INTRODUCTION

The association between type II diabetes mellitus (DM) and Parkinson’s disease (PD) has been recognized for many decades [1,2], although the evidence remains controversial [3,4]. A recent large cohort study reported that type II DM was associated with more severe motor and nonmotor symptoms [5]. Insulin resistance in the 2-h oral glucose tolerance test (OGTT) is associated with dementia in patients with PD [6]. Hyperglycemia, defined by the level of glycated hemoglobin (HbA1c), may be associated with unfavorable motor outcomes in PD patients [4]. However, DM, as defined by medical records, is not an independent predictor of unfavorable motor outcomes [4]. Conversely, these results are not consistent with those of other studies. In one meta-analysis, there was no conclusive evidence of an association between type II DM and PD [7]. DM was not a risk factor for PD in a prospective study [8]. The definition of DM (formal glucose measurement, medical record review, or self-report) was suggested as one of the reasons for these inconsistent results [9].

The findings regarding glycemic dysregulation in PD patients with formal glucose measurements have been more consistent. The fasting glucose level was a significant predictor of cognitive decline in PD patients compared to healthy controls when DM patients were excluded [10]. In an Australian longitudinal population-based study, subjects with type II DM defined by 2-h OGTT or fasting plasma glucose measurement and adults with stable high blood glucose were at risk of poorer cognition [11]. Fasting plasma glucose (FPG) variability may contribute to the risk of developing PD without DM [12]. In a biomarker study for PD, diabetes defined by FPG measurements was associated with faster motor progression and cognitive decline [13]. However, little is known about the longitudinal progression of FPG levels in PD patients and their outcomes.

HbA1c is considered a stable measure of blood glucose over the previous two to three months and can be measured without fasting. However, it can be affected by hemoglobin status and sex. HbA1c could underestimate fasting glucose in men compared to women [14]. Compared to HbA1c, FPG measures the immediate blood glucose level, making it more sensitive to diet and exercise status. FPG level might be helpful for monitoring patient status and providing precise treatment options, such as lifestyle modifications, for individual PD patients [15]. However, FPG is often considered cumbersome to analyze due to its variability.

Based on these findings, we aimed to investigate whether euglycemia, intermediate glycemia, and hyperglycemia, as defined by FPG levels, can predict longitudinal motor and nonmotor outcomes in PD patients. Cognition, cardiac sympathetic denervation, orthostatic hypotension, and postvoid residual urine volume were investigated as nonmotor outcomes.

MATERIALS & METHODS

Patients

This study was approved by the Institutional Review Board of Seoul St. Mary’s Hospital (KC17ONSI0423 and KC22RISI0610). All subjects provided written informed consent for the longitudinal data registration (KC17ONSI0423), and consent for this study to use some of these data was exempted (KC22RISI0610). This research was conducted in accordance with relevant guidelines and regulations.

This study included 201 patients who were diagnosed with PD according to the UK Brain Bank criteria between January 2015 and January 2020 [16]. Clinical information was collected regarding age, sex, body mass index (BMI), disease duration from symptom onset, hypertension status, DM status, coronary artery disease status, previous stroke status, dyslipidemia status, and smoking habits. DM was defined based on expert assessments documented in the medical records. All patients were evaluated by the Unified Parkinson’s Disease Rating Scale (UPDRS) Part I to Part III and were classified by the modified Hoehn and Yahr (H&Y) stage to measure motor symptom status. The levodopa equivalent daily dose (LEDD) was calculated by a standardized formula [17]. Postvoid residual urine volume was measured using bladder ultrasound, and a head-up tilt test was conducted to assess nonmotor burden. ¹²³I-meta-iodobenzylguanidine (¹²³IMIBG) scintigraphy was performed to assess cardiac sympathetic innervation status [18]. Apolipoprotein E (APOE) polymorphisms were evaluated, and the subjects were categorized based on the presence of the APOE ε4 allele.

The exclusion criteria were 1) a normal dopamine transporter scan based on the Movement Disorder Society Clinical Diagnostic Criteria for PD [19], 2) neurologic abnormalities related to atypical PD or secondary parkinsonism, and 3) abnormalities in the basal ganglia or cerebellum on structural magnetic resonance imaging (MRI).

Fasting plasma glucose level

Fasting was defined as abstaining from caloric intake for at least 8 hours [20]. The patients visited the outpatient clinic every 2–6 months from the time they were diagnosed with PD. Most of the FPG levels were measured at the first visit to the outpatient clinic and were followed up as needed. The closest FPG value to the day of UPDRS evaluation was used to categorize patients into three groups: euglycemia (70 mg/dL < FPG < 100 mg/dL), intermediate glycemia (100 mg/dL ≤ FPG < 126 mg/dL), and hyperglycemia (FPG ≥ 126 mg/dL) groups. No patients with hypoglycemia (FPG ≤ 70 mg/dL) were observed in our study. The time interval between the date of the closest FPG measurement and the endpoint outcome evaluation was calculated for each patient. Since each patient had a unique time interval between the FPG measurement date and outcome evaluation and the distribution of patients among FPG time gap ranges was not homogeneous, we stratified the time interval by year. We then calculated correlation coefficients for each cumulative patient total while controlling for other significant covariates, aiming to compare the correlation between FPG and the UPDRS III score or neuropsychological test score based on the FPG time interval (Supplementary Table 1 in the online-only Data Supplement).

Group-based trajectory modeling

All longitudinal FPG values of each patient’s medical records from before PD diagnosis to after diagnosis (range: -140–114 months) were collected for group-based trajectory modeling (GBTM) [21]. The longitudinal trajectory of FPG measurements was modeled using the censored normal distribution. We tested different groups and different shapes of trajectories in a stepwise manner (Supplementary Table 2 in the online-only Data Supplement) and then assessed model fit using Bayesian information criterion (BIC) values. Models with a maximum (i.e., least negative) BIC score or similar BIC scores were compared with the following criteria: 1) parsimonious model that fit the data well; 2) an average posterior probability (AvPP) value > 0.7 for each group; 3) odds of correct classification based on the posterior probability of group membership > 5 for each group; 4) lowest mismatch value; and 5) approximate group membership > 5% for each trajectory group (Supplementary Table 3 in the onlineonly Data Supplement).

Neuropsychological tests

General cognitive status and dementia severity were evaluated using the Korean version of the Mini-Mental State Examination, the Clinical Dementia Rating scale, and the Global Deterioration Scale. Five domains of cognition, including attention, language and related functions, visuospatial functions, memory, and frontal/executive functions, were assessed using the comprehensive neuropsychological tests of the Seoul Neuropsychological Screening Battery-II [22]. The attention domains were examined with the vigilance test, the digit span forward and backward test, and letter cancellation. The language domain included tests of spontaneous speech, comprehension, repetition, confrontation naming (the Korean-Boston Naming Test), reading, and writing. The language-related functions included Gerstmann’s syndrome screening (finger naming, right-left orientation, body part identification, and calculation) as well as tests for buccofacial and limb apraxia. Visuospatial function was assessed by the Rey Complex Figure Test (RCFT) and the clock drawing test. For the memory domain, the Seoul Verbal Learning Test and RCFT were utilized to assess the subdomains of immediate recall, delayed recall, and recognition. Frontal/executive function was assessed with the Controlled Oral Word Association Test, the Korean-Color Word Stroop Test, Digit Symbol Coding, the Korean-Trail Making Test-Elderly version, and several motor and copy tasks that measure motor regulation and perseveration, such as Luria loops, alternating squares and triangles, and alternating hand movements [22].

Each quantifiable neuropsychological test score was converted into a standardized score (z score) based on age-, sex-, and education-specific norms. The scores were classified as abnormal when they were more than 1.5 standard deviations (SD) from the norms. The z scores of each test were calculated for analyses. In domains with multiple z scores, the average scores of each domain were defined as the representative values [23]. Patients with mild cognitive impairment (MCI) had unimpaired functional activities of daily living, as verified by the activities of daily living assessment, and had scores 1.5 SD or more below normative data on at least two measures within at least one of the five cognitive domains according to the Movement Disorder Society Task Force Level II criteria (comprehensive assessment) [24,25]. Depending on the impaired cognitive domain, individuals with MCI were further stratified into MCI with memory impairment (amnestic MCI) and MCI without memory impairment (nonamnestic MCI) groups [26]. When the cognitive deficits were severe enough to impair daily life, dementia was diagnosed according to the Movement Disorder Society Task Force diagnostic criteria [27].

¹²³I-meta-iodobenzylguanidine myocardial scintigraphy

¹²³I-MIBG scintigraphy was conducted with a dual-head camera equipped with a low-energy, high-resolution collimator (Siemens, Munich, Germany/Infinia, GE Healthcare, Chicago, IL, USA). The data were collected 30 minutes (early) and 2 hours (delayed) after injecting 111 MBq of ¹²³I-MIBG. Tracer uptake was measured within the heart and mediastinum to calculate the heart-to-mediastinum (H/M) ratio. The myocardial ¹²³IMIBG washout rate was calculated as ([early H/M – late H/M]/early H/M) × 100 [28]. We used delayed H/M for the assessment of postganglionic presynaptic sympathetic failure. An abnormal delayed H/M ratio was defined as < 1.78 [29].

Head-up tilt test

The head-up tilt test was performed using a Manumed Special Tilt 1 Section (Enraf Nonius, Rotterdam, The Netherlands) after discontinuation of any antihypertensive medications for at least seven days. Baseline supine blood pressure (BP) was measured every 5 minutes, and standing BP at a 60° tilt was measured at 0, 3, 5, 10, 15, and 20 minutes. The mean supine BP between 5 minutes and 20 minutes was calculated, and the lowest standing BP at 3 or 5 minutes was used to diagnose orthostatic hypotension (OH). Supine hypertension (SH) was defined as a mean supine BP ≥ 140/90 mmHg [30]. The orthostatic blood pressure changes of systole (ΔSBP) and diastole (ΔDBP) were calculated. OH was defined as a decrease in BP of at least 20 mmHg systolic and/or 10 mmHg diastolic while standing. In patients with SH, ΔSBP and/or ΔDBP were ≥ 30/15 mmHg [31].

Statistical analysis

Descriptive statistics included the mean ± SD or frequency for each clinical characteristic. Analysis of variance with Tukey’s post hoc test was used for continuous variables, and the χ² test was used for categorical variables. For nonparametric covariates, the Kruskal–Wallis test was used for continuous variables, and Fisher’s exact test was used for categorical variables. The correlations between the FPG level and the UPDRS and neuropsychological domain composite z scores were analyzed by partial Spearman correlation controlling for age, sex, smoking status, BMI, DM status, education years, disease duration, APOE ε4 carrier status, and days between the closest FPG measurement and outcome evaluation (UPDRS evaluation and neuropsychological test). The time interval between FPG measurement and outcome evaluation was stratified by cumulative year, and the total number of patients in each cumulative time interval was analyzed.

Multivariate logistic regression models were fitted to examine the relationships between glycemic status and motor outcomes. In the logistic model, the event category was defined as the motor outcome (H&Y stage ≥ 2 in Model 1 and H&Y stage ≥ 3 in Model 2) or cognitive outcome (conversion from normal cognition to MCI in Model 3). Each model was adjusted for covariates, including age, sex, DM, smoking status, education years, disease duration, delayed H/M ratio (< 1.78), SH, OH, APOE ε4 carrier status, and FPG time interval. The odds ratio (OR) was adjusted for the same covariates.

Multivariate Cox proportional hazards regression analyses were performed to compare the motor and cognitive outcomes among the FPG groups after controlling for age, sex, diabetes status, smoking status, and APOE ε4 carrier status. For cognitive outcomes, the H&Y stage was also controlled. Endpoints were defined as H&Y stage ≥ 2 in Model 1 and H&Y stage ≥ 3 in Model 2. For cognitive outcome, the diagnosis of MCI was defined as the endpoint of the analysis, and the disease duration to MCI was compared among the three FPG groups. Since dementia conversion did not fit the regression model, we reanalyzed the cognitive outcome data by combining MCI and dementia data under a single label to determine whether the small number of dementia patients affected the results disproportionately. The ZPH option was used for diagnostics, based on the weighted Schoenfeld residuals, to check the proportional hazards assumption. All parameters satisfied the proportional hazard assumption except for the sex parameter in Model 2, which was controlled through stratification. Regression statistics with tolerance and variance inflation factors (VIFs) were used to exclude collinearity among the covariates. In addition, subgroup analysis of motor outcomes was conducted based on the presence or absence of DM through stratification to eliminate any potential interaction between DM and FPG level that might affect the outcome. All the statistical analyses were performed using SAS (version 9.4; SAS Institute Inc., Cary, NC, USA) statistical software, and SAS Proc Traj obtained from http://www.andrew.cmu.edu/user/bjones/index.htm was used for the GBTM. A p value < 0.05 was considered to indicate statistical significance.

Data availability statement

The data supporting the findings of this study are available from the corresponding author upon reasonable request.

RESULTS

Clinical characteristics

The clinical characteristics are summarized in Table 1. Among the 201 enrolled PD patients, 82 (40.8%) had euglycemia, 93 (46.3%) had intermediate glycemia, and 26 (12.9%) had hyperglycemia. Among diabetic patients, 4.9% were in the euglycemia group, 17.2% were in the intermediate glycemia group, and 53.8% were in the hyperglycemia group. The baseline characteristics of age, sex, BMI, disease duration, education years, smoking status, medical history of hypertension, coronary artery disease, stroke status, dyslipidemia status, APOE ε4 carrier status, LEDD, and FPG time interval were not significantly different among the FPG groups. The mean UPDRS part III score was significantly greater in the hyperglycemia group; however, neuropsychological domain z scores were not significantly different among the FPG groups. The frequencies of SH, OH, residual urine volume, early and delayed H/M ratio, and washout rate did not differ among the FPG groups.

Group-based trajectory modeling

The outcome of the GBTM is presented in Supplementary Figure 1 (in the online-only Data Supplement). A three-group model with zero-order specifications was adopted, delineating three distinct groups: Group 1 (n = 85, 40.8%, mean ± SD = 93.2 ± 7.0), Group 2 (n = 91, 45.3%, mean ± SD = 108.8 ± 10.4) and Group 3 (n = 25, 13.9%, mean ± SD = 137.8 ± 24.0) (Supplementary Table 2 in the online-only Data Supplement). The AvPP value was 0.82 or greater for each trajectory group, and the odds of correct classification indicated that the trajectory model had good accuracy. A mismatch of less than 0.01 for all groups supported the close correspondence between each trajectory group’s estimated probability and the actual proportion of individuals assigned to that particular subgroup (Supplementary Table 3 in the online-only Data Supplement).

Multivariate logistic regression models

The overall distribution of our patients by FPG time interval is presented in Supplementary Table 1 (in the online-only Data Supplement), with correlations between FPG level and outcome scores. The correlation between FPG and the UPDRS part III score was significant for the cumulative patient cohort up to 3 years (r = 0.1497, p = 0.0455). For the correlation between FPG and the neuropsychological domain composite z score, only the visuospatial function domain exhibited a significant correlation (r = 0.2123, p = 0.0157, at a cumulative FPG time interval of up to one year) (Supplementary Table 1 in the onlineonly Data Supplement).

Table 2 shows the results of the multivariate logistic regression analysis for motor and cognitive outcomes (Table 2). When the event category for unfavorable motor outcomes was defined as H&Y stage ≥ 2 (Model 1), intermediate glycemia (adjusted Or = 2.150, 95% confidence interval [CI] 1.029–4.489, p = 0.0416) and hyperglycemia (adjusted Or = 8.617, 95% CI 2.264–32.804, p = 0.0016) were identified as independent predictors of worsening motor outcomes. Additionally, age ≥ 65 years (adjusted Or = 4.811, 95% CI 1.945–11.903, p = 0.0007) and disease duration (unit = 12 months, adjusted Or = 1.255, 95% CI 1.003–1.035) were found to be independent predictors of unfavorable motor outcomes. For Model 3, cognitive outcomes, defined as a diagnosis of MCI and time to the onset of MCI, were compared among the three FPG groups. The results are presented in Table 2. Neither intermediate glycemia (adjusted Or = 1.078, 95% CI 0.501–2.321, p = 0.8473) nor hyperglycemia (adjusted Or = 0.772, 95% CI 0.237–2.517, p = 0.6672) was a significant predictor of cognitive impairment.

Survival analysis

Since the cross-sectional logistic regression model revealed that disease duration, as well as FPG level, was an independent predictor of unfavorable motor outcomes, we performed Cox survival analysis to assess the longitudinal relationship between time and covariates, including FPG level. The survival outcomes are presented in Table 3 and Figure 1. The median follow-up period was 25 months (range 1–114 months), with a mean follow-up of 30.0 ± 24.8 months.

For motor outcome, the endpoint was defined as H&Y stage ≥ 2 in Model 1 (Table 3). Ninety-six patients reached the endpoint, while 95 (49.7%) patients were censored. When the endpoint was defined as H&Y stage ≥ 3 in Model 2 for comparison (Supplementary Table 4 in the online-only Data Supplement), 19 patients reached the endpoint, and 172 patients (90.05%) were censored. For cognitive outcome, the endpoint was defined as conversion to MCI based on neuropsychological testing. Seventy-seven patients were cognitive converters, while 85 (52.5%) patients were censored (Table 3).

Cox proportional hazards regression analysis yielded adjusted hazard ratios (HRs), which are listed in Table 3. The survival curve according to FPG categories is presented in Figure 1. In Model 1, compared to the euglycemia group (reference), both the intermediate glycemia group (adjusted HR 1.747, 95% CI 1.083–2.816, p = 0.0221) and the hyperglycemia group (adjusted HR 3.864, 95% CI 1.996–7.481, p < 0.0001) showed faster motor progression, resulting in unfavorable motor outcomes (Figure 1A). Age ≥ 65 years (adjusted HR 2.304, 95% CI 1.181– 4.493, p = 0.0143) was also significantly associated with poor motor outcomes. In Model 2 (Supplementary Table 4 in the online-only Data Supplement), only the hyperglycemia group (adjusted HR 8.186, 95% CI 1.810–37.023, p = 0.0063) was associated with faster motor progression and a poor motor outcome. Comparing Model 2 to Model 1, intermediate glycemia was an independent predictor for faster motor progression at relatively early, mild stages of PD (H&Y stage < 2) but not at late, moderate stages of PD (H&Y stage < 3) in our cohort (Table 3, Supplementary Table 4 in the online-only Data Supplement, and Figure 1A). No significant collinearity was detected among the covariates, including FPG level, age, sex, BMI, smoking status, DM status, delayed H/M ratio and disease duration. The lowest tolerance was 0.6273, and the highest VIF was 1.5941. In particular, there was no significant collinearity between FPG level (tolerance = 0.7516, VIF = 1.3304) and diabetes (tolerance = 0.7701, VIF = 1.2985) in our models. Additionally, the subgroup analysis stratified by DM demonstrated that DM had no significant effect on motor outcomes in our model (Supplementary Table 5 in the online-only Data Supplement and Supplementary Figure 2 in the online-only Data Supplement).

Neither intermediate glycemia (adjusted HR 1.183, 95% CI 0.697–2.009, p = 0.5339) nor hyperglycemia (adjusted HR 1.297, 95% CI 0.601–2.800, p = 0.5078) was a significant independent predictor of early MCI (Table 3, Figure 1B). The cognitive outcomes showed similar trends between the MCI converter group (Model 3) (Table 3) and the MCI with dementia converter group combined under a single label (data not shown). Diabetes was not a significant independent predictor of cognitive outcomes in Model 3 (Table 3) and had no significant effect on cognitive outcomes (Supplementary Table 5 in the online-only Data Supplement).

DISCUSSION

In the present study, both intermediate glycemia and hyperglycemia emerged as significant predictors for faster motor progression, especially in the early and mild stages of PD. Additionally, hyperglycemia was identified as an independent predictor of poorer motor outcomes in patients with advancedstage PD. While our findings on motor outcomes are consistent with previous research, the consideration of intermediate glycemia as an independent predictor is a novel observation. It is important to note that this discovery lacks conclusive evidence for causality, and prospective clinical studies may be necessary to establish a causal relationship and demonstrate the therapeutic effects of glycemic control, particularly in the intermediate glycemia group. These studies would be beneficial for individualized PD management, incorporating active implementation of treatment options such as exercise for PD patients with impaired glucose tolerance (IGT). Exercise has been suggested to alleviate PD symptoms in the mild stages [32].

For cognitive outcome, a significant correlation was found between the FPG level and the visuospatial domain composite z score. However, neither intermediate glycemia nor hyperglycemia was associated with progressive cognitive decline in PD patients. Previous research has shown that visuospatial impairment may precede the clinical onset of Alzheimer’s disease [33], and parieto-occipital hypometabolism might predict early conversion to dementia in PD patients, regardless of initial cognitive status [34]. A recent study demonstrated positive correlations between left cerebellar cortical metabolism and executive functions, prefrontal dopaminergic tone, and working memory, as well as positive correlations between quantitative electroencephalography slowing in posterior leads and both memory and visuospatial functions [35]. In addition, 18F-fluorodeoxyglucose PET studies revealed that hyperglycemia decreased cerebral glucose metabolism in specific brain regions associated with Alzheimer’s disease [36,37]. In the present study, the FPG level was suggested to play a significant role, particularly in relation to visuospatial function, in PD patients. However, dysglycemia did not affect longitudinal early or mild cognitive decline in patients with PD in our study. This result might be influenced by several cohort characteristics. The relatively small sample size might contribute to the absence of significant results for longitudinal cognitive outcomes. Additionally, the majority of our patients were enrolled in the early stage of PD, and the mean age of each FPG group ranged from 70.1 ± 9.7 to 73.9 ± 9.1 years. These age ranges might be too young to properly assess cognitive impairment. A longer period of time is necessary for further investigation regarding the association between dysglycemia and dementia conversion in PD patients.

The present study has several limitations. First, the number of patients in the hyperglycemia group was relatively small, and the frequency of DM in the hyperglycemia group was relatively high (53.8%). However, a similar proportion of hyperglycemic patients did not have a diagnosis of DM (46.2%). For these patients, it remains unclear whether they have undiagnosed DM. Additionally, DM was defined by medical records in our model, not by FPG level, and the majority of DM (58.8%) patients had euglycemia (11.8%) or intermediate glycemia (47.0%). For these patients, it remains unclear whether they had well-controlled DM or were misdiagnosed. Our results suggested that the DM group might not be identical to the hyperglycemia group. Conducting repeated glucose measurements during the clinical follow-up of PD patients with or without DM is necessary. Information about diabetic medications or physical activity levels that could affect the motor progression of PD patients was not included in our study; therefore, the influence of these covariates cannot be fully excluded. The on-off information of each patient was not included in our model, which could be another limitation. Second, we included only idiopathic PD patients and excluded those with atypical PD. However, the diagnosis of atypical PD could be time dependent, as the initial clinical presentation might resemble that of PD. Since we included early and mild PD patients with relatively short disease durations, there is a possibility that the early stages of atypical PD might have been misdiagnosed as PD, which could be a potential confounding factor. Further investigation is needed to determine whether glycemic status has a different impact on motor outcomes between patients with idiopathic PD and patients with atypical PD. Third, although we included several nonmotor profiles, the range of profiles was limited. Glycemic status appeared to have no effect on nonmotor outcomes, including OH, SH, myocardial adrenergic function, and unexplained voiding difficulties, defined as postvoid urinary residual volume. These dysfunctions did not differ among the FPG groups. However, the effects of hyperglycemia and DM on cognition in PD patients remain unclear; therefore, comprehensive further investigations are warranted.

In summary, we provide the first evidence that IGT and hyperglycemia may be associated with faster motor progression and poorer motor outcomes in PD patients. Our findings suggest that managing IGT and hyperglycemia could be beneficial, especially during the early, mild stage of PD. We recommend routine monitoring of glucose intolerance in PD patients, and FPG level may be a reliable tool for this purpose.

Supplementary Materials

Notes

Conflicts of Interest

The authors have no financial conflicts of interest.

Funding Statement

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT and Future Planning (NRF-2017R1D1A1B06028086). This research was also supported by the Korea “National Institute of Health” research project (2021-ER1008-02).

Author Contributions

Conceptualization: Ko-Eun Choi, Joong-Seok Kim. Formal analysis: all authors. Funding acquisition: Joong-Seok Kim. Investigation: all authors. Methodology: Ko-Eun Choi, Joong-Seok Kim. Resources: Joong-Seok Kim. Writing—original draft: Ko-Eun Choi. Writing—review & editing: Dong-Woo Ryu, Yoon-Sang Oh, Joong-Seok Kim.

Acknowledgments

None

Figure 1.

Survival curve using the multivariate Cox proportional hazard model by FPG category. Survival curves were used to estimate the motor outcome after disease onset using FPG. Models controlled for age, sex, diabetes status, smoking status, and/or APOE ɛ4 carrier status. A: In Model 1, the endpoint outcome was defined as H&Y stage ≥ 2. The hyperglycemia and intermediate glycemia groups showed faster motor progression than did the euglycemia group. B: In Model 2, the endpoint outcome was defined as the conversion from normal cognition to MCI. Neither hyperglycemia nor intermediate glycemia was a significant predictor of cognitive impairment compared to euglycemia. FPG, fasting plasma glucose; APOE, apolipoprotein E; H&Y, Hoehn and Yahr; MCI, mild cognitive impairment.

REFERENCES

• 1. Hu G, Jousilahti P, Bidel S, Antikainen R, Tuomilehto J. Type 2 diabetes and the risk of Parkinson’s disease. Diabetes Care 2007;30:842–847.ArticlePubMedPDF
• 2. Xu Q, Park Y, Huang X, Hollenbeck A, Blair A, Schatzkin A, et al. Diabetes and risk of Parkinson’s disease. Diabetes Care 2011;34:910–915.ArticlePubMedPMCPDF
• 3. Noyce AJ, Bestwick JP, Silveira-Moriyama L, Hawkes CH, Giovannoni G, Lees AJ, et al. Meta-analysis of early nonmotor features and risk factors for Parkinson disease. Ann Neurol 2012;72:893–901.ArticlePubMedPMCPDF
• 4. Markaki I, Ntetsika T, Sorjonen K, Svenningsson P; BioPark Study Group. Euglycemia indicates favorable motor outcome in Parkinson’s disease. Mov Disord 2021;36:1430–1434.ArticlePubMedPDF
• 5. Athauda D, Evans J, Wernick A, Virdi G, Choi ML, Lawton M, et al. The impact of type 2 diabetes in Parkinson’s disease. Mov Disord 2022;37:1612–1623.ArticlePubMedPMCPDF
• 6. Bosco D, Plastino M, Cristiano D, Colica C, Ermio C, De Bartolo M, et al. Dementia is associated with insulin resistance in patients with Parkinson’s disease. J Neurol Sci 2012;315:39–43.ArticlePubMed
• 7. Cereda E, Barichella M, Pedrolli C, Klersy C, Cassani E, Caccialanza R, et al. Diabetes and risk of Parkinson’s disease: a systematic review and meta-analysis. Diabetes Care 2011;34:2614–2623.PubMedPMC
• 8. Palacios N, Gao X, McCullough ML, Jacobs EJ, Patel AV, Mayo T, et al. Obesity, diabetes, and risk of Parkinson’s disease. Mov Disord 2011;26:2253–2259.ArticlePubMedPMCPDF
• 9. De Iuliis A, Montinaro E, Fatati G, Plebani M, Colosimo C. Diabetes mellitus and Parkinson’s disease: dangerous liaisons between insulin and dopamine. Neural Regen Res 2022;17:523–533.ArticlePubMed
• 10. Mollenhauer B, Zimmermann J, Sixel-Döring F, Focke NK, Wicke T, Ebentheuer J, et al. Baseline predictors for progression 4 years after Parkinson’s disease diagnosis in the De Novo Parkinson Cohort (DeNoPa). Mov Disord 2019;34:67–77.ArticlePubMedPDF
• 11. Anstey KJ, Sargent-Cox K, Eramudugolla R, Magliano DJ, Shaw JE. Association of cognitive function with glucose tolerance and trajectories of glucose tolerance over 12 years in the AusDiab study. Alzheimers Res Ther 2015;7:48.ArticlePubMedPMCPDF
• 12. Chung HS, Lee JS, Kim JA, Roh E, Lee YB, Hong SH, et al. Fasting plasma glucose variability in midlife and risk of Parkinson’s disease: a nationwide population-based study. Diabetes Metab 2021;47:101195.ArticlePubMed
• 13. Pagano G, Polychronis S, Wilson H, Giordano B, Ferrara N, Niccolini F, et al. Diabetes mellitus and Parkinson disease. Neurology 2018;90:e1654–e1662.ArticlePubMed
• 14. Yang G, Au Yeung SL, Schooling CM. Sex differences in the association of fasting glucose with HbA1c, and their consequences for mortality: a Mendelian randomization study. EBioMedicine 2022;84:104259.ArticlePubMedPMC
• 15. Fiory F, Perruolo G, Cimmino I, Cabaro S, Pignalosa FC, Miele C, et al. The relevance of insulin action in the dopaminergic system. Front Neurosci 2019;13:868.ArticlePubMedPMC
• 16. Gibb WR, Lees AJ. The relevance of the Lewy body to the pathogenesis of idiopathic Parkinson’s disease. J Neurol Neurosurg Psychiatry 1988;51:745–752.ArticlePubMedPMC
• 17. Schade S, Mollenhauer B, Trenkwalder C. Levodopa equivalent dose conversion factors: an updated proposal including opicapone and safinamide. Mov Disord Clin Pract 2020;7:343–345.ArticlePubMedPMCPDF
• 18. Kim JS, Park HE, Park IS, Oh YS, Ryu DW, Song IU, et al. Normal ‘heart’ in Parkinson’s disease: is this a distinct clinical phenotype? Eur J Neurol 2017;24:349–356.ArticlePubMedPDF
• 19. Postuma RB, Berg D, Stern M, Poewe W, Olanow CW, Oertel W, et al. MDS clinical diagnostic criteria for Parkinson’s disease. Mov Disord 2015;30:1591–1601.ArticlePubMed
• 20. American Diabetes Association Professional Practice Committee. 2. Classification and diagnosis of diabetes: standards of medical care in diabetes— 2022. Diabetes Care 2022;45(Suppl 1):S17–S38.ArticlePubMedPDF
• 21. Nagin DS, Jones BL, Passos VL, Tremblay RE. Group-based multi-trajectory modeling. Stat Methods Med Res 2018;27:2015–2023.ArticlePubMedPDF
• 22. Ryu HJ, Yang DW. The Seoul Neuropsychological Screening Battery (SNSB) for comprehensive neuropsychological assessment. Dement Neurocogn Disord 2023;22:1–15.ArticlePubMedPMCPDF
• 23. Oh YS, Yoo SW, Lyoo CH, Yoo JY, Yoon H, Ha S, et al. The association of β-amyloid with cognition and striatal dopamine in early, non-demented Parkinson’s disease. J Parkinsons Dis 2021;11:605–613.ArticlePubMed
• 24. Litvan I, Goldman JG, Tröster AI, Schmand BA, Weintraub D, Petersen RC, et al. Diagnostic criteria for mild cognitive impairment in Parkinson’s disease: Movement Disorder Society Task Force guidelines. Mov Disord 2012;27:349–356.ArticlePubMedPMCPDF
• 25. Goldman JG, Holden S, Ouyang B, Bernard B, Goetz CG, Stebbins GT. Diagnosing PD-MCI by MDS Task Force criteria: how many and which neuropsychological tests? Mov Disord 2015;30:402–406.ArticlePubMedPDF
• 26. Busse A, Hensel A, Gühne U, Angermeyer MC, Riedel-Heller SG. Mild cognitive impairment: long-term course of four clinical subtypes. Neurology 2006;67:2176–2185.ArticlePubMed
• 27. Emre M, Aarsland D, Brown R, Burn DJ, Duyckaerts C, Mizuno Y, et al. Clinical diagnostic criteria for dementia associated with Parkinson’s disease. Mov Disord 2007;22:1689–1707.ArticlePubMed
• 28. Agostini D, Carrio I, Verberne HJ. How to use myocardial 123I-MIBG scintigraphy in chronic heart failure. Eur J Nucl Med Mol Imaging 2009;36:555–559.ArticlePubMedPDF
• 29. Ryu DW, Kim JS, Lee JE, Oh YS, Yoo SW, Yoo IR, et al. Initial versus follow-up sequential myocardial 123I-MIBG scintigraphy to discriminate Parkinson disease from atypical Parkinsonian syndromes. Clin Nucl Med 2019;44:282–288.ArticlePubMed
• 30. Fanciulli A, Jordan J, Biaggioni I, Calandra-Buonaura G, Cheshire WP, Cortelli P, et al. Consensus statement on the definition of neurogenic supine hypertension in cardiovascular autonomic failure by the American Autonomic Society (AAS) and the European Federation of Autonomic Societies (EFAS): endorsed by the European Academy of Neurology (EAN) and the European Society of Hypertension (ESH). Clin Auton Res 2018;28:355–362.ArticlePubMedPMCPDF
• 31. Freeman R, Wieling W, Axelrod FB, Benditt DG, Benarroch E, Biaggioni I, et al. Consensus statement on the definition of orthostatic hypotension, neurally mediated syncope and the postural tachycardia syndrome. Clin Auton Res 2011;21:69–72.ArticlePubMedPDF
• 32. van der Kolk NM, de Vries NM, Kessels RPC, Joosten H, Zwinderman AH, Post B, et al. Effectiveness of home-based and remotely supervised aerobic exercise in Parkinson’s disease: a double-blind, randomised controlled trial. Lancet Neurol 2019;18:998–1008.ArticlePubMed
• 33. Mapstone M, Steffenella TM, Duffy CJ. A visuospatial variant of mild cognitive impairment: getting lost between aging and AD. Neurology 2003;60:802–808.ArticlePubMed
• 34. Baba T, Hosokai Y, Nishio Y, Kikuchi A, Hirayama K, Suzuki K, et al. Longitudinal study of cognitive and cerebral metabolic changes in Parkinson’s disease. J Neurol Sci 2017;372:288–293.ArticlePubMed
• 35. Orso B, Arnaldi D, Peira E, Famá F, Giorgetti L, Girtler N, et al. The role of monoaminergic tones and brain metabolism in cognition in de novo Parkinson’s disease. J Parkinsons Dis 2022;12:1945–1955.ArticlePubMed
• 36. Burns CM, Chen K, Kaszniak AW, Lee W, Alexander GE, Bandy D, et al. Higher serum glucose levels are associated with cerebral hypometabolism in Alzheimer regions. Neurology 2013;80:1557–1564.ArticlePubMedPMC
• 37. Kawasaki K, Ishii K, Saito Y, Oda K, Kimura Y, Ishiwata K. Influence of mild hyperglycemia on cerebral FDG distribution patterns calculated by statistical parametric mapping. Ann Nucl Med 2008;22:191–200.ArticlePubMedPDF

Figure & Data

References

Citations

Citations to this article as recorded by  

Comments on this article