Posterior parietal cortex activity during visually cued gait: a preliminary study

32 Safe gait requires visually cued (VC) step adjustments for negotiating targets and obstacles. 33 Effective step adjustments rely on good visuospatial processing. The posterior parietal cortex 34 (PPC) is implicated in visuospatial processing, yet empirical evidence is limited for the PPC’s 35 role during gait in humans. Increased cortical control of gait is associated with higher gait 36 variability, a marker of gait performance and fall risk among older adults. However, the 37 cortical underpinnings of gait variability in visually complex environments are not well 38 established. The primary aim of this preliminary study was to assess PPC activity during VC 39 gait and VC gait with perturbations (VCP). A secondary aim was to determine how PPC 40 activity relates to gait variability during VC and VCP gait. Twenty-one healthy young adults 41 completed three treadmill gait conditions at preferred speed: non-cued (NC) gait, VC gait, 42 where stepping targets were presented in a regular pattern, and VCP gait, where stepping 43 target positions were pseudorandomly shifted. Functional near-infrared spectroscopy 44 quantified relative changes in deoxygenated and oxygenated hemoglobin (ΔHbO2) 45 concentrations in the PPC. Inertial measurement units quantified gait variability. Moderate 46 effects were observed for more positive ΔHbO2 from NC to both VC and VCP gait, likely 47 reflecting the increased visuospatial processing demands. Stride time variability was 48 positively correlated with PPC ΔHbO2 during VC gait, suggesting a potential role for the PPC 49 in modulating temporal components of VC gait. Extending these findings to older adults will 50 help to elucidate the PPC’s role in gait adaptability and fall risk with aging. 51 52

53 fNIRS; mobile neuroimaging; walking; precision stepping; gait variability; inertial 54 measurement unit 55 56 57 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 3

58 Everyday gait requires visually cued (VC) step adjustments, which are coordinated by cortical 59 and subcortical pathways, to navigate complex gait environments. Mobile functional near-60 infrared spectroscopy (fNIRS) studies report increased prefrontal cortex (PFC) activity during 61 tasks involving precision stepping (Koenraadt et al., 2014; Le et al., 2023) and obstacle 62 negotiation (Chen et al., 2017; Mirelman et al., 2017, Pelicioni et al., 2022). Greater 63 recruitment of PFC resources for VC gait contrasts with the limited PFC recruitment observed 64 for non-cued gait, highlighting the complexity of the task. Though task performance may 65 decline as task demands increase, this adaptive neural response reflects an effort to 66 maintain performance (Reuter-Lorenz & Cappell, 2008). Unfortunately, the involvement of 67 cortical regions beyond the PFC during VC gait remains comparatively understudied in 68 humans. Emerging evidence indicates that the posterior parietal cortex (PPC) may play a 69 critical role during VC gait (Drew et al., 2023; Nordin et al., 2019). 70 Interspecies studies implicate the PPC in planning VC gait modifications (Drew & Marigold, 71 2015; Nordin et al., 2019). The PPC processes and relays visual information concerning 72 environmental targets or obstacles to guide the goal-directed trajectory and foot placement 73 of the stepping limb (Buneo & Andersen, 2006; Marigold & Drew, 2017). Indeed, specific 74 cells residing in the cat PPC encode either the time or distance to contact with an obstacle 75 (Marigold & Drew, 2017). Using electroencephalography (EEG) in humans, Nordin et al. 76 (2019) demonstrated that the PPC is reliably activated preceding the step over an obstacle 77 during gait. During precision stepping paradigms with targets presented at fixed or variable 78 distances, Yokoyama et al. (2021) reported higher PPC activity in the variable condition using 79 EEG, while Le et al. (2023) reported no change in activity in the superior parietal lobule (a 80 sub-region of the PPC) using fNIRS. A notable difference between the paradigms used by 81 these authors was that the former manipulated both step length and width during the 82 variable condition, while the latter manipulated step length only. Considering that 83 mediolateral step adjustments represent a greater deviation from the habitual gait pattern 84 (Hoogkamer et al., 2015), it is conceivable that these adjustments carry greater visuospatial 85 processing demands, eliciting higher PPC activity. This idea aligns with suggestions that 86 compared to anteroposterior gait adjustments, mediolateral gait adjustments are under 87 more active control via sensory integrative feedback because the passive dynamics of gait 88 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 4 are less stable in this direction (O’Connor & Kuo, 2009). Accordingly, the removal of visual 89 information during gait disproportionately affects mediolateral gait dynamics compared to 90 anteroposterior gait dynamics (Wuehr et al., 2013). 91 Gait variability (i.e., stride-to-stride fluctuations of spatiotemporal gait cycle parameters) is 92 increasingly used as a marker for gait performance and fall risk (Lord et al., 2011). Gait 93 variability is linked to gait automaticity (i.e., limited cortical control of gait), such that higher 94 variability purportedly reflects reduced gait automaticity (Nóbrega-Sousa et al., 2020). Gait 95 variability can differentiate older adults with and without mobility and cognitive impairment 96 (Hausdorff, 1998; Verghese, 2008). Further, more variable stride time and stride length can 97 predict falls in older adults (Hausdorff et al., 2001; Maki, 1997). That gait must dynamically 98 adapt to negotiate targets and obstacles complicates the interpretation of gait variability in 99 visually complex environments. Higher stride-to-stride variability can reflect this adaptability 100 in response to environmental demands (Beauchet et al., 2009). Among both young and older 101 adults, gait variability increases during obstacle negotiation (Mirelman et al., 2017; Nóbrega-102 Sousa et al., 2020). Moreover, positive associations between gait variability and fNIRS-103 measured PFC activity during obstacle negotiation (Mirelman et al., 2017) and during 104 cognitive dual-task gait (Nóbrega-Sousa et al., 2020) highlight the importance of higher 105 order brain regions for complex gait control. Unfortunately, beyond the above-mentioned 106 associations with PFC activity (Mirelman et al., 2017; Nóbrega-Sousa et al., 2020), the 107 cortical underpinnings of gait variability remain insufficiently characterized. Delineating the 108 neural correlates of gait variability during tasks that mimic the visual processing demands of 109 real-world gait could provide critical insights for neuromotor-based interventions aimed at 110 reducing mobility impairment and fall risk in older adults. 111 While previous studies assessed PPC activity during VC gait, studies relating PPC activity to 112 gait performance are scarce. Pizzamiglio et al. (2018) reported that during dual-task gait, 113 increased PPC activity in young adults predicted lower mediolateral center of mass motion, 114 suggesting a role for the PPC in regulating gait dynamics in complex gait environments. 115 Moreover, considering that the PPC engages to register and store the spatiotemporal 116 relationship between the body and environment (Marigold & Drew, 2017), there are likely 117 important functional associations between PPC activity and gait variability. Indeed, magnetic 118 resonance imaging studies report that reduced PPC grey matter volume (Beauchet et al., 119 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 5 2014) and integrity (Tian et al., 2016) are associated with increased temporal gait variability. 120 Extending this work by using fNIRS to relate real-time PPC activity to gait variability during a 121 VC gait task is a clear next step toward understanding the cortical mechanisms responsible 122 for gait impairment and fall risk in clinical populations. Examining these relationships in 123 young adults can provide useful baseline data to help better understand cortical deficiencies 124 that occur across the lifespan. 125 The primary aim of this preliminary study is to determine PPC activity patterns during VC gait 126 in young adults using fNIRS. A secondary aim is to determine how PPC activity relates to gait 127 variability during VC gait. Since the PPC appears important for fast step adjustments 128 (Potocanac & Duysens, 2017), and inaccurate reactive step adjustments relate to fall risk 129 (Robinovitch et al., 2013), we use a VC gait task which requires reactive step adjustments. 130 Our gait conditions comprise: 1) non-cued (NC) treadmill gait, 2) VC treadmill gait, where 131 stepping targets follow a regular pattern, and 3) VC treadmill gait with perturbations (VCP), 132 where stepping targets undergo pseudorandom position shifts. As with obstacle negotiation, 133 cueing participants to make reactive step adjustments imposes gait variability. We therefore 134 anticipate that gait variability measures will be highest during the VCP gait. We hypothesize 135 that compared to NC gait, PPC activity will increase during VC gait and will increase further 136 during VCP gait. Additionally, we hypothesize that increased PPC activity will relate to 137 increased gait variability during both VC and VCP gait. 138 139

140 Participants 141 Twenty-one healthy young adults (22.2 [3.6] years, 8F) were enrolled after meeting the 142 inclusion and exclusion criteria. Inclusion criteria comprised being 18-29 years old. Exclusion 143 criteria comprised having a chronic musculoskeletal or neurological injury or disease; having 144 a cardiovascular disease or arrhythmia; surgery that would affect gait within the past year; 145 uncorrected vision impairment; or inability to understand written and spoken English. All 146 participants provided informed consent before participating. A post-hoc power analysis was 147 conducted using G*Power 3.1 to determine the achieved power for a repeated measures 148 ANOVA, based on our observed medium effect size (f=0.35), alpha of 0.05, and a sample size 149 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 6 of 21. Results indicated an achieved power of 0.93. This study was conducted in accordance 150 with the Declaration of Helsinki and was approved by the Institutional Review Board at the 151 University of Massachusetts Amherst (IRB #3501). 152 Gait assessment 153 To quantify gait variability, participants were fitted with six body-worn inertial measurement 154 units (IMUs) (Opal v2, APDM, a Clario Company, Portland, OR), sampling at 128Hz. IMUs 155 were affixed to the feet, wrists, sternum, and lumbar spine (near the natural waistline, 156 where a belt would be worn) using Velcro straps. IMUs were wirelessly synchronized through 157 Mobility Lab ™ v2 software (APDM), which was used to initialize and collect the data. Gait 158 measures reflect the full 3-minute walking period for each condition. The following 159 outcomes measures were calculated by Mobility Lab™ proprietary algorithms: gait speed 160 (m/s), stride length (m), stride length variability (m), stride time (s), stride time variability (s), 161 lumbar mediolateral range of motion (RoM) (°), and lumbar mediolateral RoM variability (°). 162 Variability measures were computed by Mobility Lab™ as the standard deviation of each 163 measure across each walking condition, from which we calculated the coefficient of variation 164 (CoV; (SD/mean) × 100) (%). CoV measures comprised the primary outcome measures from 165 the gait assessment. 166 Gait conditions 167 Participants completed three 3-minute gait conditions on a split-belt treadmill (Bertec, 168 Columbus, OH) at self-selected speed in the order listed: NC gait, VC gait, and VCP gait. 169 During the second and third condition, illuminated rectangular visual stepping targets, 170 attuned to each participant’s foot size and average step length and width, were projected 171 onto the treadmill belts, and approached the participant at belt speed (Figure 1). 172 Participants were instructed to step as accurately as possible onto the targets, aiming to hit 173 the center of each rectangular target with the imagined center of their foot sole. During the 174 VCP condition, target shifts in the anterior, posterior, or lateral direction were 175 pseudorandomly imposed every 3-7 steps, requiring a step adjustment. Target shifts 176 occurred when the target came within 130% of the participant’s average step length 177 (measured from a reflective motion capture marker worn on the sacrum) (Mazaheri et al., 178 2015). Targets were shifted by 40% of the participant’s average step length in the anterior or 179 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 7 posterior direction, or by 20% of step length in the lateral direction (Mazaheri et al., 2015). 180 Immediately prior to each 3-minute condition, there was a 20-second quiet standing period, 181 serving as a baseline for cortical activity measures. 182 183 184 Fig.1 Schematic displaying the experimental set-up for the visually cued gait conditions. Participants 185 walked on a split-belt treadmill while wearing a safety harness. Purple rectangles are stepping 186 targets, attuned to each participant’s foot size and average step length and width, presented 187 consistently during visually cued gait. The red rectangle is an example of a perturbed target during 188 perturbed visually cued gait, undergoing a lateral positional shift, upon coming within 130% of step 189 length (green) from a sacrum-worn reflective marker (not shown). Target positional shifts in the 190 anterior-posterior direction were also imposed (not shown) 191 192 fNIRS data acquisition 193 Relative changes in oxygenated hemoglobin (HbO2) and deoxygenated hemoglobin (HHb) 194 were quantified during each gait condition with a portable, 24-channel continuous-wave 195 fNIRS device sampling at 50Hz (Dual Brite MKII; Artinis Medical Systems, Netherlands). 196 Twenty-two long channels (source-detector distance: 30mm) and two short-separation 197 channels (source-detector distance: 10mm) were split evenly over the left and right 198 hemispheres. After fitting the fNIRS cap according to each participant’s head circumference, 199 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 8 a digitizer (Patriot 3D Digitizer, Polhemus, Vermont) was used to digitize optode positions 200 based on anatomical references (nasion, inion, bilateral preauricular points, and vertex). 201 Parietal positions P3 and P4 from the International 10-20 System were determined. P3 and 202 P4 demonstrably correspond to Brodmann area 7 (Homan et al., 1987). Visual inspection of 203 channel locations was performed in the Patriot system GUI. From the available channels, 204 four (two per hemisphere) channels were selected for analysis, corresponding to the P3 and 205 P4 positions (i.e., left and right PPC; Figure 2). Selected channels were consistent across all 206 participants. Two short-separation channels were employed (one per hemisphere) 207 positioned at T4 and T5 on the 10-20 system. 208 209 210 Fig.2 fNIRS optode configuration comprising 10 transmitters (yellow) and 8 receivers (blue). Key 211 channels (transmitter-receiver combinations) selected for analysis are indicated with green ellipses 212 213 fNIRS data were processed in MATLAB 2023b. First, a processing pipeline involving selected 214 functions from the HOMER3 package (Huppert et al., 2009) was applied. Raw data were 215 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 9 visually inspected, and a channel pruning function (hmr_PruneChannels) was employed to 216 flag channels with a signal to noise ratio of less than 5 (von Lühmann et al., 2020). Raw light 217 intensity was converted to optical density (hmr_Intensity2OD). A motion artifact detection 218 function (hmr_MotionArtifactByChannel; tMotion = 1.0, tMask = 1.0, STDEVthresh = 40.0, 219 AMPthresh = 5.0) was then used to identify spike-like motion artifacts based on where the 220 signal exceeded a threshold in change of standard deviation within a predefined time 221 window (Huppert et al., 2009). This threshold was set at 40 after visual inspection of the 222 dataset, to optimize artifact detection, following the same approach used by others (Cooper 223 et al., 2012; Di Lorenzo et al., 2019). A cubic spline interpolation (hmr_MotionCorrectSpline; 224 p = 0.99) corrected the artifacts flagged by the detection algorithm (Scholkmann et al., 225 2010). A wavelet filter was next applied (hmr_MotionCorrectWavelet; iqr = 1.50), in line with 226 reports that combined spline and wavelet corrections yield better motion artifact removal 227 than either function alone (Di Lorenzo et al., 2019; Gao et al., 2022). A low-pass filter with a 228 cutoff frequency of 0.14 Hz was applied to remove high-frequency noise (hmr_BandpassFilt). 229 Optical density data were then converted to HbO2 and HHb concentrations via the modified 230 Beer-Lambert law (hmr_OD2Conc), with a partial pathlength factor of 6 (Boas et al., 2004; 231 von Lühmann et al., 2020). Any remaining motion artifacts were addressed by applying the 232 gait-specific correlation-based signal improvement (CBSI) method (hmr_MotionCorrectCbsi) 233 (Cui et al., 2010). While the CBSI method is most effective for calculating an activation signal 234 that combines HbO2 and HHb, it also applies a correction (albeit nominal) on both the HbO2 235 and HHb signals separately. Applying this additional motion artifact correction step ensured 236 a more conservative motion artifact correction approach, as previously used (Martini et al., 237 2024; Vitorio et al., 2020). A short-separation channel regression was performed to remove 238 superficial hemodynamic responses from the four channels of interest (bilateral PPC). A 239 baseline correction was performed: the mean of the signal from the middle 10 seconds of 240 the 20-second quiet period was subtracted from the 3-minute walking period. We excluded 241 the first 5s of the baseline period to avoid any residual activity following the earlier verbal 242 instruction. We excluded the last 5s (i.e., immediately prior to gait initiation) to avoid 243 anticipatory responses, as others have done (de Belli et al., 2021). Finally, the four channels 244 were median averaged, such that the outcome measures comprised the median relative 245 changes in HbO2 (ΔHbO2) and HHb (ΔHHb) concentration during each gait condition. The 246 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 10 median relative ΔHbO2 and ΔHHb concentration was preferred to the mean, as it would be 247 less sensitive to a potential long tail of the hemodynamic response over the 3-minute walk. 248 Statistical analysis 249 Repeated measures ANOVAs were used to compare differences in the median relative 250 ΔHbO2 and ΔHHb across gait conditions. Partial eta squared (η²) effect size was calculated 251 and categorized: small (≥0.01), medium (≥0.06), large (≥0.14) (Cohen, 1988). Where post-hoc 252 pairwise comparisons were made, Cohen’s d effect sizes were calculated and categorized: 253 small (≥0.2), medium (≥0.5), large (≥0.8) (Cohen, 1988). Pearson’s correlation was used to 254 determine the relationships between PPC ΔHbO2 and gait variability measures within each 255 condition. Alpha was set a priori at <0.05. All statistical analyses were completed using JASP 256 software (JASP 0.19.1; Amsterdam, Netherlands). 257 258

259 Gait performance (Mean (SD)) in each condition 260 Gait Condition n = 21 Gait speed (m/s) = 1.18 (0.16) NC VC VCP Stride Time Variability (CoV %) 2.04 (0.83) 2.86 (0.53)* 5.74 (1.03)*† Stride Length Variability (CoV %) 2.48 (0.92) 3.01 (0.68)* 7.63 (2.72)*† Lumbar mediolateral RoM Variability (CoV %) 13.01 (4.45) 16.14 (6.97)* 22.20 (10.77)*† Stride Time (s) Stride Length (m) Lumbar mediolateral RoM (°) 1.12 (0.07) 1.18 (0.13) 5.38 (1.68) 1.06 (0.08)* 1.12 (0.12)* 4.69 (1.49)* 1.06 (0.08)* 1.12 (0.12)* 5.12 (1.31)† 261 Table 1. Gait performance measures during non-cued gait (NC), visually cued gait (VC), and visually 262 cued gait with perturbations (VCP). * indicates a significant difference from NC gait. For VCP gait, † 263 indicates a significant difference from VC gait 264 265 PPC activity 266 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 11

of a repeated measures ANOVA showed no significant effect of gait condition on PPC 267 relative ΔHbO2, F(2,40) = 2.36, p = 0.12. However, a medium effect size was observed (η2 = 268 0.11), driven by a more positive ΔHbO2 in the PPC during VC (Cohen’s d = 0.39) and VCP 269 (Cohen’s d = 0.34) gait as compared to NC gait. PPC relative ΔHbO2 data, for each gait 270 condition, are presented in Figure 3. Relative ΔHHb concentrations were not statistically 271 different across gait conditions (NC, +0.04 ± 0.14μM; VC, -0.03 ± 0.15μM; VCP, -0.01 ± 272 0.14μM). Group averaged relative ΔHbO2 and ΔHHb across the 3-minute walking period, for 273 each condition, are presented in Supplemental Figure 1. 274 275 276 Fig.3 Box and scatter plots of the relative change in oxygenated hemoglobin (Δ HbO2), from baseline 277 quiet standing, during each gait condition: NC, non-cued; VC, visually cued; VCP, visually cued with 278 perturbations. Solid lines represent the median and dashed lines represent the mean for each 279 condition 280 281 Associations between PPC activity and gait variability 282 Gait performance during each condition is presented in Table 1. Stride time variability 283 significantly related to the relative ΔHbO2 in the PPC during the VC condition (Pearson’s r = 284 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 12 0.58, p = 0.01), such that higher stride time variability related to increased PPC activity 285 (Figure 4). During VCP gait, stride time variability was not significantly related to the relative 286 ΔHbO2 in the PPC. Neither stride length variability nor lumbar mediolateral RoM variability 287 significantly correlated with the relative ΔHbO2 in the PPC during any gait condition. 288 289 290 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 13 Fig.4 Scatter plots of the relationships between the relative change in oxygenated hemoglobin 291 (ΔHbO2) in the PPC and the coefficient of variation for a: stride time, b: stride length, and c: lumbar 292 mediolateral range of motion (RoM) during visually cued (VC) gait (left column) and visually cued gait 293 with perturbations (VCP) (right column). Black lines represent the best fit lines 294 295

296 This preliminary study aimed to decipher PPC activity patterns during non-cued (NC) gait, 297 visually cued (VC) gait and visually cued gait with perturbations (VCP). Our hypothesis for 298 increasing PPC activity from NC to VC gait, and from VC to VCP gait, was not statistically 299 supported, though we did observe moderate effect sizes to inform future study design. We 300 also examined how PPC activity relates to gait variability during these gait conditions. Our 301 hypotheses regarding relationships between PPC activity and gait variability were partially 302 supported. Specifically, increased PPC activity (more positive ΔHbO2) significantly related to 303 increased stride time variability, though during VC gait only. These results add to our 304 understanding of cortical activity during gait and serve as a foundation for future studies 305 involving populations with mobility deficits. 306 Moderate effect sizes were observed for PPC activity increases from NC to both VC and VCP 307 gait among this sample of young adults. While not statistically significant, this increase in 308 PPC activity with visual task complexity aligns with results from other studies that recorded 309 PPC activity during similar VC gait tasks (Liu et al., 2024; Yokoyama et al., 2021). EEG studies 310 report reduced alpha power in the PPC during VC gait compared to NC gait (Wagner et al., 311 2014; Yokoyama et al., 2021). Alpha power negatively correlates with blood-oxygen level-312 dependent-signal changes, which suggests that reduced alpha power indicates greater 313 cortical activity (Moosmann et al., 2003). Further, Liu et al. (2024) demonstrated that during 314 treadmill gait, PPC alpha power decreases as terrain unevenness increases. Though 315 increasing terrain unevenness increases reliance on visual processing for guiding step 316 placements (Liu et al., 2024; Matthis et al., 2018), somatosensory feedback from the uneven 317 terrains may be a factor in the observed PPC activity. Across these VC gait paradigms, young 318 adults leveraged swift and accurate visual processing, which is critical in complex gait 319 environments. Unfortunately, visual processing slows with age (Ebaid & Crewther, 2019), 320 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 14 underscoring the importance for studies quantifying PPC activity during VC gait in older 321 adults. Results from these studies would establish the effect of slowed visual processing on 322 gait impairment and fall risk. 323 While PPC activity appears to increase from NC to both VC and VCP gait, PPC activity levels 324 during VC and VCP gait were not greatly elevated above those of baseline quiet standing. 325 The non-significant increase was driven by a negative ΔHbO2 in the PPC during NC gait, i.e., 326 PPC deactivation during walking relative to the brief standing baseline period (Figure 3). This 327 pattern of PPC deactivation indicates that among young adults, PPC engagement was not 328 necessary for NC treadmill gait, potentially because treadmill gait does not require 329 significant visuomotor integration due to limited visual flow. Consistent with our findings, 330 Lau et al. (2014) reported reduced cortical sensorimotor network involvement in young 331 adults during treadmill walking compared to standing. Young adults typically exhibit gait 332 automaticity (Clark, 2015), relying more on subcortical neural networks for generating the 333 appropriate muscle activation patterns during gait. By contrast, standing requires 334 considerable active cortical control for maintaining balance and posture (Vuillerme & Nafati, 335 2007). Combined, our results and previous evidence suggest that the PPC deactivation 336 observed during NC gait relates to a shift toward subcortical gait control in young adults. 337 We did not observe a condition effect for PPC activity when comparing VC and VCP gait 338 (Cohen’s d = 0.04). Other supraspinal structures involved in locomotor control (e.g., 339 brainstem and cerebellum) may be more critical for gait modifications in response to rapidly 340 shifting stepping target positions (Hoogkamer et al., 2017). Alternatively, processing in 341 higher-order cortical regions beyond the PPC (e.g., PFC) could be more important during VC 342 and VCP gait. Supporting this notion, Corporaal et al. (2018) reported that greater stepping 343 accuracy during VC gait was associated with increased white matter tracts connecting 344 attentional cortical regions (e.g., parietal and prefrontal cortices). Importantly, comparing NC 345 to VC gait with targets presented at either fixed or variable step lengths, Le et al. (2023) 346 reported no difference in PPC activity between conditions. However, these authors identified 347 increased functional connectivity between the PPC and PFC during VC gait with variable step 348 lengths. PPC-PFC connectivity serves as a key pathway of the frontoparietal network, which 349 promotes externally oriented attention relevant for visuomotor performance when 350 responding to unexpected stimuli (Menon & D’Esposito, 2022). Specifically, VC gait requires 351 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 15 allocation of attention to external information (i.e., the positions of visual cues), highlighting 352 a key role for the frontoparietal network. However, the absence of a significant increase in 353 PPC activity from VC to VCP gait in our study, combined with the increased PFC activity 354 observed during obstacle negotiation (Mirelman et al., 2017) and VC gait (Koenraadt et al., 355 2014), suggests that the PFC may comprise the dominant node of the frontoparietal network 356 for VC gait performance in young adults. 357 The positive relationship between PPC activity and stride time variability observed during VC 358 gait could be interpreted from two, somewhat opposing perspectives. Increased PPC activity 359 from NC to VC gait may reflect the deployment of more neural resources in response to the 360 increased environmental visuospatial processing demands of the task. This top-down 361 strategy may be detrimental to gait rhythmicity. From this perspective, higher stride time 362 variability reflects an unstable gait pattern, as observed in older adult and clinical 363 populations (Hausdorff et al., 2001; Lord et al., 2011). This interpretation aligns with the 364 Compensation-Related Utilization of Neural Circuits Hypothesis (Reuter-Lorenz & Cappell, 365 2008), and mirrors that offered for the positive association between increased cortical 366 activity and gait variability among older adults (Nóbrega-Sousa et al., 2020). Concurrently 367 quantifying PFC and PPC activity during our gait paradigm could provide evidence to support 368 this potential explanation. Alternatively, increased PPC activity could be an adaptive 369 mechanism allowing for flexible step timing adjustments for young adults, mirroring results 370 from other species (Marigold et al., 2011). From this perspective, higher stride time 371 variability during VC gait reflects a skillful adaptation strategy (Stergiou & Decker, 2011), 372 where ongoing modulation of step timing supports stepping precision (Koenraadt et al., 373 2014). Higher PPC activity in response to the visuospatial processing demands may facilitate 374 this strategy. Quantifying step accuracy will help to clarify the effect of higher PPC activity on 375 gait performance during VC gait. 376 Notably, although stride time variability increased from VC to VCP gait, a positive 377 relationship between PPC activity and stride time variability did not persist in the VCP 378 condition. That PPC activity was similar during both VC and VCP gait may suggest that for 379 these young adults, the visuospatial processing demands of both conditions were 380 comparable. VCP gait imposes a degree of gait variability, as frequent step adjustments are 381 necessary for good task performance. Applying the above argument for increased PPC 382 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 16 activity as an adaptive mechanism supporting step accuracy, it would be reasonable to 383 expect PPC activity to remain positively correlated with stride time variability during VCP 384 gait. However, VCP gait performance may rely more on executive functions (e.g., attentional 385 control, decision making) than VC gait. As previously suggested, the PFC potentially plays a 386 pivotal role during VCP gait, and PFC activity in this condition conceivably relates to gait 387 variability. Relatedly, Mirelman et al. (2017) reported a positive correlation between PFC 388 activity and gait variability among older adults during obstacle negotiation. Assessing PFC 389 activation during our gait paradigm will help to clarify the role of executive control in gait 390 adaptability to shifting visual cues. 391 That we observed a relationship between PPC activity and temporal, but not spatial, gait 392 variability aligns with the brain map of gait variability put forward by Tian et al. (2017). 393 Structural MRI findings suggest that PPC grey matter volume negatively correlates with 394 stride time variability in older adults (Beauchet et al., 2014). Using our paradigm to examine 395 associations between real-time PPC activity and spatiotemporal gait variability in older 396 adults will therefore be illuminating. Moreover, Pizzamiglio et al. (2018) indicated a role for 397 the PPC in mediolateral gait control, specifically that higher PPC activity related to lower 398 mediolateral center of mass motion during unperturbed gait. Our findings offer some 399 support for this, as a moderate correlation (r = -0.38) emerged between higher PPC activity 400 and lower lumbar mediolateral RoM variability during VC gait. Furthermore, lumbar 401 mediolateral RoM was significantly lower during VC gait than NC gait. These findings indicate 402 that in response to predictable visual cues during gait, young adults promote stability by 403 restricting movement in the mediolateral direction, and this adaptive response may be 404 supported by visuomotor integration in the PPC. Notably, evidence suggests that for older 405 adults, the control of dynamic balance is more challenging in the frontal plane than in the 406 sagittal plane (Vistamehr & Neptune, 2021). Accordingly, more falls occur in the frontal plane 407 than in the sagittal plane (Parkkari et al., 1999). Therefore, unravelling the PPC’s role in 408 mediolateral gait control during VC gait is a critical step toward understanding and mitigating 409 fall risk in older adults. 410 The results and interpretation of this preliminary study need to be considered along with the 411 following limitations. First, as Beauchet et al. (2009) insist, caution must be exercised when 412 interpreting gait variability. Depending on the circumstances, both low and high variability of 413 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 17 gait parameters may reflect good gait performance. As discussed, high variability of gait 414 parameters may be considered a marker of adaptability to the gait environment. Future 415 studies accounting for step accuracy could provide additional context to the present 416 findings. Second, cortical areas beyond the PPC may be implicated in VC gait control, 417 including the PFC (Le et al., 2023), premotor cortex (Wang et al., 2008), and supplementary 418 motor area (Koenraadt et al., 2014). A more comprehensive assessment of activity across 419 the cortex during our gait conditions could offer greater insights into the cortical 420 mechanisms underlying VC gait performance. Additionally, the hemodynamic response delay 421 following neural activity renders fNIRS unsuitable for assessing cortical activation changes 422 within different phases of the gait cycle, and for a step- or stride-level comparison. Where 423 reactive step adjustments are required, characterizing intra-stride neural dynamics with EEG 424 would help to uncover how the precise timing of cortical contributions supports gait 425 performance. Our VCP gait task design, involving target position shifts in different directions 426 every 3-7 steps, meant it was not feasible to examine a potential effect of step adjustment 427 direction on PPC activation. Finally, using the 10-20 system and visual inspection to identify 428 channels corresponding to our cortical regions of interest is not the most rigorous approach, 429 but has been successfully implemented across neuroimaging techniques (Herwig et al., 430 2003; Koenraadt et al., 2014; Shafiul Hasan et al., 2020; Velu & de Sa, 2013). 431 432

433 This preliminary study offers new insights into PPC activity during VC treadmill gait in young 434 adults. The moderate cue condition effect observed for PPC activity suggests that whether 435 reactive step adjustments are necessary or not, greater PPC recruitment is required for VC 436 gait than for NC gait. This likely reflects the increased visuomotor processing demands posed 437 by visual cues. The positive relationship between PPC activity and stride time variability 438 observed during VC gait potentially suggests that higher PPC activity supports modulation of 439 step timing in response to visual stepping targets. Considering the association between high 440 gait variability and fall risk among older adults (Hausdorff et al., 2001), future studies should 441 extend these findings to aging populations. With evidence suggesting an important role for 442 the frontoparietal network during VC gait (Corporaal et al., 2018; Le et al., 2023), expanding 443 the cortical assessment to include the PFC should offer valuable insights. Quantifying PPC 444 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 18 and PFC activity during VC and VCP gait in older adults, and relating cortical activity to gait 445 performance, could help to uncover neurophysiological signatures of increased fall risk. 446 447 Data & Code Availability 448 As this study is part of an ongoing, federally funded investigation, access to data and code is 449 subject to restrictions. Access may be provided following a formal request. 450 451 Funding 452 This study was supported by the National Institute on Aging (NIA R21AG075489). 453 454 Declaration of Competing Interests 455 The authors confirm that there is no financial or personal relationship with other individuals 456 or organizations that could inappropriately influence this work. 457 458

459 Aasted, C. M., Yücel, M. A., Cooper, R. J., Dubb, J., Tsuzuki, D., Becerra, L., Petkov, M. P ., Borsook, D., 460 Dan, I., & Boas, D. A. (2015). Anatomical guidance for functional near-infrared spectroscopy: 461 AtlasViewer tutorial. Neurophotonics, 2(2), 020801. 462 https://doi.org/10.1117/1.NPh.2.2.020801 463 Beauchet, O., Allali, G., Annweiler, C., Bridenbaugh, S., Assal, F., Kressig, R. W., & Herrmann, F. R. 464 (2009). Gait Variability among Healthy Adults: Low and High Stride-to-Stride Variability Are 465 Both a Reflection of Gait Stability. Gerontology, 55(6), 702–706. 466 https://doi.org/10.1159/000235905 467 Beauchet, O., Annweiler, C., Celle, S., Bartha, R., Barthélémy, J.-C., & Roche, F. (2014). Higher Gait 468 Variability is Associated with Decreased Parietal Gray Matter Volume Among Healthy Older 469 Adults. Brain Topography, 27(2), 293–295. https://doi.org/10.1007/s10548-013-0293-y 470 Boas, D. A., Dale, A. M., & Franceschini, M. A. (2004). Diffuse optical imaging of brain activation: 471 approaches to optimizing image sensitivity, resolution, and accuracy. Neuroimage, 23, S275-472 S288. 473 Buneo, C. A., & Andersen, R. A. (2006). The posterior parietal cortex: Sensorimotor interface for the 474 planning and online control of visually guided movements. Neuropsychologia, 44(13), 2594–475 2606. https://doi.org/10.1016/j.neuropsychologia.2005.10.011 476 Chen, Y ., Cao, Z., Mao, M., Sun, W., Song, Q., & Mao, D. (2022). Increased cortical activation and 477 enhanced functional connectivity in the prefrontal cortex ensure dynamic postural balance 478 during dual-task obstacle negotiation in the older adults: A fNIRS study. Brain and 479 Cognition, 163, 105904. 480 Chen, M., Pillemer, S., England, S., Izzetoglu, M., Mahoney, J. R., & Holtzer, R. (2017). Neural 481 correlates of obstacle negotiation in older adults: an fNIRS study. Gait & posture, 58, 130-482 135. 483 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 19 Clark, D. J. (2015). Automaticity of walking: Functional significance, mechanisms, measurement and 484 rehabilitation strategies. Frontiers in Human Neuroscience, 9, 246. 485 https://doi.org/10.3389/fnhum.2015.00246 486 Cohen, J. (1988). Statistical Power Analysis for the Behavioral Sciences (2nd ed.). Routledge. 487 https://doi.org/10.4324/9780203771587 488 Cooper, R., Selb, J., Gagnon, L., Phillip, D., Schytz, H. W., Iversen, H. K., Ashina, M., & Boas, D. A. 489 (2012). A Systematic Comparison of Motion Artifact Correction Techniques for Functional 490 Near-Infrared Spectroscopy. Frontiers in Neuroscience, 6, 147. 491 https://doi.org/10.3389/fnins.2012.00147 492 Corporaal, S. H. A., Bruijn, S. M., Hoogkamer, W., Chalavi, S., Boisgontier, M. P ., Duysens, J., Swinnen, 493 S. P ., & Gooijers, J. (2018). Different neural substrates for precision stepping and fast online 494 step adjustments in youth. Brain Structure and Function, 223(4), 2039–2053. 495 https://doi.org/10.1007/s00429-017-1586-9 496 Cui, X., Bray, S., & Reiss, A. L. (2010). Functional near infrared spectroscopy (NIRS) signal 497 improvement based on negative correlation between oxygenated and deoxygenated 498 hemoglobin dynamics. Neuroimage, 49(4), 3039-3046. 499 https://doi.org/10.1016/j.neuroimage.2009.11.050 500 de Belli, V., Orcioli-Silva, D., Beretta, V. S., Vitório, R., Zampier, V. C., Nóbrega-Sousa, P ., ... & Gobbi, L. 501 T. B. (2021). Prefrontal cortical activity during preferred and fast walking in young and older 502 adults: An fNIRS Study. Neuroscience, 473, 81-89. 503 Di Lorenzo, R., Pirazzoli, L., Blasi, A., Bulgarelli, C., Hakuno, Y ., Minagawa, Y ., & Brigadoi, S. (2019). 504 Recommendations for motion correction of infant fNIRS data applicable to multiple data sets 505 and acquisition systems. NeuroImage, 200, 511–527. 506 https://doi.org/10.1016/j.neuroimage.2019.06.056 507 Drew, T., Fortier-Lebel, N., & Nakajima, T. (2023). Cortical contribution to visuomotor coordination in 508 locomotion and reaching. Current Opinion in Neurobiology, 82, 102755. 509 https://doi.org/10.1016/j.conb.2023.102755 510 Drew, T., & Marigold, D. S. (2015). Taking the next step: Cortical contributions to the control of 511 locomotion. Current Opinion in Neurobiology, 33, 25–33. 512 https://doi.org/10.1016/j.conb.2015.01.011 513 Ebaid, D., & Crewther, S. G. (2019). Visual Information Processing in Young and Older Adults. 514 Frontiers in Aging Neuroscience, 11. 515 https://www.frontiersin.org/articles/10.3389/fnagi.2019.00116 516 Gao, L., Wei, Y ., Wang, Y ., Wang, G., Zhang, Q., Zhang, J., Chen, X., & Yan, X. (2022). Hybrid motion 517 artifact detection and correction approach for functional near-infrared spectroscopy 518 measurements. Journal of Biomedical Optics, 27(2), 025003. 519 https://doi.org/10.1117/1.JBO.27.2.025003 520 Hausdorff, J. M., Cudkowicz, M. E., Firtion, R., Wei, J. Y ., & Goldberger, A. L. (1998). Gait variability 521 and basal ganglia disorders: stride-to-stride variations of gait cycle timing in Parkinson's 522 disease and Huntington's disease. Movement disorders, 13(3), 428-437. 523 Hausdorff, J. M., Rios, D. A., & Edelberg, H. K. (2001). Gait variability and fall risk in community-living 524 older adults: A 1-year prospective study. Archives of Physical Medicine and Rehabilitation, 525 82(8), 1050–1056. https://doi.org/10.1053/apmr.2001.24893 526 Herwig, U., Satrapi, P ., & Schönfeldt-Lecuona, C. (2003). Using the international 10-20 EEG system for 527 positioning of transcranial magnetic stimulation. Brain topography, 16, 95-99. 528 Hoogkamer, W., Potocanac, Z., & Duysens, J. (2015). Quick foot placement adjustments during gait: 529 Direction matters. Experimental Brain Research, 233(12), 3349–3357. 530 https://doi.org/10.1007/s00221-015-4401-y 531 Hoogkamer, W., Potocanac, Z., Van Calenbergh, F., & Duysens, J. (2017). Quick foot placement 532 adjustments during gait are less accurate in individuals with focal cerebellar lesions. Gait & 533 posture, 58, 390-393. 534 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 20 Homan, R. W., Herman, J., & Purdy, P . (1987). Cerebral location of international 10–20 system 535 electrode placement. Electroencephalography and clinical neurophysiology, 66(4), 376-382. 536 Huppert, T. J., Diamond, S. G., Franceschini, M. A., & Boas, D. A. (2009). HomER: A review of time-537 series analysis methods for near-infrared spectroscopy of the brain. Applied Optics, 48(10), 538 D280–D298. https://doi.org/10.1364/AO.48.00D280 539 Koenraadt, K. L. M., Roelofsen, E. G. J., Duysens, J., & Keijsers, N. L. W. (2014). Cortical control of 540 normal gait and precision stepping: An fNIRS study. NeuroImage, 85, 415–422. 541 https://doi.org/10.1016/j.neuroimage.2013.04.070 542 Lau, T. M., Gwin, J. T., & Ferris, D. P . (2014). Walking reduces sensorimotor network connectivity 543 compared to standing. Journal of NeuroEngineering and Rehabilitation, 11(1), 1-10. 544 https://doi.org/10.1186/1743-0003-11-14 545 Le, D. T., Tsuyuhara, M., Kuwamura, H., Kitano, K., Nguyen, T. D., Duc Nguyen, T., Fujita, N., Watanabe, 546 T., Nishijo, H., Mihara, M., & Urakawa, S. (2023). Regional activity and effective connectivity 547 within the frontoparietal network during precision walking with visual cueing: An fNIRS 548 study. Cerebral Cortex, 33(22), 11157–11169. https://doi.org/10.1093/cercor/bhad354 549 Liu, C., Downey, R. J., Salminen, J. S., Arvelo Rojas, S., Richer, N., Pliner, E. M., Hwang, J., Cruz-550 Almeida, Y ., Manini, T. M., Hass, C. J., Seidler, R. D., Clark, D. J., & Ferris, D. P . (2024). Electrical 551 brain activity during human walking with parametric variations in terrain unevenness and 552 walking speed. Imaging Neuroscience, 2, 1–33. https://doi.org/10.1162/imag_a_00097 553 Lord, S., Howe, T., Greenland, J., Simpson, L., & Rochester, L. (2011). Gait variability in older adults: A 554 structured review of testing protocol and clinimetric properties. Gait & Posture, 34(4), 443–555 450. https://doi.org/10.1016/j.gaitpost.2011.07.010 556 Maki, B. E. (1997). Gait Changes in Older Adults: Predictors of Falls or Indicators of Fear? Journal of 557 the American Geriatrics Society, 45(3), 313–320. https://doi.org/10.1111/j.1532-558 5415.1997.tb00946.x 559 Marigold, D. S., Andujar, J.-E., Lajoie, K., & Drew, T. (2011). Chapter 6 - Motor planning of locomotor 560 adaptations on the basis of vision: The role of the posterior parietal cortex. In J. P . Gossard, R. 561 Dubuc, & A. Kolta (Eds.), Progress in Brain Research, 188, 83–100. 562 https://doi.org/10.1016/B978-0-444-53825-3.00011-5 563 Marigold, D. S., & Drew, T. (2017). Posterior parietal cortex estimates the relationship between object 564 and body location during locomotion. eLife, 6, e28143. https://doi.org/10.7554/eLife.28143 565 Martini, D. N., Mancini, M., Antonellis, P ., McDonnell, P ., Vitorio, R., Stuart, S., & King, L. A. (2024). 566 Prefrontal cortex activity during gait in people with persistent symptoms after concussion. 567 Neurorehabilitation and Neural Repair, 38(5), 364-372. 568 https://doi.org/10.1177/15459683241240423 569 Matthis, J. S., Yates, J. L., & Hayhoe, M. M. (2018). Gaze and the Control of Foot Placement When 570 Walking in Natural Terrain. Current Biology, 28(8), 1224-1233.e5. 571 https://doi.org/10.1016/j.cub.2018.03.008 572 Mazaheri, M., Hoogkamer, W., Potocanac, Z., Verschueren, S., Roerdink, M., Beek, P . J., Peper, C. E., & 573 Duysens, J. (2015). Effects of aging and dual tasking on step adjustments to perturbations in 574 visually cued walking. Experimental Brain Research, 233(12), 3467–3474. 575 https://doi.org/10.1007/s00221-015-4407-5 576 Menon, V., & D’Esposito, M. (2022). The role of PFC networks in cognitive control and executive 577 function. Neuropsychopharmacology, 47(1), 90–103. https://doi.org/10.1038/s41386-021-578 01152-w 579 Mirelman, A., Maidan, I., Bernad-Elazari, H., Shustack, S., Giladi, N., & Hausdorff, J. M. (2017). Effects 580 of aging on prefrontal brain activation during challenging walking conditions. Brain and 581 Cognition, 115, 41–46. https://doi.org/10.1016/j.bandc.2017.04.002 582 Moosmann, M., Ritter, P ., Krastel, I., Brink, A., Thees, S., Blankenburg, F., Taskin, B., Obrig, H., & 583 Villringer, A. (2003). Correlates of alpha rhythm in functional magnetic resonance imaging 584 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 21 and near infrared spectroscopy. NeuroImage, 20(1), 145–158. 585 https://doi.org/10.1016/S1053-8119(03)00344-6 586 Nóbrega-Sousa, P ., Gobbi, L. T. B., Orcioli-Silva, D., Conceição, N. R. da, Beretta, V. S., & Vitório, R. 587 (2020). Prefrontal Cortex Activity During Walking: Effects of Aging and Associations With Gait 588 and Executive Function. Neurorehabilitation and Neural Repair, 34(10), 915–924. 589 https://doi.org/10.1177/1545968320953824 590 Nordin, A. D., Hairston, W. D., & Ferris, D. P . (2019). Human electrocortical dynamics while stepping 591 over obstacles. Scientific Reports, 9(1), 4693. https://doi.org/10.1038/s41598-019-41131-2 592 O’Connor, S. M., & Kuo, A. D. (2009). Direction-Dependent Control of Balance During Walking and 593 Standing. Journal of Neurophysiology, 102(3), 1411–1419. 594 https://doi.org/10.1152/jn.00131.2009 595 Parkkari, J., Kannus, P ., Palvanen, M., Natri, A., Vainio, J., Aho, H., Vuori, I., & Järvinen, M. (1999). 596 Majority of hip fractures occur as a result of a fall and impact on the greater trochanter of 597 the femur: a prospective controlled hip fracture study with 206 consecutive 598 patients. Calcified tissue international, 65, 183-187. 599 Pelicioni, P . H. S., Lord, S. R., Okubo, Y ., & Menant, J. C. (2022). Cortical activation during gait 600 adaptability in people with Parkinson’s disease. Gait & Posture, 91, 247-253. 601 Pizzamiglio, S., Abdalla, H., Naeem, U., & Turner, D. L. (2018). Neural predictors of gait stability when 602 walking freely in the real-world. Journal of NeuroEngineering and Rehabilitation, 15(1), 11. 603 https://doi.org/10.1186/s12984-018-0357-z 604 Potocanac, Z., & Duysens, J. (2017). Online adjustments of leg movements in healthy young and old. 605 Experimental Brain Research, 235(8), 2329–2348. https://doi.org/10.1007/s00221-017-4967-606 7 607 Reuter-Lorenz, P . A., & Cappell, K. A. (2008). Neurocognitive Aging and the Compensation 608 Hypothesis. Current Directions in Psychological Science, 17(3), 177–182. 609 https://doi.org/10.1111/j.1467-8721.2008.00570.x 610 Robinovitch, S. N., Feldman, F., Yang, Y ., Schonnop, R., Leung, P . M., Sarraf, T., Sims-Gould, J., & 611 Loughin, M. (2013). Video capture of the circumstances of falls in elderly people residing in 612 long-term care: An observational study. The Lancet, 381(9860), 47–54. 613 https://doi.org/10.1016/S0140-6736(12)61263-X 614 Scholkmann, F., Spichtig, S., Muehlemann, T., & Wolf, M. (2010). How to detect and reduce 615 movement artifacts in near-infrared imaging using moving standard deviation and spline 616 interpolation. Physiological Measurement, 31(5), 649. https://doi.org/10.1088/0967-617 3334/31/5/004 618 Shafiul Hasan, S. M., Siddiquee, M. R., Atri, R., Ramon, R., Marquez, J. S., & Bai, O. (2020). Prediction 619 of gait intention from pre-movement EEG signals: a feasibility study. Journal of 620 neuroengineering and rehabilitation, 17, 1-16. 621 Stergiou, N., & Decker, L. M. (2011). Human movement variability, nonlinear dynamics, and 622 pathology: Is there a connection? Human Movement Science, 30(5), 869–888. 623 https://doi.org/10.1016/j.humov.2011.06.002 624 Tian, Q., Chastan, N., Bair, W. N., Resnick, S. M., Ferrucci, L., & Studenski, S. A. (2017). The brain map 625 of gait variability in aging, cognitive impairment and dementia—a systematic 626 review. Neuroscience & Biobehavioral Reviews, 74, 149-162. 627 Tian, Q., Resnick, S. M., Landman, B. A., Huo, Y ., Venkatraman, V. K., Gonzalez, C. E., Simonsick, E. M., 628 Shardell, M. D., Ferrucci, L., & Studenski, S. A. (2016). Lower gray matter integrity is 629 associated with greater lap time variation in high-functioning older adults. Experimental 630 Gerontology, 77, 46–51. https://doi.org/10.1016/j.exger.2016.02.009 631 Velu, P . D., & de Sa, V. R. (2013). Single-trial classification of gait and point movement preparation 632 from human EEG. Frontiers in neuroscience, 7, 84. 633 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint 22 Verghese, J., Robbins, M., Holtzer, R., Zimmerman, M., Wang, C., Xue, X., & Lipton, R. B. (2008). Gait 634 dysfunction in mild cognitive impairment syndromes. Journal of the American Geriatrics 635 Society, 56(7), 1244-1251. 636 Vistamehr, A., & Neptune, R. R. (2021). Differences in balance control between healthy younger and 637 older adults during steady-state walking. Journal of Biomechanics, 128, 110717. 638 Vitorio, R., Stuart, S., & Mancini, M. (2020). Executive control of walking in people with Parkinson’s 639 disease with freezing of gait. Neurorehabilitation and neural repair, 34(12), 1138-1149. 640 von Lühmann, A., Ortega-Martinez, A., Boas, D. A., & Yücel, M. A. (2020). Using the General Linear 641 Model to Improve Performance in fNIRS Single Trial Analysis and Classification: A Perspective. 642 Frontiers in Human Neuroscience, 14. https://doi.org/10.3389/fnhum.2020.00030 643 Vuillerme, N., & Nafati, G. (2007). How attentional focus on body sway affects postural control during 644 quiet standing. Psychological Research, 71(2), 192–200. https://doi.org/10.1007/s00426-645 005-0018-2 646 Wagner, J., Solis-Escalante, T., Scherer, R., Neuper, C., & Müller-Putz, G. (2014). It’s how you get 647 there: Walking down a virtual alley activates premotor and parietal areas. Frontiers in Human 648 Neuroscience, 8. https://doi.org/10.3389/fnhum.2014.00093 649 Wang, C., Wai, Y ., Weng, Y ., Yu, J., & Wang, J. (2008). The cortical modulation from the external cues 650 during gait observation and imagination. Neuroscience Letters, 443(3), 232–235. 651 https://doi.org/10.1016/j.neulet.2008.07.084 652 Wuehr, M., Schniepp, R., Pradhan, C., Ilmberger, J., Strupp, M., Brandt, T., & Jahn, K. (2013). 653 Differential effects of absent visual feedback control on gait variability during different 654 locomotion speeds. Experimental Brain Research, 224(2), 287–294. 655 https://doi.org/10.1007/s00221-012-3310-6 656 Yokoyama, H., Kaneko, N., Masugi, Y ., Ogawa, T., Watanabe, K., & Nakazawa, K. (2021). Gait-phase-657 dependent and gait-phase-independent cortical activity across multiple regions involved in 658 voluntary gait modifications in humans. European Journal of Neuroscience, 54(12), 8092–659 8105. https://doi.org/10.1111/ejn.14867 660 661 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted April 16, 2025. ; https://doi.org/10.1101/2025.04.10.648269doi: bioRxiv preprint