Stooping, crouching, and standing – Characterizing balance control strategies across postures

Abstract

Background

While stooping and crouching postures are critical for many activities of daily living, little is known about the balance control mechanisms employed during these postures. Accordingly, the purpose of this study was to characterize the mechanisms driving net center of pressure (COP_Net) movement across three postures (standing, stooping, and crouching) and to investigate if control in each posture was influenced by time.

Methods

Ten young adults performed the three postures for 60 s each. Kinetic signals were collected via a force platform under each foot. To quantify mechanisms of control, correlations (Correl_LR) were calculated between the left and right COP trajectories in the anterior-posterior (AP) and medio-lateral (ML) directions. To examine the potential effects of time on balance control strategies, outcomes during the first 30 s were compared to the last 30 s.

Results

Correl_LR values did not differ across postures (AP: p = 0.395; ML: p = 0.647). Further, there were no main effects of time on Correl_LR (AP: p = 0.976; ML: p = 0.105). A significant posture-time interaction was observed in the ML direction (p = 0.045) characterized by 35% decreases in Correl_LR over time for stooping (p = 0.022).

Conclusion

The dominant controllers of sway (i.e., AP: ankle plantar/dorsi flexors; ML: hip load/unload mechanism) are similar across quiet stance stooping, and crouching. Changes in ML control strategies over time suggests that fatigue could affect prolonged stooping more so than crouching or standing.

Introduction

Stooping and crouching postures are required to perform important daily tasks such as reaching for items on the floor/low shelves, and gardening (Long and Pavalko, 2004). These postures involve musculoskeletal configurations that differ greatly from upright standing (Gallagher et al., 2011, Glinka et al., 2015). Stooping consists of flexing the trunk forward and rotating the head downward while keeping the legs relatively straight. Crouching requires significant flexion at the hip, knee, and ankle while maintaining a vertically oriented trunk. While the challenges associated with stooping and crouching have been examined in a number of recent studies (Bhattacharya et al., 2009, Dionisio et al., 2008, Gallagher et al., 2011, Glinka et al., 2015, Hemmerich et al., 2006, Hernandez et al., 2010, Hernandez et al., 2013, Kuo et al., 2011), little is known about how balance is actually maintained once an individual is in these postures. It is possible that the mechanisms we employ to control the body׳s net center of pressure (COP_Net) while standing (Winter et al., 1993) are different from those used during stooping and crouching. Understanding how these mechanisms may differ could contribute to our knowledge of balance-related challenges associated with tasks that require stooping and crouching postures, and ultimately inform intervention strategies aimed at improving stooping and crouching performance.

The body׳s COP_Net is typically quantified from kinetic signals recorded using a single force platform under an individual׳s feet (Mackey and Robinovitch, 2005, Panzer et al., 1995, Prieto et al., 1996). However, toward understanding the separate mechanisms responsible for controlling COP_Net movement in the anterior-posterior (AP) and medial-lateral (ML) directions, two force platforms (one under each foot) are required (Winter et al., 1993). Using two platforms facilitated the pioneering work of Winter et al. (1993) who identified that COP_Net movement in the AP direction is achieved primarily by varying the stiffness of the ankle plantar flexors and dorsiflexors, while ML control relies on the hip ab/adductors to transfer weight from one foot to the other (Winter et al., 1993, Winter et al., 1998). This early work laid the foundation for more recent studies investigating how directional (AP vs ML) balance control during upright standing can be affected either by experimentally manipulating sensory modalities, or naturally by aging and pathological processes (Morrison et al., 2016, Peterka, 2002, Termoz et al., 2008).

As stooping and crouching postures involve body configurations that differ substantially compared to standing (e.g., joint angles, muscle lengths, joint loads), the directional control mechanisms employed (i.e., ankle plantar/dorsi flexors for AP; hip load/unload for ML) might also change (Weaver et al., 2014). It is well known that when muscles act outside of their optimal operating lengths, their ability to generate force is diminished (Edman, 1966, Zajac, 1989). During stooping, the lower back and posterior leg muscles are stretched (Burgess-Limerick, 2003). Crouching involves similar length changes, particularly to the quadriceps and ankle plantar flexor muscles, as a consequence of the significant flexion occurring at the knee and ankle joints (Burgess-Limerick, 2003, Glinka et al., 2015). Further, in the crouching posture, activating only the hip musculature may be insufficient to load/unload bodyweight from each foot. Unlike standing, in a crouch the hip and knee joints are all fully flexed, potentially necessitating activation of the quadriceps and calf musculature – in addition to the hip ab/adductors – to lift the entire lower limb and initiate a weight shift between legs. It is possible that some of these postural constraints prompt a reorganization of the motor strategies utilized to maintain balance during upright stance.

It is also of interest to determine whether time spent in stooping and crouching postures influences the control strategies employed. Borrowing from the vast literature on the lower back, prolonged trunk flexion has been shown to induce viscoelastic tissue creep (McGill and Brown, 1992, Shin et al., 2009, Shin and Mirka, 2007, Solomonow et al., 2003, Toosizadeh and Nussbaum, 2013). Accordingly, fatigue of the low back extensor muscles may also occur during stooping, as these muscles must generate more active force to compensate for the reduced extension moment contribution of creep-deformed passive tissues (Shin et al., 2009, Shin and Mirka, 2007). Research has also shown that in the lower back, muscle onset can be delayed due to tissue creep, which may in turn influence sensorimotor function (Sanchez-Zuriaga et al., 2010). Therefore, it is conceivable that creep and fatigue-related changes could also occur in the lower-limbs during periods of quasi-static stooping and crouching, resulting in changes in the control of postural sway.

Despite potential drivers of between-posture differences in control strategies (as detailed above), Weaver et al. (2014) recently demonstrated that balance control during both stooping and crouching can be well-represented using the inverted-pendulum model originally developed to characterize control strategies during quiet stance (Winter et al., 1998). Specifically, all three postures demonstrated similar relationships between center of mass (COM) acceleration and the COP_Net-COM signal (Weaver et al., 2014). However, it is currently unknown which strategies are used to control postural sway while stooping and crouching. Such information would have functional relevance as stooping and crouching postures are common during activities of daily living (ADLs) and in many occupational tasks, but performing these tasks becomes more difficult for older adults (Glinka et al., 2015, Hernandez et al., 2013, Kuo et al., 2011). Knowledge of underlying control strategies across these postures could provide evidence to support generic intervention approaches that are robust across postures, or alternatively, highlight the need to develop separate intervention approaches tailored to each specific task.

Accordingly, the purpose of this study was to determine if known strategies of COP_Net control for upright standing are applicable to stooping and crouching postures. Specifically, we hypothesized that posture would not influence: 1) the synchronicity between the COP of the left and right feet, in the (a) AP or (b) ML direction, assessed via Pearson product-moment correlations; 2) the degree of synchronicity between the COP_Net and COP components controlled by either the ankle plantar/dorsi flexor or hip load/unload mechanisms, in the (a) AP or (b) ML direction, also assessed via Pearson product-moment correlations; and 3) the magnitude of each proposed mechanism׳s (ankle plantar/dorsi flexion versus hip load/unloading) contribution to balance control in the (a) AP or (b) ML direction, assessed using root mean square amplitudes. Additionally, as a secondary objective, we investigated whether extended time spent in each posture influenced balance control characteristics. Specifically, we hypothesized that no main effects of time or interaction effects between posture and time would be observed in the metrics calculated for hypotheses 1–3 (referred to as hypothesis 4 in the remainder of the document).