Walking While Talking and Prefrontal Oxygenation in Motoric Cognitive Risk Syndrome: Clinical and Pathophysiological Aspects

Abstract

Background:

Motoric cognitive risk syndrome (MCR) combines slow gait and cognitive complaints and has been proposed as a predementia syndrome. The nature of dual-task performance in MCR has not been established.

Objective:

To assess differences in dual-task performance between participants with and without MCR and to study the prefrontal cortex (PFC)-based brain activity during dual-task using functional near-infrared spectroscopy.

Methods:

Cohort study of community-dwelling non-demented older adults included in the “Central Control of Mobility in Aging” study. Comprehensive assessment included global cognition and executive function tests along with clinical variables. Dual-task paradigm consisted in walking while reciting alternate letters of the alphabet (WWT) on an electronic walkway. We compared dual-task performance between MCR (n = 60) and No MCR (n = 478) participants and assessed the relationship of dual-task performance with cognitive function. In a subsample, we compared PFC oxygenation during WWT between MCR (n = 32) and No MCR (n = 293).

Results:

In our sample of 538 high-functioning older adults (76.6±6.5 years), with 11.2% prevalence of MCR, dual-task cost was not significantly different, compared to No MCR participants. Among MCR participants, no significant relationship was found between WWT velocity and cognitive function, whereas No MCR participants with better cognitive function showed faster WWT velocities. PFC oxygenation during WWT was higher in MCR compared to No MCR (1.02±1.25 versus 0.66±0.83, p = 0.03).

Conclusion:

MCR participants showed no significant differences in the dual-task cost while exhibiting higher PFC oxygenation during dual-task walking. The dual-task performance (WWT velocity) in MCR participants was not related to cognition.

Keywords

Cognition dual-task motoric cognitive risk syndrome near-infrared spectroscopy prefrontal cortex

INTRODUCTION

Motoric cognitive risk syndrome (MCR) has been proposed as a pre-dementia syndrome and is defined as the coexistence of slow gait and subjective cognitive complaints in the absence of dementia and significant mobility disability [1, 2]. MCR is associated with increased risk for dementia [3], both Alzheimer’s disease (AD) [4] and vascular dementia [1], as well as with disability [3], falls [5], and mortality [6]. Executive dysfunction is reported in persons with MCR [7]. While motoric impairment is key to the diagnosis of MCR, recent studies show that, in individuals diagnosed with MCR, neither the presence of clinical gait abnormalities [8] nor slowing of gait [9] predicts transition to dementia. As a possible interpretation, once a person is diagnosed with MCR, the risk of developing dementia is better explained by the cognitive dysfunction rather than by gait slowing. Alternatively, gait speed as a single motoric assessment may not be the best parameter to predict progression to dementia [10]. A more sensitive locomotion assay such as dual-task gait, that taps into both cognitive and motor processes, might be needed to identify MCR participants at higher risk of progression of dementia. Dual-task paradigms increase the attentional cognitive load during walking and, therefore, may help identify subjects with fewer cognitive resources. The decrement in dual-task performance compared to the single task is larger with increasing cognitive demand [11] as well as in people with impaired mobility [12, 13] or cognition, especially executive dysfunction [14, 15]. The PFC has been identified as a key region for dual-task neural control [16, 17] and has shown structural changes among MCR [18]. Therefore, studying the PFC might help understand dual-task performance among MCR.

The nature of dual-task performance in MCR has not been established. Hence, to address this knowledge gap, the current study aims to examine differences in dual-task performance between participants with and without MCR in a cohort of community-dwelling non-demented older adults. We hypothesized that MCR participants would show worse dual-task performance, compared to ‘No MCR’ counterparts. As a secondary aim, we examined the prefrontal cortex (PFC)-based brain activity during the dual-task paradigm using functional near-infrared spectroscopy (fNIRS) in a subsample of the same cohort. fNIRS is a non-invasive, portable optic technology that measures cortical oxy- (HbO₂) and deoxyhemoglobin concentration over the PFC regions underlying the forehead and enables to overcome the limitation of conventional neuroimaging techniques to measure brain activity during actual movement [19, 20]. A previous study in the same cohort showed that increased PFC activation in high functioning individuals predicted falls, indicating early brain activity changes may precede behavioral changes [21]. We hypothesized that PFC brain activation would be higher in MCR participants than in No MCR as a compensatory mechanisms due to underlying motor and cognitive impairments [22]. Establishing the nature of dual-task and underlying neurophysiological correlates in patients with MCR may help improve risk assessments and guide future interventions in this vulnerable population.

METHODS

Participants

Participants were enrolled in the “Central Control of Mobility in Aging” (CCMA) study, a cohort study whose primary aim was to assess cognitive and neural predictors of mobility in older adults. Recruitment and other procedures in the CCMA study have been previously described [23]. Briefly, community-dwelling, non-demented, older adults aged 65 and older living in lower Westchester county (NY, USA) area were contacted via mail, and then by telephone, to invite them to participate and to assess eligibility. Exclusion criteria included inability to communicate in English, significant audiovisual impairment, dementia, active psychiatric disorders, hemodialysis, inability to walk even with assistance, and recent or scheduled medical procedures that could affect mobility. Eligible participants were evaluated in two visits at the research center in order to collect demographic, clinical, and functional status variables. They also underwent neuropsychological assessments and a structured neurological exam [24]. Dementia diagnosis was assigned at consensus case conference according to the Diagnostic and Statistical Manual of Mental Disorders, 4th edition [25], and after reviewing all available clinical and neuropsychological data [26]. Dementia cases and participants without complete data regarding cognitive diagnosis and quantitative gait assessment were excluded from the present study. From an original sample of 591 participants, 53 were excluded. Nine participants were excluded due to dementia diagnosis, 2 due to missing data, and 42 due to missing data not allowing MCR status classification or cognitive diagnosis at case-conference. The CCMA study includes yearly follow-up visits but for the purpose of the current analysis we only included baseline assessments.

The institutional review board of Albert Einstein College of Medicine approved the research protocols and participants provided written informed consent.

Motoric cognitive risk syndrome

MCR is defined as the presence of subjective cognitive complaints and slow gait velocity in older individuals without dementia or mobility disability. MCR builds on definitions of mild cognitive impairment (MCI), substituting the objective cognitive impairment criterion with slow gait. MCR was operationalized according to previously established criteria employed in the CCMA study [1, 2]: 1) Subjective cognitive complaints were assessed by one or more of the following: a ‘yes’ response to “Do you feel that you have more problems with memory than most?” or a ‘no’ response to “Is your mind as clear as it used to be?” on the Geriatric Depression Scale (GDS) [27]; a score of≥1 on the AD-8 dementia screener [28]; presence of cognitive symptoms identified by study clinician [29]; 2) Slow gait defined as gait speed≥1 standard deviation (SD) below age and sex means (Men 60–74 years: 86.2 cm/s; men≥75 years: 76.4 cm/s; women 60–74 years: 84.7 cm/s; women≥75 years: 66.1 cm/s) [4]; 3) absence of significant disability and 4) absence of dementia. Absence of disability was defined by preserved activities of daily living and ability to walk over the walkway unassisted.

Walk protocol and quantitative gait measures

The walk protocol consisted of two single tasks and one dual-task condition that were administered by trained psychology assistants. For the cognitive single task “Alpha”, participants were asked to recite alternate letters of the alphabet out loud over 30 s starting with the letter “B” while standing. The single task “Normal Walk” (NW) consisted of walking at a self-paced walking velocity over the electronic walkway. The “Walk While Talk” (WWT) dual-task combined both single tasks (NW and Alpha) as participants were asked to walk the same trajectory while reciting the alphabet starting with the letter “B”. Participants were instructed to pay equal attention to both tasks to avoid prioritizing either of the tasks [30]. Task order was counterbalanced using a Latin-square design to avoid learning or fatigue effects.

The walk protocol was applied in two different settings:

All participants performed the above described protocol while quantitative gait parameters were measured on a straight pathway with an electronic walkway with embedded sensors (GAITRite, CIR systems, Havertown, PA) measuring 8.5 m×0.9 m with an active recording area of 6.1 m×0.61 m. Mean velocity (cm/s) was calculated (distance walked over the gait mat/ambulation time) during NW and WWT tasks consistent with previous studies [15, 31] and Dual-Task Cost (DTC) was calculated as follows: [(WWT Velocity-NW Velocity)/NW Velocity]. Negative values are interpreted as a decrement in gait velocity relative to gait velocity in normal walk. Dual-task interference parameters were calculated to account for NW velocity with the DTC since we expected NW velocity to be different between groups. For behavioral results of the alpha condition, rate of correct letter production per second were calculated as follows: i) Alpha condition: number of correct letters/30 s; ii) WWT: number of correct letters/ambulation time. A DTC for the rate of correct letter was calculated [WWT correct rate – alpha correct rate) / alpha correct rate].

A subsample of the eligible participants underwent fNIRS measurements while performing the single and dual-task conditions described above. As previously reported in the CCMA cohort [32], NW and WWT conditions were performed for three continuous loops (6 straight segments and 5 turns) while gait parameters were measured with a 1.2×4.3 m electronic walkway (Zeno electronic walkway, Zenometrics LLC, Peekskill, NY) that allowed a continuous gait assessment during the 3-loop walkway. Stride velocity, which is the ratio of stride length to stride time (cm/s), was used to assess gait speed during the fNIRS measures.

fNIRS system

Changes in PFC oxygenation parameters during the execution of the described cognitive and gait tasks were measured using a fNIRS Imager 1000 (fNIRS Devices, LLC, Potomac, MD) device. Graphic illustration and details about the fNIRS system and data processing in the CCMA cohort have been published previously [32]. Briefly, the fNIRS system consisted of 4 LED light sources and 10 detectors with a source-detector distance of 2.5 cm. Light sources (Epitex Inc. type L4X730/4X805/4X850-40Q96-I) generated peak wave-lengths at 730, 805, and 850 nm. Photodetectors (Bur Brown, type OPT101) were monolithic photodiodes with a single supply transimpedance amplifier. Sampling rate was 2 Hz. The fNIRS system was built on a flexible board, which covered the forehead of the participant with 16 channels. A standardized optode placement procedure was performed as follows: the horizontal symmetry axis central (y-axis) matched with the symmetry axis of the head (between the eyes). On the vertical axis, the bottom channel row was positioned approximately on FP1 and FP2 according to the international 10-20 system [33].

Preprocessing and hemodynamic signal extraction

Data quality of the channels was inspected and accordingly removed from analysis if saturation or dark current conditions were identified. A finite impulse response filter with cut-off frequency of 0.14 Hz was used to eliminate possible respiration and heart rate signals, and unwanted high frequency noise on the raw intensity measures at 730 and 850 nm [34]. Using the modified Beer-Lambert law, oxygenated hemoglobin (HbO₂), deoxygenated hemoglobin (Hb), oxygen index (HbO₂-Hb), and total hemoglobin (HbO² ⁺ Hb) were calculated from the raw data at 730 and 850 nm [35]. In the present study HbO₂ measures were used to assess PFC hemodynamic changes during the cognitive and motor tasks since they have shown to be more sensitive to gait-related changes in regional cerebral blood flow [36]. Relative changes in the concentration of HbO₂ in each task were calculated with a normalized baseline condition using a 10 s period, where participants counted silently at a rate of about one number per second. Baseline levels for this 10 s period were adjusted to a mean HbO₂ value of zero to calculate the relative changes in each experimental condition. Excellent internal consistency for HbO₂ measurements in all three conditions was previously reported [37].

Epoch and feature extraction

Mean HbO₂ data were extracted separately for each channel and for each task. For the walking tasks, fNIRS and gait events were synchronized to optimize task related HbO₂ acquisition [32]. Average HbO₂ levels based on the 16 channels over the duration of each task (Alpha, NW and WWT) were used for the current analysis. HbO₂ data are reported as standard deviation units.

Cognitive function

An extensive neuropsychological test battery was administered at all visits. We examined performance on the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) [38], a widely used omnibus test of general cognition as well as various cognitive domains. It is a valid and reliable tool for detecting cognitive deficits across different age levels and diagnostic groups and consists of 12 subtests making up five indices: immediate memory, delayed memory, visuospatial/constructional, language, and attention. Index scores as well as total scores were calculated; scores range from 62 to 138 with higher scores reflecting better performance. For the purpose of this study, we examined RBANS total scores. The comprehensive neuropsychological evaluation included tests of executive function. Verbal fluency (VF) tests were performed following widely used testing protocols. Participants were instructed to produce as many words as possible during one minute starting with the letters F, A, and S (phonemic VF [39]) and belonging to the semantic categories animals, fruits, and vegetables (categorical VF [39]). VF tests engage verbal production, retrieval of information, attention and executive function [40]. Both parts of the Trail Making Test (TMT) [41] were administered. During Part A of the TMT, participants connected with a line 25 digits in ascending order. For the Part B, participants had to connect in an alternate manner numbers and letters in ascending order (e.g., 1-A-2-B-3-C- ...). Participants were instructed to perform both parts as quickly as possible and since score corresponds to the time needed to complete each part, higher scores indicate worse performance. TMT has been reported to involve visual scanning, psychomotor speed, working memory, mental flexibility and executive control [42, 43]. During the 90-s Digit Symbol Substitution Test [44], subjects were asked to write the corresponding symbol under each digit according to a matched symbol-digit key displayed on the top of the test form. This test assesses mainly psychomotor speed and attention, with higher scores indicating better performance. For the Digit Symbol Substitution Test, we report scaled scores while TMT A and B and VF are reported as z-scores. Depressive symptoms were assessed with the 30-item Geriatric Depression Scale (score range 0–30), higher scores suggest more depressive symptoms [27]. Cognitive status was assessed at consensus case conference, where MCI diagnosis was defined according to established criteria [45], if participants had cognitive complaints, evidence of impairment in at least one cognitive domain (neuropsychological tests scores 1.5 SD below age or sex-specific means) and independent functioning for the activities of daily living.

Other covariates

Covariates were selected based on relevant characteristics to describe our sample and compare groups according to MCR status. Socio-demographic variables such as age, sex and years of education were collected. Self-reported previous and current disease history collected during participant interviews were used to derive a previously described comorbidity index [46] which combines the presence of diabetes mellitus, hypertension, chronic heart failure, myocardial infarction, angina, chronic obstructive pulmonary disease, stroke, Parkinson’s disease, depression, and arthritis. Functional status was evaluated using a scale that assessed activities of daily living (ADLs) such as bathing, grooming, getting dressed, feeding, toileting, getting up from a chair, and indoor walking [47, 48]. Needing assistance with or inability to perform any one of the activities was scored as disability [49], so that higher scores indicate worse functional status. Difficulty performing instrumental activities of daily living (IADL) such as driving, doing laundry, shopping, cooking, using phone, managing money and medication were also assessed using a 15-item scale modified from the Activities of Daily Living-Prevention Instrument (ADL-PI) which was developed as an IADLs assessment instrument for primary prevention studies in dementia [50]. Each item is rated according to the level of performance being the absence of difficulty to perform the activity scored with 0 points and the inability to perform the activity is scored with 3 points. Activities that were never performed by the subject previously do not score as an impairment for this specific activity and total score. Higher scores in both scales indicate more difficulty to perform these activities.

Statistical analysis

Descriptive statistics were used to calculate mean and SD for continuous variables and frequencies for categorical variables. Pearson’s coefficient was used to assess correlation between quantitative variables. Bivariate analysis with t-test for continuous variables and Chi-square test for categorical variables were performed to compare baseline characteristics including gait parameters in MCR versus the No MCR participants. In the subsample that underwent fNIRS measurements, mean oxygenation values and dual-task protocol behavioral results in MCR versus No MCR participants were compared using t-test.

Performance in dual-task gait across different cognitive scores for RBANS, VF and TMT-A was assessed in MCR and No MCR groups separately by stratifying the sample based on categorically defined cognitive test scores. We aimed to separate the sample in low, mid, and high cognitive performance. Category 1 was defined as a score of less than 1 SD below the mean, category 2 was defined as a score between 1 SD below the mean and the mean, and category 3 was defined as a score above the mean. One way-ANOVAs were conducted with WWT velocity set as the dependent variable.

To explore the relationship of dual-task performance with cognitive function, we performed a multivariate linear regression with WWT velocity as the dependent variable. Independent variables included the cognitive test score and the MCR status, to assess the effect of the latter on this relationship. To account for a possible interaction between MCR status and cognitive function, the interaction term “MCR x Cognitive Test” was added in a second model. We performed separate models for each cognitive test. A model with RBANS was performed to examine the relationship with global cognition. To assess the relationship of WWT performance with executive function, we selected the executive function tests that showed the highest Pearson’s correlation coefficient with WWT velocity. All models were adjusted for age, education, and comorbidity index.

Analyses were performed on IBM SPSS Statistics for Windows, Version 25.0 (Armonk, NY: IBM Corp.).

RESULTS

A total of 538 participants were included in this study (mean age±SD = 76.6±6.5 years, 55% women). The prevalence of MCR was 11.2% (n = 60) in this CCMA cohort. Table 1 shows baseline characteristics of the sample. Global cognition and executive function tests showed scores in normal ranges. Participants were high-functioning with functional assessment scores of 0.8 ±1.2 for the ADLs scale, 1.8±2.4 for the IADL modified score, and gait speed of 98.3±22.4 cm/s. They had few comorbidities as shown by a comorbidity index of 1.6±1.1.

Table 1

Mean demographical and clinical characteristics of the global sample and comparison of baseline characteristics and dual-task performance between MCR and No-MCR groups. MCR, Motoric Cognitive Risk Syndrome; RBANS, Repeatable Battery for the Assessment of Neuropsychological Status; VF, Verbal Fluency; TMT, Trail Making Test; DSST, Digit Symbol Substitution Test; GDS, Geriatric Depression Scale; ADL, Activities of Daily Living; IADL, Instrumental Activities of Daily Living, NW, Normal Walking; WWT, Walk While Talking; DTC, Dual-Task Cost

	Global Sample (n = 538)	MCR (n = 60)	No MCR (n = 478)	p
Age, mean±SD (y)	76.6±6.48	78.19±7.33	76.40±6.34	0.043
Female, n (%)	296 (55)	33 (55)	263 (55.02)	0.9
Education, mean±SD (y)	14.57±2.95	13.67±3.05	14.68±2.92	0.012
Cognitive function
Cognitive status, n (%)	Normal 463 (86.1)	Normal 35 (58.3)	Normal 428 (89.5)	< 0.001
	MCI 75 (13.9)	MCI 25 (41.7)	MCI 50 (10.4)
RBANS Total Index, mean±SD	91.56±11.82	84.13±11.28	92.49±11.56	< 0.001
Phonemic VF, mean±SD (Z-score)	0.12±1.15	–0.18±1.06	0.16±1.59	0.03
Categorical VF, mean±SD (Z-score)	0.21±1.27	–0.54±1.19	0.30±1.24	< 0.001
TMT A, mean±SD (Z-score)	0.28±1.21	–0.31±1.68	0.36±1.12	< 0.001
TMT B, mean±SD (Z-score)	–0.05±1.19	–0.91±1.36	0.05±1.12	< 0.001
DSST, mean±SD (scaled score)	11.09±3.02	9.38±3.01	11.30±2.96	< 0.001
GDS score, mean±SD	4.68±3.93	6.28±4.57	4.47±3.80	0.001
Functional status
ADLs, mean±SD	0.81±1.19	1.81±1.47	0.69±1.1	< 0.001
IADL modified, mean±SD	1.84±2.44	3.33±3.15	1.65±2.27	< 0.001
Comorbidity index, mean±SD	1.64±1.1	1.75±1.16	1.63±1.08	0.4
Polypharmacy, n (%)	210 (39)	22 (36.7)	188 (39.3)	0.7
Dual-task performance
NW velocity, mean±SD (cm/s)	98.28±22.45	66.32±11.81	102.29±20.14	< 0.001
WWT velocity, mean±SD (cm/s)	69.36±23.94	45.79±15.66	72.32±23.16	< 0.001
DTC, mean±SD	–0.29±0.19	–0.30±0.22	–0.29±0.19	0.7
Alpha correct rate, mean±SD (letters/s)	0.23±0.07	0.21±0.07	0.24±0.07	0.002
WWT correct rate, mean±SD (letters/s)	0.92±0.37	0.72±0.33	0.95±0.37	< 0.001
DTC alpha correct rate, mean±SD	3.15±1.84	2.58±1.40	3.22±1.88	0.01

Compared to No MCR, participants with MCR were older, had fewer years of education, worse global cognitive function, higher prevalence of MCI, worse VF, TMT as well as higher scores in the Geriatric Depression Score (Table 1). Participants with MCR had more dependence in the ADLs and IADL scores. A similar level of comorbidities was reported in both groups.

Gait and dual-task performance

Compared to the No MCR group, MCR participants walked slower during NW and WWT conditions and showed a smaller decrement in gait velocity (WWT velocity – NW velocity) while dual-tasking compared to No MCR participants (–20.5±15.1 versus –29.9±19.9, p < 0.001). However, no difference was found in the DTC between both groups. MCR participants had a lower rate of correct letter generation than the No MCR group in both the cognitive single task (Alpha) as well as the WWT. The DTC of rate of correct letter was positive in both groups, meaning that the correct letter rate was higher during WWT, but was lower among MCR participants compared to No MCR participants even though MCR participants on average had longer recording time due to slower gait velocities (Table 1).

Dual-task performance and cognition

In an ANOVA with WWT velocity as dependent variable (Table 2), only in participants without MCR, WWT velocity showed significant differences across cognitive score categories (p = 0.007). To explore the specific differences between categories, post-hoc analyses showed a statistically significant difference in WWT velocity between those in lowest RBANS score category and those in the highest RBANS score category (p = 0.04), with higher WWT velocity in the highest RBANS category than in the lowest RBANS category. Significantly higher WWT velocity was found in the group with the highest categorical VF compared to the group with intermediate VF scores (p = 0.003) and in the group with the best TMT scores compared to those with the worst TMT performance (TMT-A p = 0.004 and TMT-B p < 0.001). Among MCR participants, no statistically significant differences were found across cognitive function categories. Similar pattern of differences was obtained when NW velocity and DTC were examined as the dependent variable (data not shown).

Table 2

Relationship between WWT velocity and cognitive function

	MCR	No MCR
	WWT velocity, mean±SD (cm/s)	WWT Velocity, mean±SD (cm/s)
RBANS
RBANS₁	45.5±13.5	67.9±18.3
RBANS₂	46.2±17.3	69.3±23.6
RBANS₃	45.8±17.7	75.4±23.64^*
Categorical VF
Cat VF₁	43.2±17.1	68.8±19.9
Cat VF₂	45.5±13.2	68.4±22.3
Cat VF₃	46.8±17.2	76.1±24.1^γ
TMT-A
TMT-A₁	34.9±16.8	64.8±21.6
TMT-A₂	46.3±15.4	67.4±20.7
TMT-A₃	47.5±15.2	75.6±23.8^*
TMT-B
TMT-B₁	43.1.4±10.3	64.8±20.5
TMT-B₂	42.5±12.8	71.9±22.1
TMT-B₃	47.5±18.1	77.2±24.4^*

One-way ANOVAs performed separately for MCR and No MCR groups. Sample was stratified by cognitive scores: category 1 was defined as a score of less than 1 SD below the mean, category 2 was defined as a score between 1 SD below the mean and the mean, and category 3 was defined as a score above the mean. Cognitive scores categories for MCR: RBANS₁ n = 10; RBANS₂ n = 21; RBANS₃ n = 29; Cat VF₁ n = 9; Cat VF₂ n = 21; Cat VF₃ n = 30; TMT-A₁ n = 7; TMT-A₂ n = 11, TMT-A₃ n = 42; TMT-B₁ n = 9; TMT-B₂ n = 16; TMT-B₃ n = 33.Cognitive scores categories for No MCR: RBANS₁ n = 77; RBANS₂ n = 171; RBANS₃ n = 230; Cat VF₁ n = 74; Cat VF₂ n = 164; Cat VF₃ n = 240; TMT-A₁ n = 55; TMT-A₂ n = 123, TMT-A₃ n = 299; TMT-B₁ n = 99; TMT-B₂ n = 179; TMT-B₃ n = 186. Post-hoc analyses: ^*indicates significant difference between category 1 and 3 at a p < 0.05 level; ^γindicates significant difference between category 2 and 3 at a p < 0.05 level. MCR, Motoric Cognitive Risk Syndrome; RBANS, Repeatable Battery for the Assessment of Neuropsychological Status; VF, Verbal Fluency; TMT, Trail Making Test; WWT, Walk While Talking.

A linear regression analysis showed that both MCR and RBANS score were independently associated with WWT velocity [F (5, 531) = 29.73, p < 0.001, with an R2 of 0.22 and beta coefficients of –21.2 (95% CI [–27.2, –15.3]) and 0.4 (95% CI [0.2, 0.5]) for MCR and RBANS]. However, the interaction of these two variables was not significant in the second model, suggesting that the presence of MCR does not affect the relationship between global cognition and WWT velocity. We selected categorical VF and TMT-A and TMT-B since they showed the highest bivariate linear correlation with WWT velocity. In separate models, significant regression equations with WWT velocity showed an association with categorical VF [F (5, 531) = 27.9, p < 0.001, with an R2 of 0.21, and beta coefficients of –21.9 (95% CI [–27.8, –15.9]) for MCR and 2.8 (95% CI [1.4, 4.3]) for VF], TMT-A [F (5, 530) = 27.8, p < 0.001, with an R2 of 0.21 and beta coefficients of –22.5 (95% CI [–28.4, –16.6]) for MCR and 2.8 (95% CI [1.3, 4.4]) for TMT-A] and TMT-B [F (5, 515) = 28.08, p < 0.001, with an R2 of 0.21 and beta coefficients of –21.8 (95% CI [–27.9, –15.8]) for MCR and 3.3 (95% CI [1.6, 4,9]) for TMT-B]. Again, the interaction terms (MCR x VF, MCR x TMT-A and MCR x TMT-B) were not significant. Sensitivity analysis stratifying by age and sex showed similar results.

fNIRS subsample

A subset of 325 participants underwent the fNIRS assessment. Baseline demographic, clinical, and functional characteristics of the group that did and did not complete the fNIRS assessment were similar. Prevalence of MCR in this subsample was 9.8% (n = 32). Mean Stride velocity during the fNIRS recordings was 80.1±17.3 cm/s during NW and 65.1±18.7 cm/s during WWT. Stride velocities obtained with the Zeno walkway during the fNIRS assessment showed moderate to high correlation with velocities obtained with the GAITRite walkway during NW and WWT (r = 0.83 p < 0.001 and r = 0.77 p < 0.001, respectively). The behavioral dual-task results during the fNIRS measurements showed similar results in the comparison between MCR and no MCR groups, with slower stride velocities during NW (58.81±15.05 vs 82.44±15.91, p < 0.001) and WWT (46.72±16.45 vs 67.07±17.81, p < 0.001).

Mean HbO₂ levels during WWT were higher in MCR compared to No MCR participants (Table 3). When analyzing both sides separately, this difference was only present over the left-sided channels, with higher levels of HbO₂ during WWT in the MCR compared to No MCR participants (1.1±1.2 versus 0.7±0.9, p = 0.017). There was no statistically significant difference between groups on the Alpha or the NW conditions.

Table 3

fNIRS oxygenation and gait parameters in the subsample (N = 325) during the dual-task protocol consisting of alphabet reciting (Alpha), single-task walk (Normal Walk) and walking while reciting the alphabet (WWT). MCR, Motoric Cognitive Risk Syndrome; HbO2, oxygenated Hemoglobin; NW, Normal Walking; WWT, Walk While Talking

	MCR (n = 32)	No MCR (n = 293)	p
HbO₂ levels
Alpha, mean±SD	0.67±0.79	0.69±0.51	0.8
Normal walk, mean±SD	0.26±0.86	0.09±0.63	0.17
WWT, mean±SD	1.02±1.25	0.66±0.83	0.03
Gait parameters
NW Stride Velocity, mean±SD (cm/s)	58.81±15.05	82.44±15.91	< 0.001
WWT Stride velocity, mean±SD (cm/s)	46.72±16.45	67.07±17.81	< 0.001

DISCUSSION

In our sample of 538 physically and cognitively well-functioning community-dwelling older adults, MCR prevalence was 11.2%, which is similar to MCR prevalence reported in other cohorts [4]. Our main findings show worse gait and cognitive performance during single and dual-tasks among MCR participants while the dual-task cost was not significantly different from the No MCR group. Dual-task gait performance among MCR participants was not related to global cognitive nor executive function performance. We found differential behavioral and oxygenation findings during dual-task performance in MCR participants compared to No MCR counterparts.

Gait velocity during the two walking conditions was slower in the MCR group compared to the No MCR. This was to be expected since MCR participants have slow gait by definition. A larger dual-task-related absolute decrease in gait velocity was observed in the No MCR group, while DTC interestingly was not different between MCR and non-MCR cases. This may suggest that MCR cases are able to compensate for the additional burdens of the dual task to the same extent as non-MCR participants. However, the cognitive output shows a different picture with worse accuracy rates seen in MCR. This raises the possibility that MCR cases are prioritizing the motor component over the cognitive component. Another interpretation of our findings could be that MCR participants prioritized maintaining gait safety while No MCR participants, with more intact gait patterns and cognitive resources, focused on both cognitive and motor components of the dual-task. Further examination of DTC during WWT performance is required to address these possibilities. Note that the WWT velocity might reflect stops as well as slowing due to increasing cognitive and physical demands, as postural control might be more difficult to maintain in such slow gait velocities. Alternatively, one might hypothesize a floor effect of gait velocity during single task in the MCR group.

This is the first study to our knowledge to assess dual-task performance in MCR, however a growing body of literature addresses this issue in other cognitive impairment states. One study compared the dual-task cost in older adults without cognitive complaints with MCI and AD patients [51]. Participants with cognitive impairment (MCI and AD) showed the highest DTC. However, the use of different dual-task paradigms limits the comparability with our results [52, 53]. Moreover, our study groups differ not only in cognitive but also in motor function, which may interfere in the dual-task performance.

In the current study, dual-task performance in MCR participants was not related to cognitive function. In the No MCR group, participants with better cognitive function (global cognition and executive function) showed faster WWT velocities. Dual-task performance has been associated with cognitive function, especially executive function [54, 55]. Worse executive function has been associated with slower dual-task gait velocity among cognitively healthy older adults [56] and MCI [57, 58]. Hausdorff et al. [15] reported no association of dual-task decrements in gait velocity with executive function among cognitively healthy persons; however, higher dual-task increase in swing time variability was seen in participants with worse executive function. Gait variability measures have been associated with both global cognition and executive function in AD participants [59]. Similar to our findings, Montero-Odasso et al. reported that global cognition measured with MMSE and MoCA tests was not associated with dual-task gait velocity in MCI patients [58]. According to these results, dual-task gait velocity might not be the best gait parameter to study this relationship in our sample. The smaller sample size of the MCR group in our study is a limitation and needs to be re-examined in larger samples. We also acknowledge that the different cut-off scores for the cognitive function categories might limit the comparability of the results obtained within each MCR-status group. Furthermore, although the interaction term MCR x cognitive scores was not statistically significant, an interaction may be plausible from a biological point of view and MCR could work as an effect modifier of the relationship between RBANS score and WWT.

The higher prevalence of MCI among MCR participants (compared to the No MCR group) should be taken into consideration and should be explored in larger sample sizes to account for MCI as a possible confounder. However, in our opinion cognitive (especially executive) dysfunction previously described in MCR even accounting for MCI, might be, at least partially, responsible for the dual-task performance in MCR.

Our findings suggest a different cerebral oxygenation pattern in the MCR group compared to the No MCR group. MCR participants show higher PFC oxygenation during WWT driven by higher HbO₂ levels detected on left-sided channels. The PFC is involved in the planning and control of movement [60] and is considered a key brain region for executive functions and attention [42 , 62]. Previous studies in non-demented older adults from the same cohort have shown PFC activation during gait that increases with task difficulty, i.e., higher PFC activation during WWT compared to NW [32]. Studies including participants with cognitive impairment, such as MCI, have also reported a higher PFC activation during a walking dual-task with VF compared to single task walk [63]. A functional magnetic resonance imaging (fMRI) study that assessed brain activation during imagined NW and WWT in CCMA participants found a covariance pattern of the fMRI signal related to task difficulty. Several brain regions, including PFC, showed more activation related to task difficulty (higher during WWT) [64]. In another study from the CCMA cohort, participants with peripheral neurological gait abnormalities (NGA) showed a higher increase in HbO₂ during WWT compared to participants with central neurological gait abnormalities or normal gaits [65]. Moreover, subjects with slower gait showed a higher increase in HbO₂ during walking with obstacles compared to unobstructed walk (relative to participants with normal gait) [66]. These findings suggest an increase in PFC activation in the presence of impaired gait. According to the neural inefficiency theory [22], the higher PFC activation may be understood as an attempt to maintain dual-task performance in the MCR group relative to controls. Recent neuroimaging studies in MCR subjects support this idea. In a MRI study that assessed gray matter volume covariance patterns associated with MCR [18], brain areas that showed relatively more atrophy in MCR participants included PFC, precentral, supplementary motor, and insular regions. This is in line with findings from previous research that reported smaller gray matter volumes specifically in premotor and PFC cortices [67]. These findings support the increased oxygenation of PFC during WWT in MCR as a neural compensatory mechanism due to the reduction in available resources in key brain regions for gait and attentional control processes.

Finally, the fact that the higher PFC oxygenation in MCR during WWT was driven by a left-side hyperactivation is consistent with recent literature that identifies left postero-lateral PFC areas as key dual-task neural substrate [68 , 70]. Moreover, in a CCMA-cohort study a higher PFC activation during WWT on the left-sided fNIRS channels was related to incident falls [21]. This might reflect an underlying neural inefficiency in the participants at higher fall risk that arises under challenging circumstances, consistent with the explanation provided above for the hyperactivation in MCR and supports the idea of the left-sided PFC being linked to dual-task performance.

This is the first study to address dual-task performance and neurophysiological correlates in MCR. Moreover, our study includes reliable and valid diagnostic procedures and gait assessment protocol as well as a comprehensive characterization of the sample. Potential limitations of our study include the cross-sectional design and relatively small MCR sample size. Also, the assessment of dual-task performance using only changes in gait velocity might underestimate dual-task effects on other gait variables, such as variability of different gait parameters. Finally, fNIRS does not allow assessment of brain areas other than PFC that have been linked to motor control [71, 60].

Further research is needed to explore dual-task performance in MCR with larger samples. Assessment with different dual-task paradigms and gait parameters might provide more information about the dual-task performance and its relationship with cognitive function. A better understanding of the neural correlates of dual-task performance in MCR using functional and multi-modal neuroimaging studies could shed light on the underlying physiopathology of MCR. Furthermore, longitudinal approaches could address the ability of dual-task-based assessments to predict outcomes such as dementia in MCR populations and interventional studies could help develop strategies to tackle disability associated to MCR, such as physical exercise-based strategies [72].

CONCLUSION

In summary, compared to No MCR, MCR participants showed a worse cognitive and motor performance during dual-task but with no significant differences in the dual-task cost while exhibiting higher PFC oxygenation during dual-task walking. The dual-task performance in MCR participants was not related to cognition. Further research addressing dual-task gait performance and its neural correlates in MCR is needed.

Footnotes

ACKNOWLEDGMENTS

Research reported in this publication was supported by the National Institute on Aging grants (1R01AG057548-01A1, R01AG044007-01A1, and R01 AG036921).

Authors’ disclosures available online ().

References

Verghese

, Wang

, Lipton

, Holtzer

(2013) Motoric cognitive risk syndrome and the risk of dementia. J Gerontol A Biol Sci Med Sci 68, 412–418.

Ayers

, Verghese

(2014) Diagnosing motoric cognitive risk syndrome to predict progression to dementia. Neurodegener Dis Manag 4, 339–342.

Doi

, Shimada

, Makizako

, Tsutsumimoto

, Verghese

, Suzuki

(2017) Motoric cognitive risk syndrome: Association with incidentdementia and disability. J Alzheimers Dis 59, 77–84.

Verghese

, Annweiler

, Ayers

, Barzilai

, Beauchet

, Bennett

, Bridenbaugh

, Buchman

, Callisaya

, Camicioli

, Capistrant

, Chatterji

, De Cock

, Ferrucci

, Giladi

, Guralnik

, Hausdorff

, Holtzer

, Woong Kim

, Kowal

, Kressig

, Lim

, Lord

, Meguro

, Montero-Odasso

, Muir-Hunter

, Noone

, Rochester

, Srikanth

, Wang

(2014) Motoric cognitive risksyndrome: Multicountry prevalence and dementia risk. Neurology 83, 718–726.

Callisaya

, Ayers

, Barzilai

, Ferrucci

, Guralnik

, Lipton

, Otahal

, Srikanth

, Verghese

(2016) Motoric cognitive risksyndrome and falls risk: A multi-center study. J AlzheimersDis 53, 1043–1052.

Ayers

, Verghese

(2016) Motoric cognitive risk syndrome and risk of mortality in older adults. Alzheimers Dement 12, 556–564.

Sekhon

, Allali

, Launay

, Chabot

, Beauchet

(2017) The spectrum of pre-dementia stages: Cognitive profile of motoric cognitive risk syndrome and relationship with mild cognitive impairment. Eur J Neurol 24, 1047–1054.

Ayers

, Verghese

(2019) Gait dysfunction in motoric cognitive risk syndrome. J Alzheimers Dis 71, S95–S103.

Verghese

, Wang

, Bennett

, Lipton

, Katz

, Ayers

(2019) Motoric cognitive risk syndrome and predictors of transition to dementia: A multicenter study. Alzheimers Dement 15, 870–877.

10.

Allali

, Ayers

, Verghese

(2016) Motoric cognitive risk syndrome subtypes and cognitive profiles. J Gerontol A Biol Sci Med Sci 71, 378–384.

11.

Allali

, Kressig

, Assal

, Herrmann

, Dubost

, Beauchet

(2007) Changes in gait while backward counting in demented older adults with frontal lobe dysfunction. Gait Posture 26, 572–576.

12.

Bloem

, Grimbergen

YAM

, van Dijk

, Munneke

(2006) The ‘posture second’ strategy: A review of wrong priorities in Parkinson’s disease. J Neurol Sci 248, 196–204.

13.

Vervoort

, Heremans

, Bengevoord

, Strouwen

, Nackaerts

, Vandenberghe

, Nieuwboer

(2016) Dual-task-related neural connectivity changes in patients with Parkinson’ disease. Neuroscience 317, 36–46.

14.

Ble

, Volpato

, Zuliani

, Guralnik

, Bandinelli

, Lauretani

, Bartali

, Maraldi

, Fellin

, Ferrucci

(2005) Executive function correlates with walking speed in older persons: The InCHIANTI study. J Am Geriatr Soc 53, 410–415.

15.

Hausdorff

, Schweiger

, Herman

, Yogev-Seligmann

, Giladi

(2008) Dual-task decrements in gait: Contributing factors amonghealthy older adults. J Gerontol A Biol Sci Med Sci 63, 1335–1343.

16.

Doi

, Blumen

, Verghese

, Shimada

, Makizako

, Tsutsumimoto

, Hotta

, Nakakubo

, Suzuki

(2017) Gray matter volume and dual-task gait performance in mild cognitive impairment. Brain Imaging Behav 11, 887–898.

17.

Collette

, Olivier

, Van der Linden

, Laureys

, Delfiore

, Luxen

, Salmon

(2005) Involvement of both prefrontal and inferior parietal cortex in dual-task performance. Brain Res Cogn Brain Res 24, 237–251.

18.

Blumen

, Allali

, Beauchet

, Lipton

, Verghese

(2019) A graymatter volume covariance network associated with the motoriccognitive risk syndrome: A multicohort MRI study. J Gerontol ABiol Sci Med Sci 74, 884–889.

19.

Delpy

, Cope

(1997) Quantification in tissue near–infrared spectroscopy. Philos Trans R Soc London B Biol Sci 352, 649–659.

20.

Durduran

, Choe

, Baker

, Yodh

(2010) Diffuse optics for tissue monitoring and tomography. Rep Prog Phys 73, 076701.

21.

Verghese

, Wang

, Ayers

, Izzetoglu

, Holtzer

(2017) Brain activation in high-functioning older adults and falls: Prospective cohort study. Neurology 88, 191–197.

22.

Stern

(2009) Cognitive reserve. Neuropsychologia 47, 2015–2028.

23.

Holtzer

, Wang C Verghese

(2014) Performance variance on walking while talking tasks: Theory, findings, and clinical implications. Age 36, 373–381.

24.

Verghese

, Lipton

, Hall

, Kuslansky

, Katz

, Buschke

(2002) Abnormality of gait as a predictor of non-Alzheimer’s dementia. N Eng J Med 347, 1761–1768.

25.

American Psychiatric Association (2000) Diagnostic and statistical manual of mental disorders, 4th ed., Text Revision, Washington, DC.

26.

Holtzer

, Verghese

, Wang

, Hall

, Lipton

(2008) Within-person across-neuropsychological test variability and incident dementia. JAMA 300, 823–830.

27.

Yesavage

, Brink

, Rose

, Lum

, Huang

, Adey

, Leirer

(1983) Development and validation of a geriatric depressionscreening scale: A preliminary report. J Psychiatr Res 17, 37–49.

28.

Galvin

, Roe

, Powlishta

, Coats

, Muich

, Grant

, Miller

, Storandt

, Morris

(2005) The AD8: A brief informant interview to detect dementia. Neurology 65, 559–564.

29.

Morris

(1993) The Clinical Dementia Rating (CDR): Current version and scoring rules. Neurology 43, 2412–2414.

30.

Verghese

, Kuslansky

, Holtzer

, Katz

, Xue

, Buschke

, Pahor

(2007) Walking while talking: Effect of task prioritization in the elderly. Arch Phys Med Rehabil 88, 50–53.

31.

Rosso

, Metti

, Faulkner

, Redfern

, Yaffe

, Launer

, Shaaban

, Nadkarni

, Rosano

(2019) Complex walking tasks and risk for cognitive decline in high functioning older adults. J Alzheimers Dis 71, S65–S73.

32.

Holtzer

, Mahoney

, Izzetoglu

, Wang

, England

, Verghese

(2015) Online fronto-cortical control of simple and attention-demanding locomotion in humans. Neuroimage 112, 152–159.

33.

Ayaz

, Izzetoglu

, Platek

, Bunce

, Izzetoglu

, Pourrezaei

, Onaral

(2006) Registering fNIR data to brain surface image using MRI templates. Conf Proc IEEE Eng Med Biol Soc 2006, 2671–2674.

34.

Izzetoglu

, Chitrapu

, Bunce

, Onaral

(2010) Motion artifact cancellation in NIR spectroscopy using discrete Kalman filtering. Biomed Eng Online 9, 16.

35.

Frosting

(2002) In-vivo optical imaging of brain function. CRC PRess, Irvine, CA.

36.

Harada

, Miyai

, Suzuki

, Kubota

(2009) Gait capacity affects cortical activation patterns related to speed control in the elderly. Ex Brain Res 193, 445–454.

37.

Holtzer

, Mahoney

, Izzetoglu

, Onaral

, Verghese

(2011) fNIRS study of walking and walking while talking in young and old individuals. J Gerontol A Biol Sci Med Sci 66, 879–887.

38.

Randolph

, Tierney

, Mohr

, Chase

(1998) The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS): Preliminary clinical validity. J Clin Exp Neuropsychol 20, 310–319.

39.

Benton

(1976) Multilingual aphasia examination, University of Iowa, Iowa.

40.

Patterson

(2011) Verbal Fluency, Encyclopedia of Clinical Neuropsychology, Springer New York, 2603–2606.

41.

Reitan

(1958) Validity of the Trail Making Test as an indicator of organic brain damage. Percept Mot Skills 8, 271–276.

42.

Lezak

(2012), Neuropsychological assessment, Oxford University Press, New York.

43.

Sánchez-Cubillo

, Periáñez

, Adrover-Roig

, Rodríguez-Sánchez

, Ríos-Lago

, Tirapu

, Barceló

(2009) Construct validity of the Trail Making Test: Role of task-switching, working memory, inhibition/interference control, and visuomotor abilities. J Int Neuropsychol Soc 15, 438–450.

44.

Wechsler

(1997) Wechsler Adult Intelligence Scale, The Psychological Corporation, New York.

45.

Petersen

(2011) Clinical practice. Mild cognitive impairment. N Engl J Med 364, 2227–2234.

46.

Verghese

, Wang

, Lipton

, Holtzer

, Xue

(2007) Quantitative gait dysfunction and risk of cognitive decline and dementia. J Neurol Neurosurg Psychiatry 78, 929–935.

47.

Verghese

, Holtzer

, Lipton

, Wang

(2009) Quantitative gait markers and incident fall risk in older adults. J Gerontol A Biol Sci Med Sci 64, 896–901.

48.

Gill

, Allore

, Holford

, Guo

(2004) Hospitalization, restricted activity, and the development of disability among older persons. J Am Med Assoc 292, 2115–2124.

49.

Ayers

, Shapiro

, Holtzer

, Barzilai

, Milman

, Verghese

(2017) Symptoms of apathy independently predict incident frailty and disability in community-dwelling older adults. J Clin Psychiatry 78, e529–e536.

50.

Galasko

, Bennett

, Sano

, Marson

, Kaye

, Edland

(2006) ADCS Prevention Instrument Project: Assessment of instrumental activities of daily living for community-dwelling elderly individuals in dementia prevention clinical trials. Alzheimer Dis Assoc Disord 20, S152–169.

51.

Muir

, Speechley

, Wells

, Borrie

, Gopaul

, Montero-Odasso

(2012) Gait assessment in mild cognitive impairment and Alzheimer’s disease: The effect of dual-task challenges across the cognitive spectrum. Gait Posture 35, 96–100.

52.

Bishnoi

, Hernandez

(2021) Dual task walking costs in older adults with mild cognitive impairment: A systematic review and meta-analysis. Aging Ment Health 25, 1618–1629.

53.

Al-Yahya

, Dawes

, Smith

, Dennis

, Howells

, Cockburn

(2011) Cognitive motor interference while walking: A systematic review and meta-analysis. Neurosci Biobehav Rev 35, 715–728.

54.

Yogev-Seligmann

, Hausdorff

, Giladi

(2008) The role of executive function and attention in gait. Mov Disord 23, 329–342.

55.

Holtzer

, Verghese

, Xue

, Lipton

(2006) Cognitive processes related to gait velocity: Results from the Einstein aging study. Neuropsychology 20, 215–223.

56.

Coppin

, Shumway-Cook

, Saczynski

, Patel

, Ble

, Ferrucci

, Guralnik

(2006) Association of executive function and performance of dual-task physical tests among older adults: Analyses from the InChianti study. Age Ageing 35, 619–624.

57.

Doi

, Shimada

, Makizako

, Tsutsumimoto

, Uemura

, Anan

, Suzuki

(2014) Cognitive function and gait speed under normal and dual-task walking among older adults with mild cognitive impairment. BMC Neurol 14, 67.

58.

Montero-Odasso

, Bergman

, Phillips

, Wong

, Sourial

, Chertkow

(2009) Dual-tasking and gait in people with mild cognitive impairment. the effect of working memory. BMC Geriatr 9, 41.

59.

Sheridan

, Solomont

, Kowall

, Hausdorff

(2003) Influence of executive function on locomotor function: Divided attention increases gait variability in Alzheimer’s disease. J Am Geriatr Soc 51, 1633–1637.

60.

Mirelman

, Shema

, Maidan

, Hausdorff

(2018) Gait. In Handbook of Clinical Neurology, 159, pp. 119–134.

61.

Reuter-Lorenz

(2000) Cognitive neuropsychology of the aging brain. In Cognitive aging: A primer, Park

, Schwarz

, eds. Psychology Press, New York, pp. 93–114.

62.

Alvarez

, Emory

(2006) Executive function and the frontal lobes: A meta-analytic review. Neuropsychol Rev 16, 17–42.

63.

Doi

, Makizako

, Shimada

, Park

, Tsutsumimoto

, Uemura

, Suzuki

(2013) Brain activation during dual-task walking and executive function among older adults with mild cognitive impairment: A fNIRS study. Aging Clin Exp Res 25, 539–544.

64.

Blumen

, Holtzer

, Brown

, Gazes

, Verghese

(2014) Behavioral and neural correlates of imagined walking and walking-while-talking in the elderly. Hum Brain Mapp 35, 4090–4104.

65.

Holtzer

, Verghese

, Allali

, Izzetoglu

, Wang

, Mahoney

(2016) Neurological gait abnormalities moderate the functional brain signature of the posture first hypothesis. Brain Topogr 29, 334–343.

66.

Chen

, Pillemer

, England

, Izzetoglu

, Mahoney

, Holtzer

(2017) Neural correlates of obstacle negotiation in older adults: An fNIRS study. Gait Posture 58, 130–135.

67.

Beauchet

, Allali

, Annweiler

, Verghese

(2016) Association of motoric cognitive risk syndrome with brain volumes: Results from the GAIT Study. J Gerontol A Biol Sci Med Sci 71, 1081–1088.

68.

Dux

, Ivanoff

, Asplund

, Marois

(2006) Isolation of a central bottleneck of information processing with time-resolved fMRI. Neuron 52, 1109–1120.

69.

Filmer

, Mattingley

, Dux

(2013) Improved multitasking following prefrontal tDCS. Cortex 49, 2845–2852.

70.

Yuan

, Blumen

, Verghese

, Holtzer

(2015) Functional connectivity associated with gait velocity during walking and walking-while-talking in aging: A resting-state fMRI study. Hum Brain Mapp 36, 1484–1493.

71.

Takakusaki

(2017) Functional neuroanatomy for posture and gait control. J Mov Disord 10, 1–17.

72.

Schirinzi

, Canevelli

, Suppa

, Bologna

, Marsili

(2020) The continuum between neurodegeneration, brain plasticity, and movement: A critical appraisal. Rev. Neurosci 31, 723–742.