Twenty-two healthy adult participants (11 males and 11 females) were recruited for this laboratory-based study from the Oregon State University community via email solicitations. The means (standard deviations) of the participants’ age, weight, and height were 30.3 (7.9) years, 79.1 (17.0) kg, and 173.3 (7.9) cm. Participants were recruited into the study if they 1) were 18–49 years-old; 2) had no musculoskeletal pain in the past 7 days and no history of musculoskeletal disorders in the neck, shoulder, and back regions; 3) had no self-reported visual or vestibular deficits that may influence normal balance control; and 4) were not pregnant.

The experimental protocol was approved by Oregon State University’s Institutional Review Board. Informed consent was obtained from participants prior to experimental sessions. The experimental setup included a car seat located on a 6-degree-of-freedom high-fidelity motion platform (MB-E–6DOF/24/1800 KG; Moog Inc.; East Aurora; NY) to create realistic WBV exposures in a laboratory setting (Fig. 1 (a)). After demographic and anthropometric data were collected, the seat height was adjusted for each participant so that their feet could rest firmly on a force platform (AMTI OR6-7-1000, Watertown, MA, USA) mounted in front of the seat while their thighs were parallel to the force platform (Fig. 1 (a)). For each participant, the same seat height was maintained across all experimental conditions to control the participants’ posture and its potential impact on postural control measures (Mazzà et al., 2004; Whitney et al., 2005; Zimmermann and Cook 1997). Next, reflective markers were placed on participants’ upper and lower body to collect kinematic measures. During an initial practice session, the participants performed a series of postural control assessment tasks (Fig. 1) until they verbally confirmed their understanding of and proficiency with them and overall study protocols.

In a repeated-measures experiment, participants sat on the heavy equipment vehicle seat, mounted on the motion platform, and were exposed for a total of 4 h to each experimental condition (Fig. 2). This 4-h session was broken up into two sessions (exposure 1 and exposure 2) lasting 2 h each, with a 30-min break in between. This was repeated for four different WBV exposure conditions on four separate days: No WBV exposure, vertical-axial dominant WBV exposure with a single-axial passive suspension seat, multi-axial WBV exposure with a single-axial passive suspension seat, and multi-axial WBV exposure with a multi-axial active suspension seat intervention. A minimum 24-h gap was maintained between the experimental conditions to avoid any potential residual effects from a prior condition. The order of all conditions was randomized across participants to minimize any potential systematic bias due to the order of exposure conditions. During the two 2-h sessions, monotonic and neutral nature documentary films were displayed on a flat screen to keep participants’ neck posture similar to driving postures while minimizing boredom. Immediately before and after each 2-h exposure, postural control was assessed during sit-to-stand transitions, static stance on a foam surface, and stair descent (Fig. 1). Altogether, the postural control assessments took about 5 min. In total, each experimental session lasted for 5 h, including the study protocol briefing, practice session, the two 2-h exposure sessions with a 30-min midterm break, and postural control assessments.

The sit-to-stand transition started with participants in a seated position on the vehicle seat with their bare feet resting on the force platform in front of the seat and their arms on their chests (Fig. 1 (a)). They were then instructed to stand up without using their arms as soon as they heard an auditory cue and to remain as still as possible for approximately 5 s until receiving another auditory cue (Norman-Gerum and McPhee 2020). We did not control the speed of the sit-to-stand transition to allow participants to perform the task at a self-selected speed (Jeon et al. 2021; Norman-Gerum and McPhee 2020; Petrella et al., 2021). The sit-to-stand transition was performed with eyes open condition and repeated three times (i.e., self-perturbation). Participants then were asked to self-evaluate their perceived confidence that they did not lose balance or become unsteady while performing the sit-to-stand transition tasks and verbally announced their confidence level on a scale of 0–10 (0: no confidence, 10: completely confident) (Powell and Myers 1995).

For the static stance task (Fig. 1 (b)), participants were asked to stand on a foam pad (thickness: 7.5 cm) placed on the force platform in front of the seat and remain as still as possible for 30 s, with their bare feet close together, arms resting on their sides, head facing straight ahead, and eyes open (Lin et al., 2008). The instruction to ‘stand as still as possible’ was issued to minimize inter-individual differences (Zok et al. 2008). The static task was performed once on a foam pad, with no repetition. Considering that healthy adults, like the participants in this study, primarily depend on somatosensory information to maintain their balance, a foam surface was used to challenge postural control during a static stance (Maitre and Paillard 2016) and improve the sensitivity of detecting the changes in postural control measures related to WBV. During the sit-to-stand transitions and static stance trials, tri-axial ground reaction force (GRF) and moments were collected using the force platform at a sampling rate of 1000 Hz.

Subsequently, participants were directed to dismount the motion platform and stand on a three-step staircase constructed according to OSHA 1910/25 standards (riser height = 18 cm, tread width = 90 cm, tread depth = 20 cm) at the ground level (Fig. 1 (c)). The top of the stairs was equipped with a force platform (AMTI BMS464508–2K-SYS, Watertown, MA, USA). The stair descent task started with participants standing upright on the force platform for 5 s in a comfortable position (arms resting on their sides and no restrictions on the distance between their feet). Then, as soon as they received an auditory cue, participants stepped down the staircase. This task was performed with eyes open condition and repeated twice. During the stair descents, tri-axial GRFs and ground reaction moments were collected at 1000 Hz using the force platform to evaluate anticipatory postural adjustments (APA) preceding stair descent. To evaluate postural control during the stair descent tasks, kinematic data were collected at 100 Hz using an optical motion capture system (Flex 13; Optitrack; Natural Point, OR) and reflective markers. The reflective markers were placed bilaterally on the radius styloid process, humerus lateral epicondyle, clavicle-acromion, anterior and posterior superior iliac spines, lateral knee condyle and ankle malleolus anatomical landmarks. Additionally, four tracking markers were placed on the front center of the thighs and shins, along with markers placed on each foot at the tip of second toe, heel, and the first and fifth metatarsals to establish the base of support (BoS).

The motion planform (Fig. 1 (a)) was used to play two types of tri-axial vibration profiles collected from long-haul trucks and mining vehicles in field settings (Johnson et al., 2018; Kim et al., 2018a,b) as shown in Fig. 3. The detailed approaches for the WBV simulation are described in our previous studies (Kia et al., 2020, 2022). The vertical-dominant vibration, which had the vertical (Z) axis as the dominant axis, represented the exposures from long-haul trucks operated on paved terrain, including city streets, freeways, and interstate highways (Kim et al., 2016). The multi-axial vibration, in which WBV parameters were greater in the lateral (Y) and fore-aft (X) axes than in the vertical axis, represented the WBV exposures collected from various mining vehicles in off-road terrains and their corresponding vehicle movements (Kim et al. 2018; Dennerlein et al., 2022).

During the WBV simulations, vibration data (i.e., acceleration) were collected at the floor of the motion platform and the seat level using an 8-channel vibration acquisition system that consisted of a data logger (DA-40; Rion Co. LTD; Tokyo, Japan) and two accelerometers (Model 356B40; PCB Piezotronics; Depew, NY) at a sampling rate of 1280 Hz. The vibration data were processed and summarized per ISO 2631–1 WBV standards at the floor and seat levels (ISO, 1997) (Fig. 3).

Two different suspension seats were evaluated in this study: a vertical passive air suspension seat (control) and a multi-axial electromagnetic active suspension seat (intervention). To minimize the potential confounding influence of different seat designs and maximize the blinding effect, we used the same seat (BoseRide Prototype; Bose Corporation; Framingham, MA, USA) with two suspension settings: vertical passive air suspension and multi-axial electromagnetic active suspension. The single-axial (vertical) passive suspension seat is a current industry-standard seat with passive pneumatic suspension that uses passive components of compressed air and dampers to attenuate vertical WBV. Due to its fixed response and low resonant frequencies, the passive pneumatic suspensions have been found to be ineffective in reducing impulsive exposures that are commonly experienced in off-road vehicles (Kim et al. 2018). The multi-axial active suspension seat continuously measures both vertical and lateral vibration using a built-in inertial measurement unit (IMU). The built-in microprocessor uses seat position and acceleration data to control two highly responsive electromagnetic linear actuators. These control seat travel and counteract the road-induced vibration disturbances. Due to far greater fidelity in frequency response, this multi-axial active suspension seat can effectively attenuate not only the low frequency vibration exposures but also the higher frequency impulsive exposures. Both of these exposures can be difficult for traditional pneumatic seat suspension systems to effectively control. In addition, this seat has a passive fore-aft mechanical suspension to isolate the fore-aft vibration. These features make the multi-axial active suspension seat a particularly well-suited engineering control for off-road vehicles such as construction, forestry, and mining vehicles.

The dependent variables (i.e., outcome measures) are summarized in Table 1. The independent variables were CONDITION (4 levels: No WBV, vertical-dominant WBV, multi-axial WBV, and multi-axial WBV with intervention) and TIME (4 levels: 0, 120, 150, 270 min). A repeated-measures analysis of variance (ANOVA) was used to determine the main and interaction effects of CONDITION and TIME on each outcome measure. Sex-associated interaction effects were explored initially. As the interaction effects of CONDITION, TIME, and SEX were not significant for most of outcome measures, SEX was excluded in the final ANOVA model. Parametric model assumptions were assessed using Shapiro-Wilk tests, and data transformations were performed to meet these assumptions as relevant (Table 2). Partial eta-squared (ηp2) was used to quantify effect sizes for the main and interaction effects. Given the study goals, the subsequent presentation of results and the discussion emphasizes the interaction effects of CONDITION and TIME. When such significant interaction effects were found, linear contrast analyses were set up for specifically testing the relative impacts of vertical-dominant WBV (vs. No WBV), multi-axial WBV (vs. vertical-dominant WBV), and multi-axial WBV with intervention (vs. multi-axial WBV) on changes in each of the outcome measures following exposure 1, break, exposure 2, and total (exposure 1 + break + exposure 2). Summary results were back-transformed and are presented in the original units as least-squares means (95% confidence intervals) unless stated otherwise. All statistical analyses were completed using JMP^® Pro (v. 16.0, SAS Institute Inc., Cary, NC), with statistical significance noted when p < 0.05.

During sit-to-stand transition, significant interaction effects of CONDITION and TIME were observed in GRF peak-to-peak amplitude (Table 2). Specifically, a decrease in GRF peak-to-peak amplitude following total exposure was larger in vertical-dominant WBV than in both No WBV and multi-axial WBV conditions (Fig. 6, Table A1, and Table A2). Furthermore, multi-axial WBV with intervention led to a greater decrease in GRF peak-to-peak amplitude following both break and total exposure as compared to multi-axial WBV condition (Fig. 6 and Table A3).

During upright stance on a foam surface, CoP RMS distance in the AP direction showed significant interaction effects of CONDITION and TIME (Table 2). Linear contrast analyses showed that a decrease in CoP RMS distance in the AP direction following a break was larger in No WBV than in vertical-dominant WBV condition (Fig. 7 and Table A1). Conversely, an increase in this measure was greater in multi-axial WBV than in multi-axial WBV with intervention conditions (Fig. 7 and Table A3).

During anticipatory postural adjustments prior to stair descent, significant interaction effects of CONDITION and TIME were seen in imbalance amplitude in the ML direction only (Table 2). Specifically, vertical-dominant WBV led to a larger increase in imbalance amplitude in the ML direction following total exposure as compared to both No WBV and multi-axial WBV conditions (Fig. 8, Table A1, and Table A2). An increase in this measure following exposure 2 was also greater in vertical-dominant WBV than in No WBV condition (Fig. 8 and Table A1).

Sitting while being exposed to vertical-dominant WBV resulted in a more pronounced decrease in GRF peak-to-peak amplitude during sit-to-stand transition compared to sitting without WBV (Fig. 6). Decreased GRF amplitude is indicative of inability to generate force quickly and maintain postural stability, thereby implying impaired postural control during sit-to-stand transition (Kera et al., 2020; Lord et al., 2002). Muscular load and fatigue caused by prolonged exposure to vertical-dominant WBV (Kia et al., 2020, 2022; Kim et al., 2018a,b; Troxel et al., 2016) may have contributed to diminished force exertion capabilities and a reduction in GRF peak-to-peak force amplitude. Such changes can be indicative of a higher risk of losing balance (Cameron et al., 2003; Kera et al., 2022b; Simpkins and Yang 2022). Vertical-dominant WBV also led to a greater increase in imbalance amplitude in the ML direction compared to the control condition (No WBV) during APA prior to stair descent (Fig. 8). Increased imbalance amplitude can suggest deteriorated APA (Callisaya et al., 2016). Hence, our first hypothesis was confirmed. Also, these results suggest that vertical-dominant WBV, common in on-road vehicles, can have adverse impacts on dynamic postural control during sit-to-stand transition and APA prior to stair descent, which may lead to increased risk of falls during vehicle egress.

Nonetheless, the effects of vertical-dominant WBV on the postural control measures were relatively small. Briefly, the effects of the changes in GRF peak-to-peak amplitude during sit-to-stand transition observed in this study were small: 3% decrease in the vertical-dominant WBV compared to 1% decrease in No WBV (Fig. 6). These changes were relatively smaller than other studies analyzing clinical changes. For example, this measure was 19% lower among patients with sarcopenia (vs. non-sarcopenia) (Kera et al., 2022a), 14% lower among older males (vs. young males) (Smith et al., 2020), and 9% lower among elderly with locomotive syndrome (vs. elderly without locomotive syndrome) (Yamako et al., 2023). Moreover, vertical-dominant WBV increased imbalance amplitude in the ML direction by 21–33% during stair descent, as compared to 8–20% decrease in No WBV (Fig. 8). To the best of our knowledge, only one study, which assessed APA preceding stair descent using CoP data, did not show significant differences between individuals with patellofemoral osteoarthritis and asymptomatic controls (Wyndow et al., 2019). Given that prior work using these measures have predominantly considered clinical conditions where such differences may be amplified, whether the small but statistically significant changes observed with vibration exposure in this study hold practical meaning is not currently clear.

In relation to other postural control measures used in this study, such as those assessed during static stance on a foam surface and stair descent, vertical-dominant WBV did not lead to significant differences in postural control when compared to No WBV condition. Several potential factors may account for these observations. First, exposure to vertical-dominant WBV may not compromise sensory information or muscle strength substantially enough to affect static stance. Second, participants may have recovered from the effects of WBV exposure by the time that static stance and stair descent tasks were conducted, as these were the last tasks we tested (Kowalski et al. 2021; Johanson et al., 2011). Third, given that the stairs used in this study were designed with a wider width and lower height compared to those typically found in heavy vehicles, the stair descent task in this study may have been less demanding than typical egress from heavy vehicles.

As compared with vertical-dominant WBV, multi-axial WBV did not present more detrimental effects on postural control. Instead, vertical-dominant WBV led to a greater decrease in GRF peak-to-peak amplitude compared to multi-axial WBV during sit-to-stand transition (Fig. 6). Furthermore, imbalance amplitude in the ML direction during APA prior to stair descent increased more with vertical-dominant WBV compared to multi-axial WBV (Fig. 8). For the remaining postural control measures utilized in this study, multi-axial and vertical-dominant WBV did not have distinctly different effects on postural control. These results fail to support our second hypothesis. These findings are also somewhat contradictory to our previous findings, where we observed that multi-axial WBV led to greater impairments on postural control (than vertical-dominant WBV) during static stance on a firm surface (Park et al., 2021), as opposed to the foam surface used in this study. Hence, our results suggest that the lateral component of multi-axial WBV from off-road vehicles may not have additional adverse effects on dynamic postural control during sit-to-stand transition and stair descent when compared to vertical-dominant WBV experienced in on-road vehicles.

For multi-axial WBV exposures, using a multi-axial active suspension seat (intervention) did not demonstrate superior performance in alleviating the impact of multi-axial WBV on postural control compared to the vertical-axial passive suspension seat (the current industry standard). Rather, multi-axial WBV with multi-axial active suspension led to a larger decrease in GRF peak-to-peak amplitude during sit-to-stand transition compared to multi-axial WBV with single-axial passive suspension (Fig. 6). These results can be attributed to the fact that the multi-axial active suspension seat did not substantially outperform the single-axial passive suspension seat in reducing the lateral component (Y-axis) of WBV (Fig. 3), or due to the fact that the lateral component of vibration associated with multi-axial WBV itself did not significantly compromise postural control compared to vertical-dominant WBV. Some previous studies have also reported the limited efficacy of the multi-axial active suspension seat in mitigating lateral WBV (Kia et al., 2022; Park et al., 2021). However, interestingly, the multi-axial active suspension seat led to significantly better recovery in postural sway measures associated with static stance during the break between the successive 2-h long exposures to WBV, compared to the standard single-axial passive suspension seat (Fig. 7). Hence, while the seat did not produce differences in pre-post postural control measures during each vibration period, its use seems to have promoted recovery of postural control during the breaks.

Although these results fail to support our third hypothesis and further emphasize the need for more effective interventions that can reduce the lateral component of WBV, the potential for a multi-axial active suspension seat to promote faster recovery from the detrimental effects of WBV needs to be better understood.

Vibration exposures did not significantly affect CoP RMS distance in the AP direction during static stance when compared to No WBV (sitting without vibration). However, this measure recovered at a significantly better rate in No WBV condition than in the other WBV conditions during the 30-min break, following the first exposure (Fig. 7). This is in accordance with results from previous studies, indicating that incorporating breaks between prolonged sitting can effectively mitigate muscular fatigue (Thorp et al., 2014; Kar and Hedge 2020; Ding et al., 2020), subsequently reducing the risks associated with postural control loss (Kowalski et al. 2021; Johanson et al., 2011). These results collectively highlight the value of having periodic breaks in drivers’ work schedules to prevent potential postural control impairment, and that exposure to WBV may cause slower recovery in postural control measures compared to prolonged sitting without WBV exposure.

There were a few limitations to our study. First, this study was conducted in a controlled laboratory setting in which the environment (e.g., road conditions, noise, traffic, WBV) and tasks (i.e., driving) may be different from actual driving conditions. For example, participants in this study were not engaged in driving tasks but instructed to watch monotonic nature documentaries while exposed to WBV. Previous investigations have shown that cognitively demanding tasks can increase trunk sway (Arippa et al., 2022; Leban et al., 2017), which may in turn affect postural control. However, in an effort to make our WBV exposure more realistic, this study utilized actual field-measured vibration data collected during regular vehicle operations (from on-road semi-trucks and off-road mining vehicles), to better simulate a professional vehicle operator’s daily WBV exposures. Therefore, the WBV exposures in this laboratory study were far more realistic, especially when compared to previous WBV studies (Arora and Grenier 2013; Martin et al., 1981; Stamenkovic et al. 2014). Second, this study did not examine the long-term effects of WBV exposure on postural control. A longitudinal study may be needed to investigate changes in postural control through cumulative exposures to WBV as professional vehicle operators are regularly exposed to hours of WBV over the course of their careers. Third, because this study included a comprehensive set of postural control measures, potential balance recovery over the course of the postural control assessments may have diminished the effects of WBV on the measures collected later in the experimental sessions. A previous study showed that postural balance itself can recover when dynamic movements are performed (e.g., sit-to-transition, static standing, and gait initiation) and so impairments in balance may diminish with delayed assessment (Pollard et al., 2017). Lastly, the sit-to-stand transition in our study, adapted from clinical postural control assessments, may differ from real-world scenarios of egressing a vehicle. In vehicle egress, drivers may use their arms, have different body orientations, and may not fully stand due to the low-rise ceiling, unlike the conditions in our study and clinical assessments. Also, the dynamic postural control assessments in this study have been primarily used in clinical settings to understand the association between postural control measures and increased fall risks. Therefore, these measures may not be directly compared to those in clinical settings or linked directly to fall risks. However, they can provide valuable insights into the effect of WBV on postural control during high-risk dynamic activities such as egressing from a vehicle.