Eight participants diagnosed with idiopathic PD (mean age 67.8 ± 4.7 years; 3 males and 5 females), and 11 age- and gender-matched HOAs (mean age 68.7 ± 5.0 years) completed this study (Table 1). Participants with PD were recruited via the Iowa State University Alternative Medicine and Music for PD Lab Database, which consists of a list of individuals diagnosed with PD who have indicated interest in research opportunities. HOAs were recruited via word of mouth. The primary inclusion criterion for this group was being between the ages of 50 and 80 years (the average age range of a person with PD) (Wirdefeldt et al., 2011). Additional inclusion criteria for participants with PD included a diagnosis of PD by a neurologist. All participants were excluded if they demonstrated any cognitive impairment (Mini-Mental Status Exam < 25), had severe hearing loss, been diagnosed with a mental disorder besides anxiety or depression, had any musculoskeletal disorders or any other brain-related conditions, used tobacco, illicit drugs, or excessive amounts of alcohol, and had either a systolic blood pressure (SBP) above 140 mmHg or a diastolic blood pressure above 90 mmHg during the initial screening. One HOA was excluded during the initial blood pressure screening to avoid any potential cardiac events during the SECP task. All participants provided written informed consent. All procedures were approved by the Iowa State University Institutional Review Board (approval #: 15–499) and were performed in accordance with relevant guidelines and regulations.

Participants were scheduled for two visits within a 1-week timespan. During the initial visit, participants were screened for cognitive impairment with the Mini-Mental Status Exam (MMSE) and high blood pressure using an automated Omron blood pressure cuff. Participants then provided demographic information and completed a battery of questionnaires that are reliable and valid measures for persons with PD, including the Geriatric Depression Scale (GDS) (Ertan et al., 2005) and the State and Trait Anxiety Inventory (STAI) (Yang et al., 2019). The Perceived Stress Scale (PSS) was also collected. While this scale has not been tested for its psychometric properties in persons with PD, it has been shown to be reliable and valid in older adults (Jiang et al., 2017; Lee, 2012). Participants with PD also completed the Movement Disorders Society-Unified Parkinson’s Disease Rating Scale (MDS-UPDRS) (Goetz et al., 2008). The same trained rater completed the collection and scoring of the MMSE and MDS-UPDRS across all participants and visits. Table 1 shows the means and standard deviations of demographic and questionnaire data.

During visit 1, participants were informed about the dietary and medical restrictions they were requested to follow before their second visit. These included limiting their caffeine intake to one cup (8 oz.) of a caffeinated beverage finished at least 3 h before the start of the second visit and refraining from alcoholic beverages during the 24 h before the start of the second visit. If they were non-compliant with any of these restrictions, participants were asked to reschedule. All participants were also asked to eat 1–2 h before the second visit. Participants with PD were asked to take their normal PD medication 1 h before the beginning of the second visit. The second visit was scheduled to align with their normal medication times. The second visit’s starting times were scheduled between 12 PM and 3 PM.

During the second visit, adherence to the dietary and medical restrictions and recommendations was first reviewed to ensure compliance. Then, an electroencephalogram (EEG) and electromagnetic sensors (Ascension trackSTAR, Victoria) were fitted on the participant. Electromagnetic sensors were placed on the tip of the index finger, the metacarpophalangeal joint of the index finger, and the metacarpophalangeal joint of the pinky finger on the most affected side (Figure 1B). Next, a salivary cortisol sample was collected, and then the participants completed either the physical stress intervention (SECP task) or the control intervention (warm water hand bath). The order was counterbalanced. After completing either the SECP or control task, blood pressure was obtained with an automated Omron blood pressure cuff. Then, participants verbally reported how stressful they thought the task was on a Likert scale from 1 to 10, with respective response anchors of “not stressful at all” and “most stressful thing I can imagine.” Cortisol, blood pressure, and perceived stress were measured to assess the effectiveness of the stress and control interventions. EEG data were then collected during a paired click paradigm to examine inhibitory gating. Within 5 min of completing the SECP and control tasks, participants completed the bradykinesia and tremor tests. Twenty-five minutes after completing the paired click paradigm, a second salivary cortisol sample was collected from each participant. Following this, participants rested or read a magazine or a book for half an hour. After the rest period, participants completed the same series of steps and tasks again with the uncompleted intervention (i.e., warm water bath or SECP). See Figure 1A for the order of first and second visit procedures.

The physical stress intervention was a SECP task in which the participants placed their hands up to the wrist in an ice water bath (2–4° C) for 90 s. Participants were also made aware that this task was being video recorded and that their facial expressions would be judged by trained experimenters at a later time. During the control condition, participants placed their hand in warm water (36–38°C) for 90 s, and participants were reminded that they were not being videotaped, and the camera was moved out of sight.

Participants completed 80 trials of the paired click paradigm to examine inhibitory gating. Inhibitory gating is typically evaluated using a paired click paradigm where pairs of simple auditory stimuli elicit the EEG p50 auditory components (Lijffijt et al., 2009). During the paired-click paradigm, the first click is referred to as S1 and is followed shortly thereafter by the second click, referred to as S2. The most common inhibitory gating measure is the S2/S1 ratio of the p50 component (p50 ratio) (Boutros and Belger, 1999; Guterman et al., 1992; Jones et al., 2016). The normal gating response is indicated by a low S2/S1 ratio, with a large amplitude response to S1 and a reduced amplitude response to S2.

During this task, the participants were sitting comfortably in a semi-reclined armchair in a quiet, well-lit room. Participants were instructed to keep their movements to a minimum and their eyes open and focused on a piece of tape stuck to a desk 5 feet in front of them. Participants were also instructed to refrain from falling asleep.

To collect the potentials, a 64-electrode EEG cap was fitted according to the international 10–20 system (Biosemi, Amsterdam, Netherlands). Eye movements were monitored with electrodes placed above and below the right eye for filtering purposes. A single common reference electrode was placed over the mastoid process ipsilateral to the participant’s most affected side (HOAs were matched), and the ground electrode was placed on the forehead just above the nose and between the eyebrows. The electrode of interest was Cz, which is located in the midline at the ‘vertex’ or top of the head. Impedances were checked to make sure they are below 5 kohms before beginning the data collection. Signals were recorded at 2 kHz and amplified. During the task, auditory clicks were presented through speakers at an 80 dB hearing level. The auditory stimuli were pairs of identical auditory clicks (1,000-Hz tone) with a duration of 20 ms and an interval of 500 ms between clicks. Eighty trials were presented with a 7-s interval between trials. The time for each recording epoch was 300 ms before the first pair of clicks until 1 s following the second click. The auditory tones were collected simultaneously by the EEG data collection system (The Motion Monitor, Innsport, Chicago) to ensure synchronization. Overall, these are the recommended methods for collecting the p50 response (Dalecki et al., 2011).

The EEG data were processed using standard methods from the Krigolson Laboratory¹. Signals were re-referenced offline using a bipolar montage (Cz-Iz), baseline corrected, and filtered using a 0.5–45 Hz 4th-order dual-pass Butterworth band-pass and a 60-Hz notch filter. Next, the data were divided into smaller epochs (−200 ms to 300 ms) around each auditory tone. Finally, independent component analysis followed by inspection of the data was used to identify artifacts, any trials with voltage on any channel exceeding a 10 μV/ms gradient, and an absolute voltage difference >100 μV. These trials were eliminated, resulting in 65–70 epochs per participant entered for analysis.

The P50 ratio (S2/S1) was the measure of habituation (Dalecki et al., 2011). S1 (peak-to-peak method) was the difference in amplitudes in the positive deflection between 35 and 65 ms and the most negative deflection between 15 and 45 ms of the first stimulus. S2 was the difference of those amplitudes for the second stimulus. In addition, latency measures were also calculated for S2 and S1, where the latency value reflected the timing difference between the maximum positive deflection and the auditory stimulus. A clear p50 component was interpretable in both the S1 and S2 in both conditions (SECP and control) for 19 participants (8 PD and 11 HOAs).

Participants were seated comfortably in a semi-reclined armchair in a quiet, well-lit room. Select MDS-UPDRS motor tests for bradykinesia (sections 3.4 finger tapping, 3.5 hand movements, 3.6 pronation-supination movements of hands) and tremor (3.15 postural tremor of the hands, and 3.17 rest tremor of the hands) were examined in the same order for all participants. Only seated tests were chosen so that data could be captured in a timely manner after the SECP condition. In addition, rigidity was not collected as the lab was not equipped with the methodology to measure rigidity. Standard instructions for these motor tests were provided, with the modification that participants were instructed to complete each of the tests for 10 s to allow for the collection of kinematic data. For participants with PD, the least affected side was placed in the water (SECP or warm water control), so data were only recorded on the most affected side. For HOAs, they were matched to participants with PD and used the same counterpart for data collection.

For all movement tests, the primary outcome measures were movement rate and amplitude. Movement rate was measured in cycles per second (Hz), with the end of each cycle marked by: (1) when the index finger reached a minimum on the y-axis (i.e., touched the thumb for the finger tapping task or palm of the hand) for finger tapping and hand movements (Figure 1C) or (2) reached the maximum rotation when pointing the palm to the floor (pronation-supination of the hand) (Figure 1C). For finger tapping and hand movements, amplitude was analyzed by examining the magnitude of movement of the index finger in relation to the metacarpophalangeal joint of the index finger. For pronation-supination movements of hands, the amplitude of the movement was analyzed by examining the degree of rotation between the metacarpophalangeal joints of the index and pinky fingers. For the tests of postural and resting tremor, the amplitude of tremor was measured by examining the magnitude of tremor frequency movements (3–8 Hz) using the tip of the index finger. The largest tremor seen during the tremor recording periods was recorded as the maximum amplitude. To measure interruptions, decrements in amplitude, and slowing, variability of movement was examined by calculating the coefficient of variation (CV; standard deviation/mean) for each of the outcome measures with the exception of max amplitude.

On the day of collection, salivary cortisol samples were stored in a −20°C freezer within 30 min. Cortisol was analyzed with the Salimetrics^® Cortisol Enzyme Immunoassay Kit (RRID AB_2801306). The kit uses a competitive immunoassay in which cortisol competes with cortisol conjugated to horseradish peroxidase for the antibody binding sites. To determine the cortisol reactivity, the area under the curve was calculated using the baseline and the 25-min post intervention samples for each condition.

An independent samples t-test was used to examine any group differences on demographic and questionnaire outcome measures. A 2 condition (SECP, control) × 2 group (PD, HOA) repeated measures ANOVA was completed for measures of stress (perceived stress, blood pressure, and cortisol) to confirm that the SECP task initiated an increase in stress for participants. To test the hypothesis that the SECP task impaired inhibitory gating (higher p50 ratios) in both persons with PD and HOAs, a 2 condition (SECP, control) × 2 group (PD, HOA) repeated measures ANOVA was used to determine differences in inhibitory gating (p50 ratio). To test the hypothesis that the SECP task would negatively impact bradykinesia and tremor in persons with PD but not in HOAs, a group (PD, HOA) × 2 (SECP, control) repeated measures ANOVA was used to determine differences in each tremor (amplitude, amplitude CV, and max amplitude) and upper extremity bradykinesia (movement rate, movement rate CV, amplitude, and amplitude CV) kinematic outcome measure. For all repeated measures ANOVAs, partial eta squared (η_p²) effect sizes were calculated. Significance was set at α = 0.05. For all post hoc analyses when appropriate, a Bonferroni correction was used to interpret statistical significance.

A repeated measures Pearson’s correlation (Bakdash and Marusich, 2017) was used to examine the relationship between the p50 ratio and bradykinesia and tremor during the control and stress conditions. Repeated measures correlations were used to examine within-individual associations for measures assessed on two or more occasions for multiple individuals. Repeated measures correlations do not violate the assumption of independence of observations, and estimate the association shared among individuals. Repeated measures correlation uses an analysis of covariance (ANCOVA) to adjust for inter-individual variability by removing between-participants variance and provides the best linear fit for each participant using a parallel regression line with varying intercepts. Similar to the Pearson correlation coefficient (r), repeated measures correlation is bounded by −1 to 1 and represents the strength of the linear association between two variables. Significance was set at α = 0.05.

Although participants with PD have higher scores than HOAs on the mood questionnaires, only the STAI2 reached significance. Participants with PD had higher trait anxiety than the HOAs [t₍₂₈₎ = 3.086, p = 0.005, Mean Difference (MD) = +9.2, d = 1.126]. No demographic variables were found to be significantly different (Table 1).

For perceived stress, both participants with PD and HOAs similarly rated the SECP condition more stressful than the control condition. Results revealed a main effect of condition [F_(1,28) = 147.393, p < 0.001, η_p² = 0.840]; the SECP task was rated higher than the control condition (MD = +5.2, d = 2.256). There was no main effect of group [F_(1,28) = 0.267, p = 0.609, η_p² = 0.009], or an interaction effect for group x condition [F_(1,28) = 0.006, p = 0.938, η_p² = 0.000].

Blood pressure was elevated similarly for participants with PD and HOAs after the SECP condition compared to the control condition. There was a main effect of condition for both systolic [F_(1,28) = 8.003, p = 0.009, η_p² = 0.222] and diastolic blood pressure [F_(1,28) = 5.543, p = 0.026, η_p² = 0.165]. The participants had higher systolic blood pressure (MD = +6.4 mmHg, d = 0.518) and diastolic blood pressure (MD = +2.8 mmHg, d = 0.430) after the SECP is compared to the control condition. There was no main effect of group for systolic blood pressure [F_(1,28) = 2.138, p = 0.155, η_p² = 0.071] or diastolic blood pressure [F_(1,28) = 0.360, p = 0.553, η_p² = 0.013]. There were no interaction effects of group x condition for systolic blood pressure [F_(1,28) = 0.889, p = 0.354; η_p² = 0.031] or diastolic blood pressure [F_(1,28) = 0.767, p = 0.388; η_p² = 0.027].

Similar to the other stress outcome measures, cortisol was similarly raised in both participants with PD and HOAs after the SECP condition was compared to the control condition. There was a main effect of condition [F_(1,28) = 6.780, p = 0.015 η_p² = 0.195]; cortisol area under the curve (AUC) values were higher after the SECP was compared to the control condition (MD = +13.4 ng/dL/h, d = 0.479). There was no main effect of group [F_(1,28) = 0.187, p = 0.699, η_p² = 0.007] and no interaction effects for group x condition [F_(1,28) = 0.596, p = 0.447, η_p² = 0.021].

No differences in inhibitory gating between the participants with PD and HOA were revealed, but inhibitory gating was reduced in both participants with PD and HOA after the SECP condition was compared to the control condition. Figure 2A shows the results for inhibitory gating, and Figure 2B shows an example of the EEG waveform from one person with PD in both conditions. Results revealed a main effect of condition [F_(1,17) = 12.813, p = 0.002, η_p² = 0.430]. In general, there was less inhibitory gating following the SECP condition compared to the control condition (MD = +0.20, d = 0.760). No main effect of group [F_(1,17) = 0.736, p = 0.403, η_p² = 0.042] or interaction effects for condition x group [F_(1,17) = 1.971, p = 0.178, η_p² = 0.104] were found.

Overall, for finger tapping, no differences in movement rate or movement amplitude were revealed between participants with PD and HOAs, but amplitude variability (i.e., CV) was significantly greater in participants with PD than those with HOAs. After the SECP, amplitude CV increased in persons with PD and decreased in HOAs. Additionally, amplitude significantly decreased in both groups after the SECP. Figure 3A shows the results for finger tapping amplitude, and Figure 3B shows the results for finger tapping amplitude CV. A main effect of condition on finger tapping amplitude [F_(1,27) = 7.420, p = 0.011, η_p² = 0.216] was revealed. Participants had smaller finger taps following the SECP condition (MD = −0.4 cm, d = 0.496). No main effect of condition was found for movement rate [F_(1,27) = 2.006, p = 0.168, η_p² = 0.069], amplitude CV [F_(1,27) = 0.815, p = 0.375, η_p² = 0.029], or movement rate CV [F_(1,27) = 0.495, p = 0.488, η_p² = 0.018]. A main effect of group on amplitude CV [F_(1,27) = 7.552, p = 0.011, η_p² = 0.219] was revealed. The variability in amplitude was greater in persons with PD (MD = +3.4, d = 0.739). No main effect of group was found for movement amplitude [F_(1,27) = 3.238, p = 0.083, η_p² = 0.107], movement rate [F_(1,27) = 0.402, p = 0.531, η_p² = 0.015], or movement rate CV [F_(1,27) = 1.147, p = 0.294, η_p² = 0.041]. The results revealed a condition x group interaction for amplitude CV [F_(1,27) = 5.501, p = 0.027, η_p² = 0.169] where the finger tapping amplitude CV increased in persons with PD (MD = +4.6, d = 0.519) and decreased in HOAs (MD = −2.0, d = 0.326), following the SECP condition. Post hoc analysis found that persons with PD had a significantly higher amplitude CV than HOAs during the SECP condition [t₍₂₇₎ = 3.516, p = 0.002, MD = 7.6, d = 1.305], but amplitude CV was similar between the two groups during the control condition [t₍₂₇₎ = 0.386, p = 0.702, MD = 0.8, d = 0.175]. No condition x group interaction effects were found for amplitude [F_(1,27) = 1.161, p = 0.291, η_p² = 0.041], movement rate [F_(1,27) = 0.056, p = 0.815, η_p² = 0.002], or movement rate CV [F_(1,27) = 0.000, p = 0.989, η_p² = 0.000].

For both repetitive hand movements, there were no differences between conditions or groups. For repetitive hand movements, no main effects of condition [F_(1,28) > 0.059, p < 0.809, η_p² < 0.083], group [F_(1,28) < 3.242, p > 0.083, η_p² < 0.104], or condition × group interaction effect [F_(1,28) < 1.195, p > 0.284, η_p² < 0.041] were revealed for any outcome measure.

For pronation-supination of the hand, participants with PD were slower and smaller than HOAs, but the effect of the SECP condition was similar between groups. There were no main effects of condition on any of the pronation-supination of the hand outcome measures [F_(1,27) < 2.820, p > 0.105, η_p² < 0.095]. However, a main effect of group for amplitude [F_(1,27) = 7.148, p = 0.013, η_p² = 0.209] and movement rate [F_(1,27) = 9.086, p = 0.006; η_p² = 0.252] was revealed, where persons with PD had less pronation-supination of the hand rotation (MD = −24.4⁰, d = 0.926), and slower rotations (MD = −0.4 Hz, d = 1.001) than HOAs. No main effect of group was found for amplitude CV [F_(1,27) = 0.073, p = 0.788, η_p² = 0.003], or movement rate CV [F_(1,27) = 0.047, p = 0.829, η_p² = 0.002]. The results revealed no condition x group interaction effects [F_(1,27) < 1.535, p > 0.226; η_p² > 0.054].

Participants with PD had more postural tremor than HOAs. Postural tremor increased in both groups after the SECP condition compared to the control condition. Figure 3C shows postural tremor amplitude, and Figure 3D shows postural tremor amplitude CV. There was a main effect of condition on postural tremor amplitude [F_(1,27) = 6.676, p = 0.016, η_p² = 0.198], where amplitude was greater following the SECP condition (MD = 0.42 mm, d = 0.473). There was no effect of condition for any other measures [F_(1,27) < 3.064, p > 0.091, η_p² < 0.102]. There was a main effect of group on max amplitude [F_(1,27) = 8.727, p = 0.006, η_p² = 0.244] and amplitude CV [F_(1,27) = 5.666, p = 0.025, η_p² = 0.173], but not for amplitude [F_(1,27) = 3.302, p = 0.080, η_p² = 0.109]. In general, persons with PD had a greater amplitude (MD = 0.36 mm, d = 1.48), max amplitude (MD = 1.88 mm, d = 0.813), and amplitude CV (MD = 13.8, d = 0.677) than HOAs. No condition x group interaction effects were found [F_(1,27) < 2.494, p > 0.126, η_p² < 0.085].

Participants with PD also had more resting tremor than HOAs, but there was no effect of the SECP condition in either group. No main effect of condition for any resting tremor measures was revealed [F_(1,27) < 2.412, p > 0.132, η_p² < 0.079]. A main effect of group on amplitude [F_(1,27) = 5.440, p = 0.027, η_p² = 0.163], and max amplitude [F_(1,27) = 6.012, p = 0.021, η_p² = 0.077] was revealed. Persons with PD had a higher amplitude resting tremor (MD = 0.62 cm, d = 0.749) and a higher max amplitude resting tremor (MD = 2.32 cm, d = 0.862) than HOAs. There was no main effect of group on amplitude CV [F_(1,27) = 2.349, p = 0.137, η_p² = 0.077]. No condition x group interaction effects were found [F_(1,27) < 0.915, p > 0.347, η_p² < 0.032] (Table 2).

Table 3 shows the repeated measures Pearson’s r correlations between all movement outcome measures and the p50 ratio. Figure 4 shows the significant repeated measures correlations. Significant associations were revealed between p50 inhibitory gating and finger tapping amplitude (r = −0.806, p = 0.029) and between p50 inhibitory gating and postural tremor maximum amplitude (r = 0.787, p = 0.020). In general, as the p50 ratio increased, the finger tapping amplitude decreased, and the maximum postural tremor amplitude increased. No other correlations were significant.

During the control condition, persons with PD and HOAs had similar inhibitory gating ratios, and stress decreased inhibitory gating in both groups. While not statistically significant, there was a medium to large condition by group interaction effect size (η_p² = 0.104), where the SECP task resulted in a decrease in inhibitory gating in persons with PD that was more than double that observed in the HOAs. This suggests that in persons with PD, stress may impair inhibitory gating to a greater degree. While a larger sample size is warranted, the lack of significant difference in this study may also be explained by the stage of PD progression of the participant sample. The sample of persons with PD in this study was in the mild to moderate stages of disease severity [Hoehn and Yahr (H&Y) = 2.3]. Previous research demonstrated that compared to HOAs, persons with PD who were in the later stages of the disease (H&Y = 4 and 5) have greater impairment in inhibitory gating (Teo et al., 1997). Due to greater degeneration in various parts of the brain (e.g., locus coeruleus, prefrontal cortex, basal ganglia, etc.) with the progression of PD, stress may have a differential impact on inhibitory gating. Further research is needed to determine the impact of disease progression on inhibitory gating in PD.

As expected, compared with HOAs, persons with PD had a smaller movement amplitude for finger tapping and the pronation-supination of the hand. Persons with PD also demonstrated slower pronation-supination of the hand. Finally, persons with PD had more postural tremor and resting tremor. These results suggest that the quantitative analysis used was sensitive enough to capture PD motor symptoms of bradykinesia and tremor.

The results revealed that the SECP task decreased finger tapping amplitude in both groups, with a larger (non-significant) difference in persons with PD (MD = −0.54 cm) than HOAs (MD = −0.23 cm). Moreover, the SECP task increased finger tapping amplitude variability in persons with PD, while the amplitude variability in HOAs decreased. Previous research has shown that impairments in upper limb repetitive movements in PD are rate-dependent, with greater impairment at movement rates above 2 Hz (Freeman et al., 1993; Nakamura et al., 1978; Stegemöller et al., 2009; Yahalom et al., 2004). However, movement rate did not change for finger and hand movements during the SECP task, which suggests that any changes in amplitude mediated by the SECP task may not be explained by changes in movement rate. Furthermore, the SECP task increased amplitude variability in persons with PD and decreased amplitude variability in HOAs. This suggests that while the SECP task may decrease amplitude in both persons with PD and HOA, the underlying mechanisms may be different. Further research to delineate the differences in finger tapping performance is needed. Nonetheless, given that finger tapping performance is significantly related to the performance of more complex tasks, such as Purdue Pegboard (Müller et al., 2000), buttoning (Uzochukwu and Stegemöller, 2019), handwriting (Stegemöller et al., 2019), and postural instability (Stegemöller et al., 2016) in persons with PD, stress may further exacerbate more complex movements. Indeed, previous research has found that stress negatively impacted skilled reaching in a PD rat model (Smith et al., 2008) and that a stressful environment negatively impacted gait in persons with PD (Ehgoetz Martens et al., 2015). Yet, the SECP did not significantly impact hand movements or pronation-supination of the hand in this study, possibly due to the small sample size. Continued research to better understand the impact of stress on motor impairment is needed.

Postural tremor in both persons with PD and HOAs increased after the SECP task, with persons with PD showing a larger increase in postural tremor than HOAs. Increases in tremor were not completely unexpected, as others have shown that stress can increase postural tremor in healthy young adults (Mitchem and Tuttle, 1954). Specifically, increasing catecholamines associated with stress, such as norepinephrine, epinephrine, and acetylcholine, can increase postural tremor in healthy young adults (Marshall and Schnieden, 1966). However, resting tremor, a cardinal symptom of PD, did not significantly change with the SECP task. Evidence suggests that resting tremor may have a different underlying pathophysiology than postural tremor (Louis et al., 1999; Louis et al., 2001). Resting tremor has been shown to be mediated by the loss of dopaminergic neurons in the retrorubral area (Hirsch et al., 1992; Jellinger, 1999; Oiwa et al., 2003). While evidence is sparse, stress does not appear to impact dopaminergic output of the retrorubral area (Deutch and Roth, 1991). Thus, stress induced by the SECP task may have increased postural tremor by releasing catecholamines, while the networks and neurotransmitters that mediate resting tremor were unaffected. Future research confirming this notion is needed.

Sensory processing impairments are hypothesized to contribute to motor impairments in persons with PD (Conte et al., 2013; Konczak et al., 2009; Müller et al., 2013; Nieuwboer and Giladi, 2013). Persons with PD demonstrate impairments in inhibitory gating in early and mid-latency components (Gulberti et al., 2015; Lukhanina et al., 2009; Lukhanina et al., 2011; Teo et al., 1997). Inhibitory gating has been associated with bradykinesia symptoms but no other MDS-UPDRS motor scores (Lukhanina et al., 2011). The results of this study are in line with the previous research. However, the study by Lukhanina et al. (2011) used the subjective ordinal MDS-UPDRS rating scale, which is unable to detect small differences in specific kinematic components, such as amplitude and frequency. Thus, the results of this study add to the literature suggesting that inhibitory gating may be related to movement amplitude impairment in some motor symptoms of PD.

While the exact mechanisms of the negative impact of stress on PD inhibitory gating and motor symptoms are unknown, there are three potential pathways that could explain the results observed in this study. First, decreases in inhibitory gating may result in sensory overloading, which may impair cognitive functioning by competing for limited cognitive resources (Croft et al., 2001; Desimone and Duncan, 1995; Yadon et al., 2009). In turn, this could affect movement by negatively impacting the ability of persons with PD to use cognitive mechanisms to compensate for basal ganglia damage and the resulting movement impairments. Second, stress may also negatively affect PD motor symptoms by disrupting normal sensory processing. Finally, PD motor symptoms may be negatively impacting other inhibitory processes that are associated with inhibitory gating, such as motor inhibition (Cheng et al., 2016; Liu et al., 2011; Yadon et al., 2009), which is also impaired in persons with PD (Obeso et al., 2011). In short, there are a number of possible mechanisms by which inhibitory gating may influence PD motor symptoms. Additional research is warranted to examine these relationships.

Finally, variability in participant characteristics may have contributed to the findings. Participants with PD have higher trait anxiety scores than the HOAs in this study. Higher anxiety alone may have independently modulated both inhibitory gating and motor impairment. Participants with PD in this study were all asked to adhere to the same timing of medication administration on the study day, but the dosage and effects of the dosage likely vary from participant to participant. These differences in medication may also have the potential to modulate both inhibitory gating and motor impairment. Consideration of these additional factors is important in the interpretation of the results. Future studies should consider including larger sample sizes that will allow for adequate control and analysis of participant characteristics, such as anxiety and medication.

While our sample size was large enough to support our initial hypothesis that the task reduced inhibitory gating in both HOAs and persons with PD, we were likely underpowered to find a group by condition interaction. However, the effect sizes support the notion that stress may reduce inhibitory gating in individuals with PD to a greater degree. Moreover, significant associations between inhibitory gating and some motor impairments were revealed, contributing additional information to the literature. However, the sample size was small, increasing the risk of Type 1/11 error. Another limitation is that the SECP task included both a physical stressor (cold water) and a social evaluation. Thus, it is not possible to determine if the differences observed in this study were from nociceptive stress, social-evaluative stress, or their interaction. We also did not measure the effect of stress on more complex motor skills, such as walking, handwriting, or speaking. Whether stress negatively impacts complex motor skills in persons with PD remains to be examined. Moreover, not all cardinal symptoms, rigidity, gait, and postural instability were measured, which further limits the interpretation. However, given that bradykinesia and tremor impact functional motor skills, stress may still be considered an important modifiable factor. Finally, the potential mechanisms mediating neurotransmitters, motor symptoms, and inhibitory gating were not assessed, and future studies are needed to clarify how stress might impact these in persons with PD.