Gait initiation (GI), the transition from quiet standing to walking, is a complex task requiring precise postural control for stability (McIlroy and Maki, 1999, Rogers et al., 2001). The critical phase involves reducing the base of support (BoS) to a single stance foot as the swing foot lifts, rendering the body susceptible to mediolateral (ML) instability if the center of mass (CoM) is not appropriately positioned (Lyon and Day, 1997). The central nervous system generates anticipatory postural adjustments (APAs), which are referred to as dynamic pre-movements (Bouisset and Do, 2008, Yiou et al., 2012), to initiate volitional movements such as walking. During GI, this process includes a characteristic shift in the center of pressure (CoP) towards the swing limb (McIlroy and Maki, 1999, Yiou et al., 2017, Yiou et al., 2011). This CoP movement is not merely a strategy to maintain stability, but a mechanical necessity: it generates the moment required to displace the CoM over the stance limb, thereby unloading the swing foot and enabling forward progression (Brenière et al., 1987). The timing and shape of the CoP shift trajectory may reflect aspects of anticipatory neural control.

The margin of stability (MoS), which denotes the distance between the extrapolated CoM (XCoM) and BoS boundary, quantifies dynamic postural stability during GI (Singer et al., 2013). Studies have shown that healthy adults adapt their APA to maintain the MoS under varying conditions. For example, during GI at different self-selected speeds, individuals modulate their APA and step width such that the MoS at swing heel contact remains constant (Caderby et al., 2014). Individuals can also modify their APA by widening the CoP shift to preserve the MoS under constrained conditions (Yiou et al., 2012). However, the effectiveness of APAs varies across populations and tasks (Caderby et al., 2020), highlighting the adaptability of the central nervous system in managing dynamic postural balance. Thus, the MoS is a critical quantitative indicator of dynamic postural stability during GI.

Compared with normal GI, obstacle crossing during GI (GIObs) is a challenging task that introduces a significant constraint, as the swing foot must be raised to cross the obstacle, potentially altering the timing and magnitude of APA (Yiou et al., 2016). A lifted leg lift may require a forceful or prolonged ML CoP shift to maintain lateral stability (Zettel et al., 2002). Furthermore, crossing requires increased precision in foot placement and timing, placing a higher demand on the sensorimotor system (Deshpande et al., 2011). Although previous studies have examined obstacle crossing during continuous gait, adaptations during GI require further investigation. Obstacle crossing offers a unique context for examining adaptive mechanisms supporting balance that are not apparent in standard GI (Caderby et al., 2014).

Some studies have specifically examined the adaptive mechanisms underlying dynamic postural balance during the complex task of GIObs, particularly in adults. Although studies on anteroposterior (AP) (Ledebt et al., 1998, Malouin and Richards, 2000) and ML (Mani et al., 2019, Palluel et al., 2008) APA development during GI have proposed direction-specific developmental trajectories (Blanchet et al., 2019, Ledebt et al., 1998, Malouin and Richards, 2000), the combined demands of GI and obstacle crossing on APA and MoS modulation in children remain unexplored.

This study addressed this gap by examining APA and MoS modulation in children and adults during GI with and without an obstacle. Understanding these differences is crucial for developing interventions for the dynamic postural balance of children, considering their developing neuromotor control. Specifically, we compared the timing and magnitude of ML and AP CoP shifts, as well as the MoS, of children and adults during GI and GIObs. We hypothesized that children would exhibit smaller APA and MoS than adults, particularly during obstacle crossing. These differences may reflect adaptive strategies that are suited to their body structure and neuromuscular development, although such strategies may appear less refined or more variable owing to the ongoing maturation of postural control. This study focused on coordinated ML and AP movements and revealed the extent to which children's neuromotor control can adapt to maintain stability in the presence of obstacles. Identifying APA and MoS differences between children and adults can inform interventions for pediatric dynamic postural balance.