Neural Correlates of Balance in People With Parkinson’s Disease,Older,and Younger Adults: An fNIRS Study

Abstract

Background:

Postural instability is one of the most debilitating symptoms of Parkinson’s disease (PD). Moreover, older adults (OA) often show issues with postural control leading to increased fall-risk. However, the current understanding of the neural underpinnings of standing postural control remains limited. This study aims to investigate cortical control of postural control in OA, younger adults (YA), and people with PD (PwPD), using functional near-infrared spectroscopy (fNIRS).

Methods:

A total of four 2-minute standing conditions were performed. Postural control and cortical activity were recorded in 80 PwPD, 33 OA, and 38 YA. A wireless fNIRS system recorded changes in relative oxygenated haemoglobin (∆HbO₂) across cortical regions including the prefrontal cortex (PFC), supplementary motor area (SMA), primary motor cortex (M1), primary somatosensory cortex (S1), and primary visual cortex (V1). Six wearable sensors provided sway outcome measures (area, jerkiness velocity, root mean square, and frequency).

Results:

Sway outcomes were greater across several conditions in PwPD. Significant group effects were found with increased ∆HbO² in the PFC in PD compared to OA. Moreover, YA had increased ∆HbO₂ in the S1 compared to OA and PD.

Conclusions:

PwPD showed greater PFC recruitment, indicating reliance on executive-attentional resources for balance. In contrast, YA engaged somatosensory regions more strongly, suggesting that ageing may affect the integration of sensory information for postural control. Findings support interventions that (a) reduce executive load during balance in PD and (b) bolster somatosensory integration in ageing.

Keywords

fNIRS postural control balance impairment older adults Parkinson’s disease postural sway wearable sensors postural stability neural correlates

Introduction

Postural instability is one of the most debilitating symptoms of Parkinson’s disease (PD).¹ Individuals whose postural control is compromised are at greater risk of experiencing falls and subsequent injury, significantly impacting their quality of life.² Moreover, various age-related factors contribute to postural control impairment in healthy individuals without any neurological conditions. For example, sarcopenia, inner-ear issues, visual impairments, back pain, and fear of falling.^3,4 These are co-morbidities that further complicate the assessment and management of postural instability in people with PD (PwPD), in addition to underlying neuropathology. Some pharmacological interventions have shown significant improvements in gait metrics (eg, step length and speed). Yet dopaminergic medication generally does not improve postural control or reduce the risk of falls.^5,6

Several behavioural studies have demonstrated a positive relationship between mobility and cognition in PwPD and older adults (OA), suggesting shared neural pathways.^7-9 However, the development of postural control impairment due to ageing and pathology, and its neural underpinnings, remain poorly understood. Previous research indicates increased cortical activation during walking tasks to compensate for age-related decline,¹⁰ further increasing with PD.¹¹ This is suggested to represent a compensatory shift from lower spinal and subcortical regions, decreasing automaticity and engaging higher-order cortical resources.¹² This hypothesis helps explain why dual-task (DT) conditions, which require additional cognitive resources, often worsen performance in OA and PwPD.^7,11

However, there has been a disproportionate focus on neural correlates of gait, whereas the neural mechanisms underlying postural control are relatively underexplored.¹³ Studies that have used functional magnetic resonance imaging to investigate neural activity in relation to postural control report dissimilar neural activation patterns between PwPD and OA.⁷ However, these studies rely on resting-state correlates and mental imagery tasks in a supine position which lack ecological validity for the investigation of postural control. Similarly, electroencephalography studies have reported increased cortical connectivity in early-PD, compared to OA, suggesting pathological cortical adaptations in relation to balance.¹⁴ Moreover, increased theta power in frontal–central and frontal–parietal regions corresponded to better balance performance in PwPD.^15,16

Functional near-infrared spectroscopy (fNIRS) offers an alternative non-invasive method for assessing cortical activity during standing tasks by measuring relative changes in oxygenated haemoglobin (∆HbO₂). To our knowledge, no studies have incorporated fNIRS to examine cortical activity related to postural control across multiple cortical regions in PD. Postural control is a complex process that involves various sensory, motor, and cognitive information.^17,18 With the additional alterations in the function of these regions in relation to posture in OA and PwPD, it is imperative that we assess activity across multiple cortical regions. Additionally, wearable inertial measurement unites can be synchronised to provide real-time sway outcome measures, enabling the objective quantification of postural control. By combining these technologies, the relationship between cortical activity and postural control can be examined, providing new insights into the impact of ageing and PD on its neural underpinnings.

A greater understanding of cortical control of postural control is required in both ageing and Parkinson’s pathology. Therefore, this study examined activity across multiple cortical regions including the prefrontal cortex (PFC), supplementary motor area (SMA), primary motor area (M1), primary somatosensory motor area (S1), and primary visual cortex (V1). These regions were selected because they represent key cortical networks involved in the cognitive, motor, somatosensory, and visual processes that contribute to postural control. Activity was recorded during 4 separate standing conditions. Participants stood with their eyes open on a firm surface (EOFS), eyes closed on a firm surface (ECFS), eyes open on a foam surface (EOFOAM), and finally with their eyes closed on a foam surface (ECFOAM).

The aims of this study are to:

1. Compare differences in cortical activity during standing tasks between younger adults (YA) and OA and PwPD, highlighting potential age-related and disease-related neural adaptation, respectively.

2. Explore how different sensory conditions during standing tasks impact cortical control of postural control in YA, OA, and PwPD.

3. Examine associations between cortical activity and sway outcome measures in YA, OA, and PwPD.

First, we hypothesise that PwPD will exhibit greater postural sway compared to YA and OA, with YA demonstrating the best postural control. Second, we hypothesise that there will be differences in cortical activity across the cortex between YA, OA, and PwPD respectively, reflecting a greater reliance on cortical regions to maintain postural control with ageing, further exacerbated by PD. Third, we hypothesise that there will be associations between sway outcome measures and cortical activity in OA and PwPD.

Methods

Recruitment and Eligibility Criteria

Inclusion Criteria

Participants must meet the following criteria: a clinical diagnosis of PD made by a movement disorder specialist in accordance with the UK Brain Bank criteria¹⁹; Hoehn and Yahr (H&Y) stage I–III; aged over 50 years; able to stand and walk independently for a minimum of 2 minutes, stable medication regimen for at least 1 month prior to the study and whether any changes were anticipated over a period of 6 months. OA and YA were required to meet the same eligibility criteria, except for clinical diagnoses. The age range for YA was defined as 18 to 35 years For further details, see Vitório et al.²⁰

Exclusion Criteria

Participants were excluded if they present with: psychiatric comorbidities (eg, schizophrenia or major depressive disorder, as indicated by a Geriatric Depression Scale [GDS]-15 score ≥10)²¹; a clinical diagnosis of dementia or other severe cognitive impairment (Montreal Cognitive Assessment [MoCA] <21²²); history of neurological conditions other than PD (eg, Huntington’s disease, stroke, traumatic brain injury, multiple sclerosis, and Alzheimer’s disease); acute lower back or lower limb pain, peripheral neuropathy, rheumatic or orthopaedic disorders; unstable medical conditions, including cardiovascular events within the past 6 months (eg, angina, myocardial infarction, and pulmonary embolism); inability to comply with the testing protocol; or participation in another clinical or research intervention that could affect gait (eg, pharmaceutical, exercise, or therapy trials; Figure 1).

Demographic Assessments

Demographic data collected included height, weight, age, and years of formal education. Global cognition (MoCA),²² attention/executive function (Trail-Making-Task A/B),²³ visuo-spatial (Judgement of line orientation [JLo]²⁴ and Clock drawing exectuive test part 1 and 2 (CLOX1/2)²⁵), working memory (Forward digit span²⁶), as well as fear of falling (Falls-Efficacy-Scale [FES]²⁷) were examined. Contrast sensitivity and visual acuity were also evaluated. Depressive symptoms were assessed in both the OA and PD group with the GDS²⁸. Additional information gathered for the PD group included disease duration, medication use, and disease severity using the Movement Disorders Society-Unified Parkinson’s Disease Rating Scale part III (MDS-UPDRS III) score, and Hoehn and Yahr.

Procedures and Materials

Participants were fitted with a portable 26-channel fNIRS system (Octamon + Brite 24, Artinis Medical System, The Netherlands). Once the device was fitted and anatomical references were gathered, participants performed a series of 2-minute standing tasks. Prior to these tasks, participants were instructed to sit still and maintain their gaze on a fixed cross on the wall for 2 minutes. Participants then performed 4 consecutive standing tasks; EOFS, ECFS, EOFOAM, and ECFOAM.

Equipment

A non-invasive fNIRS system (OctaMon + Brite 24, Artinis Medical Systems, The Netherlands) recorded cortical activity during the standing tasks. The fNIRS montage included 20 regular channels (inter-optode distance of 3 cm) and 6 short-separation channels (inter-optode distance of 1 cm). The fNIRS signals were recorded at 50 Hz (Oxysoft). A 3D-digitizer (FASTRAK, Polhemus, VT, USA) was used to provide 3-dimensional coordinates of anatomical references (Cz, nasion, and left and right preauricular points) and positions of optodes.

This data was entered into MATLAB via the software package “NIRS-Statistical Parametric Mapping (NIRS-SPM).” The Spatial Registration routine (stand-alone NIRS, using a 3D-digitizer) was used to determine the correspondence between scalp locations where fNIRS measurements were performed and the underlying cortical surface where source signals were located. NIRS-SPM registered fNIRS channel data onto the Montreal Neurological Institute standard brain space. Regions of interest (ROIs) included the PFC, SMA, M1, S1, and V1. See Supplemental Materials for a montage of the average Montreal Neurological Institute coordinates for fNIRS emitter and detector optodes exported from NIRS SPM software including Brodmann areas corresponding to each optode.

Participants wore 6 Opal version 2.0 (V2; APDM Wearable Technologies, Portland, OR, USA) inertial sensors, each incorporating tri-axial accelerometers and gyroscopes sampling at 128 Hz. Sensors were placed at the lumbar spine, chest, both ankles, and both wrists and were wirelessly synchronised via Portasync (Artinis Medical Systems, The Netherlands) with the mobile fNIRS system. Five sway outcome measures were exported from the MobilityLAB version 2.0 (V2).^29,30 Instrumented Sway (ISway) plugin for automated analysis: area (m²), jerkiness (m²/s), velocity (m/s), frequency (Hz), and root mean square (m²). Sway frequency was defined as the total centroidal (anterior posterior + mediolateral) mean frequency.

Data Processing

fNIRS data processing and analysis was performed in line with previous fNIRS studies that incorporated short standing baseline periods as a reference for motor tasks.^30-32 The fNIRS data was processed using MATLAB2022b with the following steps: (1) data was low pass filtered (cut-off 0.14 Hz) removing high-frequency noise; (2) corrected for baseline (ie, removing the median of the first 20 s of the EOFS condition from the entire sensory-demanding trials; (3) reference channel correction (short 1.5 cm channels were subtracted from long 3 cm channel signals); (4) visual signal inspection; and (5) averaging across fNIRS channels for ROI.

Oxyegenated haemoglobin (HbO₂) was selected as the primary outcome as this is most commonly reported in fNIRS studies and has been suggested to be more sensitive to task-related haemodynamic changes due to its higher signal-to-noise ratio than deoxyhaemoglobin (HHb).³³ However, HHb was examined for completeness and transparency, with descriptive group-level medians and supplementary linear mixed-effects model results reported in the Supplemental Material. The median signal from the initial 20 seconds of the EOFS task was selected as the reference baseline to allow changes in cortical activity during sensory-challenging conditions to be examined relative to normal standing. This enabled task-related haemodynamic responses to be interpreted as reflecting the additional cortical involvement associated with reduced visual input, altered somatosensory feedback, or the combined challenge of standing on foam with eyes closed. This 20-second window was chosen to provide a consistent early estimate of haemodynamic activity during stable upright stance, when participants were expected to be settled in the task and performing under the least challenging sensory condition, while reducing the influence of later within-trial factors such as postural adjustments, fatigue, or signal drift.

Motion artefacts and signal quality were addressed through a combination of preprocessing and visual inspection. Although the standing tasks involved less gross movement than walking paradigms, postural sway, subtle head movements, weight shifts, and changes in optode–scalp coupling may still have affected the fNIRS signal. Therefore, signals were visually inspected to identify channels or trials with excessive noise. Channels with poor signal quality were excluded where appropriate prior to ROI averaging.

A 3D-digitizer (Polhemus Patriot) obtained morphological locations for cortical ROIs relative to scalp position and the fNIRS optodes. Optical density of the raw signal was measured and converted into HbO₂ (in Nanomolar [nM]) using Beer–Lamberts law³⁴ with the following equation (∆OD is the change in optical density and x is the path length):

\begin{array}{l} Δ ({HbO}_{2}) = - 3.6132) x Δ OA (780 nM) \\ + 1.1397 x Δ OD (805 nm) \\ + 3.0154 x Δ OD (830 nM) \end{array}

Statistical Analysis

Data were analysed using Statistical Package for the Social Sciences (v29, IBM, Chicago, IL, USA). Data normality was assessed using Kolmogorov–Smirnov tests, and parametric analyses were conducted based on Critical Central Theorem.³⁵ Descriptive characteristics were compared across groups (PD, OA, and YA) using 1-way analysis of variance; pairwise comparisons between groups were conducted using independent t-tests. A Spearman’s rank order correlation was conducted to assess the relationship between cortical activity and sway metrics. A P value of ≤ .05 was considered significant throughout the analysis.

Three separate linear mixed-effects models (LMEM) examined whether cortical activity (PFC, SMA, M1, S1, and V1) changed across the 3 sensory demanding standing tasks (EOFS was excluded as this was used as the baseline measures). An autoregressive (AR1) covariance structure was applied. The models tested the main effects of group (PD and OA/OA and YA/PD and YA), task (ECFS, EOFOAM, and ECFOAM) and their interaction. Each LMEM included a random intercept for each subject to account for repeated measures. Restricted maximum likelihood estimation was used, and degrees of freedom were calculated using Satterhaite approximation. Estimated marginal means were compared using Bonferroni-adjusted pairwise comparisons for group, task, and group × task interaction effects. Pairwise comparisons were conducted for significant effects. Similar LMEMs examined changes in sway metrics.

Results

Demographic Characteristics

A total of 155 participants were included in this study: 82 PD, 35 OA, and 38 YA. Table 1 shows PD and OA participants were similar in age: PD = 69.68 (8.14), OA = 68.15 (7.92), with the PD group having a higher number of males compared to females (P = .011). Significant differences were found in years of education (P < .001) between groups, with the PD group having the lowest mean average (13.27), compared to OA (15.03) and YA (17.32). The PD group performed significantly worse on the TMT A and B, CLOX 1 and 2, and JLo than OA and YA. PD performed worse than OA on TMT A (P < .001), TMT B (P < .001), CLOX 1 (P = .005), CLOX 2 (P < .001), and JLo (P = .010). The PD group had a significantly higher fear of falling (FES) compared to OA (P < .001) and YA (P < .001) and depression symptoms were significantly higher (P < .001) than the OA group. Visual acuity (P < .001) and contrast sensitivity (P < .001) were significantly worse in PwPD than the YA. However, compared to OA, PD differed in contrast sensitivity (P = .008), but not in visual acuity (P = .397).

Table 1.

Demographics, Cognition, Clinical, and Visual Characteristics.

		PD (n=80)	OA (n=33)	YA (n=38)	Sig. (p)
Demographics	Age Gender (M/F) Disease duration Education(years) Height (cm) Weight (kg) H&Y LEDD	69.68 (8.14) 58/24 5.34 (4.83) 13.27 (3.12) 168.79 (14.90) 79.86 (14.90) 2.12 (0.78) 582.46 (400.96)	68.15 (7.92) 15/18 - 15.03 (3.43) 170.71 (10.24) 80.58 (17.76) - -	22.55 (3.38) 18/19 - 17.32 (2.58) 175.43 (7.62) 75.43 (15.32) - -	<.001^b23 .011^a* - <.001^b123 .663^b .225^b - -
Cognition	MoCA Digit span TMT-A (s) TMT-B (s) CLOX 1 (s) CLOX 2 (s) JLO	26.78 (2.61) 5.93 (1.17) 43.56 (26.09) 109.41(78.18) 12.46 (1.49) 13.17 (1.29) 18.51 (6.23)	26.67 (1.74) 6.30 (1.40) 26.97 (8.78) 50.13 (16.04) 13.27 (1.40) 14.00 (1.25) 21.88 (5.30)	27.54 (1.87) 5.94 (1.01) 19.50 (5.42) 45.78 (20.50) 14.09 (0.79) 14.43 (0.79) 23.25 (4.01)	.227^b3 .303^b <.001^b123 <.001^b12 <.001^b123 <.001^b12 <.001^b12
Clinical	FES GDS MDS-UPDRS III	25.69 (8.80) 3.18 (2.75) 37.30 (16.64)	19.41 (3.34) 0.97 (0.97) -	21.76 (6.61) - -	<.001^b12 <.001^c1 -
Visual function	Contrast sensitivity Visual acuity	1.51 (0.17) 0.10 (0.18)	1.60 (0.14) 0.08 (0.13)	1.74 (0.06) -0.05 (0.09)	<.001^b123 <.001^b23
^aχ²		^bOne-way ANOVA		^cIndependent-samples t-test
¹PD–OA		²PD–YA		³OA–YA

Superscript numbers indicate significant pairwise group differences (p < .05): ¹PD–OA, ²PD–YA, ³OA–YA. Superscript letters indicate the statistical tests performed.

Abbreviations: ANOVA, analysis of variance; FES, Falls Efficacy Scale; GDS, Geriatric Depression Scale; H&Y, Hoehn and Yahr; JLo, Judgement of Line Orientation; LEDD, Levodopa equivalent daily dose; MDS-UPDRS III, Movement Disorders Society-Unified Parkinson’s Disease Rating Scale part III; MoCA, Montreal cognitive assessment; TMT, Trail making test; OA, older adults; PD, Parkinson’s disease; UPDRS, Unified Parkinson’s Disease Rating Scale; YA, younger adults.

Of note, 6 PwPD were unable to perform the ECFOAM condition, 3 of which were also unable to complete the EOFOAM condition. No imputation was applied; missing data were retained as linear mixed-effects models appropriately handle unbalanced data structures. See Supplemental Table 1 for comparisons of demographics, cognition, clinical, and visual characteristics between these 6 participants and the rest of the PD group. Significant differences were found including MDS-UPDRS part III (P = .021), H&Y (P = .004), and contrast sensitivity (P = .011). Worse scores were found in the group of participants that were unable to perform the foam standing tasks.

Participants unable to perform the foam standing tasks exhibited significantly greater sway during the EOFS condition, including increased sway area (1.00 ± 2.19 m² vs 0.11 ± 0.34 m²; P = .044), jerkiness (174.30 ± 380.65 m/s³ vs 6.28 ± 5.87 m/s³; P = .019), and root mean square (RMS) sway (0.26 ± 0.27 m² vs 0.12 ± 0.16 m²; P = .046) compared with participants who were able to complete the foam tasks. No significant differences were observed for sway velocity (1.10 ± 0.96 vs 0.94 ± 1.93 m/s; P = .676) or sway frequency (0.93 ± 0.30 vs 0.83 ± 0.19 Hz; P = .414; Figure 2).

fNIRS Outcome Measures

For PwPD and OA, there was a main effect of group in the PFC. PwPD showed greater PFC activity compared to OA across the 3 standing tasks (P = .033). A group effect (P = .001) in the S1 was also found for the LMEM comparing PwPD and YA, again with greater S1 activity in YA. Similarly, a significant group effect (P = .004) was found when comparing OA and YA in the S1 region with higher levels of ∆HbO² in YA. In the LMEM that included PwPD and YA, a main effect of Task was also found in the V1 region, with a significant decrease during the ECFOAM condition compared to ECFS (P = .048) and the EOFOAM condition (P = .023). (See Figure 2 for means and standard error for fNIRS outcomes per region, group, and task).

Sway Outcome Measures

Sway measures were significantly greater in PD compared to OA (sway area [P = .001], sway velocity [P = .006], RMS [P = .006], and sway frequency [P < .001]). Moreover, a significant task effect was found for models including PD and OA for sway area (P < .001), jerkiness (P < .001), velocity (P < .001), RMS (P < .001), and frequency (P = .014). Sway measures increased with increased task difficulty, that is, on foam versus firm surface.

One significant group effect was found in models including OA and YA. Namely, YA had greater sway frequency compared to OA (P < .001). Main effects for Task were found in sway area, jerkiness, velocity, RMS, and frequency (all P < .001). Pairwise comparisons demonstrated sway measures generally increased during the foam conditions compared to the firm surface conditions and with eyes closed compared to eyes open in all groups. An interaction effect was found for Sway Frequency (P < .001), with OA not showing any significant differences in the post hoc analysis. However, YA showed the following: EOFS < EOFOAM, EOFS < ECFOAM, ECFS < EOFOAM, ECFS < ECFOAM (all P < .001).

Finally, the LMEM including PD and YA showed significant group effects for sway area (P = .001), RMS (P = .014), and frequency (P < .001). Task effects were found for all outcome measures (all P < .001). People with PD had greater mean values across all sway outcome measures compared to YA. (See Figure 3 for means and standard error for sway outcomes per group and condition).

Associations Between fNIRS and Sway Metrics

Pd

During ECFS, greater SMA activity was associated with reduced postural instability, reflected by a negative correlation with RMS sway (rs = −.306, P = .032). For EOFOAM, sway jerkiness was positively associated with V1 activity (rs = .306, P = .029), whereas RMS sway showed a negative association with SMA activity (rs = −.362, P = .014).

Oa

During ECFS, SMA activity was negatively associated with sway frequency (rs = −.556, P = .032). During EOFOAM, S1 activity was negatively associated with both sway velocity (rs = −.619, P = .002) and RMS sway (rs = −.417, P = .027). For ECFOAM, sway frequency was again negatively associated with SMA activity (rs = −.588, P = .021).

Ya

During ECFS, sway area was positively associated with both M1 (rs = .480, P = .044) and V1 activity (rs = .639, P = .004). In addition, V1 activity was positively correlated with sway jerkiness (rs = .585, P = .011) and RMS (r = .536, P = .022). In ECFOAM, increased PFC activity was associated with greater sway area (rs = .444, P = .044; Figure 4).

Discussion

This is the first study, to our knowledge, that has employed fNIRS to measure cortical activity across multiple cortical regions during standing tasks, in PwPD, OA, and YA. The purpose of this study was to identify age-related and disease-related cortical adaptations involved in maintaining postural control. Across the 3 sensory-demanding standing conditions, group differences were found in the PFC with greater activity in PwPD, compared to OA. In addition, PwPD and OA both demonstrated reduced activity in the S1 region, compared to YA. These combined findings highlight potential cortical adaptation whereby PFC activity reflects an executive-attentional compensatory mechanism to mitigate deficits in automatic postural control in PwPD and an age-related reduction in somatosensory processing.

Cortical Control of Standing Balance Across Groups and Conditions

A significant group effect was observed in the PFC region with greater activity in PwPD compared to OA.^36,37 When sensory systems involved in postural control are challenged (ECFS, EOFOAM, and ECFOAM), PwPD appear to recruit additional PFC resources to maintain postural control. These findings align with previous research demonstrating individuals with Parkinsonian syndrome and increased PFC activity relative to healthy controls to maintain postural control.³⁸ Furthermore, studies investigating cortical control of gait have shown increased PFC activity during walking tasks in OA³⁹ and PwPD,^30,40 relative to standing. Together, these results suggest the PFC is recruited, possibly as compensation, for subcortical degeneration in PwPD that affects automatic postural control, particularly during sensory demanding tasks. However, ΔHbO₂ increases can also index processing inefficiency or neurovascular differences; thus, we avoid strong reverse inference and frame PFC changes as task-dependent resource reallocation.

Notably, no group or task effects were found in the M1. These findings suggest that M1 activity does not significantly change between increasingly demanding standing tasks. Moreover, age or PD does not significantly affect haemodynamic responses during balance tasks in the M1, according to these results. This is somewhat unexpected, as the M1 is a key structure involved in voluntary motor execution and control. Moreover, targeted neuromodulation of the M1 through anodal transcranial direct current stimulation has demonstrated improved effects of postural training compared to sham stimulation.⁴¹ However, M1 activity has predominantly been shown to be associated with reactive balance,⁴² for example, response to perturbations⁴³ and proactive balance, like mediating anticipatory postural adjustments (APAs), prior to gait initiation in people with PD.⁴⁴ The current study focusses on static balance and sensory integration during standing tasks, for which there is less evidence of M1 involvement. Taken together, these current findings suggest that cortical contributions to sensory demanding static balance may not be primarily mediated by the M1, with M1 engagement becoming more prominent during dynamic, anticipatory, or reactive postural control for all groups.

Greater ∆HbO² in the S1 was found in YA compared to PwPD and OA. This points to an age-related decrease in somatosensory processing. Age-related reductions in somatosensory processing have been previously shown with simple tactile stimulus.⁴⁵ These findings are consistent with previous studies highlighting diminished somatosensation with age.⁴⁶ One study reported that proprioceptive deficits had a greater detrimental effect on postural control than visual impairments in OA,⁴⁷ highlighting the importance of proprioception.

V1 activity decreased during ECFOAM relative to less challenging tasks, consistent with down-weighting vision when it is unreliable. These findings suggest a sensory reweighting with a shift to other sensory inputs such as vestibular and proprioceptive processes and reduced reliance on visual input when postural control becomes more challenging. Because this was not seen between PwPD and OA, this may suggest reduced cortical adaptability with ageing, with an increased ability to disengage V1 activity in YA. The increased reliance on visual input to maintain postural stability is well established in OA and PwPD. OA with diminished contrast sensitivity and stereopsis show poorer sway outcomes and have been identified as independent predictors of postural stability, specifically when standing on a foam surface.⁴⁸ Similarly, PwPD have been suggested to be over-reliant on visual information to maintain balance.⁴⁹ PD neuropathology disrupts the integration and processing of sensory input leading to impaired proprioceptive and vestibular feedback. As a result, this maladaptive strategy of overreliance on vision can cause issues when visual input is hindered (eg, low lighting). YA may be better able to integrate additional sensory information and inhibit their V1 activity dependent on the task demands.

Postural Stability Across Groups and Conditions

Postural stability was worse in PwPD than in OA across conditions; sway metrics were consistently higher in PwPD. This aligns with previous research demonstrating increased postural instability in PwPD,^1,50,51 and with evidence that such instability increases fall risk.⁵² Moreover, sway area, jerkiness, velocity, and RMS increased with task difficulty in YA and OA. However, in PwPD, a different pattern emerged: sway decreased on the eyes-closed firm surface (ECFS) relative to baseline (EOFS), but increased on both foam conditions (EOFOAM and ECFOAM). Similar results have been reported in previous studies. For example, whereas healthy adults and people with multiple sclerosis exhibited increased sway metrics with increasing task difficulty (side-by-side, semi-tandem, and tandem stance), PwPD did not.⁵³ Additionally, PwPD tend to reduce sway during a DT whereas OA show an increase in sway during the same condition.⁵⁴

These results underscore the impact of additional cognitive load on postural stability. Specifically, PwPD appear to overly constrain their postural adjustments as task difficulty increases. However, our results suggest that when proprioception is further challenged (ie, standing on a foam surface), this compensatory strategy is either abandoned or insufficient, leading to increased postural instability. PwPD often show delayed and inappropriately scaled reactive responses, requiring additional steps to recover when executing reactive postural responses.⁵⁵ This maladaptive strategy, combined with bradykinesia and rigidity, may heighten vulnerability to unexpected perturbations and uneven surfaces, where rapid postural adjustments are required.

Sway frequency was the only outcome measure that differed between OA and YA, with higher sway frequency observed in YA. This suggests a reduction in postural sway frequency with age. Similarly, 1 study found reduced sway frequency during volitional head movements in OA, compared to static stance, whilst YA did not change.⁵⁶ In conjunction, YA significantly increased sway velocity between the EOFS and EOFOAM conditions. Moreover, YA have been shown to exhibit more irregular fluctuation in postural acceleration compared to middle-aged adults (mean age: 51.4 ± 5.9 years).⁵⁷ More efficient sensory processing may allow YA to make larger adjustments with altered proprioceptive feedback reflecting greater efficiency or automaticity in postural control. Age-associated increase in postural variability has been linked to larger amplitude, slower oscillations in the <0.5 Hz band, consistent with a shift towards low-frequency sway components with ageing.⁴⁴ Moreover, the neuromuscular system in YA is generally more responsive and can facilitate these adaptations without compromising stability.^58-60 This may support more efficient neural processing of rapid responses in YA. These findings reflect changes in postural strategies with age. Yet, PwPD showed the highest mean sway frequency, which was significantly higher than OA across all conditions. This may be attributed to disease-specific alterations in postural strategies.

Associations Between Cortical Activity and Sway

In PwPD, greater SMA activity was associated with lower RMS during ECFS and EOFOAM. Although lower RMS is often interpreted as “better” stability, reductions in sway amplitude can also arise from stiffening/over-constraining strategies that prioritise immediate steadiness at the expense of adaptability.⁶⁰ However, if the hypothesis that PwPD are over constraining when sensory systems are challenged is correct, this may be mediated by SMA activity. The SMA is involved in generating APA’s, which help stabilise postural control in preparation for movement.⁶¹ This suggests that PwPD may rely more heavily on top–down motor planning via the SMA during sensory-demanding standing tasks, possibly as a compensatory strategy. While this may reduce sway, it could reflect a rigid or overly constrained postural strategy that limits adaptability. Because ΔHbO₂ is a haemodynamic proxy, increases do not by themselves prove compensation (they may index effort or neurovascular differences); we therefore interpret these associations cautiously and in context with sway behaviour.

Similarly, in OA, SMA activity was associated with lower sway frequency during ECFS and ECFOAM. Reduced complexity of postural sway has been associated with reduced adaptability to stressors and increased risk of falling in OA.⁶² The current results may reflect that increased SMA activity is linked to less adaptive postural control in OA when there is an increased reliance on the vestibular system. In YA, increased M1, V1 activity during ECFS and PFC activity during ECFOAM was associated with greater sway area. This may indicate that reduced subcortical governing of postural control leads to increased postural instability, even in YA. The visual system has been suggested to be the predominant sensory system used by YA for optimal postural control.⁶³ The current associations between increased V1 activity and several sway metrics during ECFS suggest there may be an inability to reduce V1 activity and reallocate cortical resources in these individuals.

Limitations

To our knowledge, this is the largest study to date to assess full cortical control of balance both in ageing and PwPD. However, there were limitations to this study. First, due to hardware and software issues, there were several missing data points across all 3 groups. To account for missing data, LMEM’s were used as this method of analysis is robust to missing data and random effects.⁶⁴ Second, it is important to acknowledge that 6 PwPD were unable to perform the ECFOAM condition, 3 of which were also unable to complete the EOFOAM condition. All 6 were classified as H&Y III. These individuals had several larger sway parameters during EOFS, worse contrast sensitivity and more severe motor impairment (MDS-UPDRS III) compared to participants who were able to complete all 4 standing trials. See Supplemental Materials Table 1 for demographics and sway outcome measure comparisons. The inability to perform this task is likely due to increased impairment of postural control. As a result, cortical and sway data in PwPD may underestimate group differences under foam standing conditions, with these individuals showing poorer performance. In addition, participants performed standing tasks for a duration of 2 minutes, which is a relatively longer trial time. During longer standing trials, participants may naturally make small weight shifts, postural adjustments, or slow drifts in centre of pressure position.^65,66 These shifts may influence traditional sway metrics, such as sway area, velocity, or frequency. Therefore, the sway outcomes should be interpreted as reflecting performance during relatively prolonged standing and may not directly generalise to shorter quiet-standing assessments. To mitigate fatigue, participants had the opportunity to take seated rest breaks in between each standing task. However, the fixed order of tasks may have introduced fatigue effects during foam standing conditions which must be acknowledged. Moreover, there were large discrepancies in sample size between PD and YA and OA groups, these imbalances are an important consideration when interpreting these results. Finally, given the exploratory nature of this study with multiple testing, these findings should be interpreted with caution. This study was designed to generate initial insights to differences in age-related and disease-related cortical adaptations in relation to postural stability with further studies required to validate our findings.

Conclusions

Findings indicate postural control is worse in PwPD compared to YA and age-matched OA. Moreover, relative to EOFS standing, increased PFC activity was found in PwPD compared to OA, suggesting that PwPD rely on prefrontal executive-attentional compensatory mechanism to mitigate deficits in automatic postural control. Furthermore, YA demonstrated greater S1 acitivty than both OA and PD, suggesting an age-related reduction in somatosensory processing. Together, these findings point to a task-dependent and region-specific pattern of cortical involvement in postural control, with age and disease influencing the extent of cortical engagement. Longitudinal fNIRS-balance studies and interventions targeting somatosensory integration and executive load could test causality.

Supplemental Material

sj-docx-1-nnr-10.1177_15459683261454945 – Supplemental material for Neural Correlates of Balance in People With Parkinson’s Disease, Older, and Younger Adults: An fNIRS Study

Supplemental material, sj-docx-1-nnr-10.1177_15459683261454945 for Neural Correlates of Balance in People With Parkinson’s Disease, Older, and Younger Adults: An fNIRS Study by Patrick Tait, Rodrigo Vitorio, Lisa Graham, Tamlyn Watermeyer, Richard Walker, Claire McDonald, Martina Mancini, Samuel Stuart and Rosie Morris in Neurorehabilitation and Neural Repair

Footnotes

Acknowledgements

This study was funded by the Parkinson’s Foundation (PI: Dr Samuel Stuart) through a Post-doctoral Fellowship for Basic Scientists (PF-FBS-1898-18-21) and a Clinical Research Award (PF-CRA-2073). Further funding was received through a Northumbria University PhD studentship (PI: Dr Rosie Morris). Dr Morris is supported by an NIHR Advanced Fellowship (NIHR303544). Dr Watermeyer is the recipient of a post-doctoral fellowship from the National Institute for Health and Care Research (NIHR) Applied Research Collaboration North East & Cumbria. This fellowship was part of an initiative funded by the NIHR and Alzheimer’s Society to support post-doctoral capacity building in applied dementia research. The views expressed are those of the author(s) and not necessarily those of the Alzheimer’s Society, NIHR, or the Department of Health.

ORCID iDs

Patrick Tait

Rodrigo Vitorio

Lisa Graham

Richard Walker

Martina Mancini

Author Contributions

Patrick Tait: Formal analysis; Writing—original draft; and Writing—review & editing. Rodrigo Vitorio: Conceptualisation; Data curation; Formal analysis; Methodology; Software; Supervision; and Writing—review & editing. Lisa Graham: Investigation and Writing—review & editing. Tamlyn Watermeyer: Supervision and Writing—review & editing. Richard Walker: Writing—review & editing. Claire McDonald: Writing—review & editing. Martina Mancini: Writing—review & editing. Samuel Stuart: Supervision; Writing—review & editing. Rosie Morris: Conceptualisation; Formal analysis; Funding acquisition; Supervision; and Writing—review & editing.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Supplementary Material

Supplemental material for this article is available online.

References

Park

Kang

Horak

FB.

What is wrong with balance in Parkinson’s disease?

J Mov Disord. 2015;8(3):109-114.

Crouse

Phillips

Jahanshahi

Moustafa

AA.

Postural instability and falls in Parkinson’s disease. Rev Neurosci. 2016;27(5):549-555.

Xia

Zhou

, et al. Factors associated with balance impairments in the community-dwelling elderly in urban China. BMC Geriatrics. 2023;23(1):545.

Sturnieks

St George

Lord

SR.

Balance disorders in the elderly. Neurophysiol Clin. 2008;38(6):467-478.

Smulders

Dale

Carlson-Kuhta

Nutt

Horak

FB.

Pharmacological treatment in Parkinson’s disease: effects on gait. Parkinsonism Relat Disord. 2016;31:3-13.

Curtze

Nutt

Carlson-Kuhta

Mancini

Horak

FB.

Levodopa is a double-edged sword for balance and gait in people with Parkinson’s disease. Mov Disord. 2015;30(10):1361-1370.

Tait

Graham

Vitorio

, et al. Neuroimaging and cognitive correlates of postural control in Parkinson’s disease: a systematic review. J NeuroEng Rehabil. 2025;22(1):24.

Xiao

Yang

Smith

, et al. Correlation between cognition and balance among middle-aged and older adults observed through a Tai Chi intervention program. Front Psychol. 2020;11:668.

Morris

Lord

Bunce

Burn

Rochester

Gait and cognition: mapping the global and discrete relationships in ageing and neurodegenerative disease. Neurosci Biobehav Rev. 2016;64:326-345.

10.

Nóbrega-Sousa

Gobbi

LTB

Orcioli-Silva

Conceição

NRD

Beretta

Vitório

Prefrontal cortex activity during walking: effects of aging and associations with gait and executive function. Neurorehabil Neural Repair. 2020;34(10):915-924.

11.

Stuart

Vitorio

Morris

Martini

Fino

Mancini

Cortical activity during walking and balance tasks in older adults and in people with Parkinson’s disease: a structured review. Maturitas. 2018;113:53-72.

12.

Palmer

Payne

Ting

Borich

MR.

Cortical engagement metrics during reactive balance are associated with distinct aspects of balance behavior in older adults. Front Aging Neurosci. 2021;13:684743.

13.

Tait

Morris

Watermeyer

, et al. Chapter 15 - Balance assessment in neurological conditions. In: Stuart

Morris

, eds. Gait, Balance, and Mobility Analysis. Academic Press; 2025:405-434.

14.

Huang

Chen

Hwang

IS.

Effects of task prioritization on a postural-motor task in early-stage Parkinson’s disease: EEG connectivity and clinical implication. Geroscience. 2022;44(4):2061-2075.

15.

Bosch

Kammermeier

Groth

, et al. Cortical and cerebellar oscillatory responses to postural instability in Parkinson’s disease. Front Neurol. 2021;12:752271.

16.

Jacob

Guerrini

Pescaglia

, et al. Adaptation strategies and neurophysiological response in early-stage Parkinson’s disease: BioVRSea approach. Front Hum Neurosci. 2023;17:1197142.

17.

Peterka

RJ.

Chapter 2 - Sensory integration for human balance control. In: Day

Lord

, eds. Handbook of Clinical Neurology. Elsevier; 2018:27-42.

18.

KZH

Bherer

Mirelman

Maidan

Hausdorff

. Cognitive involvement in balance, gait and dual-tasking in aging: a focused review from a neuroscience of aging perspective. Front Neurol. 2018;9:913.

19.

Clarke

Patel

Ives

, et al. UK Parkinson’s Disease Society Brain Bank Diagnostic Criteria. NIHR Journals Library; 2016.

20.

Vitório

Morris

Das

Walker

Mancini

Stuart

Brain activity response to cues during gait in Parkinson’s disease: a study protocol. PLoS ONE. 2022;17(11):e0275894.

21.

Aikman

Oehlert

ME.

Geriatric Depression Scale: long form versus short form. Clin Gerontol. 2001;22(3-4):63-70.

22.

Nasreddine

Phillips

Bédirian

, et al. The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc. 2005;53(4):695-699.

23.

Llinàs-Reglà

Vilalta-Franch

López-Pousa

Calvó-Perxas

Torrents Rodas

Garre-Olmo

The trail making test. Assessment. 2017;24(2):183-196.

24.

Benton

Varney

Hamsher

KD.

Visuospatial judgment. A clinical test. Arch Neurol. 1978;35(6):364-367.

25.

Royall

Cordes

Polk

CLOX: an executive clock drawing task. J Neurol Neurosurg Psychiatry. 1998;64(5):588-594.

26.

Mishra

King

PV.

Research with the wechsler digit span subtest: implications for assessment. School Psychol Rev. 1985;14(1):37-47.

27.

Yardley

Beyer

Hauer

Kempen

Piot-Ziegler

Todd

Development and initial validation of the Falls Efficacy Scale-International (FES-I). Age Ageing. 2005;34(6):614-619.

28.

Kurlowicz

The Geriatric Depression Scale (GDS). J Gerontol Nurs. 1999;25(7):8-9.

29.

Morris

Stuart

McBarron

Fino

Mancini

Curtze

Validity of Mobility Lab (version 2) for gait assessment in young adults, older adults and Parkinson’s disease. Physiol Meas. 2019;40(9):095003.

30.

Stuart

Belluscio

Quinn

Mancini

Pre-frontal cortical activity during walking and turning is reliable and differentiates across young, older adults and people with Parkinson’s disease. Front Neurol. 2019;10:536.

31.

Vitorio

Stuart

Mancini

Executive control of walking in people with Parkinson’s disease with freezing of gait. Neurorehabil Neural Repair. 2020;34(12):1138-1149.

32.

Stuart

Mancini

Prefrontal cortical activation with open and closed-loop tactile cueing when walking and turning in Parkinson disease: a pilot study. J Neurol Phys Ther. 2020;44(2):121-131.

33.

Menant

Maidan

Alcock

, et al. A consensus guide to using functional near-infrared spectroscopy in posture and gait research. Gait & Posture. 2020;82:254-265.

34.

Delpy

Cope

Zee

Pvd

Arridge

Wray

Wyatt

Estimation of optical pathlength through tissue from direct time of flight measurement. Phys Med Biol. 1988;33(12):1433.

35.

Charpentier

The SAGE Encyclopedia of Educational Research, Measurement, and Evaluation. SAGE; 2018.

36.

Silva-Batista

Liu

Vitorio

Stuart

Quinn

Mancini

The time course of changes in prefrontal cortex activity during walking in people with Parkinson’s disease. Neurorehabil Neural Repair. 2024;38(9):635-645.

37.

Ranchet

Hoang

Cheminon

, et al. Changes in prefrontal cortical activity during walking and cognitive functions among patients with Parkinson’s disease. Front Neurol. 2020;11:601686.

38.

Mahoney

Holtzer

Izzetoglu

Zemon

Verghese

Allali

The role of prefrontal cortex during postural control in Parkinsonian syndromes a functional near-infrared spectroscopy study. Brain Res. 2016;1633:126-138.

39.

Hoang

Paire-Ficout

Derollepot

Perrey

Devos

Ranchet

Increased prefrontal activity during usual walking in aging. Int J Psychophysiol. 2022;174:9-16.

40.

Nieuwhof

Reelick

Maidan

, et al. Measuring prefrontal cortical activity during dual task walking in patients with Parkinson’s disease: feasibility of using a new portable fNIRS device. Pilot Feasibil Stud. 2016;2(1):59.

41.

Jafarzadeh

Ehsani

Yosephi

Zoghi

Jaberzadeh

Concurrent postural training and M1 anodal transcranial direct current stimulation improve postural impairment in patients with chronic low back pain. J Clin Neurosci. 2019;68:224-234.

42.

Monaghan

Takla

Ofori

, et al. The neural contributions to reactive balance control: a scoping review of EEG, fNIRS, MRI, and PET studies. Brain Sci. 2025;15:12.

43.

Smith

Fisher

BE.

Anticipatory postural adjustments and spatial organization of motor cortex: evidence of adaptive compensations in healthy older adults. J Neurophysiol. 2018;120(6):2796-2805.

44.

Petersen

Rosenberg

Petersen

Nielsen

JB.

Cortical involvement in anticipatory postural reactions in man. Exp Brain Res. 2009;193(2):161-171.

45.

Brodoehl

Klingner

Stieglitz

Witte

OW.

Age-related changes in the somatosensory processing of tactile stimulation–an fMRI study. Behav Brain Res. 2013;238:259-264.

46.

Shaffer

Harrison

AL.

Aging of the somatosensory system: a translational perspective. Phys Ther. 2007;87(2):193-207.

47.

Judge

King

Whipple

Clive

Wolf son

LI.

Dynamic balance in older persons: effects of reduced visual and proprioceptive input. J Gerontol Ser A. 1995;50A(5):M263-M270.

48.

Lord

Menz

HB.

Visual contributions to postural stability in older adults. Gerontology. 2000;46(6):306-310.

49.

Wylie

van den Wildenberg

Ridderinkhof

Claassen

Wooten

Manning

CA.

Differential susceptibility to motor impulsivity among functional subtypes of Parkinson’s disease. J Neurol Neurosurg Psychiatry. 2012;83(12):1149-1154.

50.

Martini

Morris

Madhyastha

, et al. Relationships between sensorimotor inhibition and mobility in older adults with and without Parkinson’s disease. J Gerontol A Biol Sci Med Sci. 2021;76(4):630-637.

51.

Palakurthi

Burugupally

SP.

Postural instability in Parkinson’s disease: a review. Brain Sci. 2019;9(9):239.

52.

Vitorio

Mancini

Carlson-Kuhta

Horak

Shah

VV.

Should we use both clinical and mobility measures to identify fallers in Parkinson’s disease?

Parkinsonism Relat Disord. 2023;106:105235.

53.

Warmerdam

Schumacher

Beyer

, et al. Postural sway in Parkinson’s disease and multiple sclerosis patients during tasks with different complexity. Front Neurol. 2022;13:857406.

54.

Holmes

Jenkins

Johnson

Adams

Spaulding

SJ.

Dual-task interference: the effects of verbal cognitive tasks on upright postural stability in Parkinson’s disease. Parkinsons Dis. 2010;2010:696492.

55.

Bath

Wang

DD.

Unraveling the threads of stability: a review of the neurophysiology of postural control in Parkinson’s disease. Neurotherapeutics. 2024;21(3):e00354.

56.

Koceja

Allway

Earles

DR.

Age differences in postural sway during volitional head movement. Arch Phys Med Rehabil. 1999;80(12):1537-1541.

57.

Promsri

Pitiwattanakulchai

Saodan

Thiwan

Age-related changes in postural stability in response to varying surface instability in young and middle-aged adults. Sensors. 2024;24(21):6846.

58.

Shirazi

Huang

HJ.

Older adults use fewer muscles to overcome perturbations during a seated locomotor task. J Neurophysiol. 2024;131(6):1250-1259.

59.

Pavol

Runtz

Edwards

Pai

Y-C.

Age influences the outcome of a slipping perturbation during initial but not repeated exposures. J Gerontol Ser A. 2002;57(8):M496-M503.

60.

Hunter

Pereira

Keenan

KG.

The aging neuromuscular system and motor performance. J Appl Physiol (1985). 2016;121(4):982-995.

61.

Jacobs

Lou

Kraakevik

Horak

FB.

The supplementary motor area contributes to the timing of the anticipatory postural adjustment during step initiation in participants with and without Parkinson’s disease. Neuroscience. 2009;164(2):877-885.

62.

Zhou

Habtemariam

Iloputaife

Lipsitz

Manor

The complexity of standing postural sway associates with future falls in community-dwelling older adults: the MOBILIZE Boston Study. Sci Rep. 2017;7(1):2924.

63.

Grace Gaerlan

Alpert

Cross

Louis

Kowalski

Postural balance in young adults: the role of visual, vestibular and somatosensory systems. J Am Acad Nurse Pract. 2012;24(6):375-381.

64.

Wilkinson

Mazzo

Feeney

DF.

Rethinking the statistical analysis of neuromechanical data. Exerc Sport Sci Rev. 2023;51(1):43-50.

65.

Zatsiorsky

Duarte

Rambling and trembling in quiet standing. Motor Control. 2000;4(2):185-200.

66.

Duarte

Zatsiorsky

VM.

Patterns of center of presure migration during prolonged unconstrained standing. Motor Control. 1999;3(1):12-27.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

1.50 MB