Exploring the Effects of Dopamine on Sensorimotor Inhibition and Mobility in Older Adults

Abstract

Background:

Dopaminergic activity decreases in older adults (OAs) with normal aging and is further reduced in Parkinson’s disease (PD), affecting cortical motor and sensorimotor pathways. Levodopa is the prevailing therapy to counter dopamine loss in PD, though not all PD motor signs improve with levodopa. The purpose of this preliminary study was to explore the effects of levodopa on sensorimotor inhibition, gait and quiet standing in OAs and to investigate the relationships between sensorimotor inhibition and both gait and standing balance both OFF- and ON-levodopa.

Methods:

Fifteen OA males completed a gait, balance and sensorimotor assessments before and one hour after they were given a 100mg dose of levodopa. Short-latency afferent inhibition quantified sensorimotor inhibition. Wearable sensors characterized gait (two-minute walk) and standing balance (one-minute stance).

Results:

No sensorimotor inhibition, gait, or standing balance measures changed from OFF- to ON-levodopa. When OFF-levodopa, worse inhibition significantly related to increased double stance (r=0.62; p=0.01), increased jerkiness of sway (r=0.57; p=0.03) and sway area (r=0.58; p=0.02). While ON-levodopa, worse inhibition related to increased arm swing range of motion (r=0.63; p=0.01) and jerkiness of sway (r=0.53; p=0.04). The relationship between SAI and arm swing excursion significantly changed from OFF- to ON-levodopa (z = −3.05; p = 0.002; 95% confidence interval = −0.95, −0.21).

Conclusion:

Sensorimotor inhibition relationships to both gait and balance may be affected by dopamine in OAs. Cortical restructuring due to the loss of dopamine may be responsible for the heterogeneity of levodopa effect in people with PD and OAs.

1. Introduction

Dopaminergic activity decreases in normal aging, with an exacerbated decrease observed in Parkinson’s disease (PD)(Luo and Roth 2000). In PD, dopaminergic denervation may result in cortical restructuring that alters the roles of neurotransmitters in motor pathways (Bohnen et al. 2018). Evidence from rodent stroke indicate that dopamine has a role in sensorimotor integration in the somatosensory cortex (Obi et al. 2018). Altered motor and sensory pathways following a decrease in dopaminergic activity manifests as gait and balance impairments in people with PD. Levodopa, a dopamine replacement medication, is the prevailing therapy to treat motor signs and symptoms of PD. However, not all PD motor signs and symptoms improve with levodopa. Some components of gait and standing balance deteriorate from OFF- to ON-levodopa states in people with PD (Curtze et al. 2015; Wilson et al. 2020), but the reasons that levodopa negatively impacts some motor functions remain unclear.

Short-latency afferent inhibition (SAI), a surrogate for sensorimotor inhibition, was initially linked to cholinergic activity because the cortical inhibition quantified by SAI improved with a cholinergic agonist (Di Lazzaro et al. 2000). Other studies subsequently observed effects of gamma aminobutyric acid agonists and levodopa on SAI (Di Lazzaro et al. 2002; Sailer et al. 2003; Di Lazzaro et al. 2005). SAI worsens in the ON-levodopa state in people with PD (Sailer et al. 2003). Cumulatively, the implication is that SAI results from complex neurophysiological interactions controlling inhibition in the sensorimotor pathway. The reduction of sensorimotor inhibition in the ON-levodopa state may contribute to the observed heterogenous effects of levodopa on gait and balance in PD.

Although gait speed improves with dopamine in PD, postural stability during walking (double support time) and standing balance (sway area, frequency, smoothness) are unaffected or worsen with dopaminergic medication in PD (Curtze et al. 2015; Wilson et al. 2020). Both gait and standing balance relate to the amount of sensorimotor inhibition quantified by SAI in people with PD, older adults (OA), and OA fallers (Rochester et al. 2012; Pelosin et al. 2016; Martini et al. 2020; Martini et al. 2021). Specifically, abnormal postural sway or slower gait are associated with worse SAI is in these populations (Rochester et al. 2012; Pelosin et al. 2016; Martini et al. 2020; Martini et al. 2021). Despite established dopaminergic decreases with aging, the effect of levodopa on these relationships is not known (Luo and Roth 2000). Establishing the relationships among levodopa, sensorimotor inhibition, and mobility in OAs could provide greater insight into the heterogeneous effects of levodopa on mobility in PD.

The purpose of this exploratory investigation was to examine the effects of levodopa on sensorimotor inhibition and mobility in OAs. Specifically, we aimed to investigate the effects of levodopa on sensorimotor inhibition, gait, and quiet standing balance in OAs. Further, we aimed to explore the relationships between sensorimotor inhibition and both gait and standing balance in OAs OFF- and ON-levodopa. We hypothesized that sensorimotor inhibition, gait stability and standing balance will be worse, while other characteristics of gait will remain unchanged from the OFF- to ON-levodopa state, consistent with the aforementioned heterogeneous effect of levodopa on gait in the PD literature. Finally, we hypothesize that worse gait stability and standing balance measures will correlate to worse sensorimotor inhibition.

2. Methods

The Oregon Health & Science University Institutional Review Board approved this study. All participants reviewed the study purpose and procedures involved before signing the informed consent.

2.1. Participants

Fifteen healthy OA males were recruited from an ongoing Pacific Udall Center project at Oregon Health & Science University. Participants were screened for TMS eligibility before enrollment. Inclusion criteria included the ability to stand unsupported for 30 seconds. Exclusion criteria included: inability to walk for two minutes without an assistive device, any TMS contraindication, any musculoskeletal injury that would affect mobility, any neurological disorder, a failed medical screen (completed by JGN) for taking levodopa, or any cholinergic mediation. A single 100/25 mg dose of levodopa/carbidopa was given to each participant after completing the “OFF-levodopa medication state” assessment. A single 100 mg dose of levodopa was selected based on previous evidence that the same dose induced cognitive changes in younger and older adults (Linssen et al. 2014; Vo et al. 2018). Compounding this with the known relationship between worse cognitive and gait performance, a 100 mg dose was deemed large enough for an effect while mild enough to prevent adverse reactions (Verghese et al. 2008; Amboni et al. 2013; Verlinden et al. 2014; Cohen et al. 2016).

Each participant completed the following clinical assessments: Montreal Cognitive Assessment (MoCA; range = 0-30), Activities Balance Confidence (ABC; range = 0-100) scale, and the Falls Efficacy Scale (FES; range = 0-100). The participants completed the MoCA prior to the OFF-levodopa state sensorimotor inhibition and mobility assessments. Upon completing the OFF-levodopa state assessments, participants were given their single 100/25 mg tablet (levodopa/carbidopa) with carbonated water and an option of fruit. Participants waited an hour before starting the “ON-levodopa state” assessments. During the wait, participants completed the ABC and FES scales. After one hour, participants completed the ON-levodopa state sensorimotor inhibition and mobility assessment in the same order as the OFF-levodopa state assessment.

2.2. Transcranial Magnetic Stimulation (TMS)

TMS of the motor cortex was performed with a Magstim 200 (Magstim Co.). A figure-of-eight coil (external loop diameter of 9cm) was positioned over the hemisphere associated with the dominant hand. Motor-evoked potentials (MEPs) were recorded from the first dorsal interosseous muscle through disposable, Ag/AgCl surface electrodes. Samples were amplified (gain: 2000) and bandpass filtered (100Hz-5kHz) using BIOPAC MP150 system (BIOPAC Systems, Inc). Resting motor threshold was determined as the percentage of the minimum stimulator output to elicit an MEP of 50μV in five out of ten trials.

2.3. Short-Latency Afferent Inhibition

SAI was performed using a modified version of a protocol previously described (Martini et al. 2020; Martini et al. 2021). Participants were instructed to remain at rest, sitting as still as possible, and refrain from keeping their eyes closed. A peripheral, electric conditioning stimulus was applied over the median nerve followed by the central test stimulus, TMS. The intensity of the conditioning stimulus was set at the amplitude required to elicit a visible twitch of the first dorsal interosseous muscle. For the purpose of this investigation, we used 20ms as the N20 latency across all participants (Martini et al. 2020; Martini et al. 2021). The interstimulus intervals (ISI) were randomly applied from N20+0ms to N20+5ms, in 1ms increments. Ten trials were collected and the conditioned peak-to-peak MEP magnitudes were averaged for each ISI. The grand mean of the ISIs is expressed as the percentage of the unconditioned MEP magnitudes. A SAI grand mean around 100% indicates minimal or no sensorimotor inhibition (worse), as this indicates that the peripheral nerve stimulation ascending signal provides limited inhibition of the descending motor signal.

2.4. Mobility Assessment

Inertial sensors (Opals, APDM Inc.) were placed on each wrist and foot, around the waist, and over the sternum to characterize gait and standing balance using Mobility Lab software (APDM, Inc) (Mancini et al. 2011). Gait was characterized during a two-minute walk, back-and-forth over a seven-meter path, requiring 180 degree turns at the ends of the marked path. Participants were instructed to walk at a comfortable, self-selected pace. Gait variables of interest were double limb support (% of gait cycle), speed (meters/second [m/s]), stride time (s), stride length (m), arm swing excursion (degrees [d]), and turn velocity (d/s). Standing balance was characterized while participants stood quietly for one minute looking straight ahead with foot position standardized by a template. The balance variables of interest were jerkiness of sway, root mean square (RMS) relative to the mean sway, sway velocity, and sway area.

2.5. Statistical Analyses

Data were inspected for normality using histograms and the Kolmogorov-Smirnov test. Balance variables were not normally distributed, so they were natural log transformed to improve normality. Paired samples t-tests were used to compare medication state differences in SAI, gait and balance variables. Pearson’s correlations assessed the relationships between SAI and mobility variables. No corrections for multiple comparisons were run for these correlations. The magnitude of the correlation coefficient and the scatter plots are included to provide greater detail about these relationships. For significant relationships between SAI and mobility variables, the magnitude of Pearson’s correlations were statistically compared using the R statistics package cocor and Fisher’s z transformation of overlapping variables (e.g. OFF and ON-levodopa) (Meng et al. 1992; Diedenhofen and Musch 2015). Due to the sample size, Cohen’s d effect sizes were calculated to characterize the magnitude of the medication effect and were interpreted as weak (<0.50), moderate (0.50-0.79) or strong (≥0.80) (Cohen 1977). Alpha was set a priori to p < 0.05. IBM SPSS version 27 was used for statistical analyses. Data are presented as mean(standard deviation), unless otherwise indicated.

3. Results

3.1. Participants

The 15 OA males that participated in this study were 65.8(6.5; minimum-maximum 52-71) years-old, with an average height of 175.1(9.4) cm and mass of 81.1(7.8) kg. The average MoCA score was 27.8(1.5) with a minimum-maximum of 26-30, indicating no cognitive impairment. The average FES was 10.3(0.7) and the ABC was 98.3(1.9), indicating no fear of falling or lack of balance confidence, respectively. No participant reported an adverse event as a result of ingesting a single 100/25 mg dose of levodopa/carbidopa.

3.2. SAI in the OFF- versus ON-levodopa states

There was no significant difference in SAI between the OFF- and ON-medication states (t = 0.56; p = 0.59; Cohen’s d = 0.14). The average OFF-levodopa state SAI was 66.5(17.3)% and the average ON-levodopa state SAI was 64.4(16.3)% (Figure 1). Average SAI values in both OFF- and ON-levodopa states for these OAs are similar to previously reported values in OAs.(Rochester et al. 2012; Pelosin et al. 2016; Martini et al. 2020; Martini et al. 2021)

Figure 1:

Levodopa Effects on SAI. Box and scatter plot of SAI during OFF and ON levodopa states. Grey dashed lines indicate the direction of change in SAI for individual. Black dashed line indicates group mean.

3.3. Gait and Standing Balance Differences Between Medication States

Levodopa had no significant effect on gait or postural sway variables (Table 1). Though not significant, arm swing excursion had a moderate effect size (Cohen’s d = 0.50).

Table 1:

Mobility

		OFF	ON	Cohen’s d
Gait	Double Support (%)	19.1(2.6)	19.1(2.7)	0.03
	Stride Time (s)	1.1(0.1)	1.1(0.1)	0.34
	Gait Speed (m/s)	1.2(0.1)	1.2(0.1)	−0.12
	Stride Length (m)	1.3(0.1)	1.3(0.1)	0.09
	Arm Swing (d)	48.6(9.0)	45.8(10.1)	0.50
	Turn Velocity (d/s)	184.8(33.6)	179.2(31.0)	0.22
Postural Sway	Jerk	−2.3(0.8)	−2.2(0.7)	−0.11
	RMS	−2.6(0.5)	−2.6(0.4)	−0.01
	Velocity	−1.4(0.6)	−1.4(0.7)	−0.002
	Sway Area	−6.0(0.8)	−5.9(0.8)	−0.19

Mean(standard deviation). Postural sway variables are natural log transformed.

3.4. Correlations Between SAI and Mobility

Correlation analyses that explored the relationships between mobility and SAI within and between the OFF- to ON-levodopa states yielded interesting results. Worse SAI in the OFF-levodopa state significantly related to increased double limb support, jerkiness of sway, and sway area (Figure 2). SAI OFF was not significantly related to OFF: gait speed (r = −0.34), stride time (r = 0.24), stride length (r = −0.19), arm swing excursion (r = 0.16), turn velocity (r = −0.23), RMS Sway (r = 0.38), or sway velocity (r = 0.19). Worse SAI in the ON-levodopa state significantly related to increased arm swing excursion and jerkiness of sway (Figure 2). SAI ON was not significantly related to ON: double limb support (r = 0.45), gait speed (r = −0.12), stride time (r = 0.19), stride length (r = −0.01), turn velocity (r = 0.04), RMS Sway (r = 0.14), sway velocity (r = −0.27), or sway area (r = 0.23). There were no significant relationships between the change in SAI and the change in gait or balance measures from OFF- to ON-levodopa states. The relationship between SAI and arm swing excursion significantly changed from OFF- to ON-levodopa (z = −3.05; p = 0.002; 95% confidence interval for the difference in Pearson’s r = −0.95, −0.21).

Figure 2:

Top Row: Scatter plots highlighting the Pearson correlations between SAI and gait during the OFF (blue) and ON (red) levodopa states. Bottom Row: Scatter plots highlighting the Pearson correlations between SAI and standing posture during the OFF (blue) and ON (red) levodopa states. Sway data presented as natural log transformed data. Solid gray line represents the best fit line.

4. Discussion

The goal of this study was to quantify gait, standing balance, and sensorimotor inhibition (SAI) in OAs in OFF- and ON-levodopa states. We aimed to explore the impact of levodopa on mobility and SAI in healthy OAs to gain insights into the impact of dopamine on mobility and sensorimotor inhibition. Unlike findings in PD cohorts (Sailer et al. 2003), our preliminary findings suggest that levodopa has no effect on either mobility or SAI in OAs. However, levodopa-dependent relationships were observed between mobility and SAI. In the OFF-levodopa state, worse SAI significantly correlated with worse performance for specific aspects of gait and balance. Since increased double limb support reflects a decreased ability to maintain balance (Maki 1997; LaRoche et al. 2014), the OFF-levodopa state SAI-mobility relationships reflect a role for sensorimotor inhibition in gait stability of OAs. This early observation is mirrored for OFF-levodopa state balance, with worse stability (e.g. increased jerkiness of sway and sway area) related to worse sensorimotor inhibition. Unlike the OFF-levodopa state, the ON-levodopa state resulted in significant relationships between worse SAI and increased arm swing excursion (gait) and worse jerkiness of sway (balance).

The increased arm excursion with worse SAI is interesting, as the worse sensorimotor inhibition indicated by higher SAI may result in larger arm excursion while walking, but only when ON-levodopa. This observation is particularly interesting because levodopa can result in large, dyskinetic arm excursion while walking in patients with PD (Curtze et al. 2015). Further, the relationship between SAI and arm swing excursion OFF-levodopa significantly changed during the ON-levodopa state, and there was a moderate effect of medication on arm swing excursion (Cohen’s d = 0.50). The current study implemented standard TMS methods by targeting the dominant upper extremity to measure SAI as an indicator of sensorimotor inhibition. This may have contributed to the strong relationships between SAI and arm swing. While targeting a lower extremity muscle could be more relevant to other gait and balance metrics, there are methodological complications to this approach. Rather than the planar figure-of-eight coil used here, a different coil configuration (e.g., a butterfly coil) would be required for deeper penetration of the magnetic field into the lower extremity region of motor cortex (Ueno and Sekino 2021). In addition, the increased target depth would require a higher magnitude stimulation, thereby reducing the focality of the stimulus, activating a broader area of cortex and more muscles and potentially contributing to discomfort among participants. Consistent with prior work, the current findings demonstrate that upper extremity SAI relates to a breadth of gait variables specific to the lower extremity (Rochester et al. 2012; Pelosin et al. 2016; Martini et al. 2020; Pelosin et al. 2020; Martini et al. 2021). A potential avenue for future work would be to examine whether lower extremity SAI is similarly related to balance and gait variables.

There were no changes in gait or balance performance after taking a 100mg dose of levodopa. In contrast, three 100mg doses of levodopa over a 24-hour period improved fine motor skill speed in healthy OAs while producing no change to healthy young adult fine motor skill (Floel et al. 2008). The observed fine motor skill improvement in OAs suggests that exposure to levodopa can improve fine motor performance in healthy OAs (Floel et al. 2008). The dosage used to improve fine motor skill was selected based on the minimum levodopa dose to induce cognitive change in the subjects (Floel et al. 2008). However, more recent studies established that a single, 100mg dose of levodopa induced cognitive (i.e. learning and memory) deficits in both young and older adults (Linssen et al. 2014; Vo et al. 2018). The difference in effects of brief levodopa exposure on cognitive versus motor performance could be related to the different dopaminergic pathways involved in cognition and mobility. While the nigrostriatal pathway (substantia nigra to striatum) plays a role in motor function, the mesolimbic/cortical dopaminergic pathway (ventral tegmental area to amygdala/hippocampal and frontal cortex) likely plays a role in cognition (Alcaro et al. 2007; Bourdy et al. 2014). Additional research is needed to determine why a single dose of levodopa appears to differentially affect the two dopaminergic pathways in healthy OAs.

Contrary to the SAI literature in PD (Sailer et al. 2003), there appears to be no effect of levodopa on SAI in OAs without PD. The nigrostriatal dopaminergic pathway in OAs is largely intact and therefore may be less impacted by a single or by multiple doses of levodopa, while dopaminergic nigrostriatal pathway impairment is related to the motor impairments in people with PD. Further, sensorimotor inhibition, quantified by SAI, does not appear to be affected by age, but is affected by faller status and disease (Rochester et al. 2012; Young-Bernier et al. 2014; Pelosin et al. 2016; Pelosin et al. 2020; Martini et al. 2021). The OAs in this study had no fear of falling or lack of balance confidence, suggesting that they are not fallers, therefore functional sensorimotor inhibition. Although SAI values in this group were similar to those reported in several previous studies (Rochester et al. 2012; Pelosin et al. 2016; Martini et al. 2020; Martini et al. 2021), SAI values reported here, both off and on levodopa, were higher (worse) than the values reported in some previous studies (Young-Bernier et al. 2014). Compared to this previously reported OA group, the roughly 20% increase in SAI observed the OA group in this study could be an early indicator of motor impairment onset, particularly as motor impairment relates to falls. In fact, older adult fallers have significantly worse SAI than older adult non-fallers, but any SAI threshold for transition to faller status remains unknown (Pelosin et al. 2016). A longitudinal study quantifying SAI, mobility, and falls over longer periods, such as a year, may be needed to determine if worsening SAI in older adults is an early predictor of motor impairment or falls.

To date, only gait characteristics in the pace domain significantly related to sensorimotor inhibition in OAs (Pelosin et al. 2016; Martini et al. 2021). Specifically, gait speed and stride length were slower and shorter, respectively, with worse SAI in OAs without the presence of levodopa (Martini et al. 2021). The results herein expand on these findings with the observations that SAI was significantly related to spatiotemporal gait characteristics and postural sway characteristics, both in the OFF- and ON-levodopa states in healthy OAs. While no gait pace domain variable significantly related to SAI in this OA cohort, dynamic (gait) and quasi-static (quiet standing) stability variables were significantly related to SAI, such that worse SAI related to worse stability during gait and standing balance. The different relationships between mobility and SAI from OFF- to ON-levodopa states of the OAs herein could portend different relationships between SAI and mobility in the OFF-levodopa state for people with PD. To date, the relationships between SAI and mobility in people with PD are reported in the ON-levodopa state, limiting the ability to differentiate the effects of disease and drug on the relationships between SAI and mobility in people with PD. The identification of these different relationships could explain some of the heterogeneity in gait and balance responsiveness to levodopa in PD. Importantly, the significance of the observed relationships between SAI and mobility should be interpreted in light of the study limitations.

Limitations to this study relate to the sample and the dosing of levodopa. The small sample size could limit statistical power and increase the risk for Type II errors in levodopa effect. The small sample, in combination with the relatively homogeneous sample, contributed to the decision not to adjust for covariates in correlational analyses. The lack of variability in MoCA score and limited distinguishing effects of age on dopaminergic degeneration for people over 50 minimized the need to statistically control for cognition or age (Volkow et al. 1996). There were no females in this study, limiting the generalizability of these findings. Participants were all male, and did not vary greatly in terms of age, cognition, balance confidence or fear of falling. Building on these preliminary findings, future work should include larger samples, including both genders and with a greater range of age, cognitive abilities, and mobility. A single, 100/25 mg dose of levodopa/carbidopa may not have been strong enough to induce neurophysiological changes in some OAs without a known nigrostriatal dopaminergic deficit. However, this design was previously used with success in establishing changes in cognitive performance in both young and older adult populations (Linssen et al. 2014; Vo et al. 2018). Particularly if the relationship between gait and sensorimotor inhibition is mediated by such cognitive changes, a larger dose of levodopa could be needed to elicit effects of levodopa on mobility and sensorimotor inhibition in OA. The current study used a limited characterization of cognition. A comprehensive cognitive assessment could determine if levodopa has similar effects on cognition as shown previously. The ON-levodopa state was not clinically confirmed with the Movement Disorders Society’s Unified Parkinson’s Disease Ratings Scale Motor Examination; however, it is unclear if this measure, which was designed for people with PD, would be sufficiently sensitive to motor changes in response to levodopa in OA. The preliminary nature of this experimental study also contributed to methodologic limitations, such as the lack of a placebo and subsequent randomization and the lack of a control group. Finally, using neuronavigation for the TMS procedure could reduce MEP variability, resulting in greater consistency between TMS sessions (OFF- and ON- levodopa) (Julkunen et al. 2009).

5. Conclusion

A single, 100mg dose of levodopa does not appear to affect cortical sensorimotor inhibition or mobility in OA males. However, sensorimotor cortical inhibition is related to balance and gait performance in OAs, and these relationships appear to be affected by levodopa. Future investigations that implement a double-blind, crossover design with a placebo and larger single dose or protracted period of levodopa dosing, in a larger population of both older and young adults is the clear next step to explain the heterogeneity of balance and gait performance in OAs and the heterogeneous effect levodopa has on gait and balance in PD. The results of this study lay the foundation to further probe the effects of dopaminergic denervation on motor function in aging populations.

Data Availability Statement:

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.