluctuation for10 s was measured and the mean position was calculated. Themean value of five measurements was adopted as the QSP position.

Next, bilateral shoulder flexion trials were commenced. Sub-jects maintained CoPap position within the QSP range for at least3 s while hearing the buzzing sound. S1 was randomly presentedwithin 1–3 s after the experimenter stopped the buzzing sound,and then S2 was started 2 s after S1. In response to S2, subjectsflexed the arms at maximum speed, stopped voluntarily at theshoulder level, and maintained this position for 3 s. Shoulder flex-ion trials were repeated 20 times on the flat floor (no-limitationcondition, nLC), and the 50% range of most anterior displacementof the CoPap was calculated. Inclination axis of the floor was set

terior displacement of the center of foot pressure on anticipatory postural//dx.doi.org/10.1016/j.jelekin.2013.07.015


EMG

LED

Anterior deltoid

Erector spinae

Biceps femoris

Gastrocnemius

Soleus

Ground

Accelerometer

Greater trochanter

Wrist

Lateral malleous

C7

Position sensor camara

4.0 m

Inclinationboard

Forceplatform

CoPap

Force platform

Face-mounted display

Foot position

Rectus abdominis

Rectus femoris

Tibialis anterior

Lateral epicondyle

Fifth metatarsal head

QSP position

50 % range of CoPanterior displacement

Fig. 1. Experimental set-up for measurement of EMG and postural movement during shoulder flexion and for limitation of CoPap anterior displacement.

K. Fujiwara, C. Yaguchi / Journal of Electromyography and Kinesiology xxx (2013) xxx–xxx 3

at that range (mean position for 20 trials) (Fig. 1). Subjects werethen instructed not to incline the floor during the shoulder flexion(limitation condition, LC). Shoulder flexion trials under LC were re-peated 20 times. To familiarize subjects with the task, 10 practicetrials for each condition were performed before the experimentaltrials. Under each condition, experimental trials in which CoPapposition just before the shoulder flexion was beyond the QSP rangewere rejected. Furthermore, under LC, trials in which the subjectfailed to prevent inclination of the floor were also rejected. Theexperiment was conducted with a 30-s standing rest between eachtrial and a 3-min seated rest between conditions.

2.4. Data analysis

Under LC, the percentage of inclined trials among all trials wascalculated. All data were analyzed blinded to condition using sig-nal-processing software (BIMUTAS II; Kissei Comtec, Matsumoto,Japan).

We defined the first deviation in the accelerometer signal as thestart point of arm movement (Fig. 2). The end of arm movementwas defined as the end of the second burst of AD activity includedin the envelop line that first deviated below the mean +2 SDs justbefore 500 ms of arm lowering, with reference to the curves ofwrist position and arm acceleration. The interval between startpoint and endpoint of arm movement was defined as the armmovement duration.

Mean CoPap positions were calculated for the periods from�300 to �150 ms with respect to the burst onset of AD (beforearm movement period) and from 0 to +150 ms with respect tothe endpoint of arm movement (after arm movement period)(Fig. 2). Differences between these mean positions were definedas CoPap displacements.

Ankle (LE – LM – MH), knee (GT – LE – LM), and hip (C7 – GT –LE) angles in the sagittal plane were calculated using Excel 2010software (Microsoft, Tokyo, Japan) based on x- and y-coordinatesof LED targets (Fig. 2). For each angle, mean values were then cal-culated for the periods before and after arm movement. Differencesin mean angles between the two time periods were calculated anddefined as movement angles of the ankle, knee, and hip, respec-


tively (Fig. 2). Movement angles were considered positive for dor-siflexion of the ankle and flexion of the knee and hip.

EMGs were analyzed as described below with reference to aprevious study (Fujiwara et al., 2011b) (Fig. 2). To exclude electro-cardiographic and movement artifacts, all EMGs were high-pass fil-tered at 40 Hz using a seventh-order Butterworth method and thenfull-wave rectified (rEMG). Mean and SD of the amplitude for back-ground activity of each muscle was calculated during the periodfrom �150 to 0 ms with respect to S2 onset for AD and from�300 to �150 ms with respect to burst onset of AD for posturalmuscles. Burst activation of each muscle was identified when onsetwas within +100 to +300 ms after S2 onset for AD and �150 to+100 ms with respect to burst onset of AD for postural muscles,and when the envelope line of the burst activity deviated morethan the mean +2 SDs from background activity for at least50 ms. Burst onset was defined as the time point at which theabove deviation began in the EMG wave included in the envelopeline. AD reaction time was defined as the time difference betweenS2 onsets and the AD burst. The onset time of postural muscles wasdefined as the time difference between burst onsets of posturalmuscles and AD, and presented as a negative value when burst on-set of postural muscles preceded AD. Preceding EMG bursts of pos-tural muscles were observed only in the dorsal muscles.Subsequently, activities of the frontal muscles (RA, RF and TA)sometimes or scarcely occurred between bimodal bursts in thedorsal muscles. As a result, only EMG bursts in the dorsal posturalmuscles were analyzed.

To investigate EMG activity before S2, rEMG of each posturalmuscle from 500 ms before S1 to S2 was averaged for each condi-tion. Then the averaged waveform was low-pass filtered at 40 Hzto detect the envelop line of the activity. In each averaged wave-form, mean amplitude was calculated for 100 ms just before S2.

To analyze the activity level for burst activation after armmovement, EMG of each muscle in the period from �300 to+200 ms with respect to burst onset of the muscle was averagedseparately for each condition. The averaged EMG waveforms werethen smoothed using a 40-Hz low-pass filter (Fujiwara et al.,2011b) to detect the envelop line of the averaged burst activity.EMG peak amplitude from baseline and latency with respect toburst onset were measured.



100 ms

onset-300 ms

onset-150ms

End+150ms

1.0 mV

5 cm

5 G

80 cm

5 V

5°

5°

5°

Anterior deltoid

Erector spinae

Gastrocnemius

Biceps femoris

Soleus

CoPap

Acceleration of arm movement

Wrist position

S2

Rectus abdominis

Rectus femoris

Tibialis anterior

Hip joint

Knee joint

Ankle joint

S2 onset Burst onset of anterior deltoid

Endpoint of arm movement

Fig. 2. Representative waveforms of each measurement data under the limitation condition. Arrows with thin straight and dashed lines indicate onset of burst activation ofpostural muscles and start point of arm movement, respectively. Arrows with thick straight and dashed lines indicate CoPap displacement and each movement angle.

55

60

30

35

40

45

50

Limited position

start point endpoint

CoP

ap p

ositi

on (

%FL

)

no-limitation

limitation***

Fig. 3. Means and standard deviations of CoPap position at the start point andendpoint of the arm movement. ��p < 0.001.


2.5. Statistical analysis

Shapiro-Wilk tests confirmed that all data satisfied the assump-tion of a normal distribution. A one-sample t-test was used to as-sess whether burst onset of the postural muscles and angles ofthe ankle, knee and hip after arm movement differed significantlyfrom that of AD and those before arm movement, respectively. Apaired t-test was used to assess differences between conditionsin CoPap displacement, arm movement duration, each movementangle, AD reaction time, postural muscle activities before S2 andpeak amplitudes and latencies of each muscle. Two-way analysisof variance was used to assess the effect of condition (nLC andLC) and muscle on onset time of postural muscles. When a signif-icant interaction between these factors was shown, post-hoc mul-tiple-comparison analyses using Tukey’s honestly significantdifference test were performed to assess differences among mus-cles. Pearson correlations were used to evaluate the magnitude ofcorrelation between each parameter. The alpha level was set atp < 0.05. All statistical analyses were performed using SPSS soft-ware (version 14.0J, SPSS Japan, Tokyo, Japan).

Please cite this article in press as: Fujiwara K, Yaguchi C. Effects of limiting anterior displacement of the center of foot pressure on anticipatory posturalcontrol during bilateral shoulder flexion. J Electromyogr Kinesiol (2013), http://dx.doi.org/10.1016/j.jelekin.2013.07.015


40

60

-20

0

20

40

-60

-40

-20

-80

-60

Erectorspinae

Bicepsfemoris

Gastro-cnemius

Soleus

Ons

et ti

me

(ms)

Postural muscles

**

*** ***no-limitation

limitation

Fig. 5. Means and standard deviations of onset times of postural muscles withrespect to burst onset of anterior deltoid (AD) ��p < 0.01; ��p < 0.001. �Significantdifferences between burst onsets of AD and each postural muscle (p < 0.001).


3. Results

Fig. 3 shows CoPap positions at the start point and endpoint ofarm movement. Position at the endpoint of arm movement in LCwas slightly posterior (2.9%FL (2.2)) to the determined position(4.7%FL (2.0) anterior to QSP position). CoPap displacement wassignificantly smaller in LC than in nLC (a decrease of 4.9%FL (2.7),i.e., 12.0 mm (6.7)) (t12 = 6.6, p < 0.001). No significant differenceswere found in the percentage of inclined trials between first andlast halves of trials and the percentage of inclined trials amongall trials was 13.4% (8.4).

Fig. 4 shows movement angles of the ankle, knee, and hip.Angles of the ankle in LC, knee in both conditions and hip in nLCafter arm movement were significantly larger in the plantarflexionand extension directions, respectively, than before arm movement(t12 > 2.3, p < 0.05). No significant differences were found betweenbefore and after arm movement in angles of ankle in nLC and hip inLC. Significant differences were found between conditions inmovement angles of the ankle and knee (t12 > 2.9, p < 0.05). Hipmovement angle in LC tended to differ from that in nLC (t12 = 2.1,p = 0.06).

No significant differences were found between conditions in ADreaction time or arm movement duration (mean in both condi-tions: 185 ms and 498 ms, respectively). Preceding activations ofpostural muscles with respect to AD were found in ES and BF innLC and in all postural muscles in LC (t12 > 5.6, p < 0.001) (Fig. 5).A significant interaction was found between condition and musclefor onset time of postural muscles (F3,48 = 36.0, p < 0.001). Onsettimes for ES, GcM and Sol were significantly earlier in LC than innLC (p < 0.01). In nLC, ES and BF were activated significantly earlierthan GcM and Sol (p < 0.001), while in LC, no significant differencesin onset time were found for all postural muscles. For ES only, EMGmean amplitude just before S2 was significantly larger in LC(6.4 lV (SD = 5.4)) than in nLC (3.4 lV (1.8)) (t12 = 2.3, p < 0.05).Peak amplitude of BF was significantly smaller in LC (37.7 lV(28.6)) than in nLC (50.9 lV (35.6)) (t12 = 4.2, p < 0.01). Peaklatency of GcM was significantly shorter in LC (75.5 ms (42.4)) thanin nLC (99.3 ms (53.2)) (t12 = 2.5, p < 0.05).

With regard to correlations between postural movement indi-ces and onset times of postural muscles, significant correlationswere found between ankle movement angle and Sol onset time,and between hip movement angle and BF onset time (p < 0.05)(Table 1). With regard to correlations between postural movementindices and EMG mean amplitude just before S2, ankle movement

4

0

1

2

3

4

-3

-2

-1

-4

no-limitation

limitation

****

p = 0.06

HipKneeAnkle (Joint)

Mov

emen

t ang

le (

degr

ees)

† ††

†

Fig. 4. Means and standard deviations of movement angles of the ankle, knee andhip �p < 0.05; ��p < 0.001. �Significant differences between before and after the armmovement (p < 0.05).


angle correlated with mean amplitudes of ES and Sol (p < 0.01),knee movement angle correlated with BF mean amplitude(p < 0.05), and hip movement angle correlated with mean ampli-tudes of ES and Sol (p < 0.01). CoPap displacement correlated withonset times of ES, GcM and Sol (p < 0.01), and with ankle move-ment angle (p < 0.01).

4. Discussion

The present study tried to induce preceding activation of TSwith respect to AD during bilateral shoulder flexion from the bodyside position. For that purpose, a condition limiting the anteriordisplacement range of CoPap caused by shoulder flexion was setusing an inclination board. In experimental trials under LC afterthe practice trials, floor inclination was seen in 13% of trials,regardless of trial progress. By providing feedback information oninclination to subjects, CoPap anterior displacement largely re-duced (12 mm). This showed that CoPap anterior displacementcould be limited using the present method.

In LC, clear preceding activation with respect to AD was foundin TS, as well as in ES and BF. We will discuss this activation incomparison with that in nLC. In discussions about postural muscleactivation, some limitations exist to evaluating global activity ofeach muscle. It has been reported that surface potentials are morelocalized in pinnate muscles, as the muscle fibers are more obliqueto the skin (Mesin et al., 2011) and that bi-articular muscles playregion-specific functional roles (Watanabe et al., 2012). Activationpatterns of postural muscles should thus be discussed taking theselimitations into consideration.

Under nLC, the knee and hip significantly extended according toshoulder flexion, but the ankle did not move. This suggests that bybackward inclination of the trunk with relatively high mass, pos-tural disturbance would be moderated. In this case, the precedingactivation of postural muscles with respect to AD was found in ESand BF, but not in TS. ES and BF activation would generate exten-sion of the trunk and hip joint. For TS, stiffness of the contractileportion of muscle is reportedly higher than that of tendon at dor-siflexion below 0.5� and is not associated with knee movement(Loram et al., 2007). We consider that for TS under nLC, the focusof postural control might be on stiffness rather than precedingactivation.

On the other hand, under LC, the ankle and knee respectivelyplantarflexed and extended by 1.1�, with no hip movement. Equalmovement angle of the ankle and knee indicates that thigh inclina-



Table 1Significant correlations between each parameter.

Onset time of postural muscles EMG mean amplitude just before S2 Movement angle

Sol GcM BF ES Sol BF ES Ankle

Movement angleAnkle 0.47 �0.58 �0.66Knee �0.42Hip 0.42 0.62 0.60CoPap displacement 0.86 0.82 0.51 0.53


tion was maintained during shoulder flexion, as well as hip jointangle. Crenna et al. (1987) suggested that backward inclinationof the whole body pivoting at the ankles is the postural movementpattern required to effectively translate the COG backward. Thus,the postural control strategy adopted under LC would be backwardinclination of the whole body pivoting at the ankle to reduce CoPapanterior displacement. For this postural control strategy, TS wasactivated prior to AD activation and the onset time was almostthe same as those of ES and BF. Onset time of TS correlated withCoPap displacement. For burst of GcM activation, peak amplitudeshowed no significant difference between conditions, while peaktime under LC was clearly reduced. Fujiwara et al. (2011b) reportedthat peak times of BF and GcM were significantly shorter in a sim-ple-reaction task that required a rapid response than in a self-tim-ing task. These results indicate that in order to reduce CoPapanterior displacement, the focus of postural control would be onrapid activation of TS, and the ankle joint would then largelyplantarflex.

As mentioned above, movement angles of the ankle and kneeare very small (about 1.1�). Gurfinkel’ et al. (1983) reported thatthe perceptual threshold in angular displacement of the ankle jointwas extremely lower when standing (0.04–0.13�) than when sit-ting. In the present study, movement angles of the ankle and hipjoints correlated with onset times of Sol and BF, respectively. Thesesuggest that joint movements would be perceived with relativelyhigher sensitivity based on the integrated multiple sensory infor-mation, and would be controlled synergistically (Amblard et al.,1997). Clinical training of postural control should therefore be con-ducted with consideration of sensory integration via a sensory ref-erence frame.

The timing of preceding activation of ES was earlier under LCthan under nLC. In our previous study, subjects who inclined thelegs backward and the trunk forward did not show any clear pre-ceding activation of ES (Fujiwara et al., 2007). As described above,under LC, the lower legs largely inclined backward, while the hipjoint did not move. These findings suggest that under LC, the for-ward inclination of the trunk induced by inertia during backwardinclination of the lower legs would be inhibited by the precedingactivation of ES. For ES only, muscle activation just before S2 wassignificantly larger under LC than under nLC. This muscle activa-tion correlated closely with movement angles of the ankle andhip. The increment in ES stiffness would also achieve a marked ef-fect to inhibit forward inclination of the trunk.

For BF, no significant difference was found between conditionsin preceding onset time. The function of BF for the hip joint wouldbe similar to that of ES for the trunk including the pelvis. The kneesextended under both conditions, but the degree of extension waslarger under LC than under nLC. The knee has a locking structurethat prevents anterior rotation, which may be related to the lackof significant differences in muscle activation of BF just before S2between conditions and the decrease in burst activity of BF underLC.

Thus, by limiting CoPap anterior displacement, we demon-strated that clear preceding activation of TS could be induced foryoung adults, even with bilateral shoulder flexion from the body


side position. In future studies, we intend to use this limitationmethod to investigate whether preceding activation of TS can beinduced for elderly individuals and whether changes can be ob-tained with muscle training of TS. Furthermore, we will examinetransient floor translation with this limitation and investigate acti-vation patterns of TS.

5. Conclusions

By limiting CoPap anterior displacement, a postural controlstrategy of backward inclination of the entire body pivoting atthe ankles is adopted and preceding activation of TS can be in-duced. Using this method, both the role of and the training effectson TS for anticipatory postural control during shoulder flexion maybe able to be clearly investigated in the elderly.

Conflict of interest

We wish to confirm that there are no known conflicts of interestassociated with this publication and there has been no significantfinancial support for this work that could have influenced itsoutcome.

Acknowledgement

This work was supported by Grant-in-Aid for Scientific Re-search (B) (23300238).

References

Amblard B, Assaiante C, Fabre JC, Mouchnino L, Massion J. Voluntary headstabilization in space during oscillatory trunk movements in the frontal planeperformed in weightlessness. Exp Brain Res 1997;114(2):214–25.

Belen’kiı̆ VE, Gurfinkel’ VS, Pal’tsev EI. Elements of control of voluntary movements.Biofizika 1967;12(1):135–41.

Bouisset S, Zattara M. A sequence of postural movements precedes voluntarymovement. Neurosci Lett 1981;22(3):263–70.

Cordo PJ, Nashner LM. Properties of postural adjustments associated with rapid armmovements. J Neurophysiol 1982;47(2):287–302.

Crenna P, Frigo C, Massion J, Pedotti A. Forward and backward axial synergies inman. Exp Brain Res 1987;65(3):538–48.

Fujiwara K, Toyama H, Kunita K. Anticipatory activation of postural musclesassociated with bilateral arm flexion in subjects with different quiet standingpositions. Gait Posture 2003;17(3):254–63.

Fujiwara K, Maeda K, Kunita K, Tomita H. Postural movement pattern and muscleaction sequence associated with self-paced bilateral arm flexion duringstanding. Percept Mot Skills 2007;104(1):327–34.

Fujiwara K, Toyama H, Asai H, Yaguchi C, Irei M, Naka M, et al. Effects of regularheel-raise training aimed at the soleus muscle on dynamic balance associatedwith arm movement in elderly women. J Strength Cond Res2011a;25(9):2605–15.

Fujiwara K, Yaguchi C, Shen X, Maeda K, Mammadova A. Activation timing ofpostural muscles during bilateral arm flexion in self-timing, oddball andsimple-reaction tasks. J Electromyogr Kinesiol 2011b;21(4):595–601.

Goshima K. Studies on items and normal ranges in stabilometry. Equilibrium Res1986;45(4):368–87 [In Japanese with English Abstract].

Gurfinkel’ VS, Lipshits MI, Popov KE. Thresholds of kinesthetic sensation in thevertical posture. Hum Physiol 1983;8(6):439–45.

Horak FB, Nashner LM. Central programming of postural movements: adaptation toaltered support-surface configurations. J Neurophysiol 1986;55(6):1369–81.

Horak FB, Shupert CL, Mirka A. Components of postural dyscontrol in the elderly: areview. Neurobiol Aging 1989;10(6):727–38.


http://refhub.elsevier.com/S1050-6411(13)00184-3/h0005


































Lee WA, Buchanan TS, Rogers MW. Effects of arm acceleration and behavioralconditions on the organization of postural adjustments during arm flexion. ExpBrain Res 1987;66(2):257–70.

Loram ID, Maganaris CN, Lakie M. The passive, human calf muscles in relation tostanding: the non-linear decrease from short range to long range stiffness. JPhysiol 2007;584(Pt 2):661–75.

Mesin L, Merletti R, Vieira TM. Insights gained into the interpretation of surfaceelectromyograms from the gastrocnemius muscles: a simulation study. JBiomech 2011;44(6):1096–103.

Watanabe K, Kouzaki M, Moritani T. Task-dependent spatial distribution of neuralactivation pattern in human rectus femoris muscle. J Electromyogr Kinesiol2012;22(2):251–8.

Woollacott MH, Manchester DL. Anticipatory postural adjustments in older adults:are changes in response characteristics due to changes in strategy? J Gerontol1993;48(2):M64–70.

Katsuo Fujiwara received his PhD from Tsukuba Uni-versity in 1984. He is a professor at the Department ofHuman Movement and Health, Graduate School ofMedical Science, Kanazawa University (2001-Present).He is a member of the International Society of Electro-physiological Kinesiology, the International Society ofPosture and Gait Research, and the Society for Neuro-science. He serves as a president of the Japanese Societyof Health and Behavior Science. He is now analyzing theanticipatory processes of postural control in the brainby using electromyograms, evoked potentials, event-related potentials, and cerebral blood flow.


Chie Yaguchi received her PhD from Kanazawa Uni-versity in 2012. She is an assistant professor at theDepartment of Physical Therapy, Faculty of HumanScience, Hokkaido Bunkyo University (2011-Present).She is a member of the Society for Neuroscience. She iscurrently performing research into the relationshipbetween anticipatory postural control and visuo-spatialattention.


















effects of limiting anterior displacement of the center of foot pressure on anticipatory postural...

Documents