Mechanical work and energy of sit-to-stand and stand-to-sit transitions performed with traditional and dynamic office chair designs

1 Introduction

Prolonged sitting and standing are both strongly associated with low back pain development [1, 2] and adverse health outcomes [3–5]. Sit-stand workstations have been frequently implemented as an ergonomic intervention aimed at reducing sitting exposures and encouraging intermittent movement. Although dependent on the type of training [6–8] and continuity of organizational support [6], sit-stand workstations have been shown to reduce daily sitting time by 10.6% to 40% [6, 9–11], and increase the number of active movement events performed daily [12, 13]. Furthermore, improvements in psychological outcomes such as self-confidence/efficacy [14] and mood [15] have been self-reported when using sit-stand workstations. Taken together, sit-standworkstation usage can have a generally positive influence on behavioural, physiological, and psychological outcomes [16].

Despite the potential benefits, long-term adherence to sit-stand workstation designs decreased when measured by the number of daily transitions performed post-intervention [6, 17]. In an attempt to understand the reasoning underlying the disuse of sit-stand workstations, Renaud et al. [18] conducted a qualitative thematic analysis. The first theme that emerged was the reported feeling of discomfort and tiredness from a period of standing [18]. The bodily region of discomfort was not specified; however, a similar effect of standing periods on low back discomfort has been demonstrated [15, 20]. Specifically, standing-induced low back pain was observed in 55% of individuals and not completely resolved by intermittent sitting work breaks [19]. Secondary to discomfort, additional themes that emerged were related to the practicality and user interface of sit-stand workstations [18]. These included difficulty in adjusting and finding the correct workstation height and performing desk work that requires a high level of focus and concentration.

In consideration of the reported reasons for disuse, alternative chair and/or workstation configurations have the potential to ease transitions and improve adherence to sit-stand workstations. One way to evaluate the ease of transitions is to quantify the mechanical work and energy required to transition. Specifically, reports of tiredness, cumbersome workstation adjustments, and focus may be collectively addressed with a human-chair-workstation interface that minimizes the mechanical work and energy required to transition between sitting and standing postures. In eliminating much of the mechanical work required to transition, users are expected to maintain many active benefits associated with sit-stand workstation usage [12, 13]. Furthermore, if transitions that require less mechanical work lead to more frequent transitions, the potential for sitting- and standing-induced low back discomfort may be reduced, especially if performed early in the exposure [21]. To date, the biomechanical and ergonomic analyses of sit-stand workstations have largely focused on postural assessments, muscle activation, and subjective discomfort reports without explicit mechanical analyses of the transition [19, 21–23]. To address this knowledge gap, this study aimed to characterize and compare the work and total energy during sit-to-stand and stand-to-sit transitions using a traditional office chair and a novel dynamic chair designed for specific use with sit-stand workstations. The work and total energy of the whole-body, thigh, and shank centre-of-mass (CoM) were evaluated. It was hypothesized that transitions performed with the dynamic chair design would reduce the overall (i.e., absolute sum between all CoM points) mechanical work and energy.

2 Methods

2.3 Experimental procedure

Each participant completed three intermittent sit-to-stand and three stand-to-sit transitions at an interval of 3-minutes while completing a desk task. Transitions to and from a traditional office chair (Fig. 1) and a novel office chair (Movably Pro, Movably Inc., Lexington, MA, USA) designed specifically for a sit-stand workstation (Fig. 2) were examined on separate testing days and were an extension of a larger and seperate 3-hour data collection on sedentary exposures. Similar to a traditional office chair, the dynamic chair had an adjustable but fixed seat pan height. The dynamic chair seat pan was constructed with two separate pieces that moved independently between horizontal (i.e., seated) and vertical (i.e., standing) positions during transition. This design enabled a constant chair and body position with respect to a fixed desk position for sitting and standing work performed with the dynamic chair. That is, the desk height was not adjusted between sitting and standing tasks when the dynamic chair was used as required for traditional sit-stand workstation usage.

OPEN IN VIEWER

Fig. 1

Sit (A) and standing (B) portions of the traditional transitions.

OPEN IN VIEWER

Fig. 2

Transitions performed to and from the dynamic seat pan. A) sitting; B) transition; C) standing.

For both chair conditions, a familiarization period was performed to adjust the seat pan height such that a 90-degree trunk-thigh, knee, and ankle angle (90-90-90) were facilitated. For the traditional chair, separate workstation heights were marked for the sitting and standing portions of the task, and a laboratory assistant moved the desk to the pre-marked heights during each transition. A single workstation height was used for the dynamic chair condition. In all cases, the workstation height was adjusted to enable a horizontal forearm position.

Transitions to and from the dynamic seat pan were initiated following a tactile vibration prompt. From a seated position (Fig. 2A), the participant lifted one thigh to unload the ipsilateral portion of the seat pan, which initiated its retraction to a vertical position (Fig. 2B). This enabled the participant to weight bear on a single leg to offload the contralateral portion of the seat pan and assume a standing position (Fig. 2C). From a standing position, half of the seat pan returned to a horizontal position, which facilitated a transfer of weight onto the seated limb while the contralateral portion returned to horizontal.

2.4 Data processing

All accelerometer data were input into a custom Matlab program (2020b, The MathWorks Inc., Natick, MA) for processing. Raw voltage data were padded [27, 28] and then filtered using a dual-pass fourth order low-pass Butterworth filter and an effective cutoff frequency of 3 Hz [29, 30]. Using data from the six calibration trials, all accelerometer channels were calibrated with respect to gravity. Inclination of the pelvis in the sagittal (i.e., tilt) and frontal (i.e., obliquity) planes along with the sagittal plane inclination of the L1 accelerometer were determined using standard four quadrant inverse tangent functions [30]. The lumbar spine flexion-extension angle was quantified as the difference between the sagittal plane inclination obtained from the S1 and L1 accelerometers. Instantaneous pelvis and lumbar spine angles were subtracted from the respective angles derived during the standing calibration trial. Lumbar spine range-of-motion (RoM) was defined as the absolute difference in the maximum and minimum lumbar spine angle exhibited during the transition.

The interface between the participant and seat back of both chairs prevented the use of a rigid cluster to track the pelvis position and orientation. As such, pelvis tilt and pelvis obliquity angles derived from S1 accelerometer data were downsampled by a factor of 10 and used to calculate the three-dimensional position of anatomical landmarks used to reconstruct a pelvis-fixed local coordinate system [30]. This approach enabled the inclusion of a pelvis segment in the linked-segment model for the calculation of whole-body CoM position.

Marker position data, including calculated anatomical landmarks of the pelvis, were input to Visual3Dtrademark motion analysis software (Version 2020, C-Motion, Germantown, MA). Raw position data were padded [27, 28] and then smoothed with a dual-pass fourth order low-pass Butterworth filter with an effective cutoff frequency of 3 Hz [29, 30]. Marker position data acquired from the standing calibration trial were used to construct a linked-segment model containing the trunk, pelvis, thigh, and shank segments. Participant-specific anthropometrics (i.e., height, body mass) were also used for model construction. Using the defined proximal and distal segment endpoints, anatomically meaningful segment-fixed local coordinate systems were reconstructed in accordance with International Society of Biomechanics recommendations [31].

The time-varying CoM position of the linked-segment model and both thigh and shank segments were calculated in Visual3D using the segment inertial properties (i.e., computed from segment mass) and radii of the proximal and distal segment endpoints [32, 33]. The total energy of each CoM point (i.e., whole body, right and left thigh, right and left shank) was calculated as: E total = mgh + 1 2 mv 2 where, m, was equal to the body, thigh, or shank mass for the whole-body, thigh, and shank CoM, respectively. The thigh and shank CoM were derived using standard proportion methods [33, 34]. The time-varying Y-coordinates (i.e., vertical direction) of each CoM point were input as, h, and used to determine the vertical velocity, v. The work performed on each CoM point was determined as: W = F × d where the vertical displacement, d, was iteratively calculated using the Y-coordinate data and force, F, was derived from the product of the mass assigned to each CoM point and gravity. As such, a positive work value resulted from a positive vertical CoM displacement (i.e., upward) while a negative work value resulted from a negative vertical CoM displacement (i.e., downward). For each transition the work and energy quantified for the right and left thighs were iteratively summed to represent the combined work (Fig. 3) and energy (Fig. 4) of the thigh and shank segments. The maximum change in CoM work and energy between the sitting and standing portions of each transition was calculated and included in the statistical analyses.

OPEN IN VIEWER

Fig. 3

Representative work data series for the whole-body CoM, thigh CoM, and shank CoM. For the thigh and shank, the solid line = right leg, dashed line = left leg, and bold line = summed waveform.

OPEN IN VIEWER

Fig. 4

Representative total energy data series for the whole-body CoM, thigh CoM, and shank CoM. For the thigh and shank, the solid line = right leg, dashed line = left leg, and bold line = summed waveform.

Significant chair × transition interaction effects were observed for whole-body CoM work (p < 0.001), whole-body CoM total energy (p < 0.001), thigh CoM work (p < 0.001), thigh CoM total energy (p < 0.001), and shank CoM total energy (p < 0.001). There were no significant main effects or interactions observed for shank CoM work (p≥0.347) and lumbar spine ROM (p≥0.564).

As evidenced in Fig. 5 the change in work and total energy of the whole-body CoM was approximately equal in absolute magnitude but opposite in direction for transitions within a given chair. The difference in whole-body CoM work between the traditional and dynamic chairs was 13.6 J (p < 0.001) and 14.1 J (p = 0.005) for sit-to-stand and stand-to-sit transitions, respectively (Fig. 5). The whole-body CoM work was significantly different between sit-to-stand and stand-to-sit transitions performed with a traditional office chair (22.58 J, p < 0.001), but not the dynamic chair (5.12 J, p = 0.753). Further, the change in total body energy significantly differed between chairs by a magnitude of 280.6 J (p < 0.001) and 280.4 J (p < 0.001) for sit-to-stand and stand-to-sit transitions, respectively (Fig. 5). These significant findings are driven by the whole-body CoM undergoing a larger vertical displacement when transitioning from the traditional office chair.

OPEN IN VIEWER

Fig. 5

Average whole-body CoM work (left) and total energy (right). Within a transition, significant differences between chairs are indicated with an asterisk (*).

Thigh CoM work was significantly greater for traditional chair than the dynamic chair for sit-to-stand (3.4 J, p = 0.019) and stand-to-sit (3.1 J, p = 0.024) transitions (Fig. 6). A similar trend was observed for total thigh CoM energy (Fig. 6). The difference in total thigh CoM energy between chairs was 59.6 J (p < 0.001) and 59.1 J (p < 0.001) for sit-to-stand and stand-to-sit transitions, respectively. Given the similarities in thigh CoM work and total energy magnitudes associated with positive (0.1 J, 0.8 J) and negative (0.2 J, 0.3 J) changes, these significant findings are driven by the examined chairs facilitating an equal but opposite effect on thigh displacement for a given transition.

OPEN IN VIEWER

Fig. 6

Average thigh CoM work (left) and total energy (right). Within a transition, significant differences between chairs are indicated with an asterisk (*). Error bars represent one standard deviation.

The differences in shank CoM work during sit-to-stand (0.9 J, p = 0.989) and stand-to-sit (0.9 J, p = 0.997) transitions were not significant (Fig. 7). However, the average total energy of the shank CoM was significantly greater with the dynamic chair compared to the traditional during sit-to-stand (43.1 J, p = 0.001) and stand-to-sit (46.7 J, p < 0.001) transitions (Fig. 7). The total shank CoM energy did not significantly differ between transitions performed with the traditional chair. These differences in total shank CoM energy are driven by larger changes in shank displacement facilitated by the dynamic chair.

OPEN IN VIEWER

Fig. 7

Average shank CoM work (left) and total energy (right). Within a transition, significant differences between chairs are indicated with an asterisk (*). Error bars represent one standard deviation.

The lumbar spine RoM was similar between chairs and transitions (Fig. 8). On average, the dynamic chair facilitated a greater lumbar RoM by 1.6 degrees and 2.3 degrees for sit-to-stand and stand-to-sit transitions, respectively. These differences were not significantly different (p≥.968).

OPEN IN VIEWER

Fig. 8

Average lumbar spine RoM. Error bars represent one standard deviation.

This study characterized the mechanical energy and work required to transition to and from a traditional and a dynamic chair intended for sit-stand workstation use. Compared to the traditional office chair, transitions performed with the dynamic chair design, on average, reduced the whole-body CoM work and total energy by ±8.5 J and ± 214.6 J, respectively. Although nominal differences were observed for the thigh CoM (work =± 0.18 J, energy =± 0.55 J), the dynamic chair facilitated a greater change in total shank CoM energy (± 38.3 J). Differences in shank CoM work (± 1.1 J) were not significant between chairs. In support of the study hypothesis, the overall work and energy required to transition was approximately 56% and 64% less when the dynamic chair was used, despite the non-uniform work and total energy responses between the whole-body, thigh, and shank CoM points. It is interesting to note that the observed differences in mechanical work and energy occurred with similar magnitudes of lumbar spine RoM. These findings suggest that the mechanical work of sit-to-stand and stand-to-sit transitions can be modified by chair design despite similarities in biomechanical outcome measures.

While there were significant differences in work and energy requirements between the two chairs, the reduction in mechanical work required to transition from the dynamic chair is expected to have a negligible effect on daily energy expenditure. For example, a sit-stand training program instructed users to stand at least once within every 50-minute work period, resulting in approximately 7 sit-to-stand transitions and 140 minutes of standing per work day [17]. Using the whole-body CoM sit-to-stand work calculated in the current study (1 J of work is equal to 0.239 kcal), the caloric output resulting from these seven transitions is estimated to be 4.1 and 17.1 calories per day for the dynamic and traditional chairs, respectively. Given the similarity in energy expenditure between sitting and standing –1.18 kcal/min of sitting and 1.31 kcal/min of standing [35, 36] – sit-to-stand transitions, regardless of the number performed and chair used, are unlikely to appreciably influence metabolic expenditure in occupations requiring prolonged desk work. However, if minimizing the mechanical work required for sit-to-stand elicits more intermittent daily standing exposures, users may transition more frequently and therefore benefit from the improved physiological outcomes related to frequent movement events while limiting total standing [3] and sitting [37] time related to negative health outcomes.

In reference to the investigation by Renaud et al. [18], the leading causes for the disuse of sit-stand workstations were the development of discomfort, difficulty adjusting the work surface height, and concentration. With respect to productivity and focus, traditional sit-stand workstations have shown no significant effect on the volume of work completed during keying and mousing tasks compared to sitting and standing alone [23, 38]. Furthermore, early and frequent seated breaks from standing reduced perceived low back pain in known pain developers compared to prolonged standing [21]. The above evidence collectively suggests that frequent but efficient transitions, as demonstrated by the reduced mechanical work associated with dynamic chair usage, have potential to improve low back pain while maintaining work productivity. The only remaining question is the how individuals may respond to the frequency of transitions required to elicit this beneficial response. In extrapolating the 3-minute sit:stand transition interval used in the current study to a standard work day, it would elicit 75 transitions per day. Whether it is possible and feasible to complete such a number of transitions, and to address limited use of sit-stand workstations due to inefficiency, strategies that minimize the mechanical work appear to be an approach that merits further design exploration.

Minimizing the required work and energy of sit-to-stand and stand-to-sit transitions has potential to improve prolonged adherence to sit-stand workstation usage. Not only was a reduced mechanical work demonstrated with the dynamic seat design, when used in auto-mode, the transition is effectively initiated by timed tactile prompts. For example, a seated position was initiated by components of the seat pan swinging forward and in effect, prompted hip and knee flexion to initiate sitting. In this scenario it would have required more work to ignore or terminate the prompt then to simply transition between postures. This automated feature may help individuals overcome persistent habits of siting upon implementation without compromising productivity and discomfort [18]. The process of transitioning was further simplified without the need to adjust the workstation height. An additional advantage of a dynamic seat pan in favour of a height adjustable workstation is the presence of the seat back during the standing work portions (Fig. 2). Supported standing may prevent lumbar extension that is largely driven by anterior pelvis tilt and/or thoracic extension, which is typically seen in low back pain developers [19]. Of course, further research is required to understand the biomechanical outcomes during sitting and standing exposures associated with the dynamic chair design. Future work will also investigate the long-term adherence to the dynamic chair used with a sit-stand workstation.

5 Conclusion

The mechanical work and energy of sit-to-stand and stand-to-sit workstations was modified by the chair design. Overall, the dynamic chair design reduced the work and energy required to transition by 56% and 64%, respectively. Minimizing the mechanical work and energy required to transition may help mitigate current barriers to long-term sit-stand workstation usage, including habitual sitting, discomfort, manual adjustment of the work surface height, and concentration.

Ethical Approval

The study was approved by the University of Waterloo (Protocol number 43643).

Informed Consent

All participants provided written informed consent prior to the completion of experimental procedures. All informed consent documentation was approved by the Office of Research Ethics.

Conflict of interest

The authors declare that they have no conflict of interest.