Effects of postural threat on spinal stretch reflexes: evidence for ...

J Neurophysiol 110: 899 –906, 2013. First published May 29, 2013; doi:10.1152/jn.00065.2013.

Effects of postural threat on spinal stretch reflexes: evidence for increased muscle spindle sensitivity? Brian C. Horslen, Chantelle D. Murnaghan, J. Timothy Inglis, Romeo Chua, and Mark G. Carpenter School of Kinesiology, The University of British Columbia, Vancouver, British Columbia, Canada Submitted 28 January 2013; accepted in final form 26 May 2013

Horslen BC, Murnaghan CD, Inglis JT, Chua R, Carpenter MG. Effects of postural threat on spinal stretch reflexes: evidence for increased muscle spindle sensitivity? J Neurophysiol 110: 899 –906, 2013. First published May 29, 2013; doi:10.1152/jn.00065.2013.— Standing balance is often threatened in everyday life. These threats typically involve scenarios in which either the likelihood or the consequence of falling is higher than normal. When cats are placed in these scenarios they respond by increasing the sensitivity of muscle spindles imbedded in the leg muscles, presumably to increase balancerelevant afferent information available to the nervous system. At present, it is unknown whether humans also respond to such postural threats by altering muscle spindle sensitivity. Here we present two studies that probed the effects of postural threat on spinal stretch reflexes. In study 1 we manipulated the threat associated with an increased consequence of a fall by having subjects stand at the edge of an elevated surface (3.2 m). In study 2 we manipulated the threat by increasing the likelihood of a fall by occasionally tilting the support surface on which subjects stood. In both scenarios we used Hoffmann (H) and tendon stretch (T) reflexes to probe the spinal stretch reflex circuit of the soleus muscle. We observed increased T-reflex amplitudes and unchanged H-reflex amplitudes in both threat scenarios. These results suggest that the synaptic state of the spinal stretch reflex is unaffected by postural threat and that therefore the muscle spindles activated in the T-reflexes must be more sensitive in the threatening conditions. We propose that this increase in sensitivity may function to satisfy the conflicting needs to restrict movement with threat, while maintaining a certain amount of sensory information related to postural control. postural threat; tendon stretch reflex; Hoffmann reflex; arousal; muscle spindle

in which balance is challenged or threatened, they respond by increasing muscle spindle sensitivity to muscle stretch (Prochazka et al. 1985, 1988). Such changes in muscle spindle sensitivity have been attributed to an increase in ␥-motoneuron activity to the muscle spindles (Prochazka et al. 1985, 1988) independent from, or disproportionally to, the amount of ␣-motoneuron activation of extrafusal muscle fibers. Alternatively, mechanisms for increased muscle spindle sensitivity have been associated with sympathetic arousal (Hunt 1960), possibly through direct links between sympathetic neurons and muscle spindles (Barker and Saito 1981). Although early work did not support selective gamma drive (cf. Gandevia et al. 1997; Gandevia and Burke 1985) or sympathetic modulation of muscle spindle sensitivity in humans (cf. Macefield et al. 2003), these studies were limited to WHEN CATS ARE PLACED IN SCENARIOS

Address for reprint requests and other correspondence: M. G. Carpenter, School of Kinesiology, Univ. of British Columbia, Osborne Centre Unit 1, 6108 Thunderbird Blvd., Vancouver, BC, Canada V6T 1Z4 (e-mail: mark. [email protected]). www.jn.org

measuring resting spindle discharge rates in nonstanding tasks, and typically in unloaded limbs. In contrast, new evidence has emerged that supports spindle sensitivity changes in humans within certain behavioral states or when the host muscles are actively engaged. Recent research has suggested that changes in muscle spindle sensitivity occur in certain behavioral contexts, such as when people attend to discrete stimuli or engage in novel tasks. Hospod et al. (2007) and Ribot-Ciscar et al. (2009) both found that lower-limb muscle spindle firing rates were increased when subjects had to attend to passive movements compared with when the movements were ignored. Similarly, Nafati et al. (2004) found that wrist extensor singlemotor unit firing rates were increased in response to tendon stretch (T) reflexes and decreased with electrically evoked Hoffmann (H) reflexes when subjects were challenged to make isometric contractions with high visual feedback gain compared with low gain. Since both reflexes presumably use the same spinal circuitry yet only T-reflexes directly activate muscle spindles, it was concluded that the difference was related to increased spindle sensitivity. Also, Wong et al. (2011) found that wrist proprioceptive acuity was increased after learning a new motor task; this was also presumed to be related to context-specific changes to spindle sensitivity. Functionally, Dimitriou and Edin (2010) have suggested that selective fusimotor drive might indeed be used to uncouple spindle tension from active muscle contractions to better allow spindles to predict future muscle states. Improving spindle sensitivity might serve to facilitate feedback gain in novel or attention-demanding situations, better allowing the body to monitor motor performance. There is converging evidence to suggest that humans, like cats, may increase muscle spindle sensitivity when standing or walking under challenging or threatening conditions. For example, lower-limb ␥-motoneurons can be electrically activated independently from ␣-motoneurons when subjects are asked to balance with eyes closed (Aniss et al. 1990) but not when they are seated (Aniss et al. 1988). T-reflexes have also been shown to be facilitated when people stand at the edge of an elevated platform (Davis et al. 2011). Conversely, H-reflexes are found to be attenuated when people stand on an elevated surface (Sibley et al. 2007), walk along a narrow elevated beam (Llewellyn et al. 1990), or balance an inverted pendulum with their feet (McIlroy et al. 2003). It has been suggested that these H-reflex changes reflect spinal reflex inhibition, either by presynaptic mechanisms or homosynaptic postactivation depression, intended to mute the potentially destabilizing effects of larger stretch reflexes in these scenarios (Llewellyn et al. 1990). While these observations may reflect changes to spindle sensitivity via either selective gamma drive or sympathetic

0022-3077/13 Copyright © 2013 the American Physiological Society

899

900

EFFECTS OF POSTURAL THREAT ON SPINAL STRETCH REFLEXES

drive, none of the studies measured both H-reflexes and T-reflexes simultaneously to establish that these effects are occurring concurrently. As such, while the reflex changes reported across studies seem indicative of a common neurophysiological phenomenon activated under conditions of postural threat, the studies themselves are too different to aggregate. Therefore, the purpose of this study was to explore the effects of a postural threat on both H-reflexes and T-reflexes, within the same subjects, to determine whether spinal stretch reflexes are indeed altered in a manner consistent with increased muscle spindle sensitivity. We adopted two separate postural threat paradigms to explore the effects of threat on spinal reflexes. In study 1, we used a platform elevated to 3.2 m to increase postural threat by increasing the consequences of a fall (from height), thus replicating the type of threat used by both Sibley et al. (2007) and Davis et al. (2011). In study 2, we used impending balance perturbations to manipulate postural threat by increasing the likelihood of a balance disturbance. This paradigm allowed us to avoid some of the postural changes associated with standing at height, such as backward leaning and increased muscle activity related to sway stiffness (Carpenter et al. 1999, 2001; Davis et al. 2009), that may alter reflex responses if not controlled (Stein et al. 2007; Tokuno et al. 2007, 2008, 2009). In addition, this scenario may also be more generalizable to the types of postural threat experienced in everyday life. We predicted that in both postural threat scenarios we would see evidence of increased muscle spindle sensitivity and central inhibition, as evidenced by increased T-reflexes and decreased H-reflexes when the postural threat was present. While postural threat is known to influence fear, anxiety, and balance confidence, it also induces strong increases in autonomic arousal (Brown et al. 2006; Carpenter et al. 2006; Davis et al. 2009; Huffman et al. 2009), which by itself can influence T-reflexes in seated individuals (Bonnet et al. 1995; Both et al. 2005; Hjortskov et al. 2005; Kamibayashi et al. 2009). As such, a secondary objective in study 2 was determining the effect of arousal in mediating reflex changes related to postural threat by exposing subjects to arousing, non-posturally relevant stimuli. We had subjects watch affective pictures that were specifically chosen to either promote or reduce arousal. This enabled us to determine whether arousal alone could influence spindle sensitivity in quiet stance, and whether the effects of threat on reflexes could be scaled to underlying arousal levels. We hypothesized that if the reflex changes induced by postural threat were purely arousal driven, then we would see scaled reflex changes dependent on both the presence of a postural threat and either arousing or calming pictures. METHODS

Ethical Approval The procedures used in these studies were approved by the University of British Columbia Clinical Research Ethics Board. In study 1, 26 subjects (13 men, 13 women) completed the study, of whom 21 [10 men, 11 women; mean (SD) age ⫽ 22.6 (3.0) yr] met all inclusion criteria and were included in the final data set. In study 2, 16 male subjects completed the study [mean (SD) age ⫽ 23.4 (1.23) yr]. No subjects in either study reported any known neurological, orthopedic, or vestibular impairment that may have impeded their ability to complete the study. All subjects gave written informed consent prior to participation.

Protocol Postural threat. STUDY 1. A hydraulic lift (Pentalift: M419207B10H01D, Guelph, ON, Canada) was used to manipulate postural threat. Subjects stood in two threat conditions: a low-threat (LOW) condition in which they stood 0.8 m above ground and 0.6 m from the edge of the support surface and a high-threat (HIGH) condition in which they stood 3.2 m above ground and at the edge of the support surface. Subjects stood in the same place on the lift in both threat conditions, and a 0.6-m-wide table was used to extend the support surface in the LOW condition. Since height-induced postural threat is known to induce both posterior leaning and changes in background muscle activity (Carpenter et al. 2001), and since both of these factors can change H-reflex amplitudes (leaning: Tokuno et al. 2007, 2008, 2009; background activity: Stein et al. 2007), custom-made ankle braces were used to limit leaning about the ankles and keep background muscle activity constant across conditions. As there are documented order effects of height-induced postural threat (Adkin et al. 2000), all subjects experienced the LOW condition first in order to maximize threat effects. STUDY 2. For all trials subjects stood on a custom-made servocontrolled tilting platform that could be used to induce both toes-up and toes-down sagittal-plane perturbations via support surface rotations (amplitude: 7°, duration: 140 ms, angular speed: 50°/s). Threat of perturbation was manipulated by explicitly informing the subject whether or not he would be perturbed in the upcoming trial (threat of perturbation: PRESENT or ABSENT). Instructions were always congruent with the actual platform movement; no perturbations were experienced in ABSENT trials, while 10 perturbations (5 in each direction) were presented in a randomized order in PRESENT trials, such that direction and timing were not predictable. Since braces would hinder subjects’ ability to recover from the perturbations, ankle angle was instead monitored online to ensure that stimuli were presented at consistent angles. Infrared light-emitting diodes were placed on the fifth metatarsal, lateral malleolus, and tibial head of the right leg and used to define foot and shank segments, from which relative angles were calculated (Optotrak Certus, NDI; sampled at 3,000 Hz). Subjects completed a 60-s quiet standing trial before the first experimental trial, the data from which were used to set mean ⫾ 2 SD stimulation ankle angle windows for both the frontal and sagittal planes. An experimenter monitored these angles and only evoked stimuli (H-reflexes, T-reflexes, and perturbations) when subjects were within these windows. If subjects strayed from the defined bounds, they were given verbal feedback to guide them back and then stimuli were presented again once the subject had remained within bounds for a few seconds. Arousal manipulation. Affective pictures from the International Affective Picture System (IAPS) database were used to manipulate arousal in study 2. Normative IAPS valence (pleasantness) and arousal (sometimes called intensity) values were used to select pictures to create two picture groups with similar valence scores yet rated as either arousing (high arousal) or calming (low arousal). Only pleasant pictures were used in this study, as they have been previously demonstrated to cause a greater change in T-reflex amplitudes across arousal levels than neutral or unpleasant pictures (Bonnet et al. 1995). Pictures were displayed continuously on a 1 ⫻ 1.3-m screen, 3.7 m in front of the subject, throughout the trials, and each picture was displayed for 10 s. No affective pictures were used in study 1. Reflex stimulation. H-reflexes were elicited in the right soleus muscle (SOL) with 1-ms-long square-wave electrical pulses to the tibial nerve at the popliteal fossa (Grass S48 Stimulator, SIU5 Stimulation Isolation Unit, CCU1 Constant Current Unit; Grass Technologies). Prior to commencement of the study an M-wave recruitment curve was performed for each subject while he/she stood at the LOW position. These data were used to set the H-reflex stimulation intensity between 10% and 15% of maximum M-wave amplitude (Mmax). H-reflexes were elicited in pairs spaced 200 ms apart to induce

J Neurophysiol • doi:10.1152/jn.00065.2013 • www.jn.org


homosynaptic postactivation depression. However, the evoked M waves in the second reflex were highly variable, and, as such, we were not able to interpret the effects of threat on homosynaptic postactivation depression; these data will not be discussed further. T-reflexes were evoked in the right SOL with mechanical taps to the Achilles tendon. A magnetic linear motor (motor: LinMot PS01-23x80, controller: LinMot E2000-AT, software: LinMot v.1.3.12; NTI) aligned perpendicular to the tendon was used to deliver the taps. The motor was programmed to stroke out 1 cm in 26 ms and was positioned ⬃0.5 cm from the tendon. The peak strike force of each tap on the tendon was measured with a force transducer (Isotron Dynamic Force Sensor, Endevco; A/D sampling: 1,000 Hz) mounted to the linear motor. These forces were used to ensure consistent T-reflex stimulation intensity. T-reflexes and H-reflexes were presented in a randomized order within the same trial; stimuli were separated with a variable interstimulus interval with a minimum period of 10 s. In study 1, subjects underwent two 150-s trials at each height condition, in which they received five H-reflex and five T-reflex stimulations. The data from these trials were compiled for a total of 10 H reflex and 10 T-reflex stimulations per condition. In study 2, subjects received 15 H-reflex and 15 T-reflex stimulations in the two trials without perturbations and 10 H-reflex and 10 T-reflex stimulations in the two trials with perturbations. A larger number of stimulations were used in study 2 because pilot testing revealed more within-subject variability in stimulation intensity. Therefore, the number of stimulations was increased to improve the likelihood of meeting the inclusion criteria. There were 10 fewer reflexes evoked in the perturbations PRESENT trials to allow for the extra time required for the 10 platform perturbations. In all, there were always 30 stimulations (H-reflexes, T-reflexes, or perturbations) in a trial. Reflex inclusion criteria. H-reflexes were screened post hoc to ensure that only those evoked with an M wave between 10% and 15% of Mmax were included in the studies. To be consistent with Sibley et al. (2007), a minimum of five H-reflexes per threat condition was required for a subject to be included in the final data set. A similar screening protocol, based on the methods used by Davis et al. (2011), was used for T-reflexes. For each trial the range of tap forces was calculated. A mean ⫾ 2 SD window was then calculated from the trial with the smallest range. T-reflexes evoked with a tap force outside this window were excluded post hoc. A minimum of five T-reflexes per condition was required for a subject to be included in the final data set.

901

Psychological and autonomic state. Prior to each trial, participants reported their confidence that they could maintain their balance for the duration of the trial (0%: not at all confident, 100%: completely confident). After each trial participants rated their experienced fear of falling (0%: not at all fearful, 100%: extremely fearful) and anxiety (16-item questionnaire with a maximum cumulative score of 144, where higher scores indicate more anxiety). The above questionnaires were each modified from those used by Adkin et al. (2002) and have been demonstrated to have moderate to high reliability (Hauck et al. 2008); the individual questionnaire scores were averaged across trials within each condition. STUDY 1. Electrodermal activity (EDA), used to quantify physiological arousal, was recorded from the thenar and hypothenar eminences of the nondominant hand for the duration of each trial (model 2501, CED). These data were sampled at 1,000 Hz and then low-pass filtered at 5 Hz off-line. A mean value was calculated for the duration of each 150-s trial, and these values were then averaged within height conditions. STUDY 2. Since perturbations are known to induce large transient increases in skin conductance (Sibley et al. 2008), EDA was not averaged over the course of the trials in this study. EDA was sampled for 1 s ending 500 ms prior to each stimulus (perturbation, H-reflex, or T-reflex), and then all of these samples were averaged within a trial to give a single value. These values were then used for comparison between trials. Subjects also completed the Self-Assessment Manikin (SAM) arousal scale (Bradley and Lang 1994) to rate their perceptions of how arousing the pictures, as a block, were after each trial. Statistics For study 1, within-subjects paired-samples t-tests were used to test for significant changes across threat conditions (LOW, HIGH) for all variables. For study 2, 2 (threat of perturbation: PRESENT, ABSENT) ⫻ 2 (picture condition: calming, arousing) repeated-measures ANOVAs were used to test for significant differences between conditions for all variables. In all cases, the criterion for statistical significance was set to ␣ ⫽ 0.05; ␩2 (t-tests) and partial ␩2 (␩2p, ANOVAs) were used to calculate effect sizes. All mean differences (⫾SE) and statistical results for studies 1 and 2 are listed in Table 1.

RESULTS

Measures Electromyography. EMG was recorded at 3,000 Hz from the right SOL with a monopolar belly-tendon electrode configuration and analog band-pass filtered between 10 and 1,000 Hz (Telemyo 2400 RG2, Noraxon) and then A/D sampled at 3,000 Hz (Power 1401, CED). This monopolar configuration was used to measure both H and T-reflexes, as recommended by Hadoush et al. (2009). Simultaneously, bipolar belly-belly EMG was recorded at 3,000 Hz from both the right SOL and tibialis anterior muscle (TA). These data were also analog filtered (10 to 1,000 Hz), and then A/D sampled at 1,000 Hz; these data were used to quantify background muscle activity. Background muscle activity was calculated off-line for both SOL and TA as the root mean square amplitude of a 100-ms period of EMG immediately preceding a reflex. These values were then averaged within each condition. Since it is common for individuals to increase their background muscle activity when standing at height (study 1; Carpenter et al. 2001), any single reflex evoked that was preceded by background EMG in either SOL or TA that was more than twice the mean amplitude of background activity in the LOW or ABSENT condition was excluded from the analysis. H-reflexes, M waves, and T-reflexes were all calculated as peak-to-peak amplitudes from the belly-tendon SOL EMG trace. Reflexes that passed stimulation intensity screening were then averaged within threat conditions.

Study 1 Subject exclusions. Of the 21 subjects who completed the study, 20 subjects met the T-reflex and 18 the H-reflex inclusion criteria; 17 subjects met both. One subject was excluded from the final EDA analysis because of equipment malfunction. Reflexes and background muscle activity. There was a significant increase in T-reflex peak-to-peak amplitudes in the HIGH compared with LOW surface height conditions (Fig. 1 and Fig. 2A). H-reflexes, in contrast, were not significantly modulated with postural threat (Fig. 1 and Fig. 2A). There were no systematic effects of surface height on M-wave amplitude (Fig. 2A) or background EMG in either SOL or TA (Table 1). Psychological and autonomic state. Subjects felt significantly less confident in their ability to maintain balance in the HIGH compared with LOW surface height conditions. Subjects also felt significantly more fearful of falling, more anxious, and less stable in the HIGH compared with LOW surface height conditions. Finally, subjects were more aroused, as indicated by mean EDA, during the HIGH compared with LOW surface height conditions (Table 1).


902


Table 1. Summary of statistical test results and mean differences across conditions Study 1 Variable T-reflex, mV

H-reflex, mV

M wave, mV

SOL background, ␮V

TA background, ␮V

Balance confidence, %

Fear, %

Anxiety (max 144)

EDA, ␮S

SAM arousal

Statistic t19 ⫽ P⫽ ␩2 ⫽ Mean t17 ⫽ P⫽ ␩2 ⫽ Mean t17 ⫽ P⫽ ␩2 ⫽ Mean t20 ⫽ P⫽ ␩2 ⫽ Mean t20 ⫽ P⫽ ␩2 ⫽ Mean t20 ⫽ P⫽ ␩2 ⫽ Mean t20 ⫽ P⫽ ␩2 ⫽ Mean t20 ⫽ P⫽ ␩2 ⫽ Mean t19 ⫽ P⫽ ␩2 ⫽ Mean N/A

⌬ (SE)

⌬ (SE)

⌬ (SE)

⌬ (SE)

⌬ (SE)

⌬ (SE)

⌬ (SE)

⌬ (SE)

⌬ (SE)

Study 2 Threat

ⴚ2.60 0.018 0.262 0.558 (0.215) 1.57 0.135 0.120 ⫺0.278 (0.177) ⫺1.51 0.149 0.112 0.0556 (0.127) 0.16 0.877 0.001 0.207 (1.32) ⫺0.02 0.981 ⬍0.001 0.0164 (0.689) 3.16 0.005 0.333 ⴚ11.8 (3.73) ⴚ3.34 0.003 0.358 17.4 (5.19) ⴚ3.34 0.003 0.358 13.7 (4.11) ⴚ2.57 0.019 0.258 5.41 (2.10)

Statistic F1,10 ⫽ P⫽ ␩P2 ⫽ Mean ⌬ F1,12 ⫽ P⫽ ␩P2 ⫽ Mean ⌬ F1,12 ⫽ P⫽ ␩P2 ⫽ Mean ⌬ F1,15 ⫽ P⫽ ␩P2 ⫽ Mean ⌬ F1,15 ⫽ P⫽ ␩P2 ⫽ Mean ⌬ F1,14 ⫽ P⫽ ␩P2 ⫽ Mean ⌬ F1,14 ⫽ P⫽ ␩P2 ⫽ Mean ⌬ F1,14 ⫽ P⫽ ␩P2 ⫽ Mean ⌬ F1,14 ⫽ P⫽ ␩P2 ⫽ Mean ⌬ F1,15 ⫽ P⫽ ␩P2 ⫽ Mean ⌬

(SE)

(SE)

(SE)

(SE)

(SE)

(SE)

(SE)

(SE)

(SE)

(SE)

Threat

Pictures

Interaction

7.15 0.023 0.417 0.449 (0.168) 2.07 0.176 0.147 0.165 (0.115) 0.285 0.603 0.023 ⫺0.0229 (0.0429) 0.12 0.737 0.008 ⫺0.320 (0.935) 1.35 0.263 0.083 2.14 (1.84) 11.9 0.004 0.896 ⴚ17.3 (5.03) 12.3 0.003 0.468 18.2 (5.17) 6.07 0.026 0.288 6.34 (2.58) 25.72 ⬍0.001 0.648 5.16 (1.02) 1.521 0.236 0.092 0.281 (0.228)

0.005 0.946 ⬍0.001 0.00505 (0.073) 0.973 0.343 0.075 0.138 (0.14) 0.375 0.552 0.030 0.022 (0.036) 2.16 0.162 0.126 1.14 (0.778) 3.47 0.082 0.188 1.72 (0.925) 2.204 0.158 0.285 2.66 (1.79) 6.18 0.026 0.306 6.50 (2.61) 1.365 0.261 0.083 1.66 (1.42) 0.027 0.871 0.002 ⫺0.124 (0.75) 113.07 ⬍0.001 0.883 3.84 (0.361)

0.926 0.359 0.085 – 3.02 0.108 0.201 – 0.446 0.517 0.036 – 0.63 0.438 0.041 – 0.27 0.612 0.018 – 1.22 0.288 0.178 – 4.707 0.048 0.252 – 0.04 0.845 0.003 – 0.58 0.459 0.04 – 1.05 0.323 0.065 –

Study 1: threat: HIGH-LOW. Study 2: threat: perturbations PRESENT-ABSENT, pictures arousing-calming. SOL, soleus muscle; TA, tibialis anterior muscle; EDA, electrodermal activity; SAM, Self-Assessment Manikin; N/A, not applicable. Significant values are in boldface.

Study 2 Subject exclusions. Of the 16 subjects who completed the study, 5 were excluded from the final T-reflex analysis because of high variability in tendon-tap force across conditions. Three subjects were excluded from the H-reflex analysis, two subjects because of high M-wave variability and one subject could not tolerate the H-reflex stimuli. Ten subjects were included in both the T-reflex and H-reflex analyses. One subject was excluded from the EDA analysis because of equipment malfunction, and psychological state questionnaire data were not used from one subject because they did not complete the questionnaires. Reflexes and background muscle activity. T-reflex amplitudes were significantly larger when the threat of perturbation was PRESENT compared with ABSENT (Fig. 1 and Fig. 2A). There was no significant main effect of picture condition or interaction between threat of perturbation and picture condition on T-reflex amplitudes. H-reflexes and M-wave amplitudes were not significantly influenced by the main effects of threat

of perturbation (Fig. 1 and Fig. 2A) or picture condition or by their interaction. There were no main effects of threat of perturbation, picture condition, or interaction effects for either SOL or TA background EMG (Table 1). Psychological and autonomic state. Participants were significantly more aroused, as indicated by EDA, when the threat of perturbation was PRESENT compared with ABSENT. While subjects rated arousing pictures higher on the SAM arousal scale than calming pictures, there was no effect of picture condition on EDA, nor was there an interaction (Table 1). Participants felt significantly less confident before and more anxious after the threat of perturbation PRESENT compared with ABSENT trials; however, there were no significant main effects of picture condition or interactions affecting these variables. There was a significant threat of perturbation ⫻ picture condition interaction on rated fear of falling. While fear of falling was on average higher when the threat of perturbation was PRESENT compared with ABSENT, there were



Study 2 T-reflexes

Study 1 T-reflexes

1 mV

0.5 mV

903

50 ms

50 ms

Height

Perturbations

LOW

ABSENT

HIGH

PRESENT

H-reflexes

2 mV

2 mV

H-reflexes

Fig. 1. Two representative subjects’ EMG recordings of T-reflexes (top) and H-reflexes (bottom) delivered while standing on low and high heights (study 1, left) or in the presence and absence of perturbations (study 2, right). The lower threat (LOW, ABSENT) conditions are drawn in black and the higher threat (HIGH, PRESENT) conditions in gray.

50 ms

50 ms

differences in the effects of picture arousal levels between these conditions. Post hoc comparisons revealed that subjects felt significantly more fearful of falling when exposed to high-arousal compared with low-arousal pictures when the threat of perturbation was PRESENT (t14 ⫽ ⫺2.60, P ⫽ 0.021, ␩2 ⫽ 0.326) but not when the threat of perturbation was ABSENT (t14 ⬍ 0.001, P ⬎ 0.999, ␩2 ⬍ 0.001). Relationship between T and H-reflexes. We performed a Pearson product-moment correlation post hoc to determine whether or not changes in T-reflexes might be related to changes in H-reflexes. Subjects who had both T and H-reflexes were included, and data from both studies were combined to increase the sample size (Fig. 2B; n ⫽ 27). Change values were calculated to normalize the data; changes were expressed as percent change from the lowest threat condition (i.e., study 1: LOW, study 2: ABSENT). There was no relationship (r ⫽ ⫺0.0928) between changes in T and H-reflexes. DISCUSSION

Postural threat posed by the increased consequence (study 1) or likelihood (study 2) of a fall led to increases in arousal, fear, and anxiety and decreases in balance confidence. In both cases, the increased threat was accompanied by a significant increase in the T-reflex amplitude in SOL. The observation of increased T-reflex amplitudes when standing on an elevated surface (study 1) or when expecting a balance perturbation (study 2) is consistent with prior observations of increased T-reflexes under conditions of increased arousal in subjects who are sitting (Bonnet et al. 1995; Both et al. 2005; Hjortskov et al. 2005; Kamibayashi et al. 2009) or are standing on elevated surfaces (Davis et al. 2011). Our observations are unique in that the increases in T-reflex amplitude with threat are shown to occur in a posturally engaged muscle, despite the absence of any

concomitant change in H-reflex amplitude or background EMG. Taken together, this provides evidence to suggest that these changes may be occurring through increases in muscle spindle receptor sensitivity rather than reflex facilitation at the spinal level during stance. While we interpret these results as being reflective of increased muscle spindle sensitivity with increased postural threat, there are certain limitations that must be acknowledged. First, it has been argued that H-reflexes and T-reflexes are not directly comparable because 1) H-reflexes indiscriminately activate all large-diameter Ia and Ib afferents compared with T-reflexes, which are thought to preferentially activate Ia afferents, and 2) synchronous H-reflex vs. asynchronous T-reflex afferent volleys mean that the reflexes have different sensitivities to presynaptic and polysynaptic reflex changes (Burke et al. 1983). However, while the magnitude of effect of these different factors may differ between reflexes, the direction of effect should be the same. That is, if presynaptic inhibition were suppressing H-reflexes, then T-reflexes should also be suppressed, if not to the same degree. Considering the fact that there was a highly consistent increase in T-reflexes between subjects and across studies (in that most subjects showed increases with threat) and no consistent change in H-reflexes (there is a near-even distribution of subjects who demonstrated increases and decreases), we suggest that the factors that influenced T-reflexes with threat did not affect H-reflexes. A second limitation in these studies is that muscle history effects were not controlled between conditions. Muscle spindle tension, as indicated by discharge rate, can be tonically increased after isometric contraction of the parent muscle (Gregory et al. 1990; Wilson et al. 1995). This is thought to be caused by formation of stable intrafusal muscle fiber cross


904


A

0.8

Threat Source: Height (Study 1)

0.6

Perturbation (Study 2)

Mean Difference (mV)

0.4

0.2

0

-0.2

-0.4

-0.6

-0.8

B

*

n.s.

n.s.

T-reflex

M wave

H-reflex

75

Change H-reflex (%)

50

25

0

-25

r = -0.0927

-50

-75 -50

-25

0

25

50

75

100

Change T-reflex (%) Fig. 2. A: mean amplitude differences between threat conditions for T reflexes, M waves, and H reflexes for study 1 (gray circles) and study 2 (black squares). Differences are calculated as HIGH ⫺ LOW or PRESENT ⫺ ABSENT; as such, higher values indicate larger amplitudes with threat. Bars represent SE. *Significant at P ⬍ 0.05. n.s., Not statistically different. B: scatterplot of mean differences of T reflexes (x-axis) against H reflexes (y-axis) for all subjects in whom both T and H reflexes were elicited (n ⫽ 27; study 1: 17, study 2: 10). Data from both studies are presented together, and differences are expressed as % change. The % change values were calculated as follows: study 1: (HIGH ⫺ LOW)/LOW ⫻ 100, study 2: (PRESENT ⫺ ABSENT)/ABSENT ⫻ 100. Positive values indicate an increase in amplitude with threat, r value calculated with Pearson’s productmoment correlation.

bridges during the contraction (Gregory et al. 1990), a phenomenon known as intrafusal stiction, and has been reported to last as long as 4 min in resting human subjects (Wilson et al. 1995). Stiction can increase T-reflex amplitudes and has been suggested to be a major contributor to T-reflex variability when

not controlled (Wood et al. 1994). There were no changes in background muscle activity that would suggest a conditioning contraction in the threatening conditions of either of our studies; however, we cannot exclude the possibility that such a contraction occurred between trials when EMG was not recorded. Despite this, we would argue that stiction effects cannot explain the results from study 2, as the large-amplitude stretches and subsequent dynamic contractions associated with the perturbations in the perturbations PRESENT condition would have broken the stable cross bridges and reset spindle tension (Wilson et al. 1995). The final limitations that must be addressed are the potential effects of ankle bracing in study 1 and auditory feedback in study 2. It is possible that the ankle braces used to control leaning in study 1 provided additional haptic feedback or changed the locus of postural control from the ankle joints to the knees. Similarly, the use of occasional auditory feedback to guide subjects back to a constant ankle angle in study 2 may have caused subjects to adopt a more conscious control of posture than usual. While the potential for each of these issues to influence stretch reflexes is unknown, each limitation applies to only one study, and therefore it is unlikely that either of these factors would contribute to reflex changes observed across both of these studies. Since we did not observe changes in H-reflexes or background EMG, it could be argued that the changes in spindle sensitivity occurred without systematic changes in tonic ␣-motoneuron activity, which would suggest ␣-␥ decoupling. Reduced animal preparations have demonstrated that ␥-motoneuron activity can be influenced by reticulospinal and vestibulospinal projections (cf. Hulliger 1984 for review), both of which are excited in states of fear, anxiety, or arousal (Balaban 2002; Balaban and Thayer 2001). However, direct neural recordings would be required to confirm this hypothesis. Alternatively, the changes in muscle spindle sensitivity could be related to the influence of sympathetic neurons acting directly on muscle spindles (Barker and Saito 1981); sympathetic nervous system activity was increased in the present studies, as indicated by increases in EDA (Venables 1991). These findings, along with others, support the hypothesis that arousal influences muscle spindle sensitivity. Changes in muscle spindle sensitivity have been implicated in increases in Ia afferent firing rate with a kinesthetic attention task (Hospod et al. 2007; Ribot-Ciscar et al. 2009) or mental arithmetic (Ribot-Ciscar et al. 2000). Also, T-reflexes have been shown to be increased independently from H-reflexes in quiescent SOL with tasks that increase heart rate and blood pressure, such as mental arithmetic and isometric handgrip (Kamibayashi et al. 2009). These studies support the theory that balance-relevant proprioceptive inputs can be modulated to suit changes in postural context, such as when the eyes are closed (Aniss et al. 1990) or threat is increased (Davis et al. 2011; Prochazka et al. 1988). These changes have been proposed as a means of increasing “feedback gain and resolution” to supraspinal areas in challenging motor tasks (Llewellyn et al. 1990). However, it is not yet clear which supraspinal centers would use this information, as Davis et al. (2011) demonstrated that threat-induced increases in T reflexes are not associated with larger somatosensory cortical potentials. Therefore, it is more likely that a subcortical system uses this increase in afferent information.



If we presume that people use normal postural sway not just to balance but also to maintain a certain volume or quality of dynamic sensory input (Carpenter et al. 2010; Murnaghan et al. 2011), then increasing sensitivity of balance-relevant sensory sources should mean that less body sway would be required to maintain a requisite volume of sway-generated sensory information. In fact, this is exactly what happens to postural sway (Carpenter et al. 2001) and center of foot pressure (Carpenter et al. 1999; Davis et al. 2009) when people stand in high postural threat scenarios. Likewise, if the volume of sensory information evoked by a perturbation is larger than normal under conditions of postural threat, as would be suggested by an increase in spindle sensitivity, then the postural reaction to the balance perturbation may also be different from normal. This might translate into altered postural responses, as observed when perturbations are evoked at different levels of postural threat (Brown and Frank 1997; Carpenter et al. 2004). It must be acknowledged that our lack of observed changes in H-reflex amplitude with threat (studies 1 and 2) differs from those studies that have reported H-reflex inhibition under conditions of postural challenge (Llewellyn et al. 1990; McIlroy et al. 2003) or threat (Sibley et al. 2007). The most likely explanation for the discrepancy between studies is a difference in the task and/or level of threat. Llewellyn et al. (1990) observed a decrease in the peak H-reflex amplitude, primarily in stance phase of the gait cycle, when subjects walked on a 3.5-cm-wide, 34-cmhigh beam. This observation may be related to changes in gait parameters, known to occur when walking under conditions of increased threat (Brown et al. 2002; Caetano et al. 2009; Tersteeg et al. 2012) as opposed to a direct effect of threat itself. Likewise, Sibley et al. (2007) had subjects stand on a platform that was much lower than that used in study 1 (1.6 m vs. 3.2 m), and they did not restrict leaning or changes in background EMG, both of which may have potentially influenced their results (Stein et al. 2007; Tokuno et al. 2007, 2008, 2009). Alternatively, the ankle braces used in study 1 may have provided additional sensory feedback that prevented H-reflex inhibition that was not available to the subjects in the Sibley et al. (2007) study. However, the same increase in T-reflexes without a change in H-reflexes was observed when we removed the braces for study 2, suggesting that the braces did not contribute to the lack of significant H-reflex changes in these studies. It is also notable that we did not replicate the effects of affective picture-induced arousal on soleus T-reflexes (cf. Bonnet et al. 1995). This is likely due to the fact that we could not induce a change in arousal, as indicated by EDA, across picture conditions. While there were differences in subjective ratings of arousal, these effects did not translate to either EDA or T-reflexes. The failure to elicit increased EDA is in contrast with both Bonnet et al. (1995) and previous work from our research group (Horslen and Carpenter 2011). It is possible that the arousal evoked by the pictures was not sufficiently strong to induce an appreciable difference in EDA across picture conditions on top of the transient changes related to H and T-reflexes. Postural perturbations are known to evoke transient increases in EDA, known as electrodermal responses (Sibley et al. 2008). H and T-reflexes, which cause small postural disturbances and may be somewhat painful, may have overridden any potential picture effects and may therefore have masked the potential changes in T-reflexes related to arousing pictures.

905

In conclusion, we have demonstrated that postural threat, be it from increased consequences or likelihood of a fall, leads to an amplification of tendon stretch reflexes, without any concomitant change in H-reflex amplitude. The most likely explanation for this effect is through an increase in muscle spindle sensitivity, possibly due to fusimotor drive independent of tonic changes to muscle activation state, or increases in sympathetic drive. We propose that this increase in sensitivity might reflect an adaptation strategy meant to satisfy the conflicting needs to restrict movement with threat and to maintain a certain amount of sensory information related to postural control. These data further support the notion that sensory modalities can adapt to different contexts and that automatic behavioral changes seen with threat may be linked to changes in sensory monitoring of postural control. ACKNOWLEDGMENTS These data have been published in part as a thesis by B. C. Horslen. GRANTS The authors thank the Natural Sciences and Engineering Research Council (NSERC) for financial support. DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the author(s). AUTHOR CONTRIBUTIONS Author contributions: B.C.H., C.D.M., J.T.I., R.C., and M.G.C. conception and design of research; B.C.H. and C.D.M. performed experiments; B.C.H. and C.D.M. analyzed data; B.C.H., C.D.M., J.T.I., R.C., and M.G.C. interpreted results of experiments; B.C.H. prepared figures; B.C.H. drafted manuscript; B.C.H., C.D.M., J.T.I., R.C., and M.G.C. edited and revised manuscript; B.C.H., C.D.M., J.T.I., R.C., and M.G.C. approved final version of manuscript. REFERENCES Adkin AL, Frank JS, Carpenter MG, Peysar GW. Fear of falling modifies anticipatory postural control. Exp Brain Res 143: 160 –170, 2002. Adkin AL, Frank JS, Carpenter MG, Peysar GW. Postural control is scaled to level of postural threat. Gait Posture 12: 87–93, 2000. Aniss AM, Diener HC, Hore J, Burke D, Gandevia SC. Reflex activation of muscle spindles in human pretibial muscles during standing. J Neurophysiol 64: 671– 679, 1990. Aniss AM, Gandevia SC, Burke D. Reflex changes in muscle spindle discharge during a voluntary contraction. J Neurophysiol 59: 908 –921, 1988. Balaban CD. Neural substrates linking balance control and anxiety. Physiol Behav 77: 469 – 475, 2002. Balaban CD, Thayer JF. Neurological bases for balance-anxiety links. J Anxiety Disord 15: 53–79, 2001. Barker D, Saito M. Autonomic innervation of receptors and muscle fibres in cat skeletal muscle. Proc R Soc Lond B Biol Sci 212: 317–332, 1981. Bonnet M, Bradley MM, Lang PJ, Requin J. Modulation of spinal reflexes: arousal, pleasure, action. Psychophysiology 32: 367–372, 1995. Both S, Boxtel G, Stekelenburg J, Everaerd W, Laan E. Modulation of spinal reflexes by sexual films of increasing intensity. Psychophysiology 42: 726 –731, 2005. Bradley MM, Lang PJ. Measuring emotion: the Self-Assessment Manikin and the Semantic Differential. J Behav Ther Exp Psychiatry 25: 49 –59, 1994. Brown LA, Polych MA, Doan JB. The effect of anxiety on the regulation of upright standing among younger and older adults. Gait Posture 24: 397– 405, 2006. Brown LA, Gage WH, Polych MA, Sleik RJ, Winder TR. Central set influences on gait. Age-dependent effects of postural threat. Exp Brain Res 145: 286 –296, 2002.


906


Brown LA, Frank JS. Postural compensations to the potential consequences of instability: kinematics. Gait Posture 6: 89 –97, 1997. Burke D, Gandevia SC, McKeon B. The afferent volleys responsible for spinal proprioceptive reflexes in man. J Physiol 339: 535–552, 1983. Caetano MJ, Gobbi LT, Sánchez-Arias MR, Stella F, Gobbi S. Effects of postural threat on walking features of Parkinson’s disease patients. Neurosci Lett 452: 136 –140, 2009. Carpenter MG, Frank JS, Adkin AL, Paton A, Allum JH. Influence of postural anxiety on postural reactions to multi-directional surface rotations. J Neurophysiol 92: 3255–3265, 2004. Carpenter MG, Frank JS, Silcher CP. Surface height effects on postural control: a hypothesis for a stiffness strategy for stance. J Vestib Res 9: 277–286, 1999. Carpenter MG, Frank JS, Silcher CP, Peysar GW. The influence of postural threat on the control of upright stance. Exp Brain Res 138: 210 –218, 2001. Carpenter MG, Murnaghan CD, Inglis JT. Shifting the balance: evidence of an exploratory role for postural sway. Neuroscience 171: 196 –204, 2010. Carpenter MG, Adkin AL, Brawley LR, Frank JS. Postural, physiological and psychological reactions to challenging balance: does age make a difference? Age Ageing 35: 298 –303, 2006. Davis JR, Campbell AD, Adkin AL, Carpenter MG. The relationship between fear of falling and human postural control. Gait Posture 29: 275–279, 2009. Davis JR, Horslen BC, Nishikawa K, Fukushima K, Chua R, Inglis JT, Carpenter MG. Human proprioceptive adaptations during states of heightinduced fear and anxiety. J Neurophysiol 106: 3082–3090, 2011. Dimitriou M, Edin BB. Human muscle spindles act as forward sensory models. Curr Biol 20: 1763–1767, 2010. Gandevia SC, Wilson LR, Inglis JT, Burke D. Mental rehearsal of motor tasks recruits alpha-motoneurones but fails to recruit human fusimotor neurones selectively. J Physiol 505: 259 –266, 1997. Gandevia SC, Burke D. Effect of training on voluntary activation of human fusimotor neurons. J Neurophysiol 54: 1422–1429, 1985. Gregory JE, Mark RF, Morgan DL, Patak A, Polus B, Proske U. Effects of muscle history on the stretch reflex in cat and man. J Physiol 424: 93–107, 1990. Hadoush H, Tobimatsu Y, Nagatomi A, Kimura H, Ito Y, Maejima H. Monopolar surface electromyography: a better tool to assess motoneuron excitability upon passive muscle stretching. J Physiol Sci 59: 243–247, 2009. Hauck LJ, Carpenter MG, Frank JS. Task-specific measures of balance efficacy, anxiety, and stability and their relationship to clinical balance performance. Gait Posture 27: 676 – 682, 2008. Hjortskov N, Skotte J, Hye-Knudsen C, Fallentin N. Sympathetic outflow enhances the stretch reflex response in the relaxed soleus muscle in humans. J Appl Physiol 98: 1366 –1370, 2005. Horslen BC, Carpenter MG. Arousal, valence and their relative effects on postural control. Exp Brain Res 215: 27–34, 2011. Hospod V, Aimonetti J, Roll J, Ribot-Ciscar E. Changes in human muscle spindle sensitivity during a proprioceptive attention task. J Neurosci 27: 5172–5178, 2007. Huffman JL, Horslen BC, Carpenter MG, Adkin AL. Does increased postural threat lead to more conscious control of posture? Gait Posture 30: 528 –532, 2009. Hulliger M. The mammalian muscle spindle and its central control. Rev Physiol Biochem Pharmacol 101: 1–110, 1984. Hunt CC. The effect of sympathetic stimulation on mammalian muscle spindles. J Physiol 151: 332–341, 1960.

Kamibayashi K, Nakazawa K, Ogata H, Obata H, Akai M, Shinohara M. Invariable H-reflex and sustained facilitation of stretch reflex with heightened sympathetic outflow. J Electromyogr Kinesiol 19: 1053–1060, 2009. Llewellyn M, Yang JF, Prochazka A. Human H-reflexes are smaller in difficult beam walking than in normal treadmill walking. Exp Brain Res 83: 22–28, 1990. Macefield VG, Sverrisdottir YB, Wallin BG. Resting discharge of human muscle spindles is not modulated by increases in sympathetic drive. J Physiol 551: 1005–1011, 2003. McIlroy WE, Bishop DC, Staines WR, Nelson AJ, Maki BE, Brooke JD. Modulation of afferent inflow during the control of balancing tasks using the lower limbs. Brain Res 961: 73– 80, 2003. Murnaghan CD, Horslen BC, Inglis JT, Carpenter MG. Exploratory behavior during stance persists with visual feedback. Neuroscience 195: 54 –59, 2011. Nafati G, Rossi-Durand C, Schmied A. Proprioceptive control of human wrist extensor motor units during an attention-demanding task. Brain Res 1018: 208 –220, 2004. Prochazka A, Hulliger M, Trend P, Durmuller N. Dynamic and static fusimotor set in various behavioural contexts. In: Mechanoreceptors: Development, Structure, and Function, edited by Hnik P, Soukup T, Vejsada R, Zelena J. New York: Plenum, 1988, p. 417– 430. Prochazka A, Hulliger M, Zangger P, Appenteng K. “Fusimotor set”: new evidence for ␣-independent control of ␥-motoneurones during movement in the awake cat. Brain Res 339: 136 –140, 1985. Ribot-Ciscar E, Rossi-Durand C, Roll JP. Increased muscle spindle sensitivity to movement during reinforcement manoeuvres in relaxed human subjects. J Physiol 523: 271–282, 2000. Ribot-Ciscar E, Hospod V, Roll J, Aimonetti J. Fusimotor drive may adjust muscle spindle feedback to task requirements in humans. J Neurophysiol 101: 633– 640, 2009. Sibley KM, Carpenter MG, Perry JC, Frank JS. Effects of postural anxiety on the soleus H-reflex. Hum Mov Sci 26: 103–112, 2007. Sibley KM, Mochizuki G, Esposito JG, Camilleri JM, McIlroy WE. Phasic electrodermal responses associated with whole-body instability: presence and influence of expectation. Brain Res 1216: 38 – 45, 2008. Stein RB, Estabrooks KL, McGie S, Roth MJ, Jones KE. Quantifying the effects of voluntary contraction and inter-stimulus interval on the human soleus H-reflex. Exp Brain Res 182: 309 –319, 2007. Tersteeg MC, Marple-Horvat DE, Loram ID. Cautious gait in relation to knowledge and vision of height: is altered visual information the dominant influence? J Neurophysiol 107: 2686 –2691, 2012. Tokuno CD, Carpenter MG, Thorstensson A, Garland SJ, Cresswell AG. Control of the triceps surae during the postural sway of quiet standing. Acta Physiol (Oxf) 191: 229 –236, 2007. Tokuno CD, Taube W, Cresswell AG. An enhanced level of motor cortical excitability during the control of human standing. Acta Physiol (Oxf) 195: 385–395, 2009. Tokuno CD, Garland SJ, Carpenter MG, Thorstensson A, Cresswell AG. Sway-dependent modulation of the triceps surae H-reflex during standing. J Appl Physiol 104: 1359 –1365, 2008. Venables PH. Autonomic activity. Ann NY Acad Sci 620: 191–207, 1991. Wilson LR, Gandevia SC, Burke D. Increased resting discharge of human spindle afferents following voluntary contractions. J Physiol 488: 833– 840, 1995. Wong JD, Wilson ET, Gribble PL. Spatially selective enhancement of proprioceptive acuity following motor learning. J Neurophysiol 105: 2512– 2521, 2011. Wood SA, Morgan DL, Gregory JE, Proske U. Fusimotor activity and the tendon jerk in the anaesthetised cat. Exp Brain Res 98: 101–109, 1994.
