Parkinson’s disease (PD) is the second-most prevalent neurodegenerative disease in the world, with the average age of onset of approximately 60 years1. It is most consistently linked with aging, and the global burden of PD is expected to increase sharply in future as life expectancies increase and demographics change2.

It is largely defined by the cardinal symptoms of resting tremor, bradykinesia, rigidity, and impaired postural reflexes and levodopa remains as the gold standard treatment for these symptoms for several decades3. However, long-term usage of levodopa can give rise to complications, with the incidence rate and severity of those complications correlated with increased cumulative dosage3,4.

Levodopa therapy-induced complications can be characterized by: (1) dyskinesia and (2) fluctuations in medication effectiveness (i.e. ON-OFF timings). The Unified Parkinson’s Disease Rating Scale part IV (UPDRS IV) was developed to measure complications of levodopa therapy and is based upon patients’ subjective reporting of their experienced symptoms5.

At present, there is no cure for PD. Assessing the patient’s quality of life (QoL) is therefore particularly helpful in allowing clinicians to determine the effectiveness of treatment6. As such, the primary goal of PD management focuses on treating the symptoms that have significant impact on the patient’s QoL.

The primary aim of this study was to determine the impact of levodopa therapy-induced complications on the quality of life of PD patients over a 1-year follow-up period.


Demographics and clinical characteristics

The baseline demographics of patients who completed versus those who did not complete the 1-year follow-up were comparable (Table 1). There was significant difference between the baseline and follow-up visit for patients who completed 1-year follow-up for mean/median LEDD; PDQ-SI; the PDQ domains for ADL, emotional well-being, cognitive impairment, and communication; S&E ADL; NMSS; and total UPDRS III scores (Table 2).

Table 1 Baseline characteristics of PD patients.
Table 2 Characteristic at baseline and 1-year follow-up.

UPDRS IV results - Domain A (Dyskinesia)

In our overall patient sample, dyskinesia was the least common levodopa therapy-induced complication. At baseline, only 13.5% of the patients experienced dyskinesias during their waking hours. Most of the dyskinesia experienced was rated as not disabling (91.2%) and not painful (96.7%). A small proportion also experienced early morning dystonia (8.0%).

At 1-year follow-up, there was a trend suggestive of increasing dyskinesia duration (from 5.1% to 16.7%, 1.3% to 2.6%, and 1.3% to 2.3% for 1–25%, 26–50%, and 51–75% of their waking day, respectively), and more disabling (from 93.6% to 88.5% not disabling, 5.1% to 9.0% mildly disabling, and 1.3% to 2.6% as moderately disabling). An increasing proportion also experienced early morning dystonia (from 11.5% to 21.8%).

UPDRS IV results - Domain B (Clinical fluctuations)

Clinical fluctuations were the most frequent levodopa-induced complications. At baseline for the entire sample, 39.1% had OFF periods not more than 25% of the waking day, while 14.2% and 2.6% reported experiencing them for 25–50% and 51–75% of their waking day, respectively. A majority had gradual (98.2%) and predictable (54.0%) OFF periods.

For the 1-year follow-up group, there was a shift in responses to having longer duration (from 41.0% to 32.1%, 11.5% to 21.8%, and 5.1% to 7.1% for 1–25%, 26–50%, and 51–75% of the patient’s waking day, respectively), more sudden onset (from 0% to 6.4%), and unpredictable (from 1.3% to 3.3%) OFF periods.

Univariate analysis

The univariate analysis showed age, gender, employment status, disease duration, ECAQ, subsequent visit, UPDRS IV total and domain scores, LEDD, NMSS, UPDRS III, H&Y staging and S&E ADL score were statistically significant and individually predictive of change in PDQ-SI score from baseline to 1-year follow up.

Multivariable analysis

The multivariable analysis (Table 3) found that change in UPDRS IV total score was associated with a medium effect change in PDQ-SI (d = 0.671, p = 0.014). The other factors with smaller effect but stronger statistical significance included female gender (d = 0.374, p < 0.001), NMSS total score (d = 0.305, p < 0.001) and UPDRS III total score (d = 0.179, p < 0.001).

Table 3 Linear mixed model analysis: PDQ-SI as a function of potential clinical factors.


To the best of our knowledge, this short-term longitudinal study is one of the first examining the impact of levodopa therapy-induced complications on QoL in PD patients in a multi-ethnic Asian population like Singapore’s. The majority of other studies have been non-longitudinal and conducted on non-Asian populations7,8,9,10, while relatively few examined the determinants of QoL in PD patients in Chinese or predominantly Chinese populations11,12.

Our study results have shown that levodopa therapy-induced complications have an impact on patients’ QoL and is consistent with other studies conducted in both Asian and Caucasian populations7,8,9,12. In addition, this study has shown some evidence that suggest deterioration of these complications had a medium effect on deterioration of the overall QoL in PD patients for the 1-year follow-up interval. This finding, to our knowledge, has not been published elsewhere and it suggests that treatment related complications are important to be explored during follow-ups of PD patients.

Previous studies had inconsistent results regarding gender, age, and educational attainment as risk factors for worsening QoL10,11,12. This may possibly be due to cultural or ethnic differences across population groups. In our study, apart from levodopa therapy-induced complications, other lesser predictive factors included female gender, NMSS total score and UPDRS III total score. The findings suggest that patients with these factors may be anticipated to have worsening of QoL over time if their symptoms are not managed appropriately.

Longer disease duration is known to be associated with worsening QoL and this was also demonstrated in our study9,13,14,15. While disease duration is a well-studied clinical variable, fewer studies examined the impact of age of onset on QoL12. Our study, as well as other studies, did not find age of onset to have any significant correlation with QoL12,16,17.

In this study, patients’ perceived QoL – one of the key measures of effective management of PD – declined significantly even within a short 1-year period. The PDQ domain scores for ADL, emotional well-being, cognitive impairment and communication increased the most from baseline to 1-year follow-up. Therefore, when addressing patients’ complaints, clinicians may prioritize addressing treatment complications related to these domains.

Our study has several limitations. Firstly, the sample group was rather homogenous with most PD patients in the mild to moderate stages (H&Y stage < 3), therefore it is likely that the more severe motor complications were not captured in this population group. This may limit the generalizability of our results to those patients with more advanced disease stages, who are likely to have more severe motor complications. Secondly, patients who were cognitively impaired and/or unable to verbalize were excluded from the study, as the format of the study assessments was reliant on patients’ ability to recall and communicate the symptoms that they had experienced. Hence, those with poor communicative ability or cognitive issues could be under-represented in our study. In addition, less than a third of the participants completed the 1-year follow-up visit. However, this is an on-going study and patients are still continuously being assessed and followed-up. Lastly, this study lacked normal controls when evaluating differences in PDQ-SI score. However, the main focus of the study was on the longitudinal evaluation of QoL deterioration in PD patients.

In conclusion, our study has shown that levodopa therapy-induced complications have a significant adverse impact on QoL, in particular, the ADL, emotional well-being, cognition and communication domains in PD patients. Worsening of these complications had a medium effect on the worsening of the overall QoL for the 1-year follow-up interval. This substantiates the importance for clinicians to closely monitor and promptly manage the levodopa therapy-induced complications that may arise in PD patients. Further studies that involve developing strategies for effective management, delaying the onset, or even possibly preventing these complications will be valuable to explore.


This was a mono-center, observational, short-term longitudinal study.

Inclusion criteria

PD patients, diagnosed according to the UK PD Brain Bank criteria, were prospectively recruited from Singapore General Hospital neurology and movement disorders clinics18. A total of 274 patients were enrolled, of which 78 patients were reassessed at 1-year follow-up.

Exclusion criteria

Patients with significant cognitive impairment as defined by an Elderly Cognitive Assessment Questionnaire (ECAQ) score of 5 points or less were excluded to avoid unreliable responses for the many subjective patient-dependent scores. Similarly, patients with significant speech dysfunction who were not able to verbalize appropriately were excluded. Patients with chronic debilitating conditions (e.g. severe heart failure, liver failure, renal failure requiring dialysis, or other terminal illness) were also excluded.

The Singhealth Centralized Institutional Review Board SCIRB approved the study protocol. All methods were carried out in accordance with SCIRB guidelines and regulations, including informed consent from all subjects.


Demographic data, including age, gender, educational status, family history, ages of onset and diagnosis, and disease duration were obtained from all recruited patients.

Patient QoL was measured by the 39-item Parkinson’s Disease Questionnaire (PDQ-39), the most widely used disease specific health-related QoL instrument in clinical practice and research6,19,20. It incorporates eight disease domains: mobility (10 items), ADL (6 items), emotional well-being (6 items), stigma (4 items), social support (3 items), cognition (4 items), communication (3 items), and bodily discomfort (3 items). Responses were scored from a scale of 0 (never) to 4 (always). Each domain score was transformed into a 0–100 percentage scale by summing the domain items’ raw scores, dividing by the maximum possible raw score, and then multiplying by 100. A PDQ Summary Index (PDQ-SI) was obtained by calculating the mean of the eight domain scores. Higher scores reflected a poorer patient-rated QoL6.

Levodopa therapy-induced complications were assessed with the UPDRS IV questionnaire – a validated instrument that evaluates complications of levodopa therapy, and divides answers into 2 domains: domain A: dyskinesia (4 questions) and domain B: clinical fluctuations (4 questions)5.

All patients had their PD medication regime converted into the levodopa equivalent daily dose (LEDD) via the standardized formulae: ([levodopa (mg)] + [controlled release levodopa (mg) × 0.75] + [levodopa doses taken together with entacapone (mg) × 1.33] + [pramipexole (mg) × 100] + [ropinirole (mg) × 20] + [rotigotine (mg) × 30] + [piribedil (mg) × 1] + [bromocriptine (mg) × 10] + [selegiline (mg) × 10])21.

Non-motor symptoms were assessed using the Non-Motor Symptoms Scale (NMSS)22.

Other assessments included (1) Unified Parkinson’s Disease Rating Scale Part III (UPDRS III)23, (2) modified Hoehn and Yahr (H&Y) staging scale24 and (3) the Schwab and England activities of daily living (S&E ADL)25.

Assessments were conducted face-to-face on the same day of visit to the specialist clinic by movement disorder neurologists & trained research associates. The interval between baseline and follow-up visit ranged from 9 to 18 months. All subjects were provided with symptomatic treatment (pharmacological or non-pharmacological) as deemed necessary by the treating movement disorders specialist.

Statistical analysis

Continuous data were summarized as mean (SD) or median (IQR), while categorical data was expressed in terms of frequency. Differences between time points were evaluated using the non-parametric Wilcoxon signed-rank test. A p-value of <0.05 was set to be significant.

A linear mixed effects model with subject as a random effect was used to assess the change in PDQ-SI as a function of UPDRS IV over time (2 time points) and adjusting for other factors such as LEDD, UPDRS III score, NMSS score, S&E ADL score, H&Y score, age, disease duration, gender, employment status, etc. Model diagnostics was used to test model adequacy.