A longitudinal study on physical activity in early Parkinson's disease

Abstract

Physical activity (PA) remains critical in the slowing of disease progression in early Parkinson's disease (PD), although the longitudinal follow-up of such studies remain scarce. Using data from an early PD Cohort, we longitudinally examined the impact of unprescribed PA on symptoms of early PD, controlling for demographics and medications. Over five years, the reported PA in early PD declined significantly annually. When maintained, the overall PA had significant association with improved motor symptoms, cognition, and quality of life. Higher PA maintained longitudinally is associated with a slower progression of motor and non-motor symptoms, and predict better quality of life in patients with early PD.

Plain language summary

While exercise remains an integral part of treatment in Parkinson's disease (PD), the effects of exercise are often limited in part due to the lack of compliance of PD patients in maintaining these regimes. So far, most studies in PD have looked specifically at exercise interventions but few studies have considered the total overall physical activity in PD over a longer time frame. We wanted to study the effects of all types of physical activity on the movement, memory and mood symptoms of PD over a sustained period of time. Additionally, we wanted to know if higher physical activity levels were related to improvements in these fore-mentioned symptoms across five years. One hundred and ninety-five PD patients were included in the study, and disease progression was studied in relationship to their physical activity levels. We found that baseline physical activity, as well as sustained physical activity, were related to a slower disease progression. Various subtypes of physical activity were also related to the slowing of progression in various symptoms. We also found that subjects with higher physical activity levels had a slower rate of symptom progression that subjects in the lower activity group, especially in movement and quality of life measures. It is therefore important to consider all kinds of physical activity and its maintenance in the treatment regimens of patients with early PD.

While exercise trials in Parkinson's disease (PD) have been ubiquitous and performed as a component of treatment regime, these measures often do not account for compliance and the longer-term maintenance of physical activity post-studies. Naturalistic studies of regular physical activity longitudinally remain scant.^1,2

Given the importance of long-term physical activity on early PD,^3–5 we conducted a five-year prospective longitudinal study investigating the relationship between physical activity and multiple symptom domains including motor function, cognitive, non-motor symptoms, and quality-of-life measures among patients with early PD.

One hundred and ninety-five patients (63.9 ± 8.9 years, 58.9% male) were enrolled as part of a prospective study of early PD patients in Singapore.^6,7 Early PD patients were defined as having diagnosis of ≤1 year using the National Institute of Neurological Disorders and Stroke (NINDS) criteria for PD, and with symptom onset of ≤2 years, who were diagnosed and followed-up by trained movement disorder specialists. MDS-UPDRS scores were obtained on the ON medication state. Participants in this study were excluded at the point of data extraction if they were found to have an alternative (non-idiopathic PD) diagnosis or any new medical condition that could confound the results during follow-up. Decrements in study participants across the years was due to the different follow-up times of participants and time of data extraction, and not due to subject drop-outs.

Physical activity (PA) was measured using the Physical Activity Scale for the Elderly (PASE), a widely-used and well-validated 12-item scale to measure physical activity levels amongst elderly PD patients. ⁸ PASE scores can be subdivided into Leisure, Household and Work-related activities, with each score weighted for the appropriate intensity and duration. We used the PASE as a measure of patients’ PA during the study follow-up. Additionally, participants’ motor, cognitive, non-motor, and quality of life scores was assessed annually alongside PA levels. A longitudinal linear-mixed effects model analysis was conducted, with random intercept and slope to examine various patient outcome measures using physical activity scores as predictor variable and each patient as a random variable, controlling for disease duration, demographics and levodopa equivalent daily doses (LEDD).

Overall, PA declined longitudinally with overall PASE score reducing by 7.82 (95% CI [−12.66, −2.99]) points per year (Spearman rho = −0.12, 95% CI [−0.19, −0.04], p < 0.001). The quality-of-life, apathy and sleep quality scores had also significantly declined across 5 years (see Supplemental Table 2). At baseline, leisure domains of PA were significantly associated with non-motor symptom progression (see Supplemental Table 3). Linear mixed effects model adjusted for disease duration, demographics, and LEDD also found an annual interaction effect of overall annual PA on the longitudinal clinical progression of motor and cognitive symptoms, as well as the quality of life (Supplemental Table 4). It was also observed that household PASE sub-scores had significant annual interaction effects with motor scores, cognitive scores, depression, apathy, fatigue, and quality of life scores. To further investigate and validate the findings, we performed subgroup analyses of patients based on median split of PASE scores and explored differences between high and low PA activity group. There were no significant differences in scores between high and low activity levels at year 5. A linear mixed effects model found activity level group to have a significant interaction effect on symptom scores, with a the high PA group having slower progression of motor (MDS-UPDRS III: β = −1.31 [95% CI:−2.30, −0.32], p = 0.009), non-motor (NMSS: β = −3.52 [95% CI:−5.59, −1.45], p = 0.001), depression (HADS-Depression: β = −0.383 [95% CI:−0.68, −0.08], p = 0.012) scores, and better reported life quality (PDQ-8: β = −1.56 [95% CI:−2.74, −0.39], p = 0.009) than the low PA groups (Figure 1). No significant effect was found for cognition scores. The subgroup analysis confirmed the significant association between higher PA and better symptom outcomes in early PD.

Figure 1.

Linear mixed effects model regression plots with fixed variables: Age, Gender, LEDD, Year, and Disease Duration, Education (MoCA), Random Variable: Participant.

The beneficial effect of physical activity has been reported more notably in a recent paper using the Parkinson's Progression Markers Initiative Cohort (PPMI). ⁹ Several mechanisms have been reported to support the findings, including the neuroprotective effects of aerobic motion in rodent PD models on dopaminergic neurons in the substantia nigra^10–12 and the effect of exercise on increases in dopaminergic transporter and neuromelanin concentration in PD patients. ¹³ In our prospective study of PA in early PD patients over five years performed in an Asian population, we found that overall physical activity when sustained over time had significant impact in the longitudinal slowing of the progression of motor scores and cognitive decline, as well as improvements in quality of life. We also observed that different types and levels of physical activity were associated with different clinical parameters in the disease progression over an extended period of five years, noting that various forms of PA, ⁹ such as household activities, that are not specific to exercise maybe helpful in reducing symptom progression (Supplemental Tables 3 and 4). It is noteworthy to highlight that the effectiveness of interventional exercise in Parkinson's disease has also been observed to diminish progressively. ¹⁴ In our study, we also noted a gradual decline in PA throughout the years. Therefore, maintaining a high level of PA, exercise or otherwise, is essential to achieve long-term improvement in clinical outcomes. Our study demonstrated that the high PA group experienced a slower rate of progression in motor function and better quality of life scores over five years. The strength of our study includes: (1) A well characterized early PD cohort, (2) long duration of follow up and assessment of the varied effects of various forms of physical activity over a period of five years and (3) being the first study to demonstrate that the maintenance of household-related activity is associated with better with motor and cognitive functioning, lower non-motor burdens, and overall better quality of life scores; and higher work activity with better motor and non-motor symptom scores.

The findings in our study have significant implications in real world settings. Primarily, they emphasize the importance for PD patients to maintain consistent physical activity, particularly in household and work, in addition to structured exercise to preserve their physical activity levels.

To conclude, our study has shown that not only the baseline, but also the maintenance of various forms of physical activity have substantial impacts on both motor and non-motor symptoms, as well as the quality of life and progression of PD. This study offers valuable insights with a heightened focus on long term maintenance of physical activity in PD.

Supplemental Material

sj-docx-1-pkn-10.1177_1877718X251380487 - Supplemental material for A longitudinal study on physical activity in early Parkinson's disease

Supplemental material, sj-docx-1-pkn-10.1177_1877718X251380487 for A longitudinal study on physical activity in early Parkinson's disease by Samuel Yong-Ern Ng, Darakhshan Naheed, Ehsan Seyed Saffari, Nicole Shuang-Yu Chia, Xinyi Choi, Dede Liana Heng, Shermyn Xiu-Min Neo, Zheyu Xu, Kay-Yaw Tay, Wing-Lok Au, Eng-King Tan and Louis Chew-Seng Tan in Journal of Parkinson's Disease