Vol 56, No 5 (2022)
Research Paper
Published online: 2022-09-14

open access

Page views 7377
Article views/downloads 756
Get Citation

Connect on Social Media

Connect on Social Media

Carbohydrate metabolism and lipid profile in patients with Parkinson’s disease with subthalamic deep brain stimulation

Joanna Samborska-Ćwik1, Stanisław Szlufik1, Bartosz Migda2, Agata Marszalek3, Dariusz Koziorowski1
Pubmed: 36102235
Neurol Neurochir Pol 2022;56(5):441-450.

Abstract

Aim of the study. Assessment of potential effect of subthalamic nucleus deep brain stimulation (STN-DBS) on glucose metabolism in patients with Parkinson’s disease (PD).

Clinical rationale for the study. Although a valuable alternative to pharmacotherapy in advanced PD, STN-DBS is thought to negatively affect the cardiometabolic profile of patients (including body mass, lipid profile). Exacerbation of glucose metabolism dysregulation after DBS could therefore be assumed.

Material and methods. Two groups of patients with Parkinson’s disease were included: 20 treated pharmacologically (PHT) and 20 newly qualified for STN-DBS (DBS) — with the first assessment prior to surgery, and the second 11 months after surgery on average. Body mass index (BMI), plasma concentrations of total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol (HDL-C), triglycerides (TG) and glucose levels during a three-point oral glucose tolerance test were measured three times (median intervals between visits 12 and 14 months respectively).

Results. Significant differences between the two groups were noted with respect to changes in BMI, and serum concentration of TG and HDL-C over the course of the study. In the DBS group, a significant increase in BMI (26.42 vs. 27.24 kg/m2, p = 0.03) and TG level (103.8 vs. 142.8 mg/dL, p < 0.001) with a simultaneous decrease in HDL-C level (54.4 vs. 46 mg/dL, p < 0.01) was observed. Mean glucose level after oral glucose administration was lower in the DBS than in the PHT group (147.4 vs. 120.2 mg/dL, p = 0.03 after one hour and 109.9 vs. 82.3 mg/dL, p < 0.01 after two hours) during the second visit. Also inter-visit changes in fasting glucose levels (8.4 mg/dL in the PHT group and –5.8 mg/dL in the DBS group, p = 0.02) differed over the study duration.

Conclusions. Our observations are similar to previous ones indicating less favourable changes in BMI and some lipid fractions in patients treated surgically. Interestingly, such a trend was not observed for glucose metabolism parameters, suggesting that mechanisms other than simple body mass changes are involved in early biochemical changes after STN-DBS in PD patients. Clinical implications. The metabolic consequences of DBS require further investigation as an additional factor potentially affecting the outcome of therapy, and routine patient follow-up should not be limited to neurological and psychological assessments.

RESEARCH PAPER

Neurologia i Neurochirurgia Polska

Polish Journal of Neurology and Neurosurgery

2022, Volume 56, no. 5, pages: –450

DOI: 10.5603/PJNNS.a2022.0060

Copyright © 2022 Polish Neurological Society

ISSN: 0028-3843, e-ISSN: 1897-4260

Carbohydrate metabolism and lipid profile in patients with Parkinson’s disease with subthalamic deep brain stimulation

Joanna Samborska-Ćwik1Stanisław Szlufik1Bartosz Migda2Agata Marszalek3Dariusz Koziorowski1
1Department of Neurology, Faculty of Health Science, Medical University of Warsaw, Warsaw, Poland
2Diagnostic Ultrasound Lab, Department of Pediatric Radiology, Medical Faculty, Medical University of Warsaw, Warsaw, Poland
3Students’ Scientific Association of the Department of Neurology, Medical University of Warsaw, Warsaw, Poland

Address for correspondence: Joanna Samborska-Ćwik, Department of Neurology, Faculty of Health Science, Medical University of Warsaw, Kondratowicza 8 Str., 03-242 Warsaw, Poland; e-mail: josamborska@gmail.com

Received: 12.04.2022 Accepted: 25.07.2022 Early publication date: 14.09.2022

ABSTRACT
Aim of the study. Assessment of potential effect of subthalamic nucleus deep brain stimulation (STN-DBS) on glucose metabolism in patients with Parkinson’s disease (PD).
Clinical rationale for the study. Although a valuable alternative to pharmacotherapy in advanced PD, STN-DBS is thought to negatively affect the cardiometabolic profile of patients (including body mass, lipid profile). Exacerbation of glucose metabolism dysregulation after DBS could therefore be assumed.
Material and methods. Two groups of patients with Parkinson’s disease were included: 20 treated pharmacologically (PHT) and 20 newly qualified for STN-DBS (DBS) — with the first assessment prior to surgery, and the second 11 months after surgery on average. Body mass index (BMI), plasma concentrations of total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol (HDL-C), triglycerides (TG) and glucose levels during a three-point oral glucose tolerance test were measured three times (median intervals between visits 12 and 14 months respectively).
Results. Significant differences between the two groups were noted with respect to changes in BMI, and serum concentration of TG and HDL-C over the course of the study. In the DBS group, a significant increase in BMI (26.42 vs. 27.24 kg/m2, p = 0.03) and TG level (103.8 vs. 142.8 mg/dL, p < 0.001) with a simultaneous decrease in HDL-C level (54.4 vs. 46 mg/dL, p < 0.01) was observed. Mean glucose level after oral glucose administration was lower in the DBS than in the PHT group (147.4 vs. 120.2 mg/dL, p = 0.03 after one hour and 109.9 vs. 82.3 mg/dL, p < 0.01 after two hours) during the second visit. Also inter-visit changes in fasting glucose levels (8.4 mg/dL in the PHT group and –5.8 mg/dL in the DBS group, p = 0.02) differed over the study duration.
Conclusions. Our observations are similar to previous ones indicating less favourable changes in BMI and some lipid fractions in patients treated surgically. Interestingly, such a trend was not observed for glucose metabolism parameters, suggesting that mechanisms other than simple body mass changes are involved in early biochemical changes after STN-DBS in PD patients.
Clinical implications. The metabolic consequences of DBS require further investigation as an additional factor potentially affecting the outcome of therapy, and routine patient follow-up should not be limited to neurological and psychological assessments.
Key words: Parkinson’s disease, deep brain stimulation, glycaemia, carbohydrates, lipid profile
(Neurol Neurochir Pol 2022; 56 (5): –450)

Introduction

Recent decades have witnessed a debate regarding the relationship between Parkinson’s disease (PD) and diabetes mellitus (DM). Although some case-control and cross--sectional studies have found no positive correlation between PD and DM, most cohort and observational studies have reported a greater prevalence of PD in diabetic patients [1].

Various authors have postulated the involvement of convergent molecular and biochemical mechanisms in the development of DM and PD [2–6]. Commonly proposed pathomechanisms include shared genes such as the amyloid precursor protein, genes involved in lipid metabolism, or genes associated with autoimmunity [7]. Also, genes responsible for the most common monogenic forms of PD, such as SNCA, PINK1 or DJ-1 [8], are now considered to play a role in systemic glucose metabolism and the development of DM [9–11].

One possible link between diabetes mellitus type 2 (DMT2) and PD is protein misfolding. Although the dominant protein forming toxic deposits is different in each disease (α-synuclein in PD and islet amyloid polypeptide in DMT2), the co-occurrence of misfolded proteins has been found in both the pancreas and the brain of patients with neurodegenerative disorders [12]. In vitro cross-reactivity between various amyloidogenic proteins has been proven, which results in faster co-aggregation of mixed monomers than observed for each protein alone [13]. Outside the brain, α-synuclein is present e.g. in pancreatic β-cells and its accumulation may promote islet amyloid polypeptide fibrils formation [14]. Meanwhile, islet amyloid polypeptide has been found in brain cells, where it also may co-aggregate with α-synuclein, enhancing the neurodegenerative process [2].

Chronic or recurrent hyperglycaemia and insulin resistance, which are key features in DMT2, lead to numerous detrimental changes on a cellular level [1], for example persistent inflammation and impaired mitochondrial respiration resulting in overproduction of reactive oxygen species, which is particularly detrimental to cells with high energy requirements such as dopaminergic neurons [15–17].

The role of inflammation in neurodegenerative disorders has been raised by various researchers [18]. In animal models, metabolic inflammation has been shown to actually exacerbate dopaminergic neurons degeneration [19].

As a consequence of hyperglycaemia, the level of glycation agents, especially highly reactive methylglyoxal (MGO), is increased. MGO reacts with various proteins including α-synuclein. Glycation facilitates the formation of cross-links between α-synuclein monomers, thus promoting aggregation of its oligomers which are supposed to be even more toxic than larger conglomerates. Ubiquitin-dependent proteolysis of α-synuclein aggregates is inhibited by MGO due to both increased resistance of glycated proteins to proteasomal degradation and glycation of ubiquitin itself [1, 3, 20]. Moreover, the presence of advanced glycation end products is thought to contribute to neuronal death [21].

Apart from the epidemical association between DM and PD, the presence of DM, like other cardiovascular risk factors, undoubtedly negatively affects the clinical course of PD, contributing to faster progression of motor symptoms, prominent gait disturbances, and cognitive impairment [22, 23]. The underlying mechanisms are complex and still not fully understood. Animal studies have revealed for example a correlation between insulin resistance and more pronounced dopaminergic dysfunction [24]. Pagano et al. [22] found increased tau protein in cerebrospinal fluid and greater striatal dopaminergic deficits in patients with DM. As previously discussed, hyperglycaemia-induced protein glycation and cellular insulin resistance modulating aggregation of α-synuclein, β-amyloid and tau can accelerate the neurodegeneration process [20, 25]. Disrupted insulin signalling in the brain, apart from the loss of its physiological neuroprotective action, is supposed to decrease synaptic plasticity and affect cognition [1, 26]. Since orthostatic hypotension is one of the factors thought to contribute to cognitive decline in PD [27], a possible deleterious effect of DM is easy to predict: autonomic dysfunction, resulting from α-synuclein pathology in both peripheral and central autonomic nervous system in the course of PD, may be further aggravated by autonomic neuropathy, a common but underdiagnosed complication of DM [28].

As far as more pronounced gait disturbances are concerned, an overlap between exaggerated dopaminergic motor symptoms, cognitive impairment and peripheral diabetic neuropathy resulting in imbalance may be presumed [29].

Additionally, impaired glucose regulation has been reported in non-diabetic patients with PD, with higher glycaemia after glucose loading than in healthy controls [22, 30]. This could be attributed to non-sufficient function of pancreas β-cells secondary to dysautonomia progressing throughout the natural course of the disease [30], or to direct damage of β-cells by the deposition of pathological proteins as described above. As deposits of α-synuclein are early found in vagal nerve and its motor nucleus [31], and dorsal vagal complex is involved in glucoregulatory insulin action [32], the neurodegeneration in this area may be responsible for dysregulation of postabsorptive glucose metabolism.

Antiparkinsonian pharmacotherapy may also have an impact on glucose regulation, but the effect differs between different groups of drugs. Levodopa is considered to induce both hyperinsulinaemia and hyperglycaemia, whereas dopamine agonists are suspected to improve insulin sensitivity [6].

Clinical rationale for the study

Mutual interaction between hyperglycaemia, insulin resistance and neurodegeneration in the natural course of the disease may be further influenced by therapy for PD.

In our study, we tried to investigate whether subthalamic nucleus deep brain stimulation (STN-DBS) affects glucose variability in PD. Previous studies, including one conducted by our group, implied an unfavourable influence of DBS on body mass index (BMI) and the lipid profile [33, 34]. Therefore, it could theoretically be assumed that increases in body mass and triglycerides (TG) levels should be accompanied by impaired glucose tolerance.

Material and methods

Participants and methods

Data was collected from 40 patients of the Department of Neurology, Medical University of Warsaw, Poland (20 male, 20 female) who met the UK Parkinson’s disease Society Brain Bank criteria for the diagnosis of Parkinson’s disease. The mean age of participants was 56.1 ± 10.03 years, with an average disease duration of 9.2 ± 3.93 years.

The main exclusion criteria were: additional neurological disorders, other advanced antiparkinsonian therapy (such as duodopa, apomorphine, or former neurosurgical treatment), and a previous diagnosis of diabetes mellitus.

Data was collected during three consecutive visits with median intervals between visits of 12 months (V1–V2) and 14 months (V2–V3) respectively.

At the moment of inclusion (V1), all patients were receiving optimal pharmacological treatment, i.e. levodopa//dopamine agonist in monotherapy or co-administered, and//or combined with selegiline or amantadine.

Twenty of those individuals (12 males, eight females) were recruited from patients newly qualified to STN-DBS according to the CAPSIT-PD criteria (we called this the DBS group). In this group, the first assessment (V1) was prior to surgery and the second (V2) 11 months after the implantation of electrodes, on average.

The remainder (the PHT group; eight males, 12 females) only had pharmacological treatment, continued with appropriate adjustment for the whole study duration.

Patients reported no significant changes in dietary habits after inclusion.

During each visit, body mass to the nearest 0.1 kg was measured and BMI was calculated. Plasma concentrations of total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C) and TG were determined using Abbott’s Alinity CI system. Each time a three-point oral glucose tolerance test (OGTT) was conducted — glucose levels after fasting for at least eight hours and at both 60 and 120 minutes after oral administration of 75 g glucose were measured. All analytical procedures were conducted at the Clinical Laboratory of Masovian Brodnowski Hospital.

An assessment of the severity of symptoms against the Unified Parkinson’s Disease Rating Scale (UPDRS) was performed during each visit as well.

The study was approved by the Ethics Committee of the Medical University of Warsaw. The experiments were conducted in accordance with the ethical standards of the Declaration of Helsinki. All participants gave informed consent prior to their inclusion.

Statistical analysis

All calculations were performed with Statistica software (version 13.1 Dell Inc., Statsoft). In the case of some parameters with non-Gaussian distribution, logistic transformation of the data was performed and normal distribution proved for all variables with W Shapiro-Wilk Test. The data delivered represented parameters gathered on three consecutive visits. For analysis, we used ANOVA for repeated measurements. The variances homogeneity was assessed with Cochran’s C test and Bartlett’s test. The data sphericity (equality of variance differences between all experimental pairs) was assessed with Mauchley test with its “W” statistic. The smaller the “W” value, the greater the deviation from sphericity. If the assumption of sphericity was met, one-dimensional tests could be performed. Otherwise, the Greenhouse-Geisser and Huynh-Feldt corrections were applied and finally multivariate tests (MANOVA). In post-hoc analysis, the NIR Fisher test was used. In order to compare inter-visit changes in the assessed parameters (expressed as Δ equal to the value of the variable on the later visit minus the value of the variable on the earlier visit) between groups, the t-student test and the Mann–Whitney U-test (for normally and the non-normally distributed data respectively) were carried out.

In all calculations, p < 0.05 was considered to be significant. Appropriate plots of statistically significant results are presented.

Results

At the moment of inclusion, the groups were similarly distributed in terms of age (56.8 ± 11.41 years in the PHT group and 55.4 ± 8.60 years in the DBS group, p = 0.6637), disease duration (9.0 ± 3.99 in the PHT group and 9.3 ± 3.96 in the DBS group, p = 0.9254), disease stage assessed against Hoen-Yahr Scale (2.45 ± 0.51 in the PHT group and 2.85 ± 0.587 in the DBS group, p = 0.634), and motor symptom severity expressed as UPDRS part III score in “on” (patients receiving their standard pharmacotherapy) and “off” (after at least 12 h withdrawal of levodopa and 24-h without other antiparkinsonian drugs) condition (8.4 ± 5.21 “on” and 33.9 ± 9.67 “off” in the PHT group, 6.9 ± 3.45 “on” and 37.7 ± 7.87 “off” in the DBS group, p > 0.05 for both). The main data acquired is set out in Table 1.

Table 1. Descriptive statistics

Parameter

PHT (n = 20)

DBS (n = 20)

Visit

Mean (min–max)

SD ± 95% CI

Visit

Mean (min–max)

SD ± 95% CI

Age

56.8 (30–73)

11.4 ± 8.7–16.7

55.4 (40–70)

8.6 ± 6.5–12.6

Body weight

V1

81.3 (59–129.2)

17.3 ± 13.2–25.3

V1

77.9 (50.2–123.4) †

21 ± 16–30.7

V2

80.9 (58–124.9)

17.6 ± 13.4–25.8

V2

80.7 (58.8–123.8) †

20.4 ± 15.5–29.8

V3

80.4 (56–123.7)

18.8 ± 14.3–27.4

V3

80.1 (60–118.3)

17.2 ± 13.1–25.1

BMI

V1

28.5 (22.8–43.7)

5.1 ± 3.9–7.5

V1

26.4 (18.7–41.7) †§

6.4 ± 4.8–9.3

V2

28.4 (22.3–42.2)

5.2 ± 4–7.7

V2

27.4 (21.7–41.8) †

6.2 ± 4.7–9

V3

28.2 (21.5–41.7)

5.6 ± 4.3–8.2

V3

27.2 (22.3–40) §

5.2 ± 4–7.7

Carbohydrate parameters

Glucose

V1

80 (62–104) †§

10.6 ± 8–15.4

V1

88.9 (64–142)

17.4 ± 13.2–25.4

V2

87.6 (75–106) †

8.9 ± 6.7–13

V2

83.8 (64–117)

14.2 ± 10.8–20.8

V3

88.4 (60–140) §

17.1 ± 13–25

V3

83.1 (65–96)

8.4 ± 6.4–12.2

OGTT1

V1

133.8 (67–210)

42.2 ± 32.1–61.7

V1

117.9 (67–198)

29.5 ± 22.5–43.1

V2

147.4 (80–205)

37.9 ± 28.9–55.4

V2

120.2 (69–235)

40.3 ± 30.7–58.9

V3

133.4 (68–278)

43.4 ± 33–63.5

V3

134.6 (55–200)

43.7 ± 33.2–63.9

OGTT2

V1

97.5 (56–185) †§

35 ± 26.6–51.1

V1

86.8 (53–122) §

20.7 ± 15.7–30.2

V2

109.9 (46–175) †‡

35.3 ± 26.8–51.5

V2

82.3 (39–145) ‡

27.7 ± 21.1–40.5

V3

103.6 (62–300) ‡§

50.6 ± 38.5–74

V3

87.6 (57–124) ‡§

18.8 ± 14.3–27.5

HbA1c%

V1

5.4 (4.9–6)

0.3 ± 0.2–0.4

V1

5.3 (4–6.3)

0.5 ± 0.4–0.7

V2

5.4 (4.9–5.9)

0.3 ± 0.2–0.4

V2

5.4 (4.1–6)

0.4 ± 0.3–0.6

V3

5.5 (5–6.5)

0.3 ± 0.3–0.5

V3

5.5 (4.9–6.3)

0.3 ± 0.3–0.5

Lipid parameters

TC

V1

191.6 (101–281)

43.4 ± 33–63.3

V1

187.6 (130–295)

46.8 ± 35.6–68.4

V2

191.1 (130–269)

37.3 ± 28.4–54.5

V2

199.8 (140–348)

48.6 ± 37–71

V3

179.1 (80–244)

48.7 ± 37.1–71.2

V3

194.3 (125–307)

52.3 ± 39.8–76.4

HDL-C

V1

55.2 (37–78)

11.6 ± 8.9–17

V1

54.4 (31–89) §

17.4 ± 13.2–25.4

V2

55.1 (36–76)

11.4 ± 8.6–16.6

V2

51.1 (34–88) ‡

14.9 ± 11.3–21.7

V3

53.9 (31–80)

13.5 ± 10.2–19.7

V3

46 (30–72) ‡§

12.9 ± 9.8–18.8

LDL-C

V1

116.3 (39–185)

37.3 ± 28.4–54.5

V1

112.6 (51–184)

38.7 ± 29.4–56.5

V2

120.2 (78–197)

35 ± 26.6–51.1

V2

123.9 (57–254)

44.6 ± 33.9–65.1

V3

110.2 (31–173)

34.7 ± 26.4–50.7

V3

119.9 (53–229)

43.4 ± 33–63.4

TG

V1

101.6 (42–176) §

41.5 ± 31.6–60.7

V1

103.8 (48–217) †§

44.2 ± 33.6–64.6

V2

103.2 (46–208) ‡

43.7 ± 33.2–63.8

V2

120.9 (67–239) †‡

45.8 ± 34.8–66.9

V3

95.6 (43–180) ‡§

39.3 ± 29.9–57.5

V3

142.8 (60–346) ‡§

81.4 ± 61.9–118.9

Neurological examination

UPDRS III On

V1

8.4 (2–18) §

5.2 ± 4–7.6

V1

6.9 (2–19) §

3.5 ± 2.6–5

V2

8.1 (0–14) ‡

3.9 ± 3–5.7

V2

5.5 (3–18) ‡

3.3 ± 2.5–4.8

V3

6.5 (3–12) ‡§

2.4 ± 1.8–3.4

V3

5.4 (2–20) ‡§

3.9 ± 3–5.7

UPDRS III Off

V1

33.9 (17–52) †§

9.7 ± 7.4–14.1

V1

37.7 (27–55) †§

7.9 ± 6–11.5

V2

37.6 (21–56) †‡

8.4 ± 6.4–12.2

V2

41.5 (30–59) †‡

8.3 ± 6.3–12.2

V3

39.6 (25–57) ‡§

7.8 ± 5.9–11.3

V3

43.6 (32–61) ‡§

8 ± 6.1–11.7

Figure 1. UPDRS part III score in “on” and “off” conditions

In terms of neurological examination and symptom severity against UPDRS, there was significant inter-visit amelioration in motor subscale score assessed in “on” condition (p < 0.001). “On” condition was understood as normal pharmacotherapy in the PHT group and normal pharmacotherapy with the stimulation switched on in the DBS group. For both PHT and DBS patients, there was a reduction in the UPDRS part III “on” score in the later phase of the study (for V2-V3 interval – PHT p < 0.001; DBS p = 0.0048) and over the whole study duration (for V1–V3 interval — PHT p = 0.0023; DBS p = 0.0161) (Fig. 1). Meanwhile, a statistically significant deterioration in the UPDRS part III score in the “off” phase (for the PHT group as defined previously and for the DBS group with the stimulation switched off as well) was observed in both groups for each interval (V1–V2 PHT p < 0.001; DBS p < 0.001, V2–V3 PHT p = 0.0266; DBS p = 0.0172 and V1–V3 PHT p < 0.001; DBS p < 0.001), which matches the natural course of the disease (Fig. 1).

Although average levodopa equivalent daily dose (LEDD) was higher in the DBS group at inclusion, there was no significant difference in levodopa daily dose (LDD) between groups during the first visit. As predicted, both LEDD and LDD decreased over the study duration in patients treated surgically (LEDD 1,473.3 mg, LDD 1,235.0 mg at V1 and LEDD 734.5 mg, LDD 491.3 at V3, p < 0.01 for both para­meters). At the same time, reverse changes in both LEDD and LDD were noted in the PHT group (LEDD 1,167.5 mg at V1 and 1,465.5 mg at V3, p < 0.01, LDD 1,011.5 mg at V1 and 1,220.0 mg at V3, p = 0.0257), which is also typically observed in disease progression.

Longitudinal analysis was conducted to determine changes in all assessed parameters over time in both groups. We also compared inter-visit changes in the assessed parameters (expressed as Δ equal to the value of the variable on the later visit minus the value of the variable on the earlier visit) between the two groups of patients (Tab. 2).

Table 2. Comparison of changes in assessed parameters over whole study duration (V1–V3 interval) according to method of treatment of Parkinson’s disease (pharmacotherapy alone vs. pharmacotherapy and STN-DBS) expressed as Δ equal to value of variable on later visit minus value of variable on earlier visit

Group

P-value

PHT

DBS

ΔBMI_V1–V3

–0.32 ± 2.068

0.81 ± 2.245

0.0143

ΔGlu_V1–V3

8.4 ± 20.61

–5.8 ± 15.57

0.0186

ΔOGTT1_V1–V3

–0.4 ± 48.83

16.7 ± 59.01

0.3272

ΔOGTT2_V1–V3

6.1 ± 46.83

0.8 ± 26.54

0.6592

ΔHBA1c%_V1–V3

0.03 ± 0.394

0.12 ± 0.321

0.2423

ΔTC_V1–V3

–12.5 ± 40.9

6.8 ± 25.91

0.2766

ΔHDL-C_V1–V3

–1.3 ± 7.11

–8.4 ± 10.16

0.0146

ΔLDL-C_V1–V3

–6.1 ± 30.32

7.4 ± 23.65

0.3547

ΔTG_V1–V3

–6.1 ± 31.22

39.0 ± 46.78

< 0.001

We observed an increase in body weight in surgically treated patients between consecutive visits (p = 0.0074). A significant increase in body mass was noted between the first and the second visit (77.9 kg at V1 vs. 80.7 kg at V2, p = 0.0141). Consequently, BMI in this group of patients also increased significantly (p < 0.001), with the most pronounced change during the first year after implantation of electrodes and a later stabilisation (26.42 ± 6.36 kg/m2 upon V1, 27.41 ± 6.15 kg/m2 upon V2 and 27.24 ± 5.24 kg/m2 upon V3, for V1−V2 interval p = 0.0113 and V1−V3 p = 0.0338) (Fig 2). A statistically significant difference was noted for BMI changes over time depending on the therapy (ΔBMI V1−V2 –0.15 ± 1.360 kg/m2 in the PHT and 0.98 ± 1.020 kg/m2 in the DBS group, and ΔBMI V1−V3 -0.32 ± 2.068 kg/m2 in the PHT and 0.81 ± 2.245 kg/m2 in the DBS group, p < 0.001 and p = 0.0143 respectively).

Analysis concerning the lipid profile showed inter-visit variation of HDL-C and TG plasma concentration in patients from the DBS group. A significant decrease in HDL-C serum level was observed for the V2−V3 interval (p = 0.0038) and over the whole study duration (p < 0.001) (Fig. 2). A gradual increase in TG plasma concentration was noted during each consecutive visit (p < 0.001 for V1−V3 interval and p = 0.0366 and 0.0081 for V1−V2 and V2−V3 intervals respectively), whereas the reverse tendency for serum level of TG was observed in the PHT group (p < 0.001 for V1−V3 interval) with a more pronounced decrease between visits V2 and V3 (p < 0.001) (Fig. 2). Moreover, significant differences in ΔHDL-C and ΔTG over the whole study duration (ΔHDL-C –1.3 ± 7.11 mg/dL in the PHT vs. –8.4 ± 10.15 mg/dL in the DBS group, p = 0.0146 And ΔTG –6.1 ± 31.22 mg/dL in the PHT vs. 39.0 ± 46.78 mg/dL in the DBS group, p = 0.001), with a major alteration of the latter during the second phase of the study (ΔTG V2−V3 –7.7 ± 28.56 mg/dL in the PHT group vs. 21.9 ± 49.55 mg/dL in the DBS group, p = 0.0013), were noted for inter-group comparison.

The carbohydrate metabolism parameters were of particular interest to us. As far as assessed parameter variation over time in separate groups is concerned, significant changes were observed for fasting glucose level in the PHT group only (p < 0.001) and glucose concentration 120 minutes after oral administration of 75 g glucose in both groups (p = 0.0108 for the PHT group and p < 0.001 for the DBS group).

Further analysis revealed an increase in fasting glucose level within V1−V2 (p = 0.0356) and V1−V3 (p = 0.0206) intervals in the PHT group (Fig. 2). Also, inter-group comparison of fasting glucose changes between visits exposed significant differences depending on the method of treatment over the course of the study (Δglu V1−V3 8.4 ± 20.61 mg/dL in the PHT group and –5.8 ± 15.57 mg/dL in the DBS group, p = 0.0186) with major changes during the first phase of the study (Δglu V1−V2 7.6 ± 12.29 mg/dL in the PHT group and –5.1 ± 16.81 mg/dL in the DBS group, p = 0.0096).

In the Oral Glucose Tolerance Test (OGTT), the serum glucose concentration after 120 minutes in the PHT group increased significantly between the first and the second visit (p = 0.0007), with a slight reduction during further observation (p = 0.0000), but still with a significant increase over the study duration (p = 0.0066). In DBS patients, there was a minor increase in serum glucose concentration at the final point of OGTT during the V2−V3 interval (p = 0.004) and over the whole study duration (p = 0.0142). However, the mean plasma glucose levels in OGTT during the first assessment after DBS were significantly lower than in the PHT group (147.4 ± 37.94 vs. 120.2 ± 40.32 mg/dL, p = 0.0339 after one hour and 109.9 ± 35.29 vs. 82.3 ± 27.71 mg/dL, p < 0.0092 after two hours) with no such observation for other time points (Fig. 2). Results of inter-visit comparisons with post-hoc analysis are set out in Supplementary Table 3.

Discussion

Metabolic side effects of STN-DBS have been reported previously [33, 35]. The aim of this current study was to evaluate the possible impact of the method of treatment on glucose variability as another factor contributing to overall cardiovascular risk.

Our results concerning an increase in BMI in patients after the implantation of DBS with a later minor non-significant fall and stabilisation at higher than preoperative values are in good agreement with previous reports. As the majority of our patients were not underweight, the post-surgical weight gain could not be interpreted as normalisation of body mass after previous excessive loss, and therefore should be regarded rather as an adverse side-effect.

Similarly, the deterioration in the lipid profile, including a progressive increase in TG level and lowering of HDL-C plasma concentration during the first years after surgery, observed in our study population, concurs with previous findings [33, 34]. The exact underlying mechanism has not yet been established.

As previously discussed in other papers, these changes may potentially be attributed inter alia to the normalisation of energy expenditure due to reductions of muscle rigidity, tremor and levodopa-induced dyskinesia, as well as to the influence of the electric current on brain structures in proximity to the implanted electrodes [36−39]. The possible influence of levodopa dose reduction should also be taken into account. Some authors have reported that TG and TC serum levels are significantly lower in patients on levodopa than in untreated ones. These findings have been attributed to the peripheral inhibitory effect of levodopa-derived dopamine on the autonomic nervous system, resulting in improved insulin sensitivity with increased glucose uptake and reduced lipolysis [40].

There have been far fewer studies concerning the influence of DBS on glucose metabolism, and the results are inconsistent. Lammers et al. [41] reported no influence of DBS on endogenous glucose production, hepatic or peripheral insulin sensitivity, or basal plasma concentrations of glucose, insulin, or other glucoregulatory hormones. Conversely, Batisse-Lignier et al. [42] found that DBS regulates post-absorptive glucose metabolism.

Figure 2. Body mass index, serum glucose (fasting and 60 and 120 minutes after oral administration of 75 g glucose), high-density lipoprotein cholesterol, and triglycerides levels in both groups of patients during subsequent visits

In our study, a comparison of inter-visit changes in fasting glucose levels between groups (with an upward trend in the PHT group and a downward trend in the DBS group) indicates the possible influence of DBS on basal glucose metabolism.

The results of short-term glucose variability analysis during OGTT are ambiguous. The glucose level two hours after oral glucose loading tended to increase over the study duration in patients after STN-DBS. In considering the peripheral dopaminergic action described above for lipid metabolism and the reverse correlation between glucose level and LEDD described by some authors [30], this last observation could be attributed to rapid reduction in LEDD during the initial period after surgery, with subsequent stabilisation of metabolism at a new level. However, due to the opposite effects of levodopa and dopamine agonists on glucose metabolism reported in former studies [6], the role of modification of dopaminergic treatment in metabolic alteration after STN-DBS remains unclear.

On the other hand, although HbA1c%, which reflects average glycaemia over the preceding three months, did not differ significantly between groups or in particular groups over time, mean glycaemia levels after glucose loading during the first year after surgery seem to be significantly lower than in pharmacologically treated patients.

These results suggest that unfavourable changes in the lipid profile and body mass do not result in analogical alterations in glucose metabolism, implying a potential positive effect of DBS. Several mechanisms in which STN-DBS could affect glucose metabolism may be hypothesised.

The role of the hypothalamus as a metabolism regulatory centre is well established [32]. Additionally, some studies have suggested the involvement of the dorsal thalamus in glucose metabolism [43] and the subthalamic region was proven to be the anatomical localisation of the hypoglycaemia sensor [44].

Therefore, the spread of electric current outside the STN is likely to affect adjacent pathways involved in the regulation of glucose homeostasis, consequently altering peripheral glucose metabolism. As no changes in insulin levels after DBS have been found in previous studies [41, 42], the effect of local electric current on the hypothalamus-dorsal vagal complex axis, responsible among other things for the central regulation of hepatic glucose production [32], should be taken into account. The direct modulation of the STN-thalamus circuits also could not be excluded.

Although the results of recent studies are inconclusive, there is some evidence of the influence of DBS on dysautonomia in PD. As autonomic dysfunction is one of the proposed mechanisms of the impaired insulin secretion after glucose loading in PD [30], potential/possible normalisation of autonomic function after surgery could also contribute to improved glucose tolerance [45]. As autonomic dysfunction is one of the proposed mechanisms of the impaired insulin secretion after glucose loading in PD [30], the normalisation of autonomic function after surgery could also contribute to improved glucose tolerance.

An association between lower levels of brain-derived neurotrophic factor (BDNF) and insulin resistance has been described, implying its role in glucose metabolism and a potential pathogenic role in DMT2 [46]. Recent animal studies on a potential neuroprotective role of STN-DBS demonstrated that STN-DBS significantly increases striatal BDNF level and partially restores the normal corticostriatal BDNF relationship in the α-syn preformed fibril model in rats [47]. This last observation, if reproduced in human studies, could also indicate a regulatory effect of STN-DBS on postabsorptive glucose metabolism.

While taking into account the suggested role of hyper­glycaemia and insulin signalling dysregulation in the neurodegenerative process [5, 17, 26], a neutral or even favourable effect of DBS on glucose variability is a comforting finding, considering its supposed negative influence on some other cardiovascular factors.

The main limitations of our study were the small number of enrolled patients and the possible effect of additional variables that were hard to control, such as lifestyle or the dietary habits of patients. The actual effect of STN-DBS on carbohydrate metabolism may have been partially disguised by the influence of dopaminergic drugs. For ethical reasons, pharmacotherapy had to be adjusted after surgery and continued over the course of the study. Moreover, our intention was to assess the metabolic status of patients in real life situations, not in theoretical conditions.

The study population was relatively young, and glucose metabolism changes with age, which could potentially influence the results. However, DBS is the therapeutic option most often proposed to younger patients (due to exclusion criteria including cognitive dysfunction or presence of pronounced vascular changes in electrode placement trajectory) and average age at disease onset, as well as its duration, was similar in both groups (47.8 and 9.0 years respectively in the PHT group and 46.1 and 9.3 years in the DBS group) to avoid age-related differences in metabolism between groups. Also, longer life expectancy among such patients makes them more susceptible to the long-term consequences of metabolic disorders, additionally justifying efforts to identify potential metabolic side effects of the applied therapy.

Clinical implications/future directions

Our data is insufficient to conclude whether DBS norma­lises carbohydrate metabolism disturbed in the course of PD or accompanying levodopa treatment. However, it indicates that in this group of patients, unfavourable changes in BMI and the serum lipid profile are not necessarily accompanied by impaired glucose metabolism.

In the light of numerous reports of a deleterious effect of hyperglycaemia, insulin dysregulation and other cardiovascular risk factors on the clinical course of PD itself, the influence of DBS on global metabolism and autonomic function is an important issue potentially affecting the outcome of therapy.

An assessment of the actual impact of DBS on various metabolic processes requires larger population studies with longer follow-ups. Closer investigation of possible mechanisms underlying biochemical changes after DBS are necessary to identify predictors of metabolic alteration following surgery.

Conflicts of interests: None.

Funding: None.

References

  1. Cheong JLY, de Pablo-Fernandez E, Foltynie T, et al. The association between type 2 diabetes mellitus and Parkinson’s disease. J Parkinsons Dis. 2020; 10(3): 775–789, doi: 10.3233/JPD-191900, indexed in Pubmed: 32333549.
  2. Labandeira CM, Fraga-Bau A, Arias Ron D, et al. Parkinson’s disease and diabetes mellitus: common mechanisms and treatment repurposing. Neural Regen Res. 2022; 17(8): 1652–1658, doi: 10.4103/1673-5374.332122, indexed in Pubmed: 35017411.
  3. Biosa A, Outeiro TF, Bubacco L, et al. Diabetes mellitus as a risk factor for Parkinson’s disease: a molecular point of view. Mol Neurobiol. 2018; 55(11): 8754–8763, doi: 10.1007/s12035-018-1025-9, indexed in Pubmed: 29594935.
  4. Vidal-Martinez G, Yang B, Vargas-Medrano J, et al. Could α-synuclein modulation of insulin and dopamine identify a novel link between Parkinson’s disease and diabetes as well as potential therapies? Front Mol Neurosci. 2018; 11: 465, doi: 10.3389/fnmol.2018.00465, indexed in Pubmed: 30622456.
  5. Athauda D, Foltynie T. Insulin resistance and Parkinson’s disease: A new target for disease modification? Prog Neurobiol. 2016; 145-146: 98–120, doi: 10.1016/j.pneurobio.2016.10.001, indexed in Pubmed: 27713036.
  6. Camargo Maluf F, Feder D, Alves de Siqueira Carvalho A. Analysis of the relationship between type II diabetes mellitus and Parkinson’s disease: a systematic review. Parkinsons Dis. 2019; 2019: 4951379, doi: 10.1155/2019/4951379, indexed in Pubmed: 31871617.
  7. Zhang Xi, Fan Yu, Luo Y, et al. Lipid metabolism is the common pathologic mechanism between Type 2 diabetes mellitus and Parkinson’s disease. Int J Med Sci. 2020; 17(12): 1723–1732, doi: 10.7150/ijms.46456, indexed in Pubmed: 32714075.
  8. Milanowski ŁM, Ross OA, Friedman A, et al. Genetics of Parkinson’s disease in the Polish population. Neurol Neurochir Pol. 2021; 55(3): 241–252, doi: 10.5603/PJNNS.a2021.0013, indexed in Pubmed: 33539026.
  9. Deas E, Piipari K, Machhada A, et al. PINK1 deficiency in β-cells increases basal insulin secretion and improves glucose tolerance in mice. Open Biol. 2014; 4: 140051, doi: 10.1098/rsob.140051, indexed in Pubmed: 24806840.
  10. Eberhard D, Lammert E. The Role of the Antioxidant Protein DJ-1 in Type 2 Diabetes Mellitus. Adv Exp Med Biol. 2017; 1037: 173–186, doi: 10.1007/978-981-10-6583-5_11, indexed in Pubmed: 29147909.
  11. Rodriguez-Araujo G, Nakagami H, Takami Y, et al. Low alpha-synuclein levels in the blood are associated with insulin resistance. Sci Rep. 2015; 5: 12081, doi: 10.1038/srep12081, indexed in Pubmed: 26159928.
  12. Martinez-Valbuena I, Valenti-Azcarate R, Amat-Villegas I, et al. Mixed pathologies in pancreatic β cells from subjects with neurodegenerative diseases and their interaction with prion protein. Acta Neuropathol Commun. 2021; 9(1): 64, doi: 10.1186/s40478-021-01171-0, indexed in Pubmed: 33832546.
  13. Horvath I, Wittung-Stafshede P. Cross-talk between amyloidogenic proteins in type-2 diabetes and Parkinson’s disease. Proc Natl Acad Sci U S A. 2016; 113(44): 12473–12477, doi: 10.1073/pnas.1610371113, indexed in Pubmed: 27791129.
  14. Mucibabic M, Steneberg P, Lidh E, et al. α-Synuclein promotes IAPP fibril formation in vitro and β-cell amyloid formation in vivo in mice. Sci Rep. 2020; 10(1): 20438, doi: 10.1038/s41598-020-77409-z, indexed in Pubmed: 33235246.
  15. Hong CT, Chen KY, Wang W, et al. Insulin resistance promotes Parkinson’s disease through aberrant expression of α-synuclein, mitochondrial dysfunction, and deregulation of the polo-like kinase 2 Signaling. Cells. 2020; 9(3), doi: 10.3390/cells9030740, indexed in Pubmed: 32192190.
  16. Pérez-Taboada I, Alberquilla S, Martín ED, et al. Diabetes causes dysfunctional dopamine neurotransmission favoring nigrostriatal degeneration in mice. Mov Disord. 2020; 35(9): 1636–1648, doi: 10.1002/mds.28124, indexed in Pubmed: 32666590.
  17. Renaud J, Bassareo V, Beaulieu J, et al. Dopaminergic neurodegeneration in a rat model of long-term hyperglycemia: preferential degeneration of the nigrostriatal motor pathway. Neurobiol Aging. 2018; 69: 117–128, doi: 10.1016/j.neurobiolaging.2018.05.010, indexed in Pubmed: 29890391.
  18. Madetko N, Migda B, Alster P, et al. Platelet-to-lymphocyte ratio and neutrophil-tolymphocyte ratio may reflect differences in PD and MSA-P neuroinflammation patterns. Neurol Neurochir Pol. 2022; 56(2): 148–155, doi: 10.5603/PJNNS.a2022.0014, indexed in Pubmed: 35118638.
  19. Wang L, Zhai YQ, Xu LL, et al. Metabolic inflammation exacerbates dopaminergic neuronal degeneration in response to acute MPTP challenge in type 2 diabetes mice. Exp Neurol. 2014; 251: 22–29, doi: 10.1016/j.expneurol.2013.11.001, indexed in Pubmed: 24220636.
  20. Vicente Miranda H, El-Agnaf OMA, Outeiro TF. Glycation in Parkinson’s disease and Alzheimer’s disease. Mov Disord. 2016; 31(6): 782–790, doi: 10.1002/mds.26566, indexed in Pubmed: 26946341.
  21. An S, Lee B, Byun K, et al. Microglial AGE-albumin is critical for neuronal death in Parkinson’s disease: a possible implication for theranostics. International Journal of Nanomedicine. 2016; Volume 10: 281–292, doi: 10.2147/ijn.s95077.
  22. Pagano G, Polychronis S, Wilson H, et al. Diabetes mellitus and Parkinson disease. Neurology. 2018, doi: 10.1212/WNL.0000000000005475, indexed in Pubmed: 29626177.
  23. de Pablo-Fernández E, Courtney R, Rockliffe A, et al. Faster disease progression in Parkinson’s disease with type 2 diabetes is not associated with increased α-synuclein, tau, amyloid-β or vascular pathology. Neuropathol Appl Neurobiol. 2021; 47(7): 1080–1091, doi: 10.1111/nan.12728, indexed in Pubmed: 33969516.
  24. Morris JK, Bomhoff GL, Gorres BK, et al. Insulin resistance impairs nigrostriatal dopamine function. Exp Neurol. 2011; 231(1): 171–180, doi: 10.1016/j.expneurol.2011.06.005, indexed in Pubmed: 21703262.
  25. Bednarz K, Siuda J. Alzheimer’s disease and type 2 diabetes mellitus: similarities in pathomechanisms lead to therapeutic opportunities. Neurol Neurochir Pol. 2021; 55(5): 418–428, doi: 10.5603/PJNNS.a2021.0056, indexed in Pubmed: 34355790.
  26. Fiory F, Perruolo G, Cimmino I, et al. The relevance of insulin action in the dopaminergic system. Front Neurosci. 2019; 13: 868, doi: 10.3389/fnins.2019.00868, indexed in Pubmed: 31474827.
  27. Kwaśniak-Butowska M, Dulski J, Pierzchlińska A, et al. Cardiovascular dysautonomia and cognition in Parkinson’s disease - a possible relationship. Neurol Neurochir Pol. 2021; 55(6): 525–535, doi: 10.5603/PJNNS.a2021.0040, indexed in Pubmed: 34037978.
  28. Agashe S, Petak S. Cardiac autonomic neuropathy in diabetes mellitus. Methodist Debakey Cardiovasc J. 2018; 14(4): 251–256, doi: 10.14797/mdcj-14-4-251, indexed in Pubmed: 30788010.
  29. Tipton PW. Dissecting parkinsonism: cognitive and gait disturbances. Neurol Neurochir Pol. 2021; 55(6): 513–524, doi: 10.5603/PJNNS.a2021.0084, indexed in Pubmed: 34817060.
  30. Marques A, Dutheil F, Durand E, et al. Glucose dysregulation in Parkinson’s disease: Too much glucose or not enough insulin? Parkinsonism Relat Disord. 2018; 55: 122–127, doi: 10.1016/j.parkreldis.2018.05.026, indexed in Pubmed: 29866628.
  31. Braak H, Del Tredici K, Rüb U, et al. Staging of brain pathology related to sporadic Parkinson’s disease. Neurobiol Aging. 2003; 24(2): 197–211, doi: 10.1016/s0197-4580(02)00065-9, indexed in Pubmed: 12498954.
  32. Fujikawa T. Central regulation of glucose metabolism in an insulin-dependent and -independent manner. J Neuroendocrinol. 2021; 33(4): e12941, doi: 10.1111/jne.12941, indexed in Pubmed: 33599044.
  33. Rieu I, Pereira B, Derost P, et al. Does deep brain stimulation of the subthalamic nucleus induce metabolic syndrome in Parkinson’s disease? e-SPEN, the European e-Journal of Clinical Nutrition and Metabolism. 2011; 6(3): e126–e130, doi: 10.1016/j.eclnm.2011.03.002.
  34. Samborska-Ćwik J, Szlufik S, Friedman A, et al. Influence of bilateral subthalamic nucleus deep brain stimulation on the lipid profile in patients with Parkinson’s disease. Front Neurol. 2020; 11: 563445, doi: 10.3389/fneur.2020.563445, indexed in Pubmed: 33154734.
  35. Steinhardt J, Münte TF, Schmid SM, et al. A systematic review of body mass gain after deep brain stimulation of the subthalamic nucleus in patients with Parkinson’s disease. Obes Rev. 2020; 21(2): e12955, doi: 10.1111/obr.12955, indexed in Pubmed: 31823457.
  36. Růžička F, Jech R, Nováková L, et al. Weight gain is associated with medial contact site of subthalamic stimulation in Parkinson’s disease. PLoS One. 2012; 7(5): e38020–310, doi: 10.1371/journal.pone.0038020, indexed in Pubmed: 22666437.
  37. Montaurier C, Morio B, Bannier S, et al. Mechanisms of body weight gain in patients with Parkinson’s disease after subthalamic stimulation. Brain. 2007; 130(Pt 7): 1808–1818, doi: 10.1093/brain/awm113, indexed in Pubmed: 17535833.
  38. Guimarães J, Moura E, Vieira-Coelho MA, et al. Weight variation before and after surgery in Parkinson’s disease: a noradrenergic modulation? Mov Disord. 2012; 27(9): 1078–1082, doi: 10.1002/mds.25063, indexed in Pubmed: 22700383.
  39. Balestrino R, Baroncini D, Fichera M, et al. Weight gain after subthalamic nucleus deep brain stimulation in Parkinson’s disease is influenced by dyskinesias’ reduction and electrodes’ position. Neurol Sci. 2017; 38(12): 2123–2129, doi: 10.1007/s10072-017-3102-7, indexed in Pubmed: 28913772.
  40. Scigliano G, Ronchetti G, Girotti F, et al. Sympathetic modulation by levodopa reduces vascular risk factors in Parkinson disease. Parkinsonism Relat Disord. 2009; 15(2): 138–143, doi: 10.1016/j.parkreldis.2008.04.036, indexed in Pubmed: 18556236.
  41. Lammers NM, Sondermeijer BM, Twickler ThB, et al. Subthalamic nucleus stimulation does not influence basal glucose metabolism or insulin sensitivity in patients with Parkinson’s disease. Front Neurosci. 2014; 8: 95, doi: 10.3389/fnins.2014.00095, indexed in Pubmed: 24860415.
  42. Batisse-Lignier M, Rieu I, Guillet C, et al. Deep brain stimulation of the subthalamic nucleus regulates postabsorptive glucose metabolism in patients with Parkinson’s disease. J Clin Endocrinol Metab. 2013; 98(6): E1050–E1054, doi: 10.1210/jc.2012-3838, indexed in Pubmed: 23633215.
  43. Arbelaez AM, Powers WJ, Videen TO, et al. Attenuation of counterregulatory responses to recurrent hypoglycemia by active thalamic inhibition: a mechanism for hypoglycemia-associated autonomic failure. Diabetes. 2008; 57(2): 470–475, doi: 10.2337/db07-1329, indexed in Pubmed: 18003752.
  44. Cranston I, Reed LJ, Marsden PK, et al. Changes in regional brain (18)F-fluorodeoxyglucose uptake at hypoglycemia in type 1 diabetic men associated with hypoglycemia unawareness and counter-regulatory failure. Diabetes. 2001; 50(10): 2329–2336, doi: 10.2337/diabetes.50.10.2329, indexed in Pubmed: 11574416.
  45. Bellini G, Best LA, Brechany U, et al. Clinical impact of deep brain stimulation on the autonomic system in patients with Parkinson’s disease. Mov Disord Clin Pract. 2020; 7(4): 373–382, doi: 10.1002/mdc3.12938, indexed in Pubmed: 32373653.
  46. Krabbe KS, Nielsen AR, Krogh-Madsen R, et al. Brain-derived neurotrophic factor (BDNF) and type 2 diabetes. Diabetologia. 2007; 50(2): 431–438, doi: 10.1007/s00125-006-0537-4, indexed in Pubmed: 17151862.
  47. Miller KM, Patterson JR, Kochmanski J, et al. Striatal afferent BDNF is disrupted by synucleinopathy and partially restored by STN DBS. J Neurosci. 2021; 41(9): 2039–2052, doi: 10.1523/JNEUROSCI.1952-20.2020, indexed in Pubmed: 33472823.



Neurologia i Neurochirurgia Polska