Cortically-mediated muscle responses to balance perturbations increase with perturbation magnitude in older adults with and without Parkinson’s disease

Cortically-mediated muscle responses to balance perturbations increase with perturbation magnitude in older adults with and without Parkinson’s disease | bioRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-M677548'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search New Results Cortically-mediated muscle responses to balance perturbations increase with perturbation magnitude in older adults with and without Parkinson’s disease Scott E. Boebinger , Aiden M. Payne , Jifei Xiao , Giovanni Martino , Michael Borich , View ORCID Profile J. Lucas McKay , View ORCID Profile Lena H. Ting doi: https://doi.org/10.1101/2024.12.09.627582 Scott E. Boebinger 1 Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology & Emory University , Atlanta, Georgia , United States of America Find this author on Google Scholar Find this author on PubMed Search for this author on this site Aiden M. Payne 2 Department of Psychology, Florida State University , Tallahassee, Florida, United States of America Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jifei Xiao 1 Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology & Emory University , Atlanta, Georgia , United States of America Find this author on Google Scholar Find this author on PubMed Search for this author on this site Giovanni Martino 3 Department of Biomedical Sciences, University of Padova , Padua, Italy Find this author on Google Scholar Find this author on PubMed Search for this author on this site Michael Borich 1 Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology & Emory University , Atlanta, Georgia , United States of America 4 Division of Physical Therapy, Department of Rehabilitation Medicine, Emory University , Atlanta, Georgia , United States of America Find this author on Google Scholar Find this author on PubMed Search for this author on this site J. Lucas McKay 1 Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology & Emory University , Atlanta, Georgia , United States of America 5 Departments of Biomedical Informatics and Neurology, Emory University , Atlanta, Georgia , United States of America Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for J. Lucas McKay Lena H. Ting 1 Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology & Emory University , Atlanta, Georgia , United States of America 4 Division of Physical Therapy, Department of Rehabilitation Medicine, Emory University , Atlanta, Georgia , United States of America Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Lena H. Ting For correspondence: lting{at}emory.edu Abstract Full Text Info/History Metrics Preview PDF ABSTRACT We lack a clear understanding of how cortical contributions to balance are altered in aging and Parkinson’s disease (PD), which limits development of rehabilitation strategies. Processes like balance control are typically mediated through brainstem circuits, with higher-order circuits becoming engaged as needed. Using reactive balance recovery, we investigated how hierarchical neural mechanisms shape balance- correcting muscle activity across task difficulty in older adults (OAs) with and without PD. We hypothesize that feedback loops involving brainstem and cortical circuits contribute to balance control, and cortical engagement increases with challenge, aging, and PD. We decomposed perturbation-evoked agonist and antagonist muscle activity into hierarchical components based on latency using neuromechanical models consisting of two feedback loops with different delays to reflect different neural conduction and processing times. Agonist muscle activity was decomposed into two components that both increased with balance challenge in both groups. The first component occurred ∼120ms and the second occurred ∼210ms, consistent with the latencies of brainstem and transcortical circuits, respectively. Exploratory comparisons to young adults revealed larger transcortical components in OA and PD groups at lower balance challenge levels, consistent with increased cortical involvement with aging. Antagonist muscle activity included destabilizing and stabilizing components, with the destabilizing component correlating to balance ability in OAs but not in PD. These findings demonstrate that neuromechanical models can identify changes in the hierarchical control of balance without direct brain measurements. Identifying cortical contributions during balance control may complement clinical measures of balance ability to inform balance rehabilitation and assistive devices. NEW & NOTEWORTHY A neuromechanical model can decompose perturbation evoked muscle activity into components attributed to brainstem and higher-order sensorimotor feedback based on latency. In older adults with and without PD contributions from higher-order neural circuits increase with balance challenge and are related to clinical measures of balance ability. Interpreting the neural substrates of motor output for balance may reveal individual differences in the hierarchical control of balance, that could inform rehabilitation. INTRODUCTION Balance impairments are prevalent in older adults (OAs) with and without Parkinson’s disease (PD) and are associated with less automatic, brainstem-mediated control of balance ( 1 – 7 ), but contributions from higher-order circuits to motor output during balance control remain poorly understood. To maintain standing balance, multisensory information encoding body motion is processed by the central nervous system to generate balance-correcting motor commands to muscles throughout the body ( 8 ). Balance control progressively shifts from being primarily brainstem-mediated ( 9 ) to engaging higher order centers such as the basal ganglia and cortex as balance health declines, such as in aging and PD ( 1 , 10 – 15 ), or as balance challenge increases ( 16 , 17 ). Cortical engagement during balance control is typically inferred through measures of cognitive dual task interference ( 1 , 10 , 18 – 20 ) or prefrontal cortex oxygen metabolism ( 13 , 14 , 21 , 22 ), but these techniques lack the temporal resolution to mechanistically assess the effects on motor output. Reactive balance recovery is a robust paradigm to investigate shifts in sensorimotor control of balance as task difficulty can be adjusted by varying perturbation size. Furthermore, discrete nature of the perturbation and temporal resolution of electromyography (EMG) allows for evoked muscle activity to be attributed to different hierarchical circuits based on latency. Here we investigate shifts in hierarchical motor control by examining muscle activity during reactive balance control using a neuromechanical model to dissociate the contributions of subcortical and cortical components based on feedback delays. Following external disturbances to either the upper or lower limb, the agonist muscle – i.e. the muscle initially stretched by the perturbation – is activated in a stereotypical sequence of short- and long- latency responses (SLR and LLR, respectively) ( 23 – 27 ). LLRs and are mediated by higher order centers such as the brainstem and cortex, and appear after spinally-mediated SLRs due to a longer neural transmission and processing time ( 25 – 29 ). During balance control, the nervous system maintains the body’s center of mass (CoM) in equilibrium above the base of support ( 30 – 32 ). Following a perturbation to standing, LLRs in the agonist or “prime mover” muscles are patterned based on delayed sensory feedback encoding CoM motion ( 28 , 29 , 33 – 35 ). The magnitude and time course of balance correcting agonist muscle activity can be reconstructed based on a small set of feedback parameters that describe the nervous system’s sensitivity to deviations in CoM acceleration, velocity, and displacement (k a , k v , k d ) from the desired, upright stance. These scaled CoM kinematic error signals are delayed by a common time delay (λ) to account for neural transmission and processing ( 28 , 29 , 33 – 35 ). Prior neuromechanical models used a single sensorimotor feedback loop reflecting brainstem-mediated sensorimotor transforms for the agonist muscle ( 28 , 29 , 33 – 35 ). Recently, we showed balance-correcting muscle activity may reflect feedback components based on multiple sensorimotor transformations from CoM kinematic errors to muscle activity at delays consistent with the involvement of transcortical circuits ( 15 , 36 ). Using a neuromechanical model with parallel sensorimotor feedback loops acting at different delays, we decomposed agonist muscle into components termed LLR1 and LLR2 in young adults (YAs) ( 36 ). The LLR1 and LLR2 arise at latencies consistent with brainstem and cortical circuits, respectively ( 36 ). In YA, the earlier LLR1 occurred ∼100ms post-perturbation and was engaged at all levels of task difficulty, while the longer-latency LLR2 occurred ∼200ms and became progressively engaged as task difficulty increased ( 36 ). We further corroborated that the latency of the LLR2 muscle activity was about twice that of perturbation-evoked cortical activity recorded via electroencephalography ( 36 ). Therefore, perturbation-evoked muscle activity may be dissociated into hierarchical components without need for direct brain measurements. However, the dual-loop neuromechanical model has not yet been applied to agonist muscle activity from individuals requiring greater cortical contributions to balance, nor applied to antagonist muscle activity that may also arise from multiple hierarchical neural pathways. Effective balance recovery requires directionally-specific muscle activation patterns that appropriately generate torque to counteract CoM deviations from an upright stance ( 37 , 38 ). Evoked activity in the antagonist muscle – i.e. the muscle initially shortened by the perturbation – is commonly observed in OAs and individuals with PD, creating destabilizing co-contraction that opposes agonist balance-correcting muscle activity ( 15 , 39 – 45 ). Antagonist co-contraction increases joint stiffness but reduces net joint torque as the agonist and antagonist muscles resist one another ( 46 ). We recently used a neuromechanical model to identify a component of destabilizing antagonist co-contraction activation that was associated with fall history in OAs with and without PD ( 15 ). This antagonist co-contraction occurred at latencies ∼ 180 ms but its relative timing with the cortically-mediated LLR2 activity in the agonist muscle has not been investigated. Here we dissociated hierarchically mediated agonist and antagonist muscle activity in OAs and individuals with PD during control of reactive balance. We hypothesize that parallel sensorimotor feedback loops engaging brainstem and higher-order circuitry contribute to reactive balance control at different latencies and that involvement of higher-order circuits increases with challenge, aging, and PD, increasing recruitment of longer-latency agonist muscle activity and co-contraction of the antagonist muscle ( Figure 1 ). We predicted that longer-latency agonist muscle activity would increase with perturbation magnitude and occur at latencies consistent with transcortical mediation in both OAs with and without PD. We further predicted that individuals with PD would have greater longer-latency agonist muscle activity than OAs. Finally, we predicted that the destabilizing antagonist co-contraction and the LLR2 in the agonist muscle are driven by higher- order neural substrates and will therefore occur after the LLR1. Download figure Open in new tab Figure 1: Schematic of hierarchical sensorimotor feedback loops involved in reactive balance control. A) At low levels of balance challenge, balance control is primarily mediated through brainstem sensorimotor circuits (green) which generate a stereotypical EMG waveform. B) At higher levels of balance challenge or as balance health declines, higher order circuits (blue) begin to contribute to balance-correcting muscle activity at longer latencies leading to alterations in the stereotypical waveform. C) In Parkinson’s disease (PD) higher order circuits are further engaged during balance control and leads to abnormal destabilizing antagonist activity (red). MATERIALS AND METHODS Ethics Statement All experiments were approved by the Emory University Institutional Review Board and all participants gave informed written consent before participating in this experiment. Participants Nineteen older adults and twenty individuals with PD were recruited for this study. Three participants with PD were excluded from this analysis. Two were excluded due to either a brain tumor or severe peripheral neuropathy of the legs noted in their clinical record. The other opted to leave the experiment prior to balance perturbations. After exclusions, nineteen older adults (6 female, 71 ± 6 years old, 175 ± 2 cm tall, 79 ± 4 kg) and seventeen individuals with Parkinson’s disease (4 female, 69 ± 2 years old, 171 ± 3 cm tall, 84 ± 6 kg) were included in analysis. Participants were excluded prior to participating if they reported having a history of lower extremity joint pain, contractures, major sensory deficits, evidence of orthopedic, muscular, or physical disability, evidence of vestibular, auditory, or proprioceptive impairment, orthostatic hypotension, and/or any neurological insult during recruitment. Participants were recruited from the community surrounding Emory University and Emory Movement Disorders clinic through outreach events, word of mouth, flyers, and databases of prior participants from collaborating groups. Other outcome measures from this group have been reported previously ( 47 – 49 ). OFF-medications Individuals with PD were asked to forgo their dopaminergic medications for PD a minimum of 12 hours before participating in this study. Each participant’s neurologist was consulted and signed a clearance form prior to participants withholding their medications for this study. Clinical and behavioral measures were collected during this OFF-medication session. Balance Ability Participant balance ability was determined via the mini Balance Evaluation Systems Test (miniBEST) which assesses sensory orientation, dynamic gait, anticipatory postural control and reactive postural control ( 50 – 52 ). For items that scored the left and right side separately, only the lower of the two scores was considered which results in a maximum score of 28 ( 51 ), where higher scores indicate better balance ability. Parkinson’s disease motor symptom severity For individuals with PD, we examined clinical measures of disease severity to determine if these measures could explain the variance in the hierarchical components identified by our neuromechanical model. The severity of motor impairment in participants with PD was assessed via the motor subscale of the International Parkinson and Movement Disorder Society’s Unified Parkinson’s Disease Rating Scale (MDS UPDRS-III). This test was administered by A.M.P., a certified member of the Movement Disorder Society. Assessments were filmed and sent to a practicing neurologist for scoring. Parkinson’s disease duration The number of years since PD diagnosis was self-reported by participants with PD at the time of the study and verified from their clinical records when possible. Balance perturbations As previously described ( 47 – 49 ), OAs with and without PD stood barefoot on a motorized platform (Factory Automation Systems, Atlanta, GA, USA) and underwent a series of 48 translational support- surface perturbations that were delivered at unpredictable timing, direction, and magnitude. Each participant received an equal number of perturbations in the forward and backward direction at three magnitudes: a small perturbation (5.1 cm, 11.1 cm/s, 0.15 g), which was identical across participants, and two larger magnitudes (medium: 7-7.4cm, 15.2-16.1 cm/s, and 0.21-0.22g, and large: 8.9-9.8cm, 19.1- 21.0cm/s, and 0.26-0.29g). Medium and large perturbations were adjusted based on participant height to ensure perturbations were mechanically similar across participants ( 47 – 49 ) ( Figure 2 ). To limit predictability, perturbation magnitude and direction were presented in pseudorandom block order, with each of the eight blocks containing one perturbation of each magnitude and direction. Three different block-randomized orders were used across participants to randomize any effect of trial order ( 47 – 49 ). Trials in which a participant took a step were excluded from our analyses because stepping leaves the position of the base of support temporarily undefined, thereby precluding the calculation of kinematic errors between the CoM and the base of support which are required for our neuromechanical models. Successful non-stepping trials were identified using platform-mounted force plates (AMTI OR6-6). If there were under 3 trials averaged for a certain participant-condition pair it was removed from analysis. To limit fatigue, 5-minute breaks were given every 15 minutes of experimentation. As an additional, exploratory analysis we compared the neuromechanical model outputs from the OA and PD groups to those from a YA group that has been previously published ( 36 ). This YA group underwent a similar support surface perturbation paradigm ( 36 , 53 – 55 ), however there are some key differences to note. Firstly, the YA group underwent larger magnitude support-surface perturbations than the OA and PD groups. Each YA participant received an equal number of perturbations at three magnitudes: a medium perturbation (7.7 cm, 16.0 cm/s, 0.23 g), which was identical across participants, and two larger magnitudes (large: 12.6– 15.0 cm, 26.6–31.5 cm/s, 0.38–0.45 g, and extra-large: 18.4–21.9 cm, 38.7–42.3 cm/s, 0.54–0.64 g), which were adjusted based on participant height ( 36 , 54 , 55 ). Secondly, the YA group were only perturbed in the backwards direction. To allow for larger perturbations that could sufficiently challenge balance in YAs, the platform was initialized to one end of its range. Due to limitations that the platform can travel, the YA group was only able to undergo perturbations of a single direction while the OA and PD group underwent perturbations in both the backward and forward direction. Due to these differences in experimental paradigm, we only include the YA group as part of an exploratory analysis. Download figure Open in new tab Figure 2: Experimental paradigm. Translational support-surface perturbations were delivered at unpredictable timing, direction, and magnitude. Perturbation kinematics, muscle activity, and center of mass kinematics were recorded throughout the recovery of balance. Brown shaded timeseries represent forward support-surface translations, gray shaded time series represent backward support-surface translations. Darker shading indicates larger perturbation magnitudes. Center of Mass (CoM) kinematics Kinematic marker data were collected at 100 Hz and synchronized using a ten camera Vicon Nexus 3D motion analysis system (Vicon, Centennial, CO). Body segment kinematics were determined from a custom marker set covering head–arms–trunk, thigh, shank, and foot segments. CoM displacement was derived from kinematic data as a weighted sum of segmental masses. CoM velocity was taken as the derivative of CoM displacement after smoothing using a third-order Savtizky-Golay filter with a filter size of 48 samples ( 35 ) ( Figure 2 ). CoM acceleration was computed from ground reaction forces obtained by the platform-mounted force plates (AMTI OR6-6), divided by participant mass ( Figure 2 ). Electromyography (EMG) Surface EMGs (Motion Analysis Systems, Baton Rouge, LA) were collected bilaterally from the tibialis anterior (TA) and medial gastrocnemius muscle (MG) muscles ( Figure 2 ). Analysis focused on the TA and MG since this agonist-antagonist muscle pair are activated in forward and backward support surface perturbations. Prior to EMG electrode placement, skin was shaved if necessary and scrubbed with an isopropyl alcohol wipe. EMG electrodes were placed using standard procedures ( 56 ). Bipolar silver silver-chloride electrodes were used (Norotrode 20, Myotronics, Inc., Kent, WA, USA). Electromyography signals were sampled at 1000 Hz and anti-alias filtered with an online 500 Hz low-pass filter. Raw EMG signals were then high-pass filtered at 35Hz offline with a sixth-order zero-lag Butterworth filter, mean- subtracted, half-wave rectified, and subsequently low-pass filtered at 40Hz ( 15 , 34 , 55 ). EMG signals were epoched between -200ms and 1200ms relative to perturbation onset. Single-trial EMG data were normalized to a maximum value of 1 across all trials within each participant for left and right sides independently. EMG data were then averaged across trials within each perturbation magnitude for each participant. Neuromechanical models Perturbation evoked agonist and antagonist muscle activity were reconstructed using a series of delayed feedback models to investigate the relationship between sensory information and muscle activity. All neuromechanical models employed here use kinematic signals of balance error, defined as CoM displacement (d), velocity (v), and acceleration (a) relative to the base of support, as predictors to reconstruct perturbation evoke muscle activity ( 15 , 36 ). Agonist neuromechanical model The agonist neuromechanical model uses two feedback loops with different loop delays to account for the inherently different latencies between brainstem mediated and higher order mediated muscle activity as previously validated in YAs ( 36 ) ( Figure 3 ). This model is used to decompose perturbation evoked activity in the agonist muscle, defined as the muscle initially stretched by the perturbation, into components attributed to different neural substrates based on latency. Download figure Open in new tab Figure 3: Agonist muscle activity was decomposed into different hierarchical components based on latency . A) Agonist muscle activity, defined as the muscle initially lengthened by the perturbation (Medial Gastrocnemius for the backward perturbation shown here), was decomposed into shorter-latency (green) or longer-latency order (blue) components. Agonist muscle activity was reconstructed as the weighted sum of positive CoM kinematics (those that stretch the muscle) that was delayed by two separate time delays reflecting brainstem (λ 1 ) or higher order (λ 2 ) circuits. B) Group summary of the integrated output of the shorter-latency and longer-latency feedback loops for forward and backward perturbations. The shorter-latency feedback loop multiplies the CoM predictors that stretch the agonist muscle by their respective feedback gains (k d1 , k v1 , k a1 ). These weighted signals are then summed, and delayed by a time delay (λ 1 ) to account for ascending and descending neural transmission and processing ( Equation 1 ) The output from the shorter-latency feedback loop ( emg 1 (t) ) was half wave rectified with a threshold value of 0 to represent the overall non-negative net output to motor pools ( Equation 2 ) ( 15 , 36 ). The longer latency feedback loop uses the same predictors that stretch the agonist muscle as the brainstem feedback loop but weights these kinematic signals by independent feedback gains (k d2 , k v2 , k a2 ) and delayed by a longer common time delay (λ 2 ) to account for the longer latency required for ascending and descending neural transmission and processing in higher-order circuits ( Equation 3 ). The output from the longer latency feedback loop ( emg 2 (t) ) was also half wave rectified with a threshold value of 0 to represent the overall net output to motor pools ( Equation 2 ) ( 15 , 36 ). The output from both feedback loops were then summed linearly in line with previous modeling approaches of similar behavior in the upper limb ( 26 ) ( Equation 4 ). Antagonist neuromechanical model The antagonist neuromechanical model also uses two feedback loops with different delays to reconstruct antagonist muscle activity and was developed in a prior study on a separate PD group ( 15 ) ( Figure 4 ). This model was used to decompose perturbation evoked antagonist muscle activity into stabilizing and destabilizing components. Download figure Open in new tab Figure 4: Antagonist muscle activity was decomposed into stabilizing and destabilizing components. A) Antagonist muscle activity, defined as the muscle initially shortened by the perturbation, was decomposed into hierarchical components based off latency using the antagonist neuromechanical model. Destabilizing antagonist muscle activity (red) was reconstructed as the weighted sum of positive CoM kinematics (those that shorten the antagonist muscle) that was delayed to account for neural transmission and processing. Stabilizing antagonist activity (green) was reconstructed as the weighted sum of negative CoM kinematics (those that stretch the antagonist muscle) that was delayed by a separate delay to account for neural transmission and processing. B) Group summary of the integrated output of the stabilizing and destabilizing feedback loops for forward and backward perturbations. The stabilizing feedback loop multiplies CoM predictors that stretch the antagonist muscle by their respective feedback gains (k dS , k vS , k aS ). These weighted signals are then summed, and delayed by a common time delay (λ s ) to accounts for ascending and descending neural transmission and processing ( Equation 5 ) The output from the stabilizing feedback loop ( emg stabilzing (t) ) is half wave rectified with a threshold value of 0 ( Equation 2 ). The destabilizing feedback loop uses CoM predictors that shorten the antagonist muscle (-a(t), - v(t), and -d(t)) and multiplies these kinematic signals by separate feedback gains (k dD , k vD , k aD ). These weighted signals are them summed and delayed by a common time delay (λ D ) to account for ascending and descending neural transmission and processing ( Equation 6 ). The output from both the destabilizing ( emg destabilizing (t) ) feedback loops was also half wave rectified with a threshold value of 0 to represent the overall net output to motor pools ( Equation 2 ) ( 15 ). The output from both feedback loops were then summed linearly similar to the reconstructed agonist activity ( Equation 7 ). Model parameter identification Parameters for both the agonist and antagonist neuromechanical models were selected by minimizing the error between recorded EMG data and the model reconstruction. The reconstruction error was quantified as the sum of the sum squared error of the whole time series as well as the maximum error observed at any sample ( Equation 8 ). The first term penalizes squared error (e 2 ) between the recorded data and the model prediction weighted by μ s . The second term penalizes the maximum error observed between the recorded data and the model reconstruction weighted by μ m . The final term penalizes the magnitude of the gain parameters (k i ) with weight μ k in order to minimize feedback parameters that do not contribute to accurate reconstructions. The ratio of weights μ s :μ m :μ k was 1:1:1e-6. All optimizations were performed in Matlab 2022b (Mathworks, Natick, MA, USA) using the interior-point algorithm in fmincon.m , as described previously in literature ( 15 , 36 ). Separate optimizations were performed to identify model-specific feedback parameters for the individual feedback loops ( Equations 1, 3, 5, and 6 ). The output from these optimizations are then used as an initial guess when optimizing for double-looped models ( Equations 4 and 7 ) similar to what was done in prior implementations of this model ( 15 , 36 ). This allows for modifications to the single-loop optimization’s output, as model parameters may have been overestimated in order to fit muscle activity that is better explained by the other feedback loop. Lower and upper bounds for the gain parameters were ±10% of the initial guess values, lower and upper bounds for the delay parameters were ±10ms of the initial guess values. In all cases, additional parameters supplied to fmincon.m were as follows: TolX , 1e-9; MaxFunEvals , 1e5; TolFun , 1e-7, with remaining parameters set to default ( 15 , 36 ). Goodness of fit The goodness of fit for all neuromechanical model reconstructions were assessed using a coefficient of determination (R 2 ) and variability accounted for (VAF). R 2 was calculated using regress.m , a built-in function in Matlab 2022b (Mathworks, Natick, MA, USA). VAF was defined as 100*the square of Pearson’s uncentered correlation coefficient, as performed in previous studies ( 15 , 36 , 57 ). Model selection Model selection was performed using Akaike’s Information Criterion (AIC) ( Equation 9 ). N Data is the number of data points reconstructed by the neuromechanical models (1400 samples), SSR is the sum squared error of the neuromechanical model reconstruction, and N Parameters is the number of parameters included in the neuromechanical model (8 parameters). AIC values were calculated for both the agonist and antagonist neuromechanical models described above. An AIC comparison was performed to determine whether the addition of a longer-latency loop improves the model fit to the data in the agonist neuromechanical model. Differences in AIC between models ≥ 2 were considered meaningful. Statistical characterization of neuromechanical model outputs The integrated area under the curve of each neuromechanical model component ( Equations 1, 3, 5, and 6 ) was calculated via numerical integration using trapz.m , a built-in function in Matlab 2022b (Mathworks, Natick, MA, USA). Statistical tests were performed in RStudio version 1.4.1717 (R Core Team, Vienna, Austria). Comparisons of integrated components between perturbation magnitude and group were performed using a linear mixed effects model with an interaction between the fixed factors of perturbation magnitude and group with participant as a random factor. Post hoc comparisons were performed via comparisons of the estimated marginal means between perturbation magnitudes or group. Correlations between integrated components and clinical measures (miniBEST and MDS UPDRS-III scores) were performed using a linear mixed effects models with interaction between the fixed factors of clinical score and perturbation magnitude with participant as a random factor. Separate linear mixed models were applied for the OA and PD groups. Tests are considered statistically significant at p ≤ 0.05. RESULTS Agonist sensorimotor response to perturbations Perturbation evoked agonist muscle activity exhibited an LLR1 followed by an LLR2 occurring around 100 and 200 ms latency, respectively, in both OA and PD groups. Both the LLR1 and LLR2 could be explained by sensorimotor transformations of CoM kinematics via our double looped neuromechanical model ( Figure 3A ). The shorter-latency feedback loop ( Figure 3A , green trace) explaining the initial LLR1 of agonist muscle activity occurred at 119ms ± 1.6ms for OAs and 116ms ± 1.3ms for individuals with PD, with no differences between groups in forward (F(1, 34) = 1.87, p = 0.18) or backward perturbations (F(1, 34) = 0.02, p = 0.89). The longer-latency feedback loop ( Figure 3A , blue trace) reconstructing the LLR2 of agonist muscle activity occurred at 200ms ± 3.2ms in OAs and 215ms ± 3.9ms in individuals with PD, with a significant differences between groups in backward perturbations (Forward: F(1, 33.9) = 0.46, p = 0.50; Backward: F(1, 34) = 4.17, p = 0.049). Goodness of fit of agonist neuromechanical model reconstructions increased by ∼10% when using a double versus single feedback loop model in both groups ( Table 1 ) . This led the double loop model to have higher reconstruction accuracies compared to the single loop model in both perturbation directions (Forward: R 2 : t(34.5) = 11.2, p < 0.0001; VAF: t(34.7) = 11.0, p < 0.0001; Backward: (R 2 : t( 35 ) = 6.6, p < 0.0001; VAF: t( 35 ) = 6.3, p < 0.0001). View this table: View inline View popup Download powerpoint Table 1: Reconstruction accuracies for single loop and double loop agonist neuromechanical model. Contributions of both the LLR1 and LLR2 increased with perturbation magnitude in both OA and PD groups, with no difference between groups ( Figure 3B ). In both perturbation directions, there was an increase with perturbation magnitude in the LLR1 (Forward: F(2, 65.8) = 36.24, p < 0.001; Backward: F(2, 68) = 25.98, p < 0.001) and LLR2 (Forward: F(2, 64.7) = 37.6, p < 0.001; Backward: F(2, 68) = 16.97, p < 0.001). There was no difference between OA and PD groups in either the LLR1 (Forward: F(1, 34.6) = 0.37, p = 0.55; Backward: F(1, 34) = 0.31, p = 0.58) or LLR2 (Forward: F(1, 34.6) = 0.08, p = 0.78; Backward: F(1, 34) = 0.12, p = 0.73) ( Figure 3B ). There were no interactions between perturbation magnitude and group for either the LLR1 (Forward: F(2, 65.8) = 1.36, p = 0.26; Backward: F(2, 68) = 0.22, p = 0.81) or LLR2 (Forward: F(2, 64.7) = 0.06, p = 0.94; Backward: F(2, 68) = 0.34, p = 0.71). Antagonist sensorimotor response to perturbations Perturbation-evoked antagonist muscle activity could also be explained by sensorimotor transformations of CoM kinematics via a double looped neuromechanical model ( Figure 4A ). Antagonist muscle activity was exhibited destabilizing co-contraction at the beginning of the ramp and hold perturbation that could be explained by sensorimotor transformations of CoM kinematics in the destabilizing feedback loop ( Figure 4A , red trace). This co-contraction was then followed by a stabilizing response to platform deceleration that could also be explained by the sensorimotor transformation of CoM kinematics in the stabilizing feedback loop ( Figure 4A , green trace). Contributions of the destabilizing feedback loop to antagonist muscles ( Figure 4A red trace) increased with perturbation magnitude in both OA and PD groups (Forward: F(2, 63.8) = 10.15, p < 0.001; Backward: F(2, 68) = 44.12, p < 0.001) ( Figure 4B ). There were no differences between the OA and PD groups in either perturbation direction (Forward: F(1,34.1) = 0.30, p = 0.59; Backward: F(1, 34) = 2.29, p = 0.14) ( Figure 4B ). There were no interactions between perturbation magnitude and group (Forward: F(2, 6.38) = 1.59, p = 0.21; Backward: F(2, 68) = 0.39, p = 0.68). Individuals with PD exhibited a larger acceleration driven component of destabilizing muscle activity compared to OAs. We further decomposed the output from the destabilizing feedback loop to antagonist muscles ( Figure 4A red trace) into its component parts attributed to the CoM acceleration feedback component (k aD ) and the CoM velocity and displacement feedback component (k vD +k dD ). Individuals with PD had larger k aD component compared to OAs in forward perturbations (Forward: t( 34 ) = -2.05, p = 0.048; Backward: t( 34 ) = 1.174, p = 0.25). There was no difference between OA and PD groups in the contribution of the k vD +k dD component in either perturbation direction (Forward: (t( 34 ) = 0.86, p = 0.40; Backward (t(34) = 1.37, p = 0.18). Contributions from the stabilizing feedback loop to antagonist muscles ( Figure 4A green trace) that occurred at the end of the ramp-and hold perturbations increased with perturbation magnitude, but only the backward perturbation direction (Forward: F(1, 63.9) = 0.15, p = 0.86; Backward: F(2, 68) = 3.98, p = 0.023) ( Figure 4B ). There was no effect of group on the output from the stabilizing feedback loop (Forward: F(1, 33.9) = 0.95, p = 0.34; Backward: F(1, 34) = 2.20, p = 0.15) ( Figure 4B ). There were no interactions between perturbation magnitude and group (Forward: F(2, 63.9) = 1.41, p = 0.25; Backward: F(2, 68) = 0.21, p = 0.81). Destabilizing antagonist co-contraction occurred at latencies consistent with higher-order mediation The onset of the antagonist destabilizing feedback loop occurred at an intermediate latency between the onset of the agonist LLR1 and LLR2. The antagonist destabilizing co-contraction had an onset latency (λ D ) of 180 ± 3.6ms in OAs and 173 ± 2.9ms for individuals with PD. There was no difference in onset latency between OA and PD groups (Forward: t(34) = 1.76, p = 0.09; Backward: t(34) = -0.80, p = 0.43). The onset of the antagonist destabilizing co-contraction occurred at 119 ± 1.6ms in OAs and 116 ± 1.3ms in people with PD, after the initial burst of agonist muscle activity (λ 1 ), but prior to the second burst of agonist muscle activity (λ 2 ) at 200 ± 3.2ms in OAs and 215 ± 3.9ms in people with PD. Correlations to clinical measures The output from the destabilizing feedback loop was negatively correlated with miniBEST scores in the OA group for both perturbation directions (Forward: R 2 = 0.69, p = 0.043; Backward: R 2 = 0.70, p = 0.045). Individuals with PD showed an opposite trend, as there was a positive relationship between the output from the antagonist destabilizing feedback loop and higher miniBEST scores; this trend was only present in the forward perturbations and did not reach significance (Forward: R 2 = 0.48, p = 0.083; Backward: R 2 = 0.40, p = 0.90). There was a negative trend between the contribution from the longer-latency feedback loop and miniBEST in OA, but not PD. However, this trend did not reach significance in OAs (Forward: R 2 = 0.63, p = 0.088; Backward: R 2 = 0.33, p = 0.25) or individuals with PD (Forward: R 2 = 0.53, p = 0.55; Backward R 2 = 0.24, p = 0.91). None of the assessments of PD motor symptom severity (MDS UPDRS-III score, PD phenotype, PIGD-subscore, Hoehn and Yahr state) nor PD duration could explain the variance in any of the components identified by our neuromechanical models ( Table 2 ). View this table: View inline View popup Table 2: Statistical comparisons of neuromechanical model outputs and Parkinson’s disease specific clinical measures. DISCUSSION Overall, this work demonstrates that a neuromechanical model can decompose perturbation evoked muscle activity into hierarchical components attributed to different neural substrates based on latency that are relevant to clinical measures of balance ability. Our data support the hypothesis that parallel sensorimotor feedback loops engaging hierarchical circuits contribute to perturbation evoked muscle activity during reactive balance recovery in OAs with and without PD. These results provide experimental and computational evidence that longer-latency agonist muscle activity associated with the LLR2 can be explained by sensory information encoding CoM kinematics at delays consistent with transcortical feedback loops in OAs with and without PD. We further show that cortically-mediated responses increase with balance challenge in OAs with and without PD. Furthermore, in an exploratory comparison we show that the LLR2 is recruited at smaller perturbation magnitudes in OAs than YAs ( 36 ), regardless of PD status ( Figure 5 ). Our findings are in consistent with the compensation-related utilization of neural circuits hypothesis (CRUNCH) ( 58 , 59 ), as OAs have increased activation of cortical regions compared to YAs, to compensate for declining efficiency of processing in subcortical circuits. We also show that perturbation-evoked destabilizing antagonist activity occurs at latencies consistent with higher-order mediation. Both OAs and people with PD exhibited similar amounts of destabilizing antagonist muscle activity that occurred at similar onset latencies. The magnitude of this destabilizing component was correlated with an individual’s balance ability in OAs, but not individuals with PD. This implies that neuromechanical modeling of reactive balance control may provide a useful tool to index individual differences in the hierarchical control of balance, which may help clinicians optimize future rehabilitation training ( 60 , 61 ). Download figure Open in new tab Figure 5: Exploratory comparison to previously published data in young adults (YAs). We compared the outputs from the shorter- and longer-latency feedback loops of the agonist neuromechanical model ( Figure 3A ). Comparisons between groups are qualitative due to differences in the experimental paradigms between the YAs and the OAs with and without PD. Our neuromechanical model could provide a more mechanistic assessment of cortical engagement during balance control. Our current understanding of cortical contributions to balance in aging and PD is typically either inferred through dual task interference analysis ( 1 , 10 , 18 – 20 ) or measures of cortical oxygen metabolism ( 13 , 14 , 21 , 22 ). For dual task interference assessments, cortical engagement is typically inferred when primary motor task, i.e. standing balance, is performed concurrently with a secondary cognitive task, i.e. mental math ( 10 , 18 , 62 ). Decreased balance performance when the secondary task is performed concurrently to primary task are thought to arise because the secondary task diverts cortical resources, defined as the brain’s task-general information processing capacity that is shareable across concurrent tasks in a graded manner ( 63 – 65 ), away from primary task thereby impacting performance. Cortical engagement has also been assessed through optical measurements of cortical oxygen metabolism using functional near-infrared spectroscopy (fNIRS) ( 13 , 14 , 21 , 22 ). However, fNIRS is often only used to assess prefrontal oxygen metabolism as hair on the rest of the scalp prevents this technique from being used on areas over other neural structures ( 66 – 70 ). Both techniques lack temporal resolution due to inherent delay between neurovascular coupling in the case of fNIRS or between muscle activity and movement in the case of behavioral assessments. Therefore, how these findings will be reflected in muscle activity was unable to be determined. Assessments of perturbation evoked muscle activity like those uses here potentially offer a more robust way to characterize cortical contributions to balance control as muscle activity has millisecond-level temporal resolution which is necessary due to the discrete nature of balance correcting responses. Our results are consistent with prior work that demonstrates increased cortical engagement during balance with aging and impairment ( 1 , 10 , 11 , 13 , 14 , 18 – 20 ). Older adults exhibit increased activation of cortical areas relative to younger adults performing similar tasks ( 58 , 59 ). Furthermore, individuals with impaired balance such as in PD exhibit further increases in cortical activity relative to unimpaired OAs ( 1 , 10 , 13 , 14 , 18 ). In an exploratory analysis, we compared the reconstructions of agonist muscle activity from OAs with and without PD to previously published reconstructions from YAs ( 36 ) ( Figure 5 ). OAs, regardless of PD, exhibit a component of agonist muscle activity that could be explained by CoM kinematics at a delay consistent with transcortical feedback loops, even at the smallest perturbation magnitude (5.1 cm displacement), whereas for young adults this muscle activity is largely absent at a larger magnitude (7.7cm displacement) ( Figure 5 ). There are some notable differences between the perturbation paradigm used here and those used previously in the YA group which prevent direct comparisons. Namely, the YA paradigm only had perturbations in a single direction while the OA and PD paradigm had perturbations in both directions. Therefore, the YA group would have been more able to adapt their motor response to the perturbation paradigm, which has been shown to alter our neuromechanical model’s parameters ( 34 ), since adaptation could lead to balance correcting responses to be more automatic and thereby require less cortical input. Contrary to our hypothesis that individuals with PD will have increased involvement of higher- order circuits during balance control, we found that OAs with PD had similar amounts of longer-latency muscle activity as OAs without PD. Perturbation evoked responses progressively engage higher order centers at increasing latency ( 11 , 27 ). The initial burst of agonist muscle activity (LLR1) occurred at latencies around 120ms, consistent with prior iterations of this neuromechanical model ( 15 , 28 , 29 , 34 , 36 ). Muscle activity at this latency are considered to be mediated by brainstem sensorimotor circuits ( 11 ) as evidenced by decerebrate animals being able to maintain balance and exhibit intact, perturbation-specific patterns of muscle activation when exposed to balance perturbations ( 9 ). Additionally, in humans, perturbation evoked cortical activity occurs simultaneously with the LLR1, and therefore cannot contribute to this aspect of muscle activity ( 36 , 55 , 71 ). A second burst of agonist muscle activity (LLR2) occurred at a latencies around 200ms, consistent with the latencies found in prior iterations of this model ( 36 ). Muscle activity at this longer latency has the potential to be mediated by higher-order, presumably transcortical circuits ( 11 , 25 ), as this activity can be altered by transcranial magnetic stimulation, while shorter latency muscle activity cannot ( 25 , 72 ). Additionally, the latency of the LLR2 was about twice that of perturbation-evoked cortical activity recorded via electroencephalography in YAs ( 36 ), making it possible for evoked cortical activity to contribute to this aspect of balance correcting muscle activity. Our findings in reconstructions of antagonist muscle activity corroborate previous findings showing that antagonist co-contraction can be explained by sensorimotor transformations of CoM kinematic errors at delays consistent with input from higher-order neural circuits and is related to balance ability ( 15 ). The onset of the destabilizing muscle activity occurred around 180ms in both the OA and PD groups, similar to previous implementations of this model ( 15 ). Interestingly, this destabilizing co- contraction occurs after the shorter-latency, presumably brainstem-mediated, agonist muscle activity. The initial burst of agonist muscle activity is thought to be governed by reticulospinal circuits ( 33 ) and occurred ∼120ms for both groups. The onset of the destabilizing antagonist muscle activity was substantially longer (∼60ms on average), which is sufficiently long for basal ganglia involvement ( 15 ). Additionally, the onset of destabilizing antagonist co-contraction occurred prior to the onset of the longer- latency agonist muscle activity identified by our neuromechanical model. This could suggest that these two components of perturbation evoked muscle activity are mediated by different neural circuits. However, we can only speculate on the neuroanatomical substrates involved in generating motor output since we lack neuroimaging techniques and rely on the latency of muscle activity. It is important to note that while we find similar results, we only found this phenomenon in the forward perturbations, whereas the previous publication was focused on backward perturbations. This discrepancy could result from differences in the experimental paradigm between these studies. Here examined 3 perturbation magnitudes while previously only a single magnitude was used. Furthermore, we are examining data from seventeen individuals with PD whereas the previous study examined data collected from forty-four individuals with PD. The differences in experimental paradigm and number of participants may explain discrepancies between what we show here and what was previously published. Recordings of muscle activity during reactive balance recovery coupled with our neuromechanical model could offer a means to index shifts in the hierarchical control of balance, which would be useful for tracking changes in balance control throughout rehabilitation interventions and assessing the efficacy of balance-assisting devices. Reactive balance paradigms could be used to quantify the LLR1 and LLR2 components and thereby provide a metric of cortical contributions to balance control without the need for measures of brain activity. Changes in these hierarchical components could then be tracked throughout disease progression and rehabilitation interventions at different levels of balance challenge. Additionally, the effect of assistive devices on an individual’s control of balance could be inferred by the degree to which the balance control while using a device is cortically-mediated. Furthermore, our neuromechanical modeling technique can be extended to reconstruct joint torques, which would circumvent the need to record muscle activity ( 73 ). DATA AVAILABILITY All data, model code, and statistical analysis code can be found online at: https://osf.io/56u43/?view_only=d82dd6b36d5f4313ada65f4ec6203d6c GRANTS Lena H. Ting, Wallace H. Coulter Department of Biomedical Engineering, R01 AG072756 (to L.H.T). Lena H. Ting, Wallace H. Coulter Department of Biomedical Engineering, R01 HD46922 (to L.H.T). Scott E. Boebinger, Wallace H. Coulter Department of Biomedical Engineering, National Science Foundation Graduate Research Fellowship Program Grant No. 1937971 (to S.E.B). DISCLOSURES The authors have declared that no competing interests exist DISCLAIMERS The authors have no disclaimers AUTHOR CONTRIBUTIONS L.H.T and A.M.P Conceived and designed research, A.M.P performed experiments, S.E.B and J.X analyzed data, S.E.B, G.M, A.M.P, J.L.M, M.R.B, and L.H.T interpreted results of experiments, S.E.B prepared figures, S.E.B drafted manuscript, S.E.B, J.X, G.M, A.M.P, J.L.M, M.R.B, and L.H.T edited and revised manuscript, S.E.B, J.X, G.M, A.M.P, J.L.M, M.R.B, and L.H.T approved final version of manuscript. REFERENCES 1. ↵ Brown LA , Shumway-Cook A , Woollacott MH . Attentional demands and postural recovery: the effects of aging . J Gerontol A Biol Sci Med Sci . 1999 Apr ; 54 ( 4 ): M165 – 171 . OpenUrl CrossRef PubMed Web of Science 2. Allcock LM , Rowan EN , Steen IN , Wesnes K , Kenny RA , Burn DJ . Impaired attention predicts falling in Parkinson’s disease . Parkinsonism & Related Disorders . 2009 Feb 1; 15 ( 2 ): 110 – 5 . OpenUrl CrossRef PubMed 3. Camicioli R , Majumdar SR . Relationship between mild cognitive impairment and falls in older people with and without Parkinson’s disease: 1-Year Prospective Cohort Study . Gait Posture. 2010/05/04 ed . 2010 May; 32 ( 1 ): 87 – 91 . OpenUrl CrossRef 4. Schoneburg B , Mancini M , Horak F , Nutt JG . Framework for Understanding Balance Dysfunction in Parkinson’s Disease . Mov Disord . 2013 Sep 15; 28 ( 11 ): 1474 – 82 . OpenUrl CrossRef PubMed 5. Park JH , Kang YJ , Horak FB . What Is Wrong with Balance in Parkinson’s Disease? J Mov Disord . 2015 Sep ; 8 ( 3 ): 109 – 14 . OpenUrl CrossRef PubMed 6. Li KZH , Bherer L , Mirelman A , Maidan I , Hausdorff JM. Cognitive Involvement in Balance, Gait and Dual-Tasking in Aging: A Focused Review From a Neuroscience of Aging Perspective. Front Neurol [Internet]. 2018 Oct 29 [cited 2024 Sep 13];9. Available from: https://www.frontiersin.org/journals/neurology/articles/10.3389/fneur.2018.00913/full 7. ↵ Clark DJ , Manini TM , Ferris DP , Hass CJ , Brumback BA , Cruz-Almeida Y , et al. Multimodal Imaging of Brain Activity to Investigate Walking and Mobility Decline in Older Adults (Mind in Motion Study): Hypothesis, Theory, and Methods. Front Aging Neurosci [Internet] . 2020 [cited 2020 Sep 2];11. Available from: https://www.frontiersin.org/articles/10.3389/fnagi.2019.00358/full 8. ↵ Mack S , Kandel ER , Jessell TM , Schwartz JH , Siegelbaum SA , Hudspeth AJ . Principles of Neural Science , Fifth Edition. McGraw Hill Professional ; 2013 . 1761 p. 9. ↵ Honeycutt CF , Gottschall JS , Nichols TR . Electromyographic Responses From the Hindlimb Muscles of the Decerebrate Cat to Horizontal Support Surface Perturbations . Journal of Neurophysiology . 2009 Jun ; 101 ( 6 ): 2751 – 61 . OpenUrl CrossRef PubMed Web of Science 10. ↵ Shumway-Cook A , Woollacott M , Kerns KA , Baldwin M . The Effects of Two Types of Cognitive Tasks on Postural Stability in Older Adults With and Without a History of Falls . J Gerontol A Biol Sci Med Sci . 1997 Jul 1; 52A ( 4 ): M232 – 40 . OpenUrl CrossRef PubMed Web of Science 11. ↵ Jacobs JV , Horak FB . Cortical control of postural responses . Journal of Neural Transmission . 2007 Mar 29; 114 ( 10 ): 1339 . OpenUrl CrossRef PubMed Web of Science 12. Rochester L , Galna B , Lord S , Burn D . The nature of dual-task interference during gait in incident Parkinson’s disease . Neuroscience . 2014 Apr 18; 265 : 83 – 94 . OpenUrl CrossRef PubMed 13. ↵ Stuart S , Vitorio R , Morris R , Martini DN , Fino PC , Mancini M . Cortical activity during walking and balance tasks in older adults and Parkinson’s disease: a structured review . Maturitas . 2018 Jul ; 113 : 53 – 72 . OpenUrl CrossRef PubMed 14. ↵ Stuart S , Belluscio V , Quinn JF , Mancini M . 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 . 15. ↵ McKay JL , Lang KC , Bong SM , Hackney ME , Factor SA , Ting LH . Abnormal center of mass feedback responses during balance: A potential biomarker of falls in Parkinson’s disease . PLoS One . 2021 ; 16 ( 5 ): e0252119 . OpenUrl CrossRef PubMed 16. ↵ Gebel A , Lehmann T , Granacher U . Balance task difficulty affects postural sway and cortical activity in healthy adolescents . Exp Brain Res . 2020 May 1; 238 ( 5 ): 1323 – 33 . OpenUrl CrossRef PubMed 17. ↵ Mirelman A , Maidan I , Bernad-Elazari H , Shustack S , Giladi N , Hausdorff JM . Effects of aging on prefrontal brain activation during challenging walking conditions . Brain and Cognition . 2017 Jul 1; 115 : 41 – 6 . OpenUrl CrossRef PubMed 18. ↵ Woollacott M , Shumway-Cook A . Attention and the control of posture and gait: a review of an emerging area of research . Gait Posture . 2002 Aug ; 16 ( 1 ): 1 – 14 . OpenUrl CrossRef PubMed Web of Science 19. Bayot M , Dujardin K , Tard C , Defebvre L , Bonnet CT , Allart E , et al. The interaction between cognition and motor control: A theoretical framework for dual-task interference effects on posture, gait initiation, gait and turning . Neurophysiologie Clinique . 2018 Dec 1; 48 ( 6 ): 361 – 75 . OpenUrl CrossRef PubMed 20. ↵ Kahya M , Moon S , Ranchet M , Vukas RR , Lyons KE , Pahwa R , et al. Brain activity during dual task gait and balance in aging and age-related neurodegenerative conditions: A systematic review . Exp Gerontol . 2019 Dec ; 128 : 110756 . 21. ↵ Karim H , Schmidt B , Dart D , Beluk N , Huppert T . Functional near-infrared spectroscopy (fNIRS) of brain function during active balancing using a video game system . Gait & Posture . 2012 Mar 1; 35 ( 3 ): 367 – 72 . OpenUrl CrossRef PubMed Web of Science 22. ↵ Pelicioni PHS , Tijsma M , Lord SR , Menant J . Prefrontal cortical activation measured by fNIRS during walking: effects of age, disease and secondary task . PeerJ . 2019 ; 7 : e6833 . OpenUrl CrossRef PubMed 23. ↵ Tatton WG , Lee RG . Evidence for abnormal long-loop reflexes in rigid Parkinsonian patients . Brain Research . 1975 Dec 26; 100 ( 3 ): 671 – 6 . OpenUrl CrossRef PubMed Web of Science 24. Rothwell J , Obeso J , Traub M , Marsden C . The behaviour of the long-latency stretch reflex in patients with Parkinson’s disease. Journal of Neurology , Neurosurgery & Psychiatry . 1983 Jan 1; 46 ( 1 ): 35 – 44 . OpenUrl CrossRef 25. ↵ Taube W , Schubert M , Gruber M , Beck S , Faist M , Gollhofer A . Direct corticospinal pathways contribute to neuromuscular control of perturbed stance . J Appl Physiol . 2006 ; 101 ( 2 ): 420 – 9 . OpenUrl CrossRef PubMed Web of Science 26. ↵ Pruszynski JA , Kurtzer I , Scott SH . The long-latency reflex is composed of at least two functionally independent processes . J Neurophysiol . 2011 Jul ; 106 ( 1 ): 449 – 59 . OpenUrl CrossRef PubMed Web of Science 27. ↵ Reschechtko S , Pruszynski JA . Stretch reflexes . Current Biology . 2020 Sep 21; 30 ( 18 ): R1025 – 30 . OpenUrl CrossRef PubMed 28. ↵ Welch TDJ , Ting LH . A Feedback Model Reproduces Muscle Activity During Human Postural Responses to Support-Surface Translations . Journal of Neurophysiology . 2008 Feb 1; 99 ( 2 ): 1032 – 8 . OpenUrl CrossRef PubMed Web of Science 29. ↵ Welch TDJ , Ting LH . A Feedback Model Explains the Differential Scaling of Human Postural Responses to Perturbation Acceleration and Velocity . Journal of Neurophysiology . 2009 Jun 1; 101 ( 6 ): 3294 – 309 . OpenUrl CrossRef PubMed Web of Science 30. ↵ Horak FB . Adaptation of automatic postural responses . The acquisition of motor behavior in vertebrates . 1996 ; 57 – 85 . 31. Horak FB . Postural orientation and equilibrium: what do we need to know about neural control of balance to prevent falls? Age and Ageing . 2006 Sep 1; 35 (suppl_2):ii7–11. 32. ↵ Horak FB , Macpherson JM . Postural Orientation and Equilibrium. In: Comprehensive Physiology [Internet] . American Cancer Society ; 2011 [cited 2021 Feb 22]. p. 255 – 92 . Available from: http://onlinelibrary.wiley.com/doi/abs/10.1002/cphy.cp120107 33. ↵ Lockhart DB , Ting LH . Optimal sensorimotor transformations for balance . Nature Neuroscience . 2007 Oct ; 10 ( 10 ): 1329 – 36 . OpenUrl CrossRef PubMed Web of Science 34. ↵ Welch TDJ , Ting LH . Mechanisms of Motor Adaptation in Reactive Balance Control . PLOS ONE . 2014 May 8; 9 ( 5 ): e96440 . OpenUrl CrossRef PubMed 35. ↵ Safavynia SA , Ting LH . Long-latency muscle activity reflects continuous, delayed sensorimotor feedback of task-level and not joint-level error . Journal of Neurophysiology . 2013 Sep 15; 110 ( 6 ): 1278 – 90 . OpenUrl CrossRef PubMed 36. ↵ Boebinger S , Payne A , Martino G , Kerr K , Mirdamadi J , McKay JL , et al. Precise cortical contributions to sensorimotor feedback control during reactive balance . PLOS Computational Biology . 2024 Apr 17; 20 ( 4 ): e1011562 . OpenUrl CrossRef 37. ↵ Henry SM , Fung J , Horak FB . EMG Responses to Maintain Stance During Multidirectional Surface Translations . Journal of Neurophysiology . 1998 Oct ; 80 ( 4 ): 1939 – 50 . OpenUrl CrossRef PubMed Web of Science 38. ↵ Torres-Oviedo G , Ting LH . Muscle Synergies Characterizing Human Postural Responses . Journal of Neurophysiology . 2007 Oct ; 98 ( 4 ): 2144 – 56 . OpenUrl CrossRef PubMed Web of Science 39. ↵ Andrews CJ , Burke D , Lance JW . The response to muscle stretch and shortening in Parkinsonian rigidity . Brain . 1972 ; 95 ( 4 ): 795 – 812 . OpenUrl CrossRef PubMed Web of Science 40. Andrews CJ , Neilson PD , Lance JW . Comparison of stretch reflexes and shortening reactions in activated normal subjects with those in Parkinson’s disease . J Neurol Neurosurg Psychiatry . 1973 Jun ; 36 ( 3 ): 329 – 33 . OpenUrl Abstract / FREE Full Text 41. Horak FB , Nutt JG , Nashner LM . Postural inflexibility in parkinsonian subjects . Journal of the Neurological Sciences . 1992 Aug 1; 111 ( 1 ): 46 – 58 . OpenUrl CrossRef PubMed Web of Science 42. Xia R , Rymer WZ . The role of shortening reaction in mediating rigidity in Parkinson’s disease. Exp Brain Res. 2004 Jun; 156 ( 4 ): 524 – 8 . OpenUrl CrossRef PubMed Web of Science 43. Xia R , Powell D , Rymer WZ , Hanson N , Fang X , Threlkeld AJ . Differentiation between the contributions of shortening reaction and stretch-induced inhibition to rigidity in Parkinson’s disease . Exp Brain Res . 2011 Apr 1; 209 ( 4 ): 609 – 18 . OpenUrl CrossRef PubMed 44. Lang KC , Hackney ME , Ting LH , McKay JL . Antagonist muscle activity during reactive balance responses is elevated in Parkinson’s disease and in balance impairment . PLOS ONE . 2019 Jan 25; 14 ( 1 ): e0211137 . OpenUrl CrossRef PubMed 45. ↵ Rubega M , Di Marco R , Zampini M , Formaggio E , Menegatti E , Bonato P , et al. Muscular and cortical activation during dynamic and static balance in the elderly: A scoping review . Aging Brain . 2021 Jan 1; 1 : 100013 . 46. ↵ Damiano DL . Reviewing Muscle Cocontraction . Phys Occup Ther Pediatr . 1993 ; 12 ( 4 ): 3 – 20 . OpenUrl CrossRef PubMed 47. ↵ Payne AM , Palmer JA , McKay JL , Ting LH . Lower Cognitive Set Shifting Ability Is Associated With Stiffer Balance Recovery Behavior and Larger Perturbation-Evoked Cortical Responses in Older Adults . Frontiers in Aging Neuroscience . 2021 ; 13 : 860 . 48. Payne AM , McKay JL , Ting LH . The cortical N1 response to balance perturbation is associated with balance and cognitive function in different ways between older adults with and without Parkinson’s disease . Cerebral Cortex Communications . 2022 Jul 1; 3 ( 3 ):tgac030. 49. ↵ Payne AM , Ting LH , Hajcak G . The balance N1 and the ERN correlate in amplitude across individuals in small samples of younger and older adults . Exp Brain Res [Internet ]. 2023 Aug 30 [cited 2023 Sep 12]; Available from : doi: 10.1007/s00221-023-06692-9 OpenUrl CrossRef 50. ↵ Leddy AL , Crowner BE , Earhart GM . Utility of the Mini-BESTest, BESTest, and BESTest Sections for Balance Assessments in Individuals with Parkinson Disease . J Neurol Phys Ther . 2011 Jun ; 35 ( 2 ): 90 – 7 . OpenUrl CrossRef PubMed 51. ↵ Löfgren N , Lenholm E , Conradsson D , Ståhle A , Franzén E . The Mini-BESTest - a clinically reproducible tool for balance evaluations in mild to moderate Parkinson’s disease? BMC Neurol . 2014 Dec 12; 14 : 235 . 52. ↵ Magnani PE , Genovez MB , Porto JM , Zanellato NFG , Alvarenga IC , Freire RCJ , et al. Use of the BESTest and the Mini-BESTest for Fall Risk Prediction in Community-Dwelling Older Adults Between 60 and 102 Years of Age . Journal of Geriatric Physical Therapy . 2020 Dec ; 43 ( 4 ): 179 . OpenUrl CrossRef PubMed 53. ↵ Ghosn NJ , Palmer JA , Borich MR , Ting LH , Payne AM . Cortical Beta Oscillatory Activity Evoked during Reactive Balance Recovery Scales with Perturbation Difficulty and Individual Balance Ability . Brain Sci . 2020 N ov 16; 10 ( 11 ). 54. ↵ Payne AM , Ting LH . Worse balance is associated with larger perturbation-evoked cortical responses in healthy young adults . Gait & Posture . 2020 Jul 1; 80 : 324 – 30 . OpenUrl CrossRef PubMed 55. ↵ Payne AM , Ting LH . Balance perturbation-evoked cortical N1 responses are larger when stepping and not influenced by motor planning . Journal of Neurophysiology . 2020 Dec 1; 124 ( 6 ): 1875 – 84 . OpenUrl CrossRef PubMed 56. ↵ Basmajian JV . Electrode placement in EMG biofeedback . Baltimore : Williams & Wilkins ; 1980 . 57. ↵ Safavynia SA , Ting LH . Long-latency muscle activity reflects continuous, delayed sensorimotor feedback of task-level and not joint-level error . J Neurophysiol . 2013 Sep ; 110 ( 6 ): 1278 – 90 . OpenUrl CrossRef PubMed 58. ↵ Reuter-Lorenz PA , Cappell KA . Neurocognitive Aging and the Compensation Hypothesis . Curr Dir Psychol Sci . 2008 Jun 1; 17 ( 3 ): 177 – 82 . OpenUrl CrossRef Web of Science 59. ↵ Kang W , Wang J , Malvaso A . Inhibitory Control in Aging: The Compensation-Related Utilization of Neural Circuits Hypothesis. Front Aging Neurosci [Internet]. 2022 Jul 28 [cited 2024 Nov 18];13. Available from: https://www.frontiersin.org/journals/aging-neuroscience/articles/10.3389/fnagi.2021.771885/full 60. ↵ Guadagnoli MA , Lee TD . Challenge Point: A Framework for Conceptualizing the Effects of Various Practice Conditions in Motor Learning . Journal of Motor Behavior . 2004 Jul 1; 36 ( 2 ): 212 – 24 . OpenUrl CrossRef PubMed Web of Science 61. ↵ Sawers A , Allen JL , Ting LH . Long-term training modifies the modular structure and organization of walking balance control . J Neurophysiol . 2015 Dec 1; 114 ( 6 ): 3359 – 73 . OpenUrl CrossRef PubMed 62. ↵ Ozdemir RA , Contreras-Vidal JL , Lee BC , Paloski WH . Cortical activity modulations underlying age- related performance differences during posture–cognition dual tasking . Exp Brain Res . 2016 Nov 1; 234 ( 11 ): 3321 – 34 . OpenUrl CrossRef PubMed 63. ↵ Kahneman D. Attention and Effort . Vol. 1063 . Englewood Cliffs , NJ : Prentice-Hall ; 1973 . 64. Wickens CD . Multiple resources and performance prediction . Theoretical Issues in Ergonomics Science . 2002 Jan 1; 3 ( 2 ): 159 – 77 . OpenUrl CrossRef 65. ↵ Watanabe K , Funahashi S . Neural mechanisms of dual-task interference and cognitive capacity limitation in the prefrontal cortex . Nat Neurosci . 2014 Apr ; 17 ( 4 ): 601 – 11 . OpenUrl CrossRef PubMed 66. ↵ Khan B , Wildey C , Francis R , Tian F , Delgado MR , Liu H , et al. Improving optical contact for functional near-infrared brain spectroscopy and imaging with brush optodes. Biomed Opt Express , BOE . 2012 May 1; 3 ( 5 ): 878 – 98 . OpenUrl PubMed 67. Scholkmann F , Kleiser S , Metz AJ , Zimmermann R , Mata Pavia J , Wolf U , et al. A review on continuous wave functional near-infrared spectroscopy and imaging instrumentation and methodology . NeuroImage . 2014 Jan ; 85 : 6 – 27 . OpenUrl CrossRef PubMed Web of Science 68. Kassab A , Lan JL , Vannasing P , Sawan M . Functional near-infrared spectroscopy caps for brain activity monitoring: a review. Appl Opt , AO . 2015 Jan 20; 54 ( 3 ): 576 – 86 . OpenUrl 69. Yücel MA , Selb JJ , Huppert TJ , Franceschini MA , Boas DA . Functional Near Infrared Spectroscopy: Enabling routine functional brain imaging . Current Opinion in Biomedical Engineering . 2017 Dec 1; 4 : 78 – 86 . OpenUrl CrossRef PubMed 70. ↵ Kim Y , Choi J , Kim B , Park Y , Cha J , Choi J , et al. Investigating the relationship between CSAT scores and prefrontal fNIRS signals during cognitive tasks using a quantum annealing algorithm . Sci Rep . 2024 Aug 26; 14 ( 1 ): 19760 . OpenUrl CrossRef PubMed 71. ↵ Payne AM , Hajcak G , Ting LH . Dissociation of muscle and cortical response scaling to balance perturbation acceleration . Journal of Neurophysiology . 2019 Mar 5; 121 ( 3 ): 867 – 80 . OpenUrl CrossRef PubMed 72. ↵ Petersen N , Christensen LOD , Morita H , Sinkjær T , Nielsen J . Evidence that a transcortical pathway contributes to stretch reflexes in the tibialis anterior muscle in man . The Journal of Physiology . 1998 ; 512 ( 1 ): 267 – 76 . OpenUrl CrossRef PubMed Web of Science 73. ↵ Jakubowski KL , Martino G , Beck ON , Sawicki GS , Ting LH . Center of mass states render multi-joint torques throughout standing balance recovery [Internet] . bioRxiv; 2024 [cited 2024 Dec 3]. p. 2024.08.14.607976. Available from: https://www.biorxiv.org/content/10.1101/2024.08.14.607976v1 View the discussion thread. Back to top Previous Next Posted December 15, 2024. Download PDF Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Cortically-mediated muscle responses to balance perturbations increase with perturbation magnitude in older adults with and without Parkinson’s disease Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Cortically-mediated muscle responses to balance perturbations increase with perturbation magnitude in older adults with and without Parkinson’s disease Scott E. Boebinger , Aiden M. Payne , Jifei Xiao , Giovanni Martino , Michael Borich , J. Lucas McKay , Lena H. Ting bioRxiv 2024.12.09.627582; doi: https://doi.org/10.1101/2024.12.09.627582 Share This Article: Copy Citation Tools Cortically-mediated muscle responses to balance perturbations increase with perturbation magnitude in older adults with and without Parkinson’s disease Scott E. Boebinger , Aiden M. Payne , Jifei Xiao , Giovanni Martino , Michael Borich , J. Lucas McKay , Lena H. Ting bioRxiv 2024.12.09.627582; doi: https://doi.org/10.1101/2024.12.09.627582 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Areas All Articles Animal Behavior and Cognition (7647) Biochemistry (17728) Bioengineering (13921) Bioinformatics (42047) Biophysics (21490) Cancer Biology (18637) Cell Biology (25555) Clinical Trials (138) Developmental Biology (13403) Ecology (19942) Epidemiology (2067) Evolutionary Biology (24368) Genetics (15625) Genomics (22549) Immunology (17764) Microbiology (40475) Molecular Biology (17208) Neuroscience (88761) Paleontology (667) Pathology (2842) Pharmacology and Toxicology (4834) Physiology (7659) Plant Biology (15175) Scientific Communication and Education (2047) Synthetic Biology (4304) Systems Biology (9835) Zoology (2272)