Article Text

Download PDFPDF

Original research
A prospective, controlled study of non-motor effects of subthalamic stimulation in Parkinson’s disease: results at the 36-month follow-up
  1. Stefanie Theresa Jost1,
  2. Anna Sauerbier1,2,
  3. Veerle Visser-Vandewalle3,
  4. Keyoumars Ashkan2,
  5. Monty Silverdale4,
  6. Julian Evans4,
  7. Philipp A Loehrer5,
  8. Alexandra Rizos2,
  9. Jan Niklas Petry-Schmelzer1,
  10. Paul Reker1,
  11. Gereon Rudolf Fink1,
  12. Jeremy Franklin6,
  13. Michael Samuel2,
  14. Alfons Schnitzler7,
  15. Michael Thomas Barbe1,
  16. Angelo Antonini8,9,
  17. Pablo Martinez-Martin10,
  18. Lars Timmermann5,
  19. K Ray-Chaudhuri2,11,
  20. Haidar S Dafsari1
  21. On behalf of EUROPAR and the International Parkinson and Movement Disorders Society Non-Motor Parkinson's Disease Study Group
  1. 1 Department of Neurology, University Hospital Cologne, Cologne, Germany
  2. 2 Parkinson Foundation International Centre of Excellence, King's College Hospital, London, UK
  3. 3 Department of Stereotaxy and Functional Neurosurgery, University Hospital Cologne, Cologne, Germany
  4. 4 Department of Neurology and Neurosurgery, Salford Royal NHS Foundation Trust, Manchester Academic Health Science Centre, Greater Manchester, UK
  5. 5 Department of Neurology, University Hospital of Giessen and Marburg, Campus Marburg, Marburg, Hessen, Germany
  6. 6 Institute of Medical Statistics and Computational Biology (IMSB), University of Cologne, Köln, Nordrhein-Westfalen, Germany
  7. 7 Department of Neurology, Center for Movement Disorders and Neuromodulation, Heinrich-Heine-University Duesseldorf, Duesseldorf, Germany
  8. 8 Parkinson and Movement Disorders Unit, IRCCS Hospital San Camillo, Venice, Italy
  9. 9 University of Padua, Padova, Veneto, Italy
  10. 10 Center for Networked Biomedical Research on Neurodegenerative Diseases, Madrid, Spain
  11. 11 Institute of Psychiatry, Psychology and Neuroscience, King's College London, London, UK
  1. Correspondence to Dr Haidar S Dafsari, Department of Neurology, University Hospital Cologne, Cologne 50937, Germany; haidar.salimi-dafsari{at}


Objective To examine 36-month effects of bilateral subthalamic nucleus deep brain stimulation (STN-DBS) on non-motor symptoms (NMS) compared with standard-of-care medical treatment (MED) in Parkinson’s disease (PD).

Methods Here we report the 36-month follow-up of a prospective, observational, controlled, international multicentre study of the NILS cohort. Assessments included NMSScale (NMSS), PDQuestionnaire-8 (PDQ-8), Scales for Outcomes in PD (SCOPA)-motor examination, -activities of daily living, and -complications, and levodopa equivalent daily dose (LEDD). Propensity score matching resulted in a pseudo-randomised sub-cohort balancing baseline demographic and clinical characteristics between the STN-DBS and MED groups. Within-group longitudinal outcome changes were analysed using Wilcoxon signed-rank and between-group differences of change scores with Mann-Whitney U test. Strength of clinical responses was quantified with Cohen’s effect size. In addition, bivariate correlations of change scores were explored.

Results Propensity score matching applied on the cohort of 151 patients (STN-DBS n=67, MED n=84) resulted in a well-balanced sub-cohort including 38 patients per group. After 36 months, STN-DBS significantly improved NMSS, PDQ-8, SCOPA-motor examination and -complications and reduced LEDD. Significant between-group differences, all favouring STN-DBS, were found for NMSS, SCOPA-motor complications, LEDD (large effects), motor examination and PDQ-8 (moderate effects). Furthermore, significant differences were found for the sleep/fatigue, urinary (large effects) and miscellaneous NMSS domains (moderate effects). NMSS total and PDQ-8 change scores correlated significantly.

Conclusions This study provides Class IIb evidence for beneficial effects of STN-DBS on NMS at 36-month follow-up which also correlated with quality of life improvements. This highlights the importance of NMS for DBS outcomes assessments.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Deep brain stimulation (DBS) of the subthalamic nucleus (STN) improves quality of life (QoL), motor and non-motor symptoms (NMS) in patients with Parkinson’s disease (PD).1–3 NMS are a key clinical aspect of PD and have been identified as the most important factor in determining QoL in patients with PD.4 Beneficial non-motor effects in patients undergoing STN-DBS have been observed in clinician-rated as well as laboratory-based investigations of a wide range of NMS, including polysomnography for sleep, urodynamic examinations for urinary symptoms, sniffing sticks for olfaction, as well as sensory signs such as thermal detection thresholds and pain thresholds.5 6

Benefits of bilateral STN-DBS in patients with PD on a wide range of NMS have been reported for a follow-up period up to 2 years.3 It is unclear if beneficial non-motor effects of STN-DBS are sustained beyond 24 months. Non-motor effects of STN-DBS have been compared with apomorphine and levodopa infusion.7 However, little is known about the progression of non-motor symptoms in patients undergoing standard-of-care medical treatment (MED). To our knowledge, a controlled study of non-motor effects of STN-DBS and MED has not been conducted yet. We hypothesised that beneficial motor and non-motor effects of STN-DBS can be observed in a controlled study at 36-month follow-up after surgery.


Study design and ethical approval

Here we report the 36-month follow-up of an ongoing, prospective, observational, controlled, international multicentre study (Class IIb evidence). Patients were recruited between March 2011 and October 2015 as part of the DBS and medication arms of the NILS study7 which is a comprehensive study with non-motor profiling of PD as the primary outcome measure addressing the progression of NMS and treatment responses to medication and advanced treatments. The study was carried out in accordance with the Declaration of Helsinki. All patients gave written informed consent prior to any study procedures.


PD diagnosis was based on the UK Brain Bank criteria.8 As per clinical routine, patients in the DBS group were screened for DBS surgery according to the international guidelines, as described in previous publications by our group.7 9 Patients required a sufficient levodopa responsiveness (>30%) as assessed by the Unified PD Rating Scale-motor examination (UPDRS-III). Indications for DBS were based on multidisciplinary assessments by movement disorders specialists, stereotactic neurosurgeons, neuropsychologists, psychiatrists, and when necessary, speech therapists and physiotherapists. Patients with clinically relevant psychiatric diseases or neuropsychological impairment (mini-mental state examination scores<25) were excluded from this analysis. To ensure comparability between the STN-DBS and MED groups, we included only patients in the MED group, who had advanced PD with dyskinesia, ON/OFF fluctuations or medication-refractory tremor. According to the multidisciplinary assessments, patients in the MED group were considered candidates for STN-DBS but at that time preferred conservative non-surgical therapy according to published standard-of-care recommendations.10 Patients’ informed decisions were shaped by a number of parameters such as patients’ age, disease duration, dopaminergic medication requirements, and motor and non-motor symptom profiles. During the course of the study, patients in the MED group could undergo DBS at any time and patients in the DBS group could switch off neurostimulation at any time.

Clinical assessment and outcome parameters

Patients in both groups were assessed in the on-medication state (MedON) at baseline and 36-month follow-up. The STN-DBS group was assessed in the medication and stimulation ON state (MedON/-StimON) at follow-up.

The principal outcome measure was the Non-motor Symptom Scale (NMSS).11 The 30-item NMSS evaluates nine dimensions of NMS in PD: cardiovascular, sleep/fatigue, mood/apathy, perceptual problems/hallucinations, attention/memory, gastrointestinal symptoms, urinary symptoms, sexual function and miscellaneous symptoms including items for pain, inability to smell/taste, weight change and excessive sweating. The total score ranges from 0 (no impairment) to 360 (maximum impairment).

Furthermore, assessments included the following secondary outcome measures:

QoL was quantified by means of the PD Questionnaire (PDQ)-8, a short-form of the PDQ-3912 covering eight dimensions contributing to QoL. The scale is recommended for QoL assessments by the movement disorders society scales committee and has been commonly used in PD and in patients undergoing DBS.13 14 The PDQ-8 is reported as Summary Index (PDQ-8 SI) ranging from 0 (no impairment) to 100 (maximum impairment).

The Scales for Outcomes in PD (SCOPA) was used to assess motor examination, activities of daily living and motor complications. The scale is a well-established, validated short version of the UPDRS,15 highly correlates with the corresponding UPDRS parts16 and its assessment time is approximately four time shorter than in the MDS-UPDRS.15 17 SCOPA subscales range from 0 (no impairment) to 42 (motor examination), 21 (activities of daily living) and 12 (motor complications).3

The levodopa equivalent daily dose (LEDD) was computed according to the method by Tomlinson et al. 18

Statistical analysis

Propensity score matching was used to create a pseudo-randomised19 sub-cohort balancing the following baseline demographic and disease parameters between the STN-DBS and MED groups: age at baseline, disease duration since diagnosis, LEDD and SCOPA total score. The matching was conducted with propensity score matching for SPSS (V.3.04)20 and SPSS V.25.0 (IBM Corporation). We implemented a 1:1 ratio nearest-neighbour matching algorithm with a 0.25 calliper without replacement and conducted balance diagnostics based on Cohen’s effect size d|<0.250.21

The assumption of normality was assessed with the Shapiro-Wilk test. Differences of baseline characteristics between the two treatment groups for dichotomous variables were analysed using the χ2-test and for continuous variables using Mann-Whitney U tests or unpaired t-tests, when parametric criteria were fulfilled. To determine outcome changes from baseline to follow-up within each group, Wilcoxon signed-rank or paired samples t-tests were calculated. Mann-Whitney U tests of change scores between STN-DBS and MED groups (mean testbaseline − mean testfollow-up) were conducted to analyse differences of outcome parameter changes. Multiple comparisons resulting from the use of seven scales were corrected with the Benjamini-Hochberg procedure to account for multiple testing. All p values presented are two-sided and were adjusted to give a global family-wise significance threshold p<0.05. To determine clinical relevance of the responses, we calculated relative change ((mean testbaseline − mean testfollow-up) / mean testbaseline) and Cohen’s d effect size with a method by Morris for pretest-posttest-control group designs.22 Furthermore, the relationship between NMSS and PDQ-8 change scores was explored using Spearman correlation.


Of 317 patients screened, 151 were treated with STN-DBS or MED and met inclusion criteria (figure 1). Of 151 patients in the final sample (100 males) with a mean age of 63.8±9.8 years, 67 patients underwent STN-DBS and 84 MED. The mean time to follow-up was 3.1 years. None of the patients in the MED group underwent STN-DBS during the 36-month follow-up period.

Baseline characteristics in the original and matched cohort

In the original cohort, patients in the STN-DBS group were significantly younger (61.3±8.5 years vs 65.8±10.4 years) with longer duration of PD (10.8±4.9 years vs 7.5±5.0 years) and had greater dopaminergic medication requirements (LEDD 1199.1±587.5 mg/day vs 778.1±407.1 mg/day) and more severe impairment in most of the clinical scales (table 1).

Table 1

Baseline characteristics in the original and matched sub-cohort

Propensity score matching resulted in a sub-cohort of 76 patients (38 patients in each group). Balance diagnostics indicated a good matching between the two groups, that is, no significant differences were found for the main demographic and clinical outcome parameters. Differences in NMSS domain scores were not significant except for sleep/fatigue. Baseline motor subscores and mini-mental state examination scores of the matched cohort are available in online supplementary table e–1.

Supplemental material

The results reported in this manuscript relate to the matched cohort. In addition, outcome changes in the original cohort are reported in online supplementary table e–2.

Clinical outcomes at baseline and 36-month follow-up

The time of follow-up assessment did not differ significantly between the groups (STN-DBS=3.1 ±0.24 years, MED=3.1 ±0.47 years; p=0.580). Longitudinal within-group changes and between-group differences are reported in table 2. The STN-DBS group significantly improved in NMSS total score, PDQ-8 SI, SCOPA-motor examination, and -complications. No significant longitudinal change was found for SCOPA-activities of daily living. In the MED group, outcome parameters did not change significantly. As expected, LEDD was significantly reduced in the STN-DBS group, while it remained stable in the MED group.

Table 2

Outcomes at baseline and 36-month follow-up in the matched cohort

Post-hoc exploratory analyses of NMSS domains (see table 2 and figure 2) resulted in significant beneficial effects of STN-DBS for the sleep/fatigue, urinary, sexual function and miscellaneous domains. The latter was driven by pain (baseline, 2.2±3.3; follow-up, 0.6±1.5, p=0.008), inability to smell/taste (baseline, 3.5±4.0; follow-up, 1.6±3.2, p=0.008) and weight changes (baseline, 1.8±3.1; follow-up, 0.6±1.5, p=0.025). A significant worsening in the STN-DBS group was found in the cardiovascular domain. In the MED group, the domains sleep/fatigue and urinary worsened significantly.

Figure 2

Non-motor Symptom Scale domains at baseline (blue) and 36-MFU (red) for the STN-DBS and MED groups in clustered box plots (A) and radar charts (B). Significant within-group changes of NMSS domains from baseline to 36-MFU are highlighted with a black star and significant between-group differences (STN-DBS vs MED) with a cross. (A) Outliers are illustrated with dots (2–3 SD) and extreme outliers with coloured stars (>3 SD). (B) NMSS domain scores are illustrated as percentage of maximum scores. Bigger areas represent more severe NMS impairment. In the STN-DBS group, NMSS sleep/fatigue, urinary, and miscellaneous domains significantly improved, and the cardiovascular domain significantly worsened from baseline to 36-MFU. In the MED group, NMSS sleep/fatigue and urinary domains significantly worsened from baseline to 36-MFU. Differences in NMSS domain scores were significant for sleep/fatigue, urinary, and miscellaneous domains. 36-MFU, 36-month follow-up; MED, standard-of-care medical treatment; STN-DBS, subthalamic nucleus deep brain stimulation.

Favouring STN-DBS, significant differences between change scores of STN-DBS and MED groups were found for the NMSS total score and the sleep/fatigue, urinary and miscellaneous domains. The latter was driven by the inability to smell/taste and pain. Furthermore, favouring STN-DBS, significant differences were observed for PDQ-8 SI, LEDD, SCOPA-motor examination and -complications.

Table 3 shows change scores and relative changes from baseline to 36-month follow-up for both groups and Cohen’s d effect size for the differences in change scores between the two treatment groups. In the DBS group, we observed six serious adverse events in five patients (skin perforation over battery site in two patients, disturbed wound healing, dopamine agonist withdrawal syndrome, suicide attempt, mania) which all resolved without major sequelae and 68 neurological adverse events in 40 patients and 25 psychiatric adverse events in 20 patients.

Table 3

Change scores, relative changes and effect sizes for matched cohorts

Spearman correlation analyses

There was a significant correlation of ‘moderate’ strength between change scores of NMSS total and PDQ-8 SI (see table 4). Explorative correlation analyses between change scores in NMSS domains and PDQ-8 SI revealed significant correlations for the domains sleep/fatigue (‘moderate’), mood/apathy, attention/memory, urinary and miscellaneous (all ‘weak’). The correlation within the miscellaneous domain was driven by the inability to smell/taste (rs=0.354, p=0.002) and weight changes (rs=0.231, p=0.045).

Table 4

Spearman correlations

In addition, online supplementary table e–3 reports Spearman correlations between change scores of PDQ-8 SI and NMSS for the separate treatment groups. In the STN-DBS and the MED groups, LEDD was not correlated to PDQ-8 SI or NMSS (all p>0.05). There was a ‘weak’ correlation between change scores in PDQ-8 SI and SCOPA-motor examination (rs=0.333, p=0.005), -activities of daily living (rs=0.367, p=0.002) and -motor complications (rs =0.316, p=0.006).


This study reports Class IIb evidence for beneficial effects of STN-DBS on a wide range of NMS in a controlled design at 36-month follow-up. Patients treated with STN-DBS experienced a better outcome of total NMS burden and specific non-motor aspects, such as sleep/fatigue, urinary symptoms, inability to smell/taste and pain, than patients treated with MED.

Motor symptoms, LEDD and (serious) adverse events

In line with previous studies, STN-DBS resulted in a significant improvement of motor aspects of PD.2 Confirming results from a study by Weaver et al, we observed an improvement of motor complications 36 months after STN-DBS.23 In accordance with this study, we observed a significant sustained 30% LEDD reduction at 36-month follow-up. In the MED group, standard-of-care practice10 stabilised motor examination, activities of daily living and motor complications over the 36-month course of the study with no significant increase of dopaminergic medication. As expected, changes of medication requirements differed significantly between the two groups. Based on the sum of patient years, this result is well within the range of studies with long-term follow-up periods.24

Non-motor symptoms

In line with a study by Holmberg et al reporting no significant difference of cardiovascular outcomes between the STN-DBS and MED group at 12-month follow-up,25 we observed no significant differences in the NMSS cardiovascular domain outcome at 36-month follow-up. However, when regarding longitudinal changes, we observed a significant worsening of cardiovascular symptoms in patients undergoing STN-DBS. The worsening of cardiovascular symptoms was unlikely to result from potential side effects of dopaminergic medication as these were significantly reduced postoperatively. In the STN-DBS group of the original cohort, cardiovascular symptoms at baseline were higher than in the matched sub-cohort and did not worsen at 36-month follow-up. Further studies are needed to investigate long-term effects of STN-DBS on cardiovascular symptoms in patients with high and low baseline cardiovascular impairments.

This is the first controlled study reporting an improvement of subjective sleep/fatigue symptoms. Beneficial effects of STN-DBS on subjective sleep symptoms have been demonstrated in studies with follow-up periods up to 3 years.6 7 26 27 Our results are in line with a recent study by Choi et al that found improvements in sleep dysfunction at 3-year follow-up.26 Our study adds to the evidence on beneficial effects on quality of sleep by demonstrating improvements in a controlled design. In contrast, Lilleeng et al reported no changes of sleep symptoms but a significant worsening of fatigue.28 However, this result was limited by the fact that the medication requirements at postoperative follow-ups remained high (LEDD at 1.0–1.5 years: 810 mg and at 6.0–9.0 years: 910 mg) and sleep and fatigue symptoms are common side effects of dopaminergic medication.29 As described in previous studies,30 31 possible other mechanisms of STN-DBS effects on sleep/fatigue, besides LEDD reduction, are a direct modulation of basal ganglia-thalamo-cortical loops resulting in improved nocturnal motor symptoms and a spread of current to regions near the STN, for example, the pedunculopontine nucleus.

In accordance with previous evidence, we observed no significant changes of mood/apathy in the STN-DBS group at 36month follow-up.23 However, this finding needs to be confirmed, as improved depression at 3-year follow-up after STN-DBS has been reported in Class IV evidence.32 Future studies addressing depression and its interplay with other neuropsychiatric symptoms such as anxiety, hypomania and alexithymia with longer follow-up are needed.33

Perceptual problems/hallucinations remained unchanged from baseline to 36-month follow-up in this study. In one of the few available studies on this issue, Yoshida et al reported improved hallucinations in STN-DBS at 6-month follow-up.34 Future studies are needed to investigate the relationship of hallucination outcome and dopaminergic medication, psychotropic co-medication and the co-dependency with other neuropsychiatric aspects of PD.

In our study, we observed no significant changes in attention/memory from baseline to 36-month follow-up and no inter-group differences between STN-DBS and MED. This is in line with a study by Funkiewiez et al that found no significant change of global cognitive functions and attention and memory subscales 3 years after surgery.32 This is also in line with a study by Zangaglia et al with multiple follow-up visits up to 3 years which reported no significant worsening in memory tasks after STN-DBS, whereas verbal fluency performance deteriorated in the STN-DBS group and logical executive function tasks were only impaired transiently at 12-month follow-up.35

In our cohort we found no significant change of gastrointestinal symptoms at 36-month follow-up. A study by Zibetti et al reported a significantly lower prevalence of constipation at 24-month follow-up.36 However, the authors did not employ validated scales and retrospectively extracted dichotomised data on the presence of constipation from patient charts.

To our knowledge, this is the first study to report beneficial 36-month effects of STN-DBS on urinary symptoms compared with a MED control group. This is in line with previous studies with shorter follow-up periods.3 Herzog et al reported ameliorations of bladder dysfunctions along with a modulation of blood flow of the posterior thalamus and the insular cortex.37 Nonetheless, long-term effects are unclear as Yamamoto et al found no significant change in urodynamic parameters 3 years after surgery, which may have resulted from a relatively small sample size (n=13).38

In this study, sexual functions improved significantly in the STN-DBS group. This is in line with a study which reported an improvement of sexual function in 21 male patients undergoing STN-DBS in PD at a 9–12 month follow-up.39 In contrast, Kurcova et al found no significant changes of sexual function in four female patients at a 4-month follow-up.40 Therefore, the effect of STN-DBS on sexual function may depend on demographic parameters such as sex of the patient.

This study shows beneficial effects on the miscellaneous domain. In our study, STN-DBS significantly improved pain at 36-month follow-up and this beneficial effect was also significant in the between-group comparison. These results are consistent with previous reports of beneficial effects of STN-DBS on pain in a 1–8-year follow-up after surgery.6 41 The mechanisms underlying the impact of STN-DBS on pain are not fully understood. Sensory gating has been discussed as a possible mechanism.31 To our knowledge, this study is also the first to report beneficial effects of STN-DBS on subjective olfactory function at 36-month follow-up. This is in line with studies showing beneficial immediate effects on odour identification6 and a PET study by Cury et al provided evidence for an increased glucose metabolism in the midbrain and right frontal lobe in patients with improved olfactory function following STN-DBS.42 Our study extends the time frame of beneficial effects to 36 months after STN-DBS. We found no significant difference between the STN-DBS and the MED group for excessive sweating and for weight changes. In the literature, only few studies have reported significant changes of these non-motor aspects of PD in patients undergoing STN-DBS and for both symptoms the sample sizes of available studies were small and the follow-up periods short.6 A study by Petry-Schmelzer et al provides a good overview of possible mechanisms of action (location-specific and general) which may mediate beneficial non-motor effects of STN-DBS.31

Quality of life

Confirming the results of earlier studies, we observed a significant QoL improvement at 36-month follow-up in the within-group analysis of the STN-DBS group.23 In the MED group, standard-of-care practice stabilised QoL over the 36-month course of the study. In line with previous studies with shorter follow-up periods up to 2 years,1 we observed a significantly greater QoL improvement after STN-DBS compared with MED at 36-month follow-up. QoL and NMS total burden correlated significantly in the overall matched cohort as well as in the STN-DBS and MED groups separately. QoL correlated significantly with sleep/fatigue, mood/apathy, attention/memory, urinary and the miscellaneous NMSS domains. QoL and motor examination were also significantly associated, but the correlation was weaker than with non-motor aspects, which is in line with previous studies and highlights the relative importance of NMS.43 QoL changes were significantly correlated to mood/apathy and attention/memory changes, indicating that although group level clinical effects per se were not significant, these non-motor outcomes were significantly related with postoperative QoL outcome on an individual level.


There are several limitations in our study to be considered. While propensity score matching is a well-established tool to precisely match baseline characteristics between two groups, this method can only be applied to parameters assessed clinically. Therefore, this method does not consider potentially relevant parameters, which were not measured, for example impulse control disorders. To account for this limitation, comparisons between the matched groups in this study were only carried out conservatively using independent samples tests. Diagnostic statistics indicated a well-balanced matching for the selected matching parameters, and this also led to a good balance for most NMSS domains. A notable difference, however, was the mean NMSS sleep/fatigue domain, which was higher at baseline in the STN-DBS group than in the MED group and may have contributed to the greater improvement observed in the STN-DBS group. Conversely, we found low baseline impairment in the STN-DBS group, for example, for the cardiovascular domain which may have left little room for additional improvements (floor effect). While propensity score matching cannot, therefore, replace randomised controlled trials, it can increase causal inference in observational trials and is an important tool in situations in which the randomised controlled design may not be suitable (eg in this study, investigations of long-term effects which would otherwise have resulted in withholding an effective therapy from severely affected patients for 3 years). In this context, one has to acknowledge that one of the reasons why patients in the MED group might have decided against DBS therapy at that time could have been their short disease duration (mean 7.5 years in the original cohort) which was even shorter than in the EARLYSTIM study (7.7 years).1

Furthermore, in the STN-DBS group, the matching resulted in a selection of less severely affected patients with shorter mean disease duration, as there were too few matching partners from the MED group within the defined calliper. This is important, as the observed effects of STN-DBS may be different in patients with very severe PD. Further studies are required to investigate the dependence of STN-DBS outcomes on the levels of baseline impairments and predict long-term non-motor outcomes. We chose a conservative calliper (0.25) and conducted strict balance diagnostics21 to implement a precise matching between the two groups. A narrower calliper would have resulted in smaller matched cohort sizes. Nonetheless, to our knowledge, this cohort including 76 matched patients was one of the biggest in studies of its kind. The multicentre design of our study may reduce bias caused by single centre studies and thus increase external validity. Another limitation was that the medication changes were not determined by an independent external panel as, for example, in the EARLYSTIM study.1 However, the standard-of-care clinical procedures of movement disorder specialists in each centre were based on the same criteria10 and clinical outcomes such as PDQ-8 SI, SCOPA-motor examination, -activities of daily living and -motor complications, and NMSS total burden did not worsen significantly in the MED group over the 36-month follow-up period, indicating a successful treatment of these aspects of PD. Furthermore, we analysed Cohen’s d effect size to help the interpretation of our results with regard to the clinical relevance of the observed changes. However, the ideal method for this purpose would have been the use of minimal clinically important changes, which, to our knowledge, has not been published for the NMSS and its domains yet.


Our controlled study with a 36-month follow-up provides Class IIb evidence for a beneficial effect of STN-DBS on NMS total burden and a wide range of NMS such as sleep/fatigue, urinary symptoms, pain and olfactory functions. The clinical relevance of these non-motor outcomes is highlighted by their significant correlation with QoL improvements after STN-DBS. Studies comparing QoL, motor and non-motor effects of different treatment options, for example, for STN-DBS and standard-of-care medical therapy as investigated in this study, will help to provide a basis on personalised medicine to patient’s individual PD profiles.



  • Contributors STJ: data analysis, drafting of the manuscript and figures; ASa, MSi, PAL, AR, JNP-S, PR, GRF, MTB and AA: data acquisition, critical revision of the manuscript; VV-V, KA and JE: surgical intervention, critical revision of the manuscript; JF, MSa and ASc: critical revision of the manuscript; PM-M, LT, KR-C and HSD: study concept and design, data acquisition, critical revision of the manuscript.

  • Funding This paper is independent research funded by the German Research Foundation (Grant KFO 219).

  • Competing interests STJ has received a travel grant from the German Academic Exchange Service. ASa reports no financial disclosures. VV-V is a member of the advisory boards and reports consultancies for Medtronic, Boston Scientific and St. Jude Medical. She received a grant from SAPIENS Steering Brain Stimulation. KA has received honoraria for educational meetings, travel and consultancy from Medtronic, St. Jude Medical and Boston Scientific. MSi has received honoraria from Bial, Britannia and Medtronic. JE reports no financial disclosures. PAL was funded by the Konrad-Adenauer-Foundation and the Koeln-Fortune Program. He reports travel grants from AbbVie. AR has received honorarium from UCB and was supported by a grant from Medtronic. JNP-S has received travel grants from Boston Scientific. PR has received a travel grant from AbbVie. GRF reports no financial disclosures. JF reports no financial disclosures. MSa has received honoraria for educational meetings/travel/accommodation from Medtronic, St. Jude Medical and UCB, grants from Parkinson's UK and Ipsen, and has acted as a consultant for Medtronic and St. Jude Medical. ASc reports personal consultancy and lecture fees from Medtronic Inc, Abbott and Boston Scientific, lecture fees from UCB, Teva Pharma, Bial and Zambon, and through his institution funding for his research from the German Research Council, the German Federal Ministry of Education and Health. MTB reports grants from Boston Scientific and Medtronic. AA reports personal consultancy fees from Zambon, AbbVie, Boehringer Ingelheim, GE, Neuroderm, Biogen, Bial, EVER Neuro Pharma, Therevance, Vectura grants from Chiesi Pharmaceuticals, Lundbeck, Horizon 2020 - PD_Pal Grant 825785, Ministry of Education University and Research (MIUR) Grant ARS01_01081, owns Patent WO2015110261-A1, owns shares from PD Neurotechnology Limited. PM-M has received honoraria from Editorial Viguera and Movement Disorder Society for lecturing in courses; from AbbVie for speaking in experts' meetings and from AbbVie and Zambon for participating in the advisory board of epidemiological studies. License fee payments for the King’s Parkinson’s Disease Pain Scale, and grants from the International Parkinson and Movement Disorder Society for development and validation of the MDS-Non-motor Symptom Scale. LT reports grants, personal fees and non-financial support from SAPIENS Steering Brain Stimulation, Medtronic, Boston Scientific and St. Jude Medical. KR-C has received funding from Parkinson's UK, National Institute of Health Research, UCB and the European Union; he received honoraria from UCB, Abbott, Britannia, US Worldmeds and Otsuka Pharmaceuticals, and acted as a consultant for AbbVie, UCB and Britannia. HSD work was funded by the Prof. Klaus Thiemann Foundation and the Felgenhauer Foundation and has received honoraria by Boston Scientific and Medtronic.

  • Patient consent for publication Not required.

  • Ethics approval Medical ethics committees of the participating centres approved the study protocol (master votes for Germany: Cologne, study number: 12-145, German Clinical Trials Register: DRKS00006735, and for the United Kingdom: National Research Ethics Service South East London REC 3, 10/H0808/141; National Institute of Health Research portfolio, number: 10084).

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data availability statement Data are available upon reasonable request. The data included in this study are available on request to the corresponding author. The data are not publicly available due to their containing information that could compromise the privacy of the participants.