1 Introduction

Pushing is a common manual material activity and one risk factor of working environment is the floor condition [1]. It has been suggested that working posture is extremely dependent upon the quality of the foot-to-floor interface [2]. Slipping occurs when the ratio of the friction force to the normal force at the feet exceeds the maximum coefficient of friction and the risk of slipping was associated with the maximum friction at the feet in pushing [3, 4]. When pushing on the difference surfaces, changes in posture of knee flexion and trunk extension has been observed [5]. In addition, changes in anticipatory and compensatory postural adjustments are observed in the center of pressure (COP) displacements when pushing in the anterior-posterior direction; the backward COP displacements were seen prior to the pushing followed by forward COP displacements after the movement [6]. In the medial-lateral direction, the COP displacement is affected by the placements of lower extremities on the floor [6, 7] and by the magnitude of the weight to be pushed: larger COP displacements were reported when pushing heavier weight [8]. Meanwhile, the orientations of the COP displacements and vertical torques were corresponding to the upper and lower extremities positions in pushing [8]. However, all previous studies only focused on pushing tasks itself and did not take subsequent perturbations to postural control and balance into considerations.

Performing a voluntary movement while pushing an object or responding to the movements of an unstable surface (that are considered as a primary task); these tasks are performed simultaneously with maintenance of vertical posture. The central nervous system (CNS) uses sensory information from the vestibular, visual, and somatosensory systems in control of upright posture as well as in dealing with the primary task. In addition, many daily life activities that are associated with carrying out motor tasks are performed simultaneously (e.g. holding a cup of tea or talking on the cell phone while walking). Another example is standing wearing rollerskates and performing voluntary movements. It was reported that people who use rollerskates and perform fast arm flexion movements showed larger shank muscle activities and muscle co-contraction patterns. In addition, the inverted direction of COP displacement in the anterior-posterior direction was recorded as compared to standing wearing regular shoes [9]. It has been shown in experiments involving both perturbations (pushing and externally induced forward and backward translations of the platform) that the onset time of the tibialis anterior muscle in such conditions depends on the direction of moving platform and is different compared to pushing only (while standing on a non-movable platform) [10]. Thus, dealing with two perturbations simultaneously modifies patterns of muscle activities and COP displacement as compared to experiencing a single perturbation [10, 11].

Pushing as a perturbation resulted in different performing strategies was observed in elderly compared to healthy young adults [12]. Aging population is growing unprecedentedly worldwide and it has been reported that a clear relationship between working capacity and workers’ age [13]. A better understanding of how sliding surface affects pushing and balance performances could provide new information crucial in designing working environment. The objective of the present study was to investigate how a combination of voluntary pushing movement and translational perturbation affects pushing performance and balance. To do that the experimental paradigm involved two body perturbations: one was induced by the upper body performing the pushing an object while standing on a free to move surface and the other perturbation was induced when the sliding board moved. We hypothesized that onsets of the COP displacement would be affected by the presence of a secondary perturbation; we also expected to see a decrease in the magnitudes of the COP displacements due to the translation perturbation.

2 Methods

2.1 Participants

Eight young male volunteers (age = 26.50 ± 1.70 years, height = 1.68 ± 0.02 m, mass = 72.43 ± 2.49 kg) participated in the experiment. All participants were free from any musculoskeletal disorder and neurologic disease that could affect performing the experimental tasks. The project was approved by the University of Illinois at Chicago Institutional Review Board, and all participants provided written informed consent before taking part in the experimental procedures.

2.2 Instrumentation and Procedure

The customized sliding board was 0.06 m height and made of two layers connected by a linear bearing system. The sliding board had a lock mechanism allowing the top layer (lengths 0.50 m and width 0.45 m) to be free to slide in the anterior-posterior direction or remain stationary. The bottom layer was 0.60 m length and width, which was positioned on top of the force platform (model OR-5, AMTI, USA). Three accelerometers were used in the experiment. The moment of the handle moving away (Thandle) was detected and denoted time zero by an accelerometer (model 208CO3, PCB Piezotronics Inc, USA) attached to the handle. The moment of trunk movement (Ttrunk) was detected by an accelerometer (model 1356a16, PCB Piezotronics Inc, USA) attached to the dorsal surface of the subject at the level of L5S1. The moment of the sliding board movement (Tboard) was detected by an accelerometer (model 208CO3, PCB Piezotronics Inc, USA) attached underneath the top layer of the sliding board.

The participants were instructed to stand barefoot on the top layer of the sliding board with their upper arms at 90° of elbow flexion and wrist extension, and palms slightly contacting the load cell (model LSB300, Futek Inc, USA) extending from the wooden handle as well as their feet shoulder width apart and in parallel to push the horizontal flat wooden handle (62 × 9 × 2 cm) attached to an aluminum pendulum. The pendulum was affixed to the ceiling with additional loads of 30% individual body weight; the height of the wooden handle was adjusted to each individual’s hand position accordingly. Feet position was marked on the sliding board and was reproduced across the trials. Two conditions were performed: the not moving sliding board would be referred to as the “locked condition”, while the free-moving sliding board would be referred to as the “unlocked condition”. Each participant was given two practice trials prior to data collection to allow familiarization with the task and awareness of the locked or unlocked experimental condition. Five trials were collected in each condition.

2.3 Data Processing

All data were processed offline using MATLAB software (MathWorks, Natick, MA, USA). The onsets of the accelerometer signals were detected Thandle, Ttrunk, and Tboard using the Teager-Kaiser onset time detection method [14]. The ground reaction forces and the moments of forces were filtered with a 20 Hz low-pass, 2nd order, zero-lag Butterworth filter. Time-varying COP traces in the anterior-posterior direction were calculated using the approximations described in the literature [15]. The onset of the COP moved away from the baseline was detected by the Teager-Kaiser method (COPonset).

2.4 Statistics

The dependent t-tests were performed to evaluate effects of sliding perturbation on the onset of Ttrunk, COPonset, and maximal pushing force. Means and standard errors are presented in the results and figures. Significant difference was set at p < 0.05.

3 Results

Figure 1 shows a typical example of changes in the acceleration of the handle (Thandle), of the trunk (Ttrunk), and the sliding board (Tboard). The onset of Thandle was set as time 0 and the first changes in related to time 0 was observed in Tboard and followed by Ttrunk. Regardless of the sliding board being locked or unlocked, the participants performed pushing task with similar peak acceleration of Thandle (t(7) = 2.168, p = 0.067). The first changes in the acceleration of the unlocked board were seen at 142.03 ± 30.48 ms prior to the 0; the sliding board did not move in the locked condition. Additionally, the orientation of the accelerometer attached to the board indicated that the sliding perturbation (backward movement) was directed opposite to trunk movement (forward movement).

Fig. 1.
figure 1

A typical example of the changes in the signals obtained from the three accelerometers. Thandle is the acceleration of the handle, Ttrunk is the acceleration of the trunk, and Tboard is the acceleration of the sliding board. The vertical lines indicate the moment of the start of the handle moving away (0). Gray lines reflect changes in the acceleration while standing on the locked sliding board and black lines are for the unlocked sliding board.

Full size image

The dependent t-test revealed that the onset time of trunk movement were not significantly different between the locked and unlocked conditions (t(7) = 0.254, p = 0.807). The onsets of COP were initiated prior to Thandle and it was 279.10 ± 35.22 ms preceding Thandle in the locked condition and 189.93 ± 15.71 ms in the unlocked condition (Fig. 2). Pushing while standing on the unlocked sliding board significantly affected the onset of COP (t(7) = −2.379, p = 0.049). The magnitudes of the COP displacement (t(7) = 3.064, p = .018) were significantly smaller when standing on the unlocked board (0.05 ± 0.01 m) than the locked condition (0.03 ± 0.01 m). Furthermore, the maximum pushing force was significantly affected by the board mobility (t(7) = −2.386, p = 0.048) and it was 135.51 ± 18.79 N in the locked condition and 113.31 ± 17.49 N in the unlocked condition.

Fig. 2.
figure 2

The onset time of trunk movement and center of pressure (COP) in the locked (white boxes) condition and the unlocked (black boxes) conditions.

Full size image

4 Discussions

The current study investigated effects of dual perturbations on pushing and balance performances when pushing while standing on the sliding surface. The later onset of COP was observed when pushing with translation perturbations (the unlocked condition). While in the unlocked condition, the magnitudes of the maximal exerted force and the COP displacement were significantly smaller compared to the locked condition. Thus, our hypotheses that the onset time and magnitudes of muscle activities and the COP displacement would be affect by the movability of the sliding board was supported.

The unlocked sliding board in the current study was similar with a skateboard. Pushing in the unlocked condition was comparable with that one person on a skateboard pushes another person on the other skateboard and results in both skateboarders are pushed away in the opposite directions. In the unlocked conditions, the sliding board was detected backward movement at the moment of starting to move by the accelerometer, which indicated that the translation perturbation in the unlocked condition was followed and provoked by the pushing movement. In the current study, both maximal acceleration of the pendulum and the trunk were not statistic significant different between two conditions. It implied that participants performed the same pushing task and changes in these postural adjustments due to in response to translation perturbation.

When standing on a movable surface, the delayed COP onset and the backward board movement induced by pushing movement suggested that CNS treated the pushing task as the primary perturbation and the translation perturbation as the collateral perturbation to postural control. It has been reported that the CNS prioritizes performance of the task when balance is not in imminent danger [25, 26], but the CNS reverses the priorities to maintain postural balance over motor tasks when one or two tasks involve postural control [23] or threat to the balance [24]. Thus, changes in the onset time of the tibialis anterior muscle were observed when that pushing an object while standing on a moving platform [10]. On the other hand, when simultaneously experiencing translation perturbations but imposed externally by a mechanical device increases in exerted force were observed in both pushing force and grip force when performing pushing [10] or holding an object [27]. The outcomes of these prior studies indicate that the CNS prioritizes the performance of the motor task (focal arm movements in pushing) [10] or griping reactions [27] to maintain vertical posture and overcome the effect of the translation perturbations.

In the current study, while the smaller magnitudes of the COP displacement and exerted hand force were observed in the unlocked condition. The backward sliding board movement was in the same direction of body thrust in pushing [10]. Activation of shank muscles in standing is important as it allows generating the ground reaction torque around the ankles and related COP displacement needed for control of vertical posture. For example, activation of ventral muscle serving ankle joint in standing allows moving COP backward and activation of dorsal muscle allows moving COP forward [16,17,18]. In order to assist trunk bending movement, anticipatory activation of either ventral muscles or inhibition of dorsal muscles could be used to shift COP backward [19, 20]. In addition, activation of dorsal muscles plays an important role for backward translation perturbation [21] and to shift COP forward. Hence, trunk inertia could be used a tool for increasing pushing force to reach the sufficient exerted force [22] and counteracting the destabilizing force of gravity for the smaller COP displacement in the unlocked condition compared to the locked condition.

Some limitations should be mentioned and noted. The current findings might not generate to older adults since older adults already showed muscle co-contraction and inefficient pushing strategies when performing pushing tasks while standing on a solid surface. In case, older adults utilize the same strategy of remaining too strong muscle co-contraction under the sliding surface and performing the pushing task. It has been reported that segments with stiffness affect postural adjustments corresponding to forthcoming perturbations [24], which might subsequently put elderly at higher risk of falling. However, the further study should be conducted with middle aged or aging workers to evaluate their pushing and balance responses when standing on the moving surface.

5 Conclusion

Exerted pushing force induced backward sliding board movement when standing on the unlocked surface. Standing on the unlocked sliding board resulted in the delay of the COP onset time and decreases in the magnitudes of the COP displacement. The pushing task was considered as the primary perturbation and the moving surface was the collateral perturbation to balance control. These results reveal that the CNS is capable of controlling the motor task and postural component when dealing with two self-initiated perturbations: primary (pushing) and subsequent (translational). In addition, the inertia of the trunk was used to assist in generation of the pushing force needed to suffice the pushing task supported a decreased magnitude of the pushing force in the unlocked condition. Meanwhile, body inertia allowed counteracting the destabilizing effect of force of gravity as smaller COP displacements were seen when the surface was movable. The findings also provide a basis for future studies focused on the development of potential working environment design for middle aged or aging workers by modifying movability of the standing platform.