The Influence of Motor and Nonmotor Factors on Ambulatory Activity in People With Parkinson Disease Authors: Cielita Lopez-Lennon, Jennifer Hale, Anson Rosenfeldt, Erin Suttman, A. Elizabeth Jansen, Angela Weston, Blessing S. Ofori-Atta, Jay Alberts, Leland E. Dibble Published: 2026 ABSTRACT Background and Purpose:Reduction in ambulation levels has been described as a clinical “red flag” that represents the emergence of disability in individuals with Parkinson disease (PD). Multiday ambulatory activity measures may capture disability not detected by standard clinical assessments. However, the influence of disease severity and nonmotor symptoms on ambulatory activity remains unclear. This study aimed to examine the association between daily ambulatory activity and motor disease severity, assessed in both medication states, and to identify which non-motor symptoms were associated with daily step accumulation in individuals with mild to moderate PD.Methods:Mean daily steps were calculated over 7 days in 198 participants with PD. Motor and nonmotor symptoms were characterized using Movement Disorders Society—Unified Parkinson Disease Rating Scale (MDS-UPDRS) Part I, III, and IV. Descriptive statistics were calculated at baseline to characterize participant demographics and clinical characteristics. Linear regressions modeled the association between motor/nonmotor signs/symptoms and ambulatory activity, adjusting for age, race, and body mass index.Results:Median daily step count was 5180 (interquartile range: 3521-7199). Standardized off-medication MDS-UPDRS III scores were significantly associated with fewer mean daily steps, while on-medication scores were not, though the magnitude of this association was modest. Among nonmotor symptoms, a self-reported complaint of lightheadedness upon standing was associated with approximately 2000 fewer mean daily steps.Conclusion:A self-report of lightheadedness upon standing showed a clinically meaningful association with reduced ambulatory activity, while motor severity, even assessed in the standardized off-medication state, showed only a modest association. These findings highlight the value of nonmotor symptom assessment, particularly cardiovascular autonomic screening, in understanding real-world disability in PD. INTRODUCTION A reduction in ambulatory activity has been described as a clinical “red flag” that represents the emergence of disability in individuals with Parkinson disease (PD).1,2 Multiday ambulatory activity measures capture progressing disability not readily apparent in clinical measures of gait or disease severity. Even early in the disease process, individuals with de novo PD averaged just over 5000 steps/d, those with mild PD (Hoehn & Yahr I & II) took 30% fewer steps per day than the controls and spent less time in sustained walking.3,4 Longitudinal studies further suggest that ambulatory activity may decline earlier and more rapidly than traditional clinical measures of motor severity, which is often assessed in the on-medication state.5,6 Importantly, motor disease severity and walking capacity explain only a small proportion of the variance in real-world ambulatory activity in individuals with PD and many individuals demonstrate low step counts despite preserved gait capacity.6,7 Walking endurance and fast gait speed accounted for just 7% to 13% of the variance in daily step counts and moderate intensity walking minutes.7 Motor disease severity explains just 4% of the variance in individuals with de novo PD.8 Additionally, prior work suggests that medication state transitions and motor fluctuations do not significantly influence daily walking quantity, despite a clear medication effect on gait quality.9 Together, these studies suggest that commonly assessed clinical measures may not fully explain real-world ambulatory behavior in PD. Despite this body of literature, a specific and important gap remains unaddressed. Motor disease severity, as assessed by the Movement Disorders Society Unified Parkinson Disease Rating Scale motor subscale (MDS-UPDRS III),10 has been examined almost exclusively in the on-medication state in prior ambulatory activity studies, reflecting routine rehabilitation practice in which patients are evaluated after taking their medication. For example, Cavanaugh et al5,6 assessed motor disease severity in a stable on-medication state when examining predictors of ambulatory activity decline. It is essential to distinguish such a stable on-medication or off-medication state from motor fluctuations, which refer to within-day variability in motor function that occurs as dopaminergic medication effects peak and wane during routine medication cycles.11 These are conceptually distinct: the standardized off-medication state involves deliberate dopaminergic medication withdrawal for at least 12 hours before clinical examination and captures an individual’s worst (most impaired) motor function under controlled conditions.10 Indeed, Curtze et al12 reported that mobility measures taken while patients were off medication provided a more accurate reflection of their perceived mobility disability and balance confidence than those obtained in the on-medication state, suggesting that off-medication assessment captures clinically meaningful information not fully reflected in on-medication evaluations. While the general association between greater disease severity and reduced ambulatory activity has been established,5 no prior study has directly compared whether this association differs depending on the medication state in which the motor severity is assessed. This distinction is clinically meaningful because rehabilitation assessments conducted exclusively in the on-medication state may not capture the motor state that may have the most direct influence on how individuals navigate their community daily. To our knowledge, no prior study has directly compared the association between on- and standardized off-medication motor severity and ambulatory activity, representing a novel contribution of this present study. In addition to motor severity, nonmotor symptoms (NMS) may also play a critical role in affecting daily movement patterns and physical activity levels.13 Nonmotor deficits in individuals with PD can manifest in various ways, producing symptoms of cardiovascular dysfunction, cognitive impairment, anxiety, depression, sleep disturbances, fatigue, or gastrointestinal and urinary tract dysfunction.14 At least one NMS has been reported in up to 88.5% to 99.7% of individuals with PD.14,15 NMS contribute substantially to disability in PD,15 and research has shown an association between NMS, including cognitive impairment, depression, and sleep disturbances, with physical inactivity in individuals with PD.16 However, these have not been examined specifically in relation to daily step accumulation.17 Physical inactivity has been associated with an increased risk of PD-related mild cognitive impairment and dementia.18,19 Depression and total energy expenditure have been identified as predictors of overall health-related quality of life,20 and sleep disturbances have been associated with higher sedentary behavior.21 Despite this, it remains unclear which specific NMS are most directly associated with daily step accumulation in individuals with PD. To address these gaps in the literature, this study aimed to evaluate the association between motor disease severity across medication states and NMSs on multiday ambulatory activity in a large cohort of individuals with PD. Based on prior evidence documenting ambulatory decline in PD,5 it was hypothesized that (1) motor disease severity assessed in the standardized off-medication state would be significantly associated with daily ambulatory activity while motor severity in the on-medication state would not and (2) specific NMSs would be associated with lower daily ambulatory activity. METHODS Participants The participants in this project included a cohort of individuals with PD enrolled in a large randomized clinical trial (RCT), CYClical Lower Extremity Exercise Trial for Parkinson’s disease - II (CYCLE-II), examining the role of aerobic exercise in disease progression (clinicaltrials.gov registration number NCT04000360). Details of the RCT have been published previously and are briefly summarized below.22 Participants were recruited from outpatient movement disorder clinics and community support groups at the Cleveland Clinic (Cleveland, OH) and the University of Utah (Salt Lake City, UT). Inclusion criteria were adults with idiopathic PD on a stable dose of antiparkinsonian medication, Hoehn and Yahr stage I-III,23 and in-home Wi-Fi to transmit exercise and activity monitor data. Participants were excluded if they enrolled in another disease-modifying PD trial, had a diagnosis of dementia, history of deep brain stimulation or focused ultrasound, musculoskeletal issues limiting the ability to exercise, neurological disease other than PD, or current cardiac arrhythmia. The American College of Sports Medicine Preparticipation Health Screen24,25 was used to ensure participants were safe to cycling from a cardiovascular perspective; all participants demonstrated the ability to safely mount/dismount the upright stationary cycle. Study Design This study represents a secondary analysis of baseline data collected before exercise initiation in the CYCLE-II randomized controlled trial. The specific aims examined here were not addressed in the larger trial. Data were gathered during an observational period of at least 7 days preceding exercise initiation. Demographic data were gathered from all participants. Following the informed consent process approved by the Cleveland Clinic and the University of Utah, participants underwent 2 baseline assessments. Participants were assessed in random order, separated by at least 24 hours, in the on- and off-medication states. The on-medication state was defined as at least one hour following medication dosage; the off-medication state was defined as at least 12 hours without antiparkinsonian medication. While motor tests were assessed in both medication states by a blinded rater, quality of life questionnaires reflect upon the past week and are not dependent on the acute medication state. For participant comfort, the questionnaires were collected in the on-medication state. Movement Disorders Society—Unified Parkinson Disease Rating Scale The MDS-UPDRS was used to comprehensively assess motor and nonmotor PD signs and symptoms.10 Specifically, Section I (Nonmotor Experiences of Daily Living), Section III (Motor Examination), and Section IV (Motor Complications) were evaluated for this project. Individual MDS-UPDRS I items were examined rather than the total score to identify which specific NMS domains were associated with ambulatory activity. The MDS-UPDRS I patient questionnaire items 1.1 to 1.5, which cover cognitive impairment, hallucinations, depressed mood, anxious mood and apathy and are collected by a clinician. Item 1.6 (Features of Dopamine Dysregulation Syndrome) was excluded from analyses due to limited variability (n = 14). Items 1.7 to 1.13 were completed by the participant, including sleep problems, daytime sleepiness, pain, urinary and constipation problems, lightheadedness upon standing, and fatigue.26 Section III of the MDS-UPDRS provided a standardized measure of PD motor signs assessed in both the on- and standardized off-medication states.27 To account for the influence of daily motor fluctuations on ambulatory activity, MDS-UPDRS Part IV (item 4.3)10 was evaluated as a potential covariate. This item reflects the number of hours per day an individual reports experiencing low functioning or shaking periods despite having taken their medication.10 Physical Activity Monitoring Ambulatory activity was captured using Garmin Vivofit4® (Garmin, Olathe, KS). Instructions were provided to wear the device during waking hours for the study duration. For the purposes of this report, daily steps were collected over an observational period of 7 days. Adherent days were operationally defined as those days when the participants had more than 500 steps recorded by the monitor over the day.28 While there is no current consensus for minimal step counts within the literature,29 500 steps per day has previously been used as a minimum value to determine wear adherence.28 Participants who wore the device at least 4 of the 7 days were included in this analysis.28,30 Statistical Analysis Descriptive statistics were calculated at baseline to characterize participant demographics and clinical characteristics. Normally distributed continuous variables were summarized using mean and standard deviations, and nonnormal continuous variables were summarized using median and interquartile range. Categorical factors were summarized as frequencies and percentages at each level of the factor. As the proportion of missing data was less than 5% for all variables, complete case analysis (listwise deletion) was used. To examine whether participants’ motor severity assessed in the standardized off-medication state and motor severity assessed in the on-medication state were each associated with daily ambulatory activity, separate models were examined utilizing the MDS-UPDRS III scores obtained during both states. Using single and multiple linear regression, the MDS-UPDRS III was modeled as a continuous variable in linear regression models with mean daily steps as an outcome. On-medication and standardized off-medication models were examined separately to determine whether MDS-UPDRS III scores in each medication state were independently associated with mean daily steps. No formal statistical test was conducted to directly compare the relative strength of these associations across medication states. To evaluate whether more severe NMS would be associated with lower levels of ambulatory activity, individual items from the MDS-UPDRS Part 1 were examined. NMS were modeled as categorical variables to examine their association with mean daily steps. Categories (normal, slight, mild, moderate, and severe) with fewer than 10 observations were collapsed with the adjacent category in the direction of greater disease severity (for example, if a severe category only had 3 observations, it was collapsed with moderate and a new category called moderate to severe was created). This approach was used to preserve the ordinal nature of the scale while maintaining adequate cell size for regression models. Normal was the reference category. All NMS were adjusted for age, race, and body mass index. Preliminary linear regression models were fit and diagnostics were performed until all assumptions were met. Residuals were not normally distributed; thus, transformation of the dependent variables was explored. Preliminary models were fit following log transformation of the dependent variable (mean daily steps), and the assumptions of normality were satisfied. A sensitivity analysis was performed to compare the linear model to the log-transformed models for mean daily steps and found agreement in the direction of association and model inference. Therefore, we report results for the untransformed models. To assess whether the primary findings were sensitive to influential observations, a secondary sensitivity analysis was conducted by removing the 3 participants with the greatest influence on the regression line. Results were materially unchanged (β = −30.4, SE = 14.1, P = 0.032), confirming that the observed association was not attributable to a small number of unusual observations. Model building used a purposeful selection approach consistent with established methods for covariate selection in multivariable regression. For each candidate covariate, an unadjusted bivariate model was first fit against mean daily steps. Variables meeting a prespecified screening threshold of P <0.20 in the unadjusted model were considered candidates for inclusion in adjusted models. Age is a known confounder and was present in all models regardless of the screening result. Cardiovascular comorbidities and time spent in the off state (MDS-UPDRS Part IV, item 4.3) were evaluated as potential covariates using unadjusted bivariate screening. Depression met the screening threshold and was evaluated in the adjusted model. Variables were retained in the final adjusted models if they were independently associated with mean daily steps or produced a meaningful change in the primary coefficient of interest. All final models were adjusted for age, race, and body mass index. All analyses were conducted using Stata version 16.1 (StataCorp, College Station, TX). RESULTS Participant Characteristics and Baseline Ambulatory Activity Data were gathered from a sample of the larger RCT. The baseline characteristics of participants are provided in Table 1. Medication State and Daily Steps Prior to the primary analysis, cardiovascular comorbidities and MDS-UPDRS Part IV, Item 4.3 were each evaluated in unadjusted bivariate models. Neither met the prespecified P <0.20 screening threshold (cardiovascular comorbidities: P = 0.75, Item 4.3: P = 0.79) and neither was included in the adjusted models. When modeled as a continuous variable, standardized off-medication MDS-UPDRS III score was associated with significantly fewer mean daily steps (β: −30, SE = 14.6, P = 0.04), but on-medication MDS-UPDRS III score was not (β: -23, SE = 16.1, P = 0.15). For every increase in the standardized off-medication MDS-UPDRS III by one point, participants had an average of 30 fewer steps per day. Interpreted clinically, an individual with moderate to severe PD who has an off-medication MDS-UPDRS III score of 50 would accumulate, on average, 900 fewer steps per day, compared to someone with mild PD who had an MDS-UPDRS III score of 20. When depression was additionally evaluated as a covariate in the motor severity model, the association between off-medication MDS-UPDRS III and mean daily steps was attenuated (β: −21, SE = 15, P = 0.159). Depression met the prespecified screening threshold in the unadjusted bivariate model (overall P = 0.128). Given the small number of participants with clinically meaningful depressive symptoms (n = 17), this estimate was underpowered and is reported for transparency. The fully adjusted models are reported in Table 2. Nonmotor Symptoms and Daily Steps Of the 12 NMSs that were tested for their association with mean daily steps, only lightheadedness in standing was associated with significantly fewer mean daily steps. Participants who reported mild to moderate lightheadedness in standing had an average of 1899 fewer mean daily steps compared to participants who reported no lightheadedness in standing (β: −1899, SE = 679, P = 0.006), as shown in Table 3. DISCUSSION The present study used commercially available activity monitors to characterize the relationship between motor disease severity across medication states, NMSs and real-world ambulatory activity in a large cohort of individuals with PD. Prior studies examining the relationship between motor severity and ambulatory activity, have assessed motor function almost exclusively in the on-medication state, reflecting routine clinical practice. Novel contributions of this study are the examination of on- and standardized off-medication MDS-UPDRS III scores and the influence of NMS on ambulatory activity in relation to real-world ambulatory activity. Findings indicated that standardized off-medication motor severity was significantly associated with mean daily steps, while on-medication severity was not associated with daily step count. However, the magnitude of the association was modest given our large sample size; it is possible that this finding reached statistical significance without having strong clinical significance. Although the cross-sectional design precludes causal inference, this pattern may suggest the possibility that some individuals with PD may pace their community activity according to their most impaired motor state. Future research is needed to determine the clinical significance of this finding. A more clinically meaningful finding emerged from the NMS analysis. Lightheadedness on standing is a sign of cardiovascular autonomic dysfunction26 which has been reported in 48% to 68% of people with PD.31 Participants who reported at least mild lightheadedness on standing accumulated on average nearly 2000 fewer steps than those who denied lightheadedness. This reduction exceeds recently established PD-specific MCIDs for daily step count (580-1590 steps/d), indicating that the impact of lightheadedness on daily ambulation is clinically meaningful.32 Notably, other NMS commonly linked to physical inactivity in PD, including depression, sleep disturbances, fatigue, and cognitive symptoms were not significantly associated with daily step accumulation in this sample. This suggests that cardiovascular autonomic dysfunction may represent a more direct barrier to community ambulation than mood or sleep-related symptoms. The association between lightheadedness and lower community ambulation, therefore, underscores the importance of active monitoring and routine clinical assessment of autonomic dysfunction in individuals with PD. Clinically, providers should proactively screen for orthostatic tolerance and arterial baroreflex function (active stand test33 and 10-minute NASA Lean Test34) and consider that the presence of orthostatic hypotension may influence community ambulatory activity. Whether addressing blood pressure abnormalities through non-pharmacologic35 or pharmacologic interventions improves ambulatory activity should be investigated further. When considered in the context of contemporary physical activity recommendations for people with PD, the ambulatory activity levels observed in this sample were notably low. Across the sample, participants did not meet the recommended threshold of 7000 to 10,000 steps per day.36 This sample, comprising individuals across H&Y stage I-III, represents one of the largest cohorts to date examined for multiday ambulatory activity in people with PD. Future studies should examine changes in daily ambulatory activity longitudinally and incorporate objective measures of autonomic function, such as positional blood pressure testing to better characterize the relationship between cardiovascular autonomic dysfunction and daily ambulatory activity and explore strategies to support daily ambulatory activity across varying levels of motor and nonmotor disease burden. LIMITATIONS Several limitations of this study warrant consideration. Individuals who volunteer for exercise-based research may differ systematically from the broader PD population in terms of motivation, health literacy, and access to research facilities. Individuals with a diagnosis of dementia were excluded from participation, which may limit the generalizability of these findings to individuals across the full spectrum of PD-related cognitive decline. Numerous factors not captured in this dataset, including self-efficacy, fear of falling, and social support, are known to influence physical activity in PD and may account for meaningful variance in daily steps.37,38 Finally, NMS were assessed using items from the MDS-UPDRS Part I, which, while validated, may lack the sensitivity to detect subtle differences in symptom burden compared to more comprehensive or objective assessment tools. Depression met the screening threshold for covariate consideration in preliminary analyses but was not retained in the final motor severity model. The small number of participants with clinically meaningful depressive symptoms limited the stability of this estimate, and the cross-sectional design precluded determination of whether depression acts as an independent predictor or as part of the pathway linking motor severity to reduced ambulatory activity. The relationship between depression, motor severity, and daily ambulatory activity in PD warrants further examination in larger longitudinal samples. CONCLUSIONS The presence of lightheadedness and greater motor disease severity in the standardized off-medication state were associated with fewer daily steps in individuals with mild to moderate PD. These findings highlight the clinical value of assessing NMS and considered off-medication motor disease severity when evaluating real-world ambulatory activity.