E ﬀ ects of a Wearable Carriage Aid on Whole-Body Physiological Measures and Balance

: Carriage tasks are common and can lead to shoulder and lower back injuries. Wearable carriage aids have shown mixed e ﬀ ects on local physical demand measures. This study examined the impact of a wearable carriage aid on whole-body physiological measures (normalized oxygen consumption, minute ventilation, respiratory rate, and heart rate) to obtain a more comprehensive assessment regarding aid e ﬀ ectiveness. Additionally, this study investigated the e ﬀ ect of wearing the device on perceived balance. The potential moderating e ﬀ ect of carried load mass was considered. The examination was conducted while walking on a treadmill at a constant speed (2 km / h) for 5 min and was completed by 16 participants. Wearing the device reduced normalized oxygen consumption (~14%), minute ventilation (~7%), and heart rate (~3%), while substantially improving perceived balance (~61%). These e ﬀ ects were consistent across examined carried load levels. Although this study highlighted the potential for the developed aid, future studies are required for more diverse and realistic testing conditions.


Introduction
Manual material handling (MMH) tasks are still frequent in many occupational sectors despite developments made in production technologies [1,2]. Potential motives to keep these tasks manual include benefiting from the worker skill, experience, and movement flexibility [3]. Interventions targeting physically demanding tasks can be broadly classified as administrative or engineering. A frequently used engineering approach is to automate targeted tasks. However, this solution can be practically impossible or requires unreasonably high capital investments. Wearable assistive devices are increasingly emerging for occupational use and may be a suitable alternative or aid to existing intervention methods [4]. This is because using these devices can require relatively minor changes to a present workplace and can be more cost-efficient.
However, the described wearable assistive devices (also known as exoskeletons) have been mostly assessed for military, rehabilitation, or medical purposes as an aid for patients who are injured, weak, or disabled [5,6]. A study showed that these devices can considerably reduce physical demands for occupational use [7]. However, wearing these devices can also have unintended influences such as raising physical demands on other body regions while reducing demands on targeted body parts. For example, wearing a passive exoskeleton reduced back and leg demands but increased chest discomfort in a simulated assembly task that involved a truck [8].
Load carriage is a common MMH task and is considered a combination of lifting and pulling or pushing tasks [9]. Load carriage was considered as one of the three most frequently performed MMH tasks in the U.S. [9]. For the U.S. Army soldiers, carrying and lifting tasks were specified as the most common physically demanding tasks [10]. Golriz and Walker [11] showed evidence that carrying tasks can lead to lower back and shoulder pain. To highlight the economic influences of these injuries, the annual direct costs for lower back and shoulder injuries were estimated to be at least $50 billion [12] and $7 billion [13], respectively.
Generally, carriage wearable aids redistribute the load over the wearer's body to reduce demands on the lower back and shoulder regions [14]. A few studies have specifically investigated the effectiveness of these devices in reducing carrying task physical demands. Muslim and Nussbaum [15] found that a simple wearable carriage aid can reduce lumbosacral moments after load carriage tasks. Investigating these wearable devices is commonly completed using regional physical demand measures. While these measures can provide specific effects, knowledge on the global (or whole-body) physical demand measures might be limited. Evaluating the whole-body effects is critical as per the described common and potential unintended consequences of wearing these devices. Therefore, this study examined the influence of a wearable carriage aid on whole-body physiological measures (normalized oxygen consumption, respiratory rate, minute ventilation, and heart rate) while considering the potential moderating influences of carried load mass. Additionally, this study assessed the effect of wearing this device on perceived postural balance. The proposed aid was examined earlier using regional measures and it showed potential to reduce shoulder and arm demands while slightly increasing lower back demands (cf., [14]). To completely understand the effectiveness of the developed aid, this study was conducted as per the whole-body measure. We hypothesized that the aid would improve all physiological measures and perceived postural balance regardless of the carried load mass.

Participants
A convenience sample of 16 male participants was randomly selected for this study. The required sample size was calculated based on the effect size (partial eta-squared: η 2 p ) for an essential measure in a similar study [15]. More specifically, using the obtained η 2 p of 0.088, Type I error of 0.05, and a power of 0.8, the study needed sixteen participants as calculated by G*Power software [16]. All the participants were from the university population, reported having no current or previous history of lower back pain, and were physically active. Participants signed an informed consent form that was approved by the Human Participants Review Sub-Committee of the Institutional Review Board of King Saud University (research project # E-18-3533). A symbolic amount of money ($13/h) was given to the participants to compensate for their time. The sample mean (SD) age, weight, and height was 34.63 (5.54) years, 69.97 (7.85) kg, and 164.66 (4.89) cm, respectively.

Task Description
The task was to carry a load bimanually and anteriorly while walking on a treadmill at a fixed speed (2 km/h) for 5 min. These specific experimental conditions were used in similar studies [17,18] and were assumed to exemplify common scenarios from a practical perspective.

Independent Variables
This study comprised two within-subject factors (Device and Load). The former factor comprised two conditions (with and without wearing the device). The device was designed to reduce the physical demands of the carriage task by redistributing the load of the carried masses on the wearer's body and by making the wearer's posture closer to neutrality. The device weighed 2.5 kg (for more specific details on the device, please refer to [14]). This study examined the Load factor to understand if the hypothesized effects of the device were moderated by the amount of carried load. Two Load conditions were tested and set relative to each participant's body mass: light (15%) and heavy (30%). Pilot testing was conducted to confirm participants' capability of completing the task under all examined conditions. The carried load was modified using 3 kg masses by the experimenter. Figure 1 shows the experimental setup for the two Device conditions.

Procedures and Data Collection
The study used a repeated measures design. More specifically, participants completed the described task under all four combinations of Device and Load factors. The experiment was conducted in one session that lasted for almost 2.5 h and was done in a laboratory environment. To counterbalance the order of presenting the four conditions, 4 × 4 Balanced Latin Squares were used. Participants rested for at least 10 min between conditions to reduce any effects from residual muscle fatigue.
Furthermore, the experimental conditions were described to the participants at the beginning of the session. Next, they signed the informed consent forms. After obtaining their anthropometric measures, they practiced walking on the treadmill for almost 5 min at the specified speed. After at least 10 min of rest, they were set up to perform the first testing condition as determined by the noted counterbalancing procedure. Before the first testing, they wore the oxygen consumption measurement mask and heart rate monitoring electrodes (Mega Electronics Ltd., Kuopio, Finland). Alcohol swabs and Ag/AgCl adhesive electrodes were used as detailed in [19]. Participants rested for at least 10 min after completing each condition as indicated earlier. At the start of this rest period, participants completed a perceived balance questionnaire (explained below).

Dependent Variables
As whole-body physiological metrics, normalized oxygen consumption (VO2: mL/kg), minute ventilation (VE: 1/min), respiratory frequency (Freq: breath/min), and heart rate (HR: beats/min) were collected. The mean values of these measures were determined over the testing duration for each condition. The questionnaire that was developed by Chiou et al. [20] was used to evaluate the effect of wearing the device on perceived postural balance. The questionnaire comprised four simple questions. These questions were: (1) "How much did you feel your body sway?", (2) "Did you have any difficulty maintaining balance?", (3) "Did you feel at any time that you would fall?", and (4) "What was the overall difficulty of this task?" For each of these questions, participants rated on a visual analog scale from 0 "not at all" to 10 "a lot". The perceived instability (PI) was computed as the summation of the four ratings. The used questionnaire validity and repeatability were investigated earlier under different experimental conditions (cf., [19]). The described balance metric was used earlier by Simeonov et al. [21] to examine the impact of footwear characteristics on perceived instability.

Statistical Analysis
To evaluate the influences of Device and Load, separate 2 × 2 repeated measures analyses of variance were used. Parametric model assumptions were examined and appropriate data transformation was used to meet these assumptions as required. The order of presenting the four conditions was insignificant on any examined dependent measures. Simple-effects tests were used for post hoc comparisons. Partial eta-squared (η 2 p ) was calculated to quantify the effect sizes. All statistical tests were considered significant when p < 0.05. MATLAB 2015a (The Math Works Inc., Natick, MA, USA) was used for data processing and Minitab statistical software (Minitab v.18, Minitab Inc., State College, PA, USA) was used for the statistical analyses.

Perceived Postural Instability
Wearing the device significantly improved the perceived balance by~61% (Figure 3). Additionally, carrying a heavier load led to a significantly higher perceived balance (~245%) with this effect being independent of the device being worn or not (Table 1).

Discussion
This study investigated the effectiveness of a wearable carriage aid using whole-body physiological measures while considering the potential moderating influence of carried load mass. Additionally, this study examined whether wearing the device impacts perceived postural balance. Overall, wearing the device reduced normalized oxygen consumption, minute ventilation, and HR during the carriage task while walking on a treadmill for 5 min. Additionally, wearing the device increased perceived postural balance. The indicated effects were consistent across both Load levels.
Wearing the device led to a lower normalized oxygen consumption, minute ventilation, and HR ( Figure 2). Generally, these findings highlight that the reduction of the indicated physiological metrics due to aid assistance exceeded the additional demands due to wearing the device. However, it should be recalled that the magnitude of reduction ranged from 3-14%. While it is unclear if the noted magnitude of reduction is practically meaningful in terms of injury risk, the consistency of these whole-body measures points positively toward the effectiveness of the developed aid reducing carriage task physical demands. In line with these whole-body measures, the developed aid reduced median and static shoulder muscle activity during a carriage task [14]. The design of wearable aids dictated its influence on whole-body physiological measures. Supporting this, Gregorczyk et al. [22] examined a lower-body wearable assistive device for a carriage task while walking and found it to increase oxygen consumption. Additionally, a passive trunk exoskeleton increased metabolic cost while walking but decreased it during lifting [23]. Finally, Theurel et al. [24] showed that wearing an upper-body exoskeleton did not change the task cardiac cost while walking and carrying a load.
Carrying a heavier load led to increasing all physiological metrics examined here (Table 1) regardless of whether the device was worn. Supporting this pattern of results, Gregorczyk et al. [22] showed that increasing carried load (as defined by three levels: 20, 40, and 55 kg) increased metabolic cost regardless of whether a lower-body assistive device was worn. Unlike results found here, Gao et al. [17] found that influences of increasing carried load (as defined by three conditions: 5, 10, and 15 kg) on metabolic demands can be lower when wearing a load-carrying belt. These findings generally suggest that the effectiveness of wearable assistive devices can depend on the carried load, with this interaction effect being dependent on the worn device's design characteristics (e.g., load redistribution mechanism and device weight).
Perceived postural balance was substantially improved when wearing the device (Figure 3). This is potentially because participants were more conscious about controlling their postural balance when wearing the device. Conformably, a lower-body carriage assistive device reduced static postural sway [25]. Inconsistent with the positive impact on balance found here and using objective balance metrics, an upper-body exoskeleton increased the mean center of pressure velocity in the anteroposterior direction in simulated overhead work [26]. These findings show that the wearable aid designs can impact postural control differently.
This study has a few limitations. While conducting a laboratory study can ensure relatively high internal validity, the level of external validity is unclear. The extent to which the short testing duration used (5 min for each condition) can reflect long-term effects is unknown. Furthermore, the study used male participants only. It is unclear if the results can be generalized to females, particularly because gender affects center of gravity [27] and metabolic capacity [28,29]. Additionally, the device's effectiveness was examined when positioning the carried load on one location of the body. Muslim and Nussbaum [15] found that the effectiveness of carriage aid can be moderated by the carried load location. Therefore, it is unclear if positioning the loads on different locations of the body will give a similar pattern of results. Furthermore, walking speed is a critical factor that determines carriage task physical demands [30]. While the study examined one walking speed that is thought to be practically common, it is unclear if the results can be generalized to other walking speeds. Additionally, there are other practically-relevant issues of the device that need evaluation. Among these, the loading/unloading and donning/doffing requirements were not examined here. Lastly, the study did not investigate if wearing the device constrains the wearer's movement. Altering movement can influence the wearer's ability to respond to slip and trip risks [26].

Conclusions
A laboratory-based study was conducted to evaluate the impact of a wearable carriage aid on whole-body physiological measures and perceived postural balance during walking on a treadmill for 5 min. The potential moderating influence of carried load mass was also considered. While greatly increasing perceived postural control, wearing the device reduced normalized oxygen consumption, minute ventilation, and heart rate. The found effects were consistent across the two examined carried load levels. Increasing the carried load mass increased all the tested physiological measures while also improving perceived balance. Despite the limitations discussed above, the study showed a potential for the simple developed device to reduce carriage task physical demands. However, more research is needed under more realistic and diverse conditions.