GRAPHICAL ABSTRACT

INTRODUCTION

Parkinson’s disease (PD) is a neurological disorder characterized by motor features, such as bradykinesia, tremor at rest, rigidity, and nonmotor features, including cognitive impairment and depression [1]. There is no available modality for modifying the course of PD; therefore, complementary strategies, including exercise, have been advocated for managing PD [2]. Exercise induced mitophagy in neurons is expected to reduce the accumulation of toxic α-synuclein aggregates and may attenuate the progression of PD [3,4]. Eventually, exercise and other types of physiotherapies have been proposed as adjuvants to pharmacological treatments [5].

However, exercise interventions in patients with PD have shown inconsistent results. A study of resistance training in 20 patients with PD reported improvements in balance and quality of life [6], whereas another study reported no reduction in fall frequency after 6 weeks of strength exercise [7]. In a meta-analysis addressing the effects of exercise on motor features, studies revealed heterogeneity in changes in the Unified Parkinson’s Disease Rating Scale Part III (UPDRS III) score [8].

Patients with PD show considerable variability in response to therapeutic modalities [1]. Studies using functional magnetic resonance imaging (fMRI) have revealed inconsistent changes in functional connectivity following strength exercise [9,10]. The clinical progression of PD varies depending on individual motor and nonmotor features, including axial motor deficits, cognitive impairment, and autonomic dysfunction [11]. Therefore, it is possible that the motor or nonmotor characteristics of individual patients with PD might be associated with the response to exercise intervention. However, only a few studies have explored such factors. A recent tele-exercise study revealed that patients with worse motor deficits showed greater improvement following a yoga program [12]. In another study using multimodal physiotherapy, differences in postexercise fMRI activation patterns were noted between the tremor-dominant and akinetic rigid subtypes [10]. However, it is still unclear whether individual characteristic of patients with PD are associated with improvements following exercise intervention. We performed secondary analyses of a previous exercise intervention and explored the clinical factors related to improvements in motor severity.

MATERIALS & METHODS

Participants

The design and results of the original study have been previously presented [13]. First, the participants were randomized into three exercise groups: resistance training, Tai Chi, and yoga groups. To ensure center bias, all sessions were performed at the same location. Participants who were unable to attend group exercise sessions owing to a long distance between home and the exercise center were assigned to the home exercise group. Consequently, we enrolled 99 patients with PD from five tertiary hospitals, and the patients were classified into four exercise groups: home exercise, strength training, Tai Chi, and yoga groups. The subjects participated in exercise sessions twice weekly for 12 weeks, followed by a subsequent 12-week period. The exclusion criteria were as follows: 1) being at risk of falling during exercise, 2) having other diseases affecting gait and balance, including orthopedic disease, neuromuscular disease, vestibular disorder and stroke, 3) having PD dementia according to the Movement Disorder Society (MDS) clinical criteria [14], and 4) having significant changes in motor features within 3 months of enrollment [13]. The study protocol was approved by the Institutional Review Board of Ethics at Dong-A University Hospital (DAUHIRB-23-155), and informed consent was obtained in accordance with the recommendations of the Declaration of Helsinki.

Exercise protocols

We included four types of exercises: home exercise, strength training, Tai Chi, and yoga. The home exercise program consisted of body weight exercises that could be performed at home with video instructions without the need for special equipment. The strength exercise program included resistance training using elastic bands and dumbbells. The main strength-training session was set at a rate of perceived exertion of 13–15, with progressive increase in the resistance of the bands, weight of the dumbbells, number of repetitions, and number of sets. Tai Chi was selected because of its presumed effectiveness in improving the gait and balance of patients with PD. We adopted the same protocol as in our previous study [15]. The yoga program was designed to focus on the therapeutic effects of gentle yoga and comprised three stages: breathing, asanas, and meditation. The phase of the asanas progressed by expanding from small movements in the extremities to larger motions involving the major joints, increasing the dynamic and centered movements.

Clinical measurements

Demographic and other clinical characteristics were evaluated at baseline and 12 weeks after the exercise intervention. Demographic data, including age, sex, disease duration, and body mass index (BMI), were recorded. The levodopa equivalent daily dose (LEDD) was calculated using the formula suggested by Jost et al. [16]. The motor severity of the participants was evaluated using the MDS-UPDRS score and modified Hoehn and Yahr (H&Y) stage. The MDS-UPDRS III subscores for tremor, bradykinesia, rigidity, and axial motor deficits were calculated [17]. We assessed self-efficacy for mobility using the fear of falling question on the Korean version of falls efficiency scale (FES-K) and the new freezing of gait questionnaire. To evaluate the functional capacity for gait and balance, the short physical performance battery (SPPB) and balance evaluation systems test (BESTest) were performed. The physical activity of daily living (ADL) score was measured by the short-form health survey 36 (SF36) and the Parkinson’s Disease Questionnaire 39 (PDQ39). We surveyed beliefs about patients’ ability to exercise via the self-efficacy for exercise (SEE) scale. The Montreal Cognitive Assessment and Mini-Mental State Examination (MMSE) were performed to screen for cognitive impairment. The severity of other nonmotor features was assessed using the MDS-UPDRS I, Non-Motor Symptoms Scale for Parkinson’s disease (NMSS), fatigue severity scale (FSS), Beck Depression Index (BDI), and Beck Anxiety Index (BAI). All clinical measurements were performed without withdrawal of dopaminergic medications.

In the present study, we investigated the baseline characteristics related to improvements in motor severity. To avoid potential bias from motor severity at baseline, we evaluated the effects of exercise as a percentage change in the MDS-UPDRS III total score ([baseline - follow-up]/baseline ×100%).

Statistical analyses

The distribution of continuous variables was determined via the D’Agostino‒Pearson test. Differences in continuous variables between the exercise groups were compared by analysis of variance or the Kruskal–Wallis test. Post hoc analyses were performed using the independent t-test or the Mann–Whitney U test. Correction for multiple testing was performed using the Bonferroni correction. In this manuscript, the corrected p values represent the p values after correction using the Bonferroni method. Categorical variables were compared using the chi-square test. In the frequentist statistics, statistical significance was defined as p<0.05.

Before the associations between baseline characteristics and the percentage improvement in motor severity were examined, a principal component analysis (PCA) was used to address potential collinearity issues. Varimax (orthogonal) rotation was applied, and the principal components (PCs) were retained until the cumulative explained variance reached approximately 70%. Clinical variables with absolute values of factor loading >0.3 were interpreted as being associated with the PCs. Using PCs as predictors, we employed a multivariate Bayesian regression model to investigate the associations between clinical factors and the percentage improvement in the MDS-UPDRS III score. The Bayesian regression model included covariates, including age, sex, disease duration, and LEDD. There were no significant associations in the pairwise correlation test of posterior draws between regression coefficients (Supplementary Table 1 in the online-only Data Supplement). We consider that intercorrelation between PCs exists; however, it does not cause a significant bias in the Bayesian regression model [18]. In the exercise groups, we performed additional Bayesian regression analyses, including each clinical variable as a predictor. The Bayesian models were estimated using four Markov chains, each of which had 2,000 iterations, including 1,000 warm-ups. Statistical analyses were performed using the R Statistical Software version 4.2.2 (R Foundation for Statistical Computing).

RESULTS

Baseline characteristics

The demographic and clinical characteristics at baseline are summarized in Table 1 and Supplementary Table 2 (in the online-only Data Supplement). The participants showed a 16.0±36.0% improvement in the MDS-UPDRS III score (Table 1). At baseline, there were significant differences in the MMSE, NMSS domain 5, SF36 domains 8 and 9, and SPPB total scores between the exercise groups. After Bonferroni correction, statistical significance in the NMSS domain 5 and SPPB total scores was maintained. The home exercise group (5, 6.3±5.7) had significantly higher scores on NMSS domain 5 than did the strength training group (2.3±2.8; Kruskal–Wallis test, corrected p<0.001). The total SPBB score was lower in the home exercise (3.7±0.6) and Tai Chi (3.9±0.5) groups than in the strength training (11.0±2.2) and yoga groups (10.9±1.5; Kruskal–Wallis test, corrected p<0.001).

Associations between PCA-driven factors and percentage improvement in motor severity

We extracted ten PCs with 69.0% of the cumulative explained variance (Supplementary Table 3 in the online-only Data Supplement). In the multivariate Bayesian regression model, which included age, sex, disease duration, LEDD, and PCs, three PCs were associated with the percentage improvement in the MDS-UPDRS III score without their 90% posterior intervals overlapping with zero (Figure 1). PC3 included high H&Y stage, impaired mobility and postural control (MDS-UPDRS III axial subscores, BESTest total scores, and FES-K total scores), and limitations in physical ADLs (MDS-UPDRS II and PDQ39 domains 1 and 2). In addition, the tremor subscores were associated with PC3 (Supplementary Table 3 in the online-only Data Supplement). In the multivariate Bayesian regression model, PC3 was inversely associated with the percentage improvement in the MDS-UPDRS III score (estimates of posterior interval [mean±standard deviation]: -3.9±1.7; 90% posterior intervals -6.73 to -1.10); therefore, the model suggested that patients with impaired mobility and postural instability experienced less improvement after exercise intervention (Table 2 and Figure 1).

PC2 was also associated with severe motor deficits (H&Y stage and MDS-UPDRS III total scores), reduced physical activity, and urinary discomfort (NMSS domain 7) at baseline. We considered this PC to represent motor severity unrelated to impairments in gait and posture control (Supplementary Table 3 in the online-only Data Supplement). High PC2 was positively correlated with the percentage improvement in motor scores (estimates of the posterior interval: 2.5±1.5; 90% posterior intervals 0.02 to 5.03); thus, patients with worse motor features but preserved mobility may experience greater improvement after exercise (Table 2 and Figure 1).

In contrast, PC5 did not compromise any variables associated with motor features or physical ADLs. However, nonmotor features such as depression (BDI), anxiety (BAI), and poor mood symptoms (NMSS domain 3, PDQ39 domain, and SF36 domain 7) were included in PC5 (Supplementary Table 3 in the online-only Data Supplement). In addition, poor cardiovascular (NMSS domain 1) and gastrointestinal (NMSS domain 6) symptoms, sleep (NMSS domain 2), fatigue (FSS total scores), and subjective cognitive impairment (NMSS domain 6) were associated with PC5. Therefore, the multivariate Bayesian model revealed that nonmotor features at baseline were associated with improvements in motor features after the exercise intervention (PC5, estimates of the posterior interval: 3.3±1.7; 90% posterior intervals 0.58–5.98) (Table 2 and Figure 1).

Bayesian regression analyses in exercise groups

We conducted additional Bayesian regression analyses using age, sex, disease duration, and LEDD as covariates to examine the baseline characteristics associated with the percentage improvement in the MDS-UPDRS III score following exercise. In the home exercise group, the percentage improvement in the MDS-UPDRS III score was associated with high BMI, BDI, BAI total score, tremor subscores, and low FES-K total score. The strength training group also showed an association between improvements in the MDS-UPDRS III score and tremor subscores. In addition, in the strength training group, motor deficits, including MDS-UPDRS III total scores and rigidity subscores, were associated with percentage changes in MDS-UPDRS III scores (Table 3).

In the Tai Chi group, motor features were not associated with the change rate of the MDS-UPDRS III score. However, participants who did not consider themselves at risk of falling at baseline showed greater improvement after the intervention. Nonmotor features, including perceptual problems (NMSS domain 4), subjective deficits in attention and memory (NMSS domain 5 and PDQ39 domain 6), and fatigue (FSS total scores and SF36 domain 6), were also associated with percentage improvement. Finally, in the yoga group, better physical ADLs (SF36 domains 2, 3, 5, 6, and 10 and SPPB total scores) were associated with improvements in motor features (Table 3).

DISCUSSION

The present study suggests that symptomatic improvement after exercise intervention can vary according to the individual characteristics of patients with PD. According to the Bayesian regression model using PCA-derived factors, patients with functional impairments in gait and postural control showed less improvement in motor deficits following exercise interventions. In contrast, in the case of participants without functional impairment in gait and postural control, severe motor deficits were associated with a greater percentage improvement in the MDS-UPDRS III score. In addition, nonmotor features, including depression, anxiety, cognition, and fatigue, were associated with the percentage improvement in motor deficits. In subsequent analyses for each exercise group, motor features were associated with improvements in the home exercise and strength training groups, whereas nonmotor features were related to responses in the home exercise and Tai Chi groups. Each exercise may have different effects on the nervous system; therefore, the clinical features associated with improvements in motor deficits may vary.

Our dataset includes a large number of clinical measurements relative to the sample size. The sample size in the four exercise groups is not sufficient to apply frequentist statistical methods that assume a Gaussian distribution. A reduced sample size decreases the statistical power and increases the risk of type II errors (false-negatives). However, Bayesian estimation does not rely on the asymptotic properties of large samples [19] and loses less statistical power than maximum likelihood estimation does, even with small datasets [20]. Therefore, it is a suitable method for exploring the effects of multiple variables in datasets with small sample sizes.

Although the Bayesian regression model using PCA-derived factors yielded a significant result, the association between baseline motor severity and percentage improvement after interventions was shown in only the strength training group. The exercises in the present study may have distinct effects on parkinsonian motor deficits; therefore, we are unable to conclude that motor severity may be a significant predictor of the response to other types of exercise. However, a recent trial of a tele-exercise intervention employing yoga and core muscle training reported that worse motor severity at baseline was associated with greater improvements in motor deficits [12]. Motor deficits in PD are associated with not only striatal dopaminergic dysfunction but also functional integrity of the hippocampal network [21] and sensorimotor circuit [22]. In addition, pharmacological interventions for depression and cognitive impairment in patients with PD have resulted in improved overall motor severity and gait parameters [23,24]. The benefits of exercise in patients with PD are linked to striatal dopaminergic signaling, cholinergic innervation, and neural plasticity [9,12]. Exercise may modulate the response of the entire central nervous system to dopaminergic neuronal loss in PD patients, and the effect of exercise may be prominent in patients with severe motor and nonmotor deficits.

In contrast, associations of motor improvement with physical ADLs and concern for falls were observed in the home training, Tai Chi, and yoga groups. Most previous studies on exercise interventions in patients with PD have included patients with early-stage disease; therefore, the extent of symptomatic improvement following exercise in individuals with advanced-stage disease is unclear. A study utilizing multimodal exercise in 231 patients with PD reported no significant reduction in fall frequency in the advanced PD group [25]. In the present study, features suggesting functional impairment in gait and balance, such as the presence of fear of falling and freezing of gait, poor physical ADLs, and severe axial motor deficits, were associated with less improvement in motor deficits. Patients in the home exercise, Tai Chi, and yoga groups, who experienced less burden from fear of falling, showed a greater percentage change in MDS-UPDRS III scores. In PD, postural instability and gait disturbances are associated with a high frequency of amyloid pathology, loss of microstructural integrity of the cholinergic projection, and extensive frontotemporal atrophy [26-28]. Underlying neurodegenerative changes may reduce the extent of improvement following exercise. Our results suggest that a specialized exercise protocol is needed for patients with functional impairments in gait and postural control.

Interestingly, tremor subscores in the home exercise and strength training groups were associated with improvements in motor deficits. The participants in the Tai Chi and yoga groups did not show such an association; therefore, we are unable to conclude that patients with PD with a tremor-dominant phenotype would experience prominent improvement after exercise intervention. However, our results suggest that each motor phenotype responds differently to exercise protocols. There are significant differences in nigrostriatal dopaminergic deficits [29] and intrinsic brain activity [30] between tremor and akinetic rigid phenotypes. In a study employing multimodal physiotherapy, patients with PD with the tremor-dominant phenotype presented increased recruitment in the sensorimotor cortex and bilateral thalami, whereas those with the nontremor-dominant phenotype presented increased activation in the cerebellar hemispheres [10]. Therefore, patients with PD may require different types of exercise depending on their motor phenotype.

Our study has several limitations. First, we included four exercise interventions with different effects on parkinsonian motor deficits. Therefore, our results may be influenced by merging participants into a single group. However, regular physical activity—regardless of exercise type—has been shown to improve symptoms in PD patients. Long-term engagement in regular physical activity is associated with slow progression of axial motor deficits and preserved functional connectivity in patients with PD [31,32]. In a recent systematic review of brain-derived neurotrophic factor after exercise interventions, no significant associations were observed by exercise type [33]. Although we cannot draw a definite conclusion due to the heterogeneity of exercise interventions, our findings suggest that individual factors may play a role in symptomatic improvement through increased physical activity. Second, the participants in the study were not fully randomized. As noted above, patients in the home exercise group were highly motivated to participate despite limited access to the exercise center. This suggests that they may have been more motivated than participants in other groups. Additionally, owing to the nature of home-based exercise, we were unable to monitor whether the duration, frequency, and intensity of the exercise were consistently maintained throughout the study period. An analysis of resistance training has shown that the effects of exercise can vary with the intensity and dose of the exercise protocol [34]. Third, as stated above, the sample size is small relative to the variables in the present study. Therefore, our results should be reproduced in further studies with a sufficient sample size and a unified exercise protocol.

In summary, our results suggest that patients with PD with severe motor and nonmotor features, but without functional impairments in gait and posture control, may experience greater improvements in parkinsonian motor deficits. Conversely, individuals with limitations in gait and postural control showed less improvement following the exercise intervention, suggesting that specialized exercise protocols may be necessary for patients with PD with balance impairment. Clinicians should consider individual motor phenotypes and the severity of the motor and nonmotor features of PD when recommending exercise to patients. Further studies are needed to develop exercise protocols tailored for PD patients with PD.

Supplementary Materials

Notes

Conflicts of Interest

The authors have no financial conflicts of interest.

Funding Statement

The present study was supported by the “Korea National Institute of Health” research project (2022-ER1005-00) and a grant of the Korea Health Technology R&D Project through the Korean Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (RS-2023-00265377).

Acknowledgments

None

Author Contributions

Conceptualization: Myung Jun Lee, Jinse Park, Sang-Myung Cheon. Data curation: Myung Jun Lee, Jinse Park, Dong-Woo Ryu, Dallah Yoo. Formal analysis: Myung Jun Lee. Funding acquisition: Sang-Myung Cheon. Investigation: Myung Jun Lee, Jinse Park, Dong-Woo Ryu, Dallah Yoo. Project administration: Sang-Myung Cheon. Supervision: Jinse Park, Dong-Woo Ryu, Dallah Yoo, Sang-Myung Cheon. Writing—original draft: Myung Jun Lee. Writing—review & editing: all authors.

Figure 1.

Posterior intervals of the regression coefficients for the percentage improvement in the MDS-UPDRS III score. The thick segments represent the 50% posterior intervals, the thin outer lines represent the 90% posterior intervals, and the circles represent the posterior medians. PC, principal components derived from principal component analysis; MDS-UPDRS III, Movement Disorder Society sponsored Unified Parkinson’s Disease Rating Scale Part III.

REFERENCES

• 1. Deng X, Mehta A, Xiao B, Ray Chaudhuri K, Tan EK, Tan LC. Parkinson’s disease subtypes: approaches and clinical implications. Parkinsonism Relat Disord 2025;130:107208.ArticlePubMed
• 2. Monteiro-Junior RS, Cevada T, Oliveira BR, Lattari E, Portugal EM, Carvalho A, et al. We need to move more: neurobiological hypotheses of physical exercise as a treatment for Parkinson’s disease. Med Hypotheses 2015;85:537–541.ArticlePubMed
• 3. Goenawan H, Kiasati S, Sylviana N, Megantara I, Lesmana R. Exercise-induced autophagy ameliorates motor symptoms progressivity in Parkinson’s disease through alpha-synuclein degradation: a review. Neuropsychiatr Dis Treat 2023;19:1253–1262.ArticlePubMedPMCPDF
• 4. Magaña JC, Deus CM, Giné-Garriga M, Montané J, Pereira SP. Exercise-boosted mitochondrial remodeling in Parkinson’s disease. Biomedicines 2022;10:3228.ArticlePubMedPMC
• 5. Bouça-Machado R, Rosário A, Caldeira D, Castro Caldas A, Guerreiro D, Venturelli M, et al. Physical activity, exercise, and physiotherapy in Parkinson’s disease: defining the concepts. Mov Disord Clin Pract 2019;7:7–15.ArticlePubMedPMCPDF
• 6. Dibble LE, Hale TF, Marcus RL, Gerber JP, LaStayo PC. High intensity eccentric resistance training decreases bradykinesia and improves quality of life in persons with Parkinson’s disease: a preliminary study. Parkinsonism Relat Disord 2009;15:752–757.ArticlePubMed
• 7. Morris ME, Taylor NF, Watts JJ, Evans A, Horne M, Kempster P, et al. A home program of strength training, movement strategy training and education did not prevent falls in people with Parkinson’s disease: a randomised trial. J Physiother 2017;63:94–100.ArticlePubMed
• 8. Yang CL, Huang JP, Wang TT, Tan YC, Chen Y, Zhao ZQ, et al. Effects and parameters of community-based exercise on motor symptoms in Parkinson’s disease: a meta-analysis. BMC Neurol 2022;22:505.ArticlePubMedPMCPDF
• 9. Beall EB, Lowe MJ, Alberts JL, Frankemolle AM, Thota AK, Shah C, et al. The effect of forced-exercise therapy for Parkinson’s disease on motor cortex functional connectivity. Brain Connect 2013;3:190–198.ArticlePubMedPMC
• 10. Chen K, Wang S, Wen Q, Jin Z, Wang Y, Meng D, et al. Rehabilitation response in tremor- and non-tremor-dominant Parkinson disease: a taskfMRI study. Brain Behav 2024;14:e70102. PubMedPMC
• 11. Hähnel T, Raschka T, Sapienza S, Klucken J, Glaab E, Corvol JC, et al. Progression subtypes in Parkinson’s disease identified by a data-driven multi cohort analysis. NPJ Parkinsons Dis 2024;10:95.PubMedPMC
• 12. Ha J, Park JH, Lee JS, Kim HY, Song JO, Yoo J, et al. Effectiveness of live-streaming tele-exercise intervention in patients with Parkinson’s disease: a pilot study. J Mov Disord 2024;17:189–197.ArticlePubMedPMCPDF
• 13. Park J, Cheon SM, Lee MJ, Rhu DW, Yoo D. Comparison of the impact of various exercise modalities on Parkinson’s disease. J Mov Disord 2025;18:222–230.ArticlePubMedPMCPDF
• 14. Emre M, Aarsland D, Brown R, Burn DJ, Duyckaerts C, Mizuno Y, et al. Clinical diagnostic criteria for dementia associated with Parkinson’s disease. Mov Disord 2007;22:1689–1707.quiz 1837. ArticlePubMed
• 15. Cheon SM, Chae BK, Sung HR, Lee GC, Kim JW. The efficacy of exercise programs for Parkinson’s disease: Tai Chi versus combined exercise. J Clin Neurol 2013;9:237–243.ArticlePubMedPMC
• 16. Jost ST, Kaldenbach MA, Antonini A, Martinez-Martin P, Timmermann L, Odin P, et al. Levodopa dose equivalency in Parkinson’s disease: updated systematic review and proposals. Mov Disord 2023;38:1236–1252.ArticlePubMed
• 17. Adams C, Suescun J, Haque A, Block K, Chandra S, Ellmore TM, et al. Updated Parkinson’s disease motor subtypes classification and correlation to cerebrospinal homovanillic acid and 5-hydroxyindoleacetic acid levels. Clin Park Relat Disord 2023;8:100187.ArticlePubMedPMC
• 18. Curtis SM, Ghosh SK. A Bayesian approach to multicollinearity and the simultaneous selection and clustering of predictors in linear regression. J Stat Theory Pract 2011;5:715–735.Article
• 19. Rognli EW, Zahl-Olsen R, Rekdal SS, Hoffart A, Bertelsen TB. Editorial perspective: Bayesian statistical methods are useful for researchers in child and adolescent mental health. J Child Psychol Psychiatry 2023;64:339–342.ArticlePubMedPMCPDF
• 20. van de Schoot R, Broere JJ, Perryck KH, Zondervan-Zwijnenburg M, van Loey NE. Analyzing small data sets using Bayesian estimation: the case of posttraumatic stress symptoms following mechanical ventilation in burn survivors. Eur J Psychotraumatol 2015;6:25216.PubMed
• 21. Chung SJ, Kim HR, Jung JH, Lee PH, Jeong Y, Sohn YH. Identifying the functional brain network of motor reserve in early Parkinson’s disease. Mov Disord 2020;35:577–586.ArticlePubMedPDF
• 22. Johansson ME, Toni I, Kessels RPC, Bloem BR, Helmich RC. Clinical severity in Parkinson’s disease is determined by decline in cortical compensation. Brain 2024;147:871–886.ArticlePubMedPMCPDF
• 23. Kulisevsky J, Pagonabarraga J, Pascual-Sedano B, Gironell A, García-Sánchez C, Martínez-Corral M. Motor changes during sertraline treatment in depressed patients with Parkinson’s disease. Eur J Neurol 2008;15:953–959.ArticlePubMed
• 24. Henderson EJ, Lord SR, Brodie MA, Gaunt DM, Lawrence AD, Close JC, et al. Rivastigmine for gait stability in patients with Parkinson’s disease (ReSPonD): a randomised, double-blind, placebo-controlled, phase 2 trial. Lancet Neurol 2016;15:249–258.ArticlePubMed
• 25. Canning CG, Sherrington C, Lord SR, Close JC, Heritier S, Heller GZ, et al. Exercise for falls prevention in Parkinson disease: a randomized controlled trial. Neurology 2015;84:304–312.ArticlePubMedPMC
• 26. Zuo LJ, Piao YS, Li LX, Yu SY, Guo P, Hu Y, et al. Phenotype of postural instability/gait difficulty in Parkinson disease: relevance to cognitive impairment and mechanism relating pathological proteins and neurotransmitters. Sci Rep 2017;7:44872.ArticlePubMedPMCPDF
• 27. Nazmuddin M, van Dalen JW, Borra RJH, Stormezand GN, van der Horn HJ, van der Zee S, et al. Postural and gait symptoms in de novo Parkinson’s disease patients correlate with cholinergic white matter pathology. Parkinsonism Relat Disord 2021;93:43–49.ArticlePubMed
• 28. Yin L, Zhu Z, Fu J, Zhou C, Liu Z, Li Y, et al. Differences in gray matter atrophy and functional connectivity between motor subtypes of Parkinson’s disease. Acta Neurol Belg 2024;124:1899–1911.ArticlePubMedPDF
• 29. Eggers C, Kahraman D, Fink GR, Schmidt M, Timmermann L. Akineticrigid and tremor-dominant Parkinson’s disease patients show different patterns of FP-CIT single photon emission computed tomography. Mov Disord 2011;26:416–423.ArticlePubMedPDF
• 30. Zhu J, Zeng Q, Shi Q, Li J, Dong S, Lai C, et al. Altered brain functional network in subtypes of Parkinson’s disease: a dynamic perspective. Front Aging Neurosci 2021;13:710735.ArticlePubMedPMC
• 31. Tsukita K, Sakamaki-Tsukita H, Takahashi R. Long-term effect of regular physical activity and exercise habits in patients with early Parkinson disease. Neurology 2022;98:e859–e871.ArticlePubMedPMC
• 32. Donahue EK, Venkadesh S, Bui V, Tuazon AC, Wang RK, Haase D, et al. Physical activity intensity is associated with cognition and functional connectivity in Parkinson’s disease. Parkinsonism Relat Disord 2022;104:7–14.ArticlePubMed
• 33. Paterno A, Polsinelli G, Federico B. Changes of brain-derived neurotrophic factor (BDNF) levels after different exercise protocols: a systematic review of clinical studies in Parkinson’s disease. Front Physiol 2024;15:1352305.ArticlePubMedPMC
• 34. Zhou X, Zhao P, Guo X, Wang J, Wang R. Effectiveness of aerobic and resistance training on the motor symptoms in Parkinson’s disease: systematic review and network meta-analysis. Front Aging Neurosci 2022;14:935176.ArticlePubMedPMC

Figure & Data

References

Citations

Citations to this article as recorded by  

Comments on this article